Survey
* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project
* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project
CLASSIFICATION OF POWER QUALITY PROBLEMS USING WAVELET BASED ARTIFICIAL NEURAL NETWORK| 1 CHAPTER1 INTRODUCTION 1.1 Introduction The increased requirements on supervision, control and performance in modern power systems make power quality monitoring a common practice for utilities. Studies of power quality phenomena have emerged as an important subject in recent years due to renewed interest in improving the quality of the electric supply. As sensitive electronic equipment continues to proliferate the studies of power quality will be further emphasized. New tools are required to extract all relevant information from the recordings in an automatic way. It is well known that the main power quality deviations are caused by shortcircuits, harmonic distortions, notchings, voltage sags and swells, as well as transients due to load switching. In order do correct such problems, it is required, in general that, firstly, they should be detected and identified. Nevertheless, whenever the disturbance lasts for only for a few cycles, a simple observation of the waveform in a busbar may not be enough to allow one to recognize that there is a problem in there or, more difficult yet, to identify the sort of the problem. Switching phenomena results in oscillatory transients in the electrical supply, for example capacitors switching contributes considerably to power quality (PQ) disturbances. In addition, high power non-linear loads add to the generation of current and voltage harmonic components. Among the different voltage disturbances that can produce the most noteworthy and critical power quality problems are voltage sags. Short-term voltage drops (sags) can trip electrical drives or more sensitive equipment, leading to interruptions of vital production. For these reasons, from the consumer point of view, power quality issues turn out to be key factor for satisfying good productivity. On other part, for electrical supply industry, the quality of power delivered will be one of the distinguishing factors intended for ensuring customer loyalty in the present competitive and deregulated market. International standards define power quality as the physical characteristics of the electrical supply provided under normal operating conditions that do not disrupt or upset the customer’s process Dept of EEE KMCET - DVK CLASSIFICATION OF POWER QUALITY PROBLEMS USING WAVELET BASED ARTIFICIAL NEURAL NETWORK| 2 . As a result, power quality problems arise with the association of any deviation in voltage, current or frequency resulting into an undesirable performance of customer’s equipment. Though, it is important to notice that the quality of power supply implies basically voltage quality and supply reliability PQ disturbances result in complete interruption or force many applications to halt working. So PQ problems relate to the basic system design, system maintenance issues, ensuring equipment protection within customer facilities. The impact of PQ problems is aggregated through proliferation of extremely sensitive computerized equipment. As the design of most conventional equipments such as lights, constant speed motors and power systems are based on the assumption that the voltage source are sinusoidal and the loads are linear, so that the resulting waveforms are also sinusoidal. As demands on accuracy have increased and nonlinear loads have become more common, this approximation is now no longer valid. Consequently, power quality monitoring and classification has become an essential service which many utilities perform for their customers. The monitoring and analysis tools must be able to detect, identify, and localize the disturbances on the supply lines and make proper system decisions. The whole process should be completed before these cause widespread customer complaints; equipments malfunction and even equipment failures. So fast and accurate identification and classification of power quality events is extremely important at all levels of the electrical power system and is a value for both power distributor and power consumer. However, there is still further scope for research work to identify and classify the power quality problems with very high accuracy so as to keep the losses due to power quality problems at a minimum. The wavelet transform is a mathematical tool like Fourier Transform in analyzing a signal that decomposes a signal into different scales with different levels of resolution. Santoso et al proposed wavelet transform technique for the detection and localization of the actual power quality disturbances. They explored the potential of wavelet transform as a new tool for automatically classifying power quality disturbances. Heydt and Galli proposed wavelet techniques for the identification of the power system transient signals. Fuzzy logic control technique has been discussed by Hiyama et al to enhance power system stability using static VAR compensator. The proposed control scheme is simple and suitable for on-line implementation using a microcontroller. Santoso et al combined wavelet transform with Fourier transform Dept of EEE KMCET - DVK CLASSIFICATION OF POWER QUALITY PROBLEMS USING WAVELET BASED ARTIFICIAL NEURAL NETWORK| 3 for the characterization of the power quality events. The Fourier transform has been used to characterize steady state phenomena, whereas the wavelet transform has been applied to transient phenomena. A hybrid scheme using a Fourier linear combiner and a fuzzy expert system for the classification of transient disturbance waveforms in a power system has been presented by Dash et al. Olivier et al investigated the use of a continuous wavelet transform to detect and analyze voltage sags and transients. They developed an efficient and simple algorithm for detecting and measuring power quality analysis. Thebasic unit of the architecture is the wavelet network which combines the ability of the wavelet transform for analyzing non stationary signals with the classification capability of artificial neural networks Styvaktakis et al developed an expert system to classify different types of power system events and offer useful information in terms of power quality. Huang et al presented a neuralfuzzy technology based classifier for the recognition of power quality disturbances. The proposed recognition system provides a promising approach applicable in power quality monitoring. Driesen and Belmans proposed further alternatives for power quantification based on formulations and time frequency domain described by wavelet bases. They used both the real and complex valued wavelets. He and Starzyk proposed a novel approach for power quality disturbances classification based on wavelet transform and self organizing learning array system. Wavelet transform has been utilized here to extract feature vectors for various power quality disturbances based on multi resolution analysis. Owing to various reasons mentioned above an attempt has been made in the present work to classify the most commonly occurring power quality disturbances using wavelet analysis, Multi-resolution signal decomposition (MSD) using wavelet transformation (WT) is used as a basis for feature vector for obtaining the training and testing data. 1.2Literature survey A.K. Chandel, G. Guleria, and R. Chandel [2] has explained about the various power quality disturbances, and how they are classified using wavelet analysis. In the pre-processing voltage signals are captured during a power quality problem. Due to the absence of field results different power quality problems have been simulated in Dept of EEE KMCET - DVK CLASSIFICATION OF POWER QUALITY PROBLEMS USING WAVELET BASED ARTIFICIAL NEURAL NETWORK| 4 the MATLAB/SIMULINK environment. The PQ data thus obtained matches the field data. Subsequently, using the Wavelet toolbox of the MATLAB, multi signal decomposition on the captured signals is performed. Depending on the selected resolution levels, the PQ signal is decomposed into a number of wavelet levels. If the resolution level is defined as n, after decomposing the signal, there will be one level of approximation coefficients with n level of detail coefficients. The information contained in the coefficients of details and approximations is very useful for further processing M. H. Bollen [1] has explained ,what is power quality, what is quality of supply, and IEEE definitions for power quality. He explained about two types of power quality problems .they are voltage sag and interruption. The causes for this problems and their mitigation methods. Resende, J.W., Chaves, M.L.R., Penna, C. [3] This paper presents an approach that is able to provide the detection and location in time as well as the identification of power quality problems present in both transient and steady-stable signals. The method was developed by using the discrete wavelet transform (DWT) analysis. The given signal is decomposed through wavelet transform and any change on the smoothness of the signal is detected at the finer wavelet transform resolution levels. Later, the energy curve of the given signal is evaluated and a relationship between this energy curve and the one of the corresponding fundamental component is established. The paper shows that each power quality disturbance has unique deviations from the pure sinusoidal waveform and this is adopted to provide a reliable classification of the type of disturbance. Sudipta Nath, Arindam Dey and Abhijit Chakrabarti[4] presented the features that characterize power quality disturbances from recorded voltage waveforms using wavelet transform. The discrete wavelet transform has been used to detect and analyze power quality disturbances. The disturbances of interest include sag, swell, outage and transient. A power system network has been simulated by Electromagnetic Transients Program. Voltage waveforms at strategic points have been obtained for analysis, which Dept of EEE KMCET - DVK CLASSIFICATION OF POWER QUALITY PROBLEMS USING WAVELET BASED ARTIFICIAL NEURAL NETWORK| 5 includes different power quality disturbances. Then wavelet has been chosen to perform feature extraction. The outputs of the feature extraction are the wavelet coefficients representing the power quality disturbance signal. Wavelet coefficients at different levels reveal the time localizing information about the variation of the signal. S. El Safty, and A. El-Zonkoly [8] a novel algorithm is presented to detect the fault and identify its type using wavelet entropy energy. The main principle of wavelet entropy is first presented. Then algorithm used in fault detection and identification is discussed. Simulation of the proposed system in order to obtain different current signals for testing of the proposed technique has been done using PSCAD program. 1.3 Organization of the thesis Chapter 1 explains about background and overview of power quality wavelet analysis Chapter 2 explains about power quality and power quality problems and their mitigation methods Chapter 3 explains about classification of loads Chapter 4 explains about an overview of wavelet analysis Chapter 5 explains about simulink results and conclusion Dept of EEE KMCET - DVK CLASSIFICATION OF POWER QUALITY PROBLEMS USING WAVELET BASED ARTIFICIAL NEURAL NETWORK| 6 CHAPTER 2 INTRODUCTION TO POWER QUALITY 2.1 POWER QUALITY: International standards define power quality as the physical characteristics of the electrical supply provided under normal operating conditions that do not disrupt or upset the customer’s process. Now days, power quality is a key issue for the producers, distributors and consumers. The most common problems, like harmonics, short term voltage variations (sags, swells and interruptions), long term voltage variations (under voltages, over voltages and interruptions), transients, unbalance, frequency variations and others, can cause several problems to the consumers which require high levels of power quality for their industrial processes or home use. Power quality studies are the necessary first step in order to determine what is wrong, so that measures can be taken to solve the problems. Because many of the commercially available equipments are either too expensive, or have too many limitations, it was decided to develop a new low-cost power quality monitor that could be an alternative to the equipment in the market. The utility is most commonly perceived by the user as the source of power quality and reliability problems. This is understandable because the effects of events in the mains, like lightning or breaker clearing, are fairly noticeable in a facility’s performance. However, these events represent a small percentage of total events, although they may be true in some specific areas such as weak grids. The reality is that a large percentage of problems are generated within a user’s own facility. Even though it may seem unbelievable, the main source of power disturbances in a facility is its own activity. Normal utilization of the facility and normal operation of its own equipment, including turn-on and turn-off, may give rise to undesirable events which affect the sensitive equipment within the facility if it has not been taken into account in the design process. The increasing spread of information technology and automation to practically every business and activity is mainly responsible. The Dept of EEE KMCET - DVK CLASSIFICATION OF POWER QUALITY PROBLEMS USING WAVELET BASED ARTIFICIAL NEURAL NETWORK| 7 increasing application of electronics to all kind of appliances makes possible real-time communication and/or continuous operation. This improves facilities performance levels through effective data and energy management. But electronic devices are non-linear loads and they introduce more disturbances into the same facility they are fed by. Therefore, facilities and sensible equipment are becoming more vulnerable and dependent on both the quality and continuity of the electric power supply because of their operation. Typical examples of power quality problems are overloaded circuits and transformers, and unwanted operation of protection caused by the presence of harmonics . Therefore, power quality, reliability and availability must be considered as a complete set from the design stage because they are closely related. Equipment is designed for normal operation within a rated range of power supply values. Small deviations from these values can result in increased losses, poor efficiency and unpredictable operation. Large deviations can cause protective devices to trip or the failure of a component. In addition, components always break down from time to time because they have a limited lifetime. Thus, both power quality and reliability influence the availability of any facility. Availability is easier to estimate and quantify than reliability, so users are more concerned about the former. Nevertheless, the way to increase the availability is through power quality and reliability improvements. But high availability means much more than high power quality and reliability. At first, it requires a proper design to reduce failures or disturbances in the power supply. The reliability levels wanted or needed for the equipment of a business or an activity are known at the outset. So, if the reliability level must be higher than the one provided by the utility according to power quality and availability standards, additional investment is needed in order to achieve the expected power quality and availability for the facility. 2.2 QUALITY OF SUPPLY According to the Council of European Regulators (CEER) Working Group on Quality of Supply, the following definitions apply: • Customer service. • Continuity of supply (typically referred to as reliability in the USA). Dept of EEE KMCET - DVK CLASSIFICATION OF POWER QUALITY PROBLEMS USING WAVELET BASED ARTIFICIAL NEURAL NETWORK| 8 • Voltage quality (typically referred to as power quality in the USA). 2.2.1 Customer Service Customer service typically relates to the nature and quality of the service provided to electricity consumers. Customer service (referred to as commercial quality in Europe) indicators are typically quantitative operational measures that are used extensively in existing performance-based ratemaking (PBR) schemes throughout the world. Some examples of customer service indicators utilized include: • Customer satisfaction. • Customer complaints. • Customer meter and billing accuracy. • Response to customer enquiries. • Appointments met. • Customer call/wait times. • Time required for new service connections. • Emergency/storm response. • Safety/health. • Estimating charges for work. Although important in assessing the overall quality of supply, customer service is not the focus of this research and will largely not be addressed in this chapter. There is a variety of information and resources available to policy makers and utilities regarding this component of the quality of supply, some of which are referenced at the end of this chapter. 2.2.2 Continuity of Supply Continuity of supply is characterized by the number and duration of interruptions. Several indicators are used to evaluate the continuity of supply in transmission and distribution networks. Regulation can aim to compensate customers for very long supply interruptions, keep restoration times under control and create Dept of EEE KMCET - DVK CLASSIFICATION OF POWER QUALITY PROBLEMS USING WAVELET BASED ARTIFICIAL NEURAL NETWORK| 9 incentives to reduce the total number and duration of interruptions (and disincentives to increase them). Different methods and accuracies of measuring interruptions and in assigning liability for each of them create problems in regulating continuity of supply. 2.2.3 Voltage Quality The third and final component of the quality of supply is voltage quality. Voltage quality is the quantitative form of describing power quality and includes both steady-state power quality variations and momentary disturbances that may impact loads. Categories of voltage quality include: Power frequency Magnitude of the supply voltage Harmonics and inter harmonics Voltage unbalance Flicker Voltage sags and momentary interruptions (r.m.s. variations) Transients 2.3 Power Quality Deterioration: 2.3.1 Voltage Dips The main causes of voltage dips are: short-circuits in the supply network, start-up of large loads, faulted grounding system and failures of the customer facilities Fig 2.1 voltage dip Dept of EEE KMCET - DVK CLASSIFICATION OF POWER QUALITY PROBLEMS USING WAVELET BASED ARTIFICIAL NEURAL NETWORK| 10 Sources of Voltage Dips The primary source of voltage dips is the electrical short circuit occurring on the power supply system. The switching of large loads, starting of large motors connected to the end of a long supply line power fluctuations of great magnitude (particularly of reactive power) are characteristic of some categories of loads and installations, such as variable load and/or speed drives, arc furnaces, welding equipment, etc., 2.3.2 Short Supply Interruptions Short interruptions are typically associated with switchgear operation related to the occurrence and termination of short circuits in the system or installations connected to it .The operation of a circuit-breaker or fuse disconnects part of the system from the source of energy. In the case of a radial circuit, this interrupts the supply to all downstream parts of the system. In the case of a meshed network, disconnections at more than one point are necessary in order to clear the fault. Electric power users within the disconnected segment of network suffer an interruption of supply. Automatic reclosing sequences are often applied in the supply system. Their purpose is to restore the circuit to normal operation with the minimum of delay in the event that the fault proves to be a transient one. The reclosing operation may be attempted several times (depending on the adopted practice of fault clearing) until self-clearance of the fault, or the circuit-breaker remains in the open position if the fault is a permanent one. Interruptions having duration up to 1 min (or, in the case of some reclosing systems, up to 3 min) are classified typically as short interruptions. Dept of EEE KMCET - DVK CLASSIFICATION OF POWER QUALITY PROBLEMS USING WAVELET BASED ARTIFICIAL NEURAL NETWORK| 11 Fig.2.2 short supply interruptions: r.m.s voltage characteristics Fig. 2.3 Short supply interruptions: voltage waveform 2.3.3 Long interruptions The importance of labor costs increases, reaching some 20% on average. This is twice as great as the equivalent figure for voltage dips and some 40% higher than with short interruptions. In addition there are instances where some of the other cost categories take on greater importance and these tend to be related to the long-term economic consequences created by penalties (commercial or statutory), loss of brand equity or the need for unanticipated business investment to regain lost sales/market share. Dept of EEE KMCET - DVK CLASSIFICATION OF POWER QUALITY PROBLEMS USING WAVELET BASED ARTIFICIAL NEURAL NETWORK| 12 2.3.4 Voltage fluctuations: Voltage fluctuations are caused when loads draw currents having significant sudden or periodic variations. The fluctuating current that is drawn from the supply causes additional voltage drops in the power system leading to fluctuations in the supply voltage. Loads that exhibit continuous rapid variations are thus the most likely cause of voltage fluctuations. Examples of loads that may produce voltage fluctuations in the supply include Arc furnaces Arc welders Installations with frequent motor starts (air conditioner units, fans) Motor drives with cyclic operation (mine hoists, rolling mills) Equipment with excessive motor speed changes (wood chippers, car shredders) 2.3.5 Harmonics: A harmonic is defined as a component with a frequency that is an integer multiple (the so called order of harmonic n) of the fundamental frequency (Figure 7.1). The harmonic number indicates the harm onic frequency: the first harmonic is the fundamental frequency (50 or 60 Hz), the second harmonic is the component with frequency two times the fundamental (100 or 120 Hz), and so on. Fig.2.4 harmonics Dept of EEE KMCET - DVK CLASSIFICATION OF POWER QUALITY PROBLEMS USING WAVELET BASED ARTIFICIAL NEURAL NETWORK| 13 Effects due to harmonics: 1.Thermal effects on transformers 2.Neutral conductor overloading 3.Over stressing of capacitor brands 4.Unexpected fuse operation 5.Abnormal operation of electronic relays 6.Telephonic interference 7.Thermal effects on rotating machines 8.Pulsating torque in rotating machines Sources of Current Harmonics Among the sources of harmonic voltages and currents in power systems three groups of equipment can be distinguished: magnetic core equipment, like transformers, electric motors, generators, etc. arc furnaces, arc welders, high-pressure discharge lamps, etc. electronic and power electronic equipment. 2.3.6 transient over voltages Transient over voltages are brief, high-frequency increases voltage on AC mains. Brodly speaking here are two different types of transient over voltages, low frequency transients with frequency components in the few hundred hertz region typically caused by capacitor switching, and high frequency transients with frequency components in the few hundred kilo-hertz region typically caused by lighting and inductive loads. low frequency transients are often called” capacitor switching transients”, high frequency transients are often called “impulse” ,”spike”, or surge. Fig 2.5 transient over voltages Dept of EEE KMCET - DVK CLASSIFICATION OF POWER QUALITY PROBLEMS USING WAVELET BASED ARTIFICIAL NEURAL NETWORK| 14 2.3.7 Voltage Sag A decrease in voltage to between 10 % to 90% of nominal voltage for onehalf cycle to one minute . Fig.2.6voltagesag 2.3.8 Under Voltage Under voltage is a decrease in voltage below 90% of its nominal value for more than one minute. Fig.2.7 under voltage 2.3.9 Swell A swell is the opposite of a sag an increase in voltage above 110% of nominal voltage for one- half cycle to one minute. Fig.2.8 voltage swell Dept of EEE KMCET - DVK CLASSIFICATION OF POWER QUALITY PROBLEMS USING WAVELET BASED ARTIFICIAL NEURAL NETWORK| 15 2.3.10 Flicker Flicker is a low-frequency phenomenon in which the magnitude of voltage or frequency changes, that is perceptible to human eye. Flicker frequency of 8 to 10 hz. Voltage variation of 0.3 to 0.4%. 2.3.11 Over Voltage: Over voltage is an increase in voltage above 110% of nominal for more than one minute. Reasons of cause: - Load Rejection Effects: - Overheat of electronic equipments Possible Solution: Power conditioner Fig.2.9 over voltage 2.4 Cost of poor power quality: While the consequences are not limited to electrical systems in non-residential environments and can harm domestic systems and appliances as well as mainly introducing electrical danger into the home, the residential sector is not addressed Dept of EEE KMCET - DVK CLASSIFICATION OF POWER QUALITY PROBLEMS USING WAVELET BASED ARTIFICIAL NEURAL NETWORK| 16 here. It is only by knowing what the cost consequences are that effective plans can be drawn up to reduce or eliminate these unnecessary productivity and efficiency losses and to design in solutions that make economic as well as operational sense.. The economic impacts of power quality are usually divided into three broad categories: Direct economics impacts: loss of production; unrecoverable downtime and resources (e.g. raw material, labor, capital); process restart costs; spoilage of (semi-)finished production; equipment damage; direct costs associated with human health and safety; financial penalties incurred through non-fulfillment of contract; environmental financial penalties; utility costs associated with the interruption. Indirect economic impacts: the costs to an organization of revenue/income being postponed; the financial cost of loss of market share; the cost of restoring brand equity. Social economic impacts: uncomfortable building temperatures as related to reduction in efficient working/health and safety; personal injury or fear, also as related to reduction in efficiency and health and safety; evacuating neighboring residential buildings as an indirect social impact in the event of failure of industrial safety, as it relates to the additional costs incurred by an organization that has to carry out these measures. The harmonics category may be further subdivided to focus on particular impacts that were treated as separate cases, defined as follows: Dept of EEE KMCET - DVK CLASSIFICATION OF POWER QUALITY PROBLEMS USING WAVELET BASED ARTIFICIAL NEURAL NETWORK| 17 Overstressing insulation, effects on electrical equipment including transformers, capacitors , motors; the consequences of not measuring TRMS (True Root Mean Square); additional losses. Overheating of the neutral conductor (e.g. burn-off and subsequent disruption or damage to electrical equipment). Nuisance tripping of protective devices. Malfunction of equipment control systems due to additional zero crossing. To calculate the total cost of each PQ disturbance the following PQ cost categories are to be considered for each disturbance category. 2.5 Other disturbances: There have been no other major surveys or studies on costs of PQ phenomena apart from the Leonardo PQ survey already mentioned above. Individual case studies, however, do exist, which cover such phenomena as: Inter harmonics Unbalance Problems with earthing and high-frequency phenomena These are responsible for substantial PQ costs within particular sites and regions. The approach here, however, is limited rather to analyzing historical events, recording the consequences mainly in terms of damaged equipment and associated costs. 2.6. Over view of mitigation methods To understand the various ways of mitigation, the mechanism leading to an equipment trip needs to be understood. The underlying event of the equipment trip is a short-circuit fault, a low-impedance connection between two or more phases, or between one or more phases and ground. At the fault position the voltage drops to a low value. The effect of short circuit at other positions in the system is an event of a certain magnitude and duration at the interface between the equipment and the power system. The short circuit fault will always causes a voltage sag for some customers. If the fault takes place in a radial part of the system, the protection intervention clearing the fault will also leads to an interruption. Dept of EEE KMCET - DVK CLASSIFICATION OF POWER QUALITY PROBLEMS USING WAVELET BASED ARTIFICIAL NEURAL NETWORK| 18 2.6.1 Reducing the number of short-circuit faults. Reducing the number of short-circuit faults in a system not only reduces the sag frequency but also the frequency of sustained interruptions. This is a very effective way of improving the quality of supply and many customers suggest this as the obvious solution when a voltage sag or a short interruption problem occurs. Some examples of fault mitigation are Replace overhead lines by underground cables Use covered wires for overhead line Implement a strict policy of tree trimming Install additional shielding wires Increase the insulation level Increase maintenance and inspection frequencies 2.6.2 Reducing the fault-clearing time Reducing the fault-clearing time does not reduce the number of events but only their severity. Faster fault-clearing does also not affect the number of voltage sags but it can significantly limit the sag duration. The ultimate reduction in fault clearing time is achieved by using current limiting fuses. Current limiting fuses are able to clear a fault with in one half-cycle, so that the duration of a voltage will rarely exceed one cycle. The recently introduced static circuit breakers also gives a fault clearing time within one half-cycle, but it is obviously much more expensive than a current limiting fuse. Some options for the reduction of the fault clearing time in transmission systems are In some cases faster circuit breakers could be of help. A certain reduction in grading margin is probably possible. Faster backup protection is one of the few effective means of reducing fault clearing time in transmission systems. Dept of EEE KMCET - DVK CLASSIFICATION OF POWER QUALITY PROBLEMS USING WAVELET BASED ARTIFICIAL NEURAL NETWORK| 19 2.6.3 Changing the power system By implementing changes in the supply system, the severity of the event can be reduced. But the costs become very high, especially for the transmission and sub transmission voltage levels. Some mitigation methods especially directed toward voltage sags are Install a generator the sensitive load the generators will keep the voltage up during a sag due to a remote fault Split buses or substations in the supply path to limit the number of feeders in the exposed area. Install current limiting coils at strategic paces in the system to increase the ‘clearing distance” to the fault. Feed the bus with the sensitive equipment from two or more substations. The best mitigation effect is by feeding from two different transmission substations. 2.6.4 installing mitigation equipment The most commonly applied method of mitigation is the installation ofaditional equipment at the system-equipment interface. The popularity of mitigation equipment is explained by it being the only place where the customer has control over the situation. Voltage- source converter: most modern voltage-sag mitigation methods at the system-equipment interface contain a so called voltage source converter. A voltage source converter is a power electronic device which can generate a sinusoidal voltage at any required frequency, magnitude, and phase angle. In voltage mitigation it is used to temporarily replace the supply voltage or to generate the part of the supply voltage which is missing. Series voltage controllers-DVR: The series voltage controller consists of a voltage source converter in series with the supply voltage. The dynamic voltage Dept of EEE KMCET - DVK CLASSIFICATION OF POWER QUALITY PROBLEMS USING WAVELET BASED ARTIFICIAL NEURAL NETWORK| 20 restorer(DVR) is commonly used instead of series voltage controller. The amount of energy storage depends on the power delivered by the converter and on the maximum duration of a sag. The controller is typically designed for certain maximum sag duration and a certain minimum sag voltage. combined shunt and series controllers: The series controller cannot mitigate any interruptions and that it is normally not designed to mitigate very deep sags(much below 50% of remaining voltage). A series-connected converter injects the missing voltage, and a shunt-connected converter takes a current from the supply. The power taken by the shunt controller must be equal to the power injected by the series controller. The change in stored energy in the capacitor is determined by the difference between the power injected by the series converter and the power taken from the supply by the shunt converter. Back up sources-SMESS.BESS: One of the main disadvantage of a series controller is that it can not operate during an interruption. A shunt controller operates during an interruption, but its storage requirements are higher. Various forms of energy storage have been proposed. A so called “ super conducting magnetic energy storage (SMESS)” stores electrical energy in a super conducting coil. A BESS or “battery storage system uses a large battery bank to store the energy. Cascade connected voltage controllers-UPS: The main device used to mitigate voltage sags and interruptions at the interface is the so-called “ uninterruptible power supply (UPS)” the popularity of the UPS is based on its low costs and easy use. Motor-generator set: motor generator sets are often depicted as noisy and as needing much maintenance. But industrial environments noisy equipment and maintenance on rotating machines are rather normal. Large battery blocks also require maintenance, expertise on which is much less available. 2.6.5 improving equipment immunity Improvement of equipment immunity is probably the most effective solution against equipment trips due to voltage sags. But it is often not suitable as a short time solution. Dept of EEE KMCET - DVK CLASSIFICATION OF POWER QUALITY PROBLEMS USING WAVELET BASED ARTIFICIAL NEURAL NETWORK| 21 The immunity of consumer electronics, computers, and control equipment can be significantly improved by connecting more capacitance to the internal dc bus. Single-phase low power equipment can also be improved by using a more sophisticated dc/dc converter The main source of concern are adjustable-speed drives. 2.7. Summary This chapter has explained about the power quality, and the quality of supply. In addition to this the various power quality problems are also studied, they are like capacitor switching, sag, swell, interruption etc. The mitigating methods for various power quality problems are studied. Dept of EEE KMCET - DVK CLASSIFICATION OF POWER QUALITY PROBLEMS USING WAVELET BASED ARTIFICIAL NEURAL NETWORK| 22 CHAPTER 3 3. CLASSIFICATION OF LOADS 3.1 Introduction Unbalanced operating conditions in an electric power system are caused mainly by the operation of unbalanced loads. Most low-voltage loads and certain medium-voltage ones, e.g. an electric traction motors, are single-phase appliances. Operation of such equipment in the three-phase system results in unbalanced load currents. Consequently, unsymmetrical voltage drops in individual phases of the supply system are produced, thus voltage at nodes of the network becomes unbalanced. Three-phase loads which may introduce unbalance in the power system are arc furnaces. The disturbance results from different impedances of high-current paths of the furnace and not equal phase loads being the effect of the physical nature of the melting process, i.e. variations in the arc impedance. As arc furnaces are the devices of relatively large power (tens or even hundreds of MVA), the furnace load unbalance may result in significant voltage unbalance in the supply system. The sources of unbalance can also be three-phase components of the transmission system, in particular overhead lines. Due to different tower geometries the conductors of individual phases are not simultaneously at the same location as each other and to earth. Following this, the line has different values of phase parameters, and also the values of a voltage loss in individual phases are different. Effects: The negative-sequence and zero-sequence currents flowing in an electric power system result in • Additional losses of power and energy; • Additional heating, the consequence of which is the limitation of line transmission capability for positive-sequence currents; • Voltage unbalance at nodes of the network. Voltage unbalance adversely affects the operation of loads. Asynchronous motors, synchronous generators and rectifiers are the most sensitive loads in this respect. Dept of EEE KMCET - DVK CLASSIFICATION OF POWER QUALITY PROBLEMS USING WAVELET BASED ARTIFICIAL NEURAL NETWORK| 23 3.2 Classification 3.2.1 Linear Loads Linear loads are those in which voltage and current signals follow one another very closely, such as the voltage drop that develops across a resistance, which varies as a direct function of the current that passes through it. Linear loads can be classified as follows: Resistive Elements Incandescent Lighting Electric Heaters Inductive Elements Induction motors Current limiting reactors Induction generators Damping reactors used to attenuate harmonics Turning reactors in harmonic filters Capacitive Elements Power factor correction capacitor banks Underground cables Insulated cables Capacitors used in harmonic filters 3.2.2Non-Linear loads Non linear loads are loads in which the current waveform does not resemble the applied voltage waveform due to a number of reasons, for example the use of electronic switches that conduct load current only during a fraction of the power frequency period Power Electronics Power converters Variable frequency drives DC motor controllers Cycloconverters Cranes Elevators Steel mills Dept of EEE KMCET - DVK CLASSIFICATION OF POWER QUALITY PROBLEMS USING WAVELET BASED ARTIFICIAL NEURAL NETWORK| 24 Power supplies UPS Battery Chargers ARC devices Fluorescent lighting ARC furnaces Welding machines 3.2.3. Other Loads Unbalance can also affect the operation of other three-phase loads, changing the electric power, exploitative characteristics and their service life. Moreover, voltage unbalance associated with a change in voltage magnitude has an effect on the operation of single-phase loads. Some of them may be under the influence of a supply voltage that is too high or too low. Disturbances in the functioning of control systems can also occur, resulting in disturbing and even in interrupting the operation of equipment. Harmonics have always been present in power systems. Recently, due to the widespread use of power electronic systems resulting in an increase in their magnitude, they have become a key issue in installations. Harmonic disturbances come generally from equipment with a non-linear voltage/current characteristic. Nowadays a large part of industrial, commercial and domestic loads is non-linear, making the distortion level on the low-voltage (but not only) supply network a serious concern. Linear loads are comparatively rare today: the only examples which can be considered as common are undimmed filament bulbs and unregulated heaters. Non-linear loads represent a large percentage of the total loads. Under these conditions, total harmonic distortion (THD) may become very high and therefore dangerous for the system. Harmonic distortion can be considered as a sort of pollution of the electric system which can cause problems if the sum of the harmonic currents exceeds certain limits. Knowledge of electromagnetic disturbances associated with this phenomenon is still developing; for this reason harmonics are currently an issue of great interest. Dept of EEE KMCET - DVK CLASSIFICATION OF POWER QUALITY PROBLEMS USING WAVELET BASED ARTIFICIAL NEURAL NETWORK| 25 CHAPTER 4 4. WAVELET ANALYSIS 4.1 Introduction Wavelet theory is the mathematics associated with building a model for nonstationary signal, with a set of components that are small waves, called wavelets. Informally, a wavelet is a short- term duration wave. These functions have been proposed in connection with the analysis of signals, primarily transients in a wide range of applications. A wavelet is a waveform of effectively limited duration that has an average value of zero. It gives a tool for the analysis of transient, non-stationary or time varying phenomena. The basic concept in wavelet transform is to select an appropriate wavelet function “mother wavelet” and then perform analysis using shifted and dilated versions of this wavelet. Wavelet can be chosen with very desirable frequency and time characteristics. According to Fourier theory, signal can be expressed as a sum of possibly infinite series of sines and cosines. This sum is referred to as Fourier expansion. The big disadvantage of Fourier expansion is, it has only frequency resolution and no time resolution. This means it determines all the frequencies present in the signal but it does not tell at what time they are present. To overcome this problem Wavelet transform is proposed. It provides time and frequency information simultaneously, hence giving a time frequency representation of the signal. In the wavelet analysis, the use of a fully scalable modulated window solves the signal-cutting problem. The window is shifted along the signal and for every position the spectrum is calculated. Then this process is repeated many times with a slightly shorter (or longer) window for every new cycle. In the end, the result will be a collection of time representation of the signal, all with different resolutions. The basis functions used in Fourier analysis, sine waves and cosine waves, are precisely located in frequency information of a signal calculated by the classical Fourier transform is an average over the entire time duration of the signal. Thus, if Dept of EEE KMCET - DVK CLASSIFICATION OF POWER QUALITY PROBLEMS USING WAVELET BASED ARTIFICIAL NEURAL NETWORK| 26 there is a local transient over some small interval of time in the total duration of the signal, the transient will contribute to the Fourier transform but its location on the time axis will be lost. Although the short –time Fourier transform overcomes the time location problem to a large extent, it does not provide multiple resolutions in time and frequency, which is an important characteristic for analyzing transient signal containing both high and low frequency components. Wavelet analysis overcomes the limitations of Fourier methods by employing analyzing functions that are local both in time and frequency. Unlike Fourier analysis, which uses one basis function, wavelet analysis uses a number of basis functions of a rather wide functional form. The wavelet functions are generated in the form of translation and dilation of fixed function. The basis wavelet is termed as a mother wavelet. The basic difference is, Short time Fourier transform uses a single analysis window where as wavelet transform uses short windows at high frequencies at high frequencies and long windows at low frequencies. The basis functions in WT employ time compression or dilation rather than a variation in time frequency of the signal. 4.2 Capabilities Of Wavelet Analysis: One major advantage afforded by wavelets is the ability to perform local analysis — that is, to analyze a localized area of a larger signal. Consider a sinusoidal signal with a small discontinuity — one so tiny as to be barely visible. Such a signal easily could be generated in the real world, perhaps by a power fluctuation or a noisy switch. Dept of EEE KMCET - DVK CLASSIFICATION OF POWER QUALITY PROBLEMS USING WAVELET BASED ARTIFICIAL NEURAL NETWORK| 27 Fig 4.1 fourier coefficients and wavelet coefficients Wavelet analysis is capable of revealing aspects of data that other signal analysis techniques miss, aspects like trends, breakdown points, discontinuities in higher derivatives, and self-similarity. Furthermore, because it affords a different view of data than those presented by traditional techniques, wavelet analysis can often compress or de-noise a signal without appreciable degradation. Indeed, in their brief history within the signal processing field, wavelets have already proven themselves to be an indispensable addition to the analyst’s collection of tools and continue to enjoy a burgeoning popularity today. 4.3.Wavelet Analysis Now that we know some situations when wavelet analysis is useful, it is worthwhile asking “What is wavelet analysis?” and even more fundamentally, “What is a wavelet?” A wavelet is a waveform of effectively limited duration that has an average value of zero. Compare wavelets with sine waves, which are the basis of Fourier analysis. Sinusoids do not have limited duration — they extend from minus to Dept of EEE KMCET - DVK CLASSIFICATION OF POWER QUALITY PROBLEMS USING WAVELET BASED ARTIFICIAL NEURAL NETWORK| 28 plus infinity. And where sinusoids are smooth and predictable, wavelets tend to be irregular and asymmetric. Fig.4.2 sine wave and wavelet Fourier analysis consists of breaking up a signal into sine waves of various frequencies. Similarly, wavelet analysis is the breaking up of a signal into shifted and scaled versions of the original (or mother) wavelet. Just looking at pictures of wavelets and sine waves, you can see intuitively that signals with sharp changes might be better analyzed with an irregular wavelet than with a smooth sinusoid, just as some foods are better handled with a fork than a spoon. It also makes sense that local features can be described better with wavelets that have local extent. Fourier analysis consists of breaking up a signal into sine waves of various frequencies. Similarly, wavelet analysis is the breaking up of a signal into shifted and scaled versions of the original (or mother) wavelet. Just looking at pictures of wavelets and sine waves, you can see intuitively that signals with sharp changes might be better analyzed with an irregular wavelet than with a smooth sinusoid, just as some foods are better handled with a fork than a spoon. It also makes sense that local features can be described better with wavelets that have local extent. Thus far, we’ve discussed only one-dimensional data, which encompasses most ordinary signals. However, wavelet analysis can be applied to two-dimensional data (images) and, in principle, to higher dimensional data. This toolbox uses only oneand two-dimensional analysis techniques. 4.3.1 scaling: The parameter scale (or dilation) in the wavelet analysis is similar to the scale Dept of EEE KMCET - DVK CLASSIFICATION OF POWER QUALITY PROBLEMS USING WAVELET BASED ARTIFICIAL NEURAL NETWORK| 29 used in maps. As in the case of maps, high scales correspond to a non-detailed global view (of the signal), and low scales correspond to a detailed view. Similarly, in terms of frequency, low frequencies (high scales) correspond to a global information of a signal (that usually spans the entire signal), whereas high frequencies (low scale) correspond to a detailed information of a hidden pattern in the signal. Fortunately in practical applications, low scales (high frequencies) do not last for the entire duration of the signal, unlike those shown in the figure, but they usually appear from time to time as short bursts, or spikes. Scaling, as a mathematical operation, either dilates or compresses a signal. Larger scales correspond to dilated (or stretched out) signals and small scales correspond to compressed signals. All of the signals given in the figure are the dilated and compressed versions of the same function. In the figures shown, a = 0.25 is the smallest scale anda=l is the largest scale. However, in the definition of the wavelet transform, the scaling term is used in denominator, and therefore, the opposite of the above statement holds, i.e., scales a>1 dilates the signals whereas scales a<1, compresses the signal. This interpretation of scale will be used throughout this text. It is clear from the figure that scale factor, a is inversely related to the frequency of the signal. Dept of EEE KMCET - DVK CLASSIFICATION OF POWER QUALITY PROBLEMS USING WAVELET BASED ARTIFICIAL NEURAL NETWORK| 30 Fig4.3.wavelet and sine wave at different frequencies 4.4 Classification of wavelets We can classify wavelets into two classes: (a) orthogonal and (b) biorthogonal. Based on the application, either of them can be used. (a)Features of orthogonal wavelet filter banks The coefficients of orthogonal filters are real numbers. The filters are of the same length and are not symmetric. The low pass filter, G0 and the high pass filter, H0 are related to each other by H0 (z) = z -N -1 G0 (-z ) The two filters are alternated flip of each other. The alternating flip automatically gives double-shift orthogonality between the lowpass and highpass filters [1], i.e., the scalar product of the filters, for a shift by two is zero. i.e., ΣG[k] H[k-2l] = 0, where k,lЄZ [4]. Filters that satisfy equation 2.4 are known as Conjugate Mirror Filters (CMF). Perfect reconstruction is possible with alternating flip. Also, for perfect reconstruction, the synthesis filters are identical to the analysis filters except for a time reversal. Orthogonal filters offer a high number of vanishing moments. This property is useful in many signal and image processing Dept of EEE KMCET - DVK CLASSIFICATION OF POWER QUALITY PROBLEMS USING WAVELET BASED ARTIFICIAL NEURAL NETWORK| 31 applications. They have regular structure which leads to easy implementation and scalable architecture. (b)Features of biorthogonal wavelet filter banks In the case of the biorthogonal wavelet filters, the low pass and the high pass filters do not have the same length. The low pass filter is always symmetric, while the high pass filter could be either symmetric or anti-symmetric. The coefficients of the filters are either real numbers or integers. For perfect reconstruction, biorthogonal filter bank has all odd length or all even length filters. The two analysis filters can be symmetric with odd length or one symmetric and the other antisymmetric with even length. Also, the two sets of analysis and synthesis filters must be dual. The linear phase biorthogonal filters are the most popular filters for data compression applications. 4.4.1 Wavelet families: There are a number of basis functions that can be used as the mother wavelet for Wavelet Transformation. Since the mother wavelet produces all wavelet functions used in the transformation through translation and scaling, it determines the characteristics of the resulting Wavelet Transform. Therefore, the details of the particular application should be taken into account and the appropriate mother wavelet should be chosen in order to use the Wavelet Transform effectively. Figure 4.4 Wavelet families (a) Haar (b) Daubechies4 (c) Coiflet1 (d) Symlet2 (e) Meyer (f) Morlet (g) Mexican Hat. Dept of EEE KMCET - DVK CLASSIFICATION OF POWER QUALITY PROBLEMS USING WAVELET BASED ARTIFICIAL NEURAL NETWORK| 32 Figure 4.4 illustrates some of the commonly used wavelet functions. Haar wavelet is one of the oldest and simplest wavelet. Therefore, any discussion of wavelets starts with the Haar wavelet. Daubechies wavelets are the most popular wavelets. They represent the foundations of wavelet signal processing and are used in numerous applications. These are also called Maxflat wavelets as their frequency responses have maximum flatness at frequencies 0 and π. This is a very desirable property in some applications. The Haar, Daubechies, Symlets and Coiflets are compactly supported orthogonal wavelets. These wavelets along with Meyer wavelets are capable of perfect reconstruction. The Meyer, Morlet and Mexican Hat wavelets are symmetric in shape. The wavelets are chosen based on their shape and their ability to analyze the signal in a particular application. Haar: Any discussion of wavelets begins with Haar wavelet, the first and simplest. Haar wavelet is discontinuous, and resembles a step function. It represents the same wavelet as Daubechies db1. Biorthogonal: This family of wavelets exhibits the property of linear phase, which is needed for signal and image reconstruction. By using two wavelets, one for decomposition (on the left side) and the other for reconstruction (on the right side) instead of the same single one, interesting properties are derived. Dept of EEE KMCET - DVK CLASSIFICATION OF POWER QUALITY PROBLEMS USING WAVELET BASED ARTIFICIAL NEURAL NETWORK| 33 Fig:4.5different types of Biorthogonal wavelets Symlets: The symlets are nearly symmetrical wavelets proposed by Daubechies as modifications to the db family. The properties of the two wavelet families are similar. Fig 4.6 different types of Symlets wavelets Dept of EEE KMCET - DVK CLASSIFICATION OF POWER QUALITY PROBLEMS USING WAVELET BASED ARTIFICIAL NEURAL NETWORK| 34 Morlet: This wavelet has no scaling function, but is explicit. Fig 4.7 Morelet wavelet Mexican Hat: This wavelet has no scaling function and is derived from a function that is proportional to the second derivative function of the Gaussian probability density function. Fig 4.8 Maxican Hat wavelet Other Real Wavelets Some other real wavelets are available in the toolbox: Reverse Biorthogonal Gaussian derivatives family FIR based approximation of the Meyer wavelet Complex Wavelets Dept of EEE KMCET - DVK CLASSIFICATION OF POWER QUALITY PROBLEMS USING WAVELET BASED ARTIFICIAL NEURAL NETWORK| 35 Some complex wavelet families are available in the toolbox: Gaussian derivatives Morlet Frequency B-Spline Shannon 4.5 THE DISCRETE WAVELET TRANSFORM The transform of a signal is just another form of representing the signal. It does not change the information content present in the signal. The Wavelet Transform provides a time-frequency representation of the signal. It was developed to overcome the short coming of the Short Time Fourier Transform (STFT), which can also be used to analyze non-stationary signals. While STFT gives a constant resolution at all frequencies, the Wavelet Transform uses multi-resolution technique by which different frequencies are analyzed with different resolutions. A wave is an oscillating function of time or space and is periodic. In contrast, wavelets are localized waves. They have their energy concentrated in time or space and are suited to analysis of transient signals. While Fourier Transform and STFT use waves to analyze signals, the Wavelet Transform uses wavelets of finite energy The wavelet analysis is done similar to the STFT analysis. The signal to be analyzed is multiplied with a wavelet function just as it is multiplied with a window function in STFT, and then the transform is computed for each segment generated. However, unlike STFT, in Wavelet Transform, the width of the wavelet function changes with each spectral component. The Wavelet Transform, at high frequencies, gives good time resolution and poor frequency resolution, while at low frequencies, the Wavelet Transform gives good frequency resolution and poor time resolution. 4.5.1 The Continuous Wavelet Transform and the Wavelet Series The Continuous Wavelet Transform (CWT) is provided by equation 2.1, where x(t) is the signal to be analyzed. ψ(t) is the mother wavelet or the basis function. All the wavelet functions used in the transformation are derived from the mother wavelet through translation (shifting) and scaling (dilation or compression). Dept of EEE KMCET - DVK CLASSIFICATION OF POWER QUALITY PROBLEMS USING WAVELET BASED ARTIFICIAL NEURAL NETWORK| 36 The mother wavelet used to generate all the basis functions is designed based on some desired characteristics associated with that function. The translation parameter τ relates to the location of the wavelet function as it is shifted through the signal. Thus, it corresponds to the time information in the Wavelet Transform. The scale parameter s is defined as |1/frequency| and corresponds to frequency information. Scaling either dilates (expands) or compresses a signal. Large scales (low frequencies) dilate the signal and provide detailed information hidden in the signal, while small scales (high frequencies) compress the signal and provide global information about the signal. Notice that the Wavelet Transform merely performs the convolution operation of the signal and the basis function. The above analysis becomes very useful as in most practical applications, high frequencies (low scales) do not last for a long duration, but instead, appear as short bursts, while low frequencies (high scales) usually last for entire duration of the signal. The Wavelet Series is obtained by discretizing CWT. This aids in computation of CWT using computers and is obtained by sampling the time-scale plane. The sampling rate can be changed accordingly with scale change without violating the Nyquist criterion. Nyquist criterion states that, the minimum sampling rate that allows reconstruction of the original signal is 2ω radians, where ω is the highest frequency in the signal. Therefore, as the scale goes higher (lower frequencies), the sampling rate can be decreased thus reducing the number of computations. 4.5.2 The Discrete Wavelet Transform The Wavelet Series is just a sampled version of CWT and its computation may consume significant amount of time and resources, depending on the resolution required. The Discrete Wavelet Transform (DWT), which is based on sub-band coding is found to yield a fast computation of Wavelet Transform. It is easy to implement and reduces the computation time and resources required. The foundations of DWT go back to 1976 when techniques to decompose discrete time signals were devised [5]. Similar work was done in speech signal coding Dept of EEE KMCET - DVK CLASSIFICATION OF POWER QUALITY PROBLEMS USING WAVELET BASED ARTIFICIAL NEURAL NETWORK| 37 which was named as sub-band coding. In 1983, a technique similar to sub-band coding was developed which was named pyramidal coding. Later many improvements were made to these coding schemes which resulted in efficient multiresolution analysis schemes. In CWT, the signals are analyzed using a set of basis functions which relate to each other by simple scaling and translation. In the case of DWT, a time-scale representation of the digital signal is obtained using digital filtering techniques. The signal to be analyzed is passed through filters with different cutoff frequencies at different scales. 4.5.3 DWT and Filter Banks: 4.5.3.1 Multi-Resolution Analysis using Filter Banks Filters are one of the most widely used signal processing functions. Wavelets can be realized by iteration of filters with rescaling. The resolution of the signal, which is a measure of the amount of detail information in the signal, is determined by the filtering operations, and the scale is determined by up sampling and down sampling (sub sampling) operations[5]. The DWT is computed by successive lowpass and highpass filtering of the discrete time-domain signal as shown in figure 2.2. This is called the Mallat algorithm or Mallat-tree decomposition. Its significance is in the manner it connects the continuous-time multi resolution to discrete-time filters. In the figure, the signal is denoted by the sequence x[n], where n is an integer. The low pass filter is denoted by G0 while the high pass filter is denoted by H0. At each level, the high pass filter produces detail information, d[n], while the low pass filter associated with scaling function produces coarse approximations, a[n]. Fig. 4.9 Three-level wavelet decomposition tree. Dept of EEE KMCET - DVK CLASSIFICATION OF POWER QUALITY PROBLEMS USING WAVELET BASED ARTIFICIAL NEURAL NETWORK| 38 At each decomposition level, the half band filters produce signals spanning only half the frequency band. This doubles the frequency resolution as the uncertainty in frequency is reduced by half. In accordance with Nyquist’s rule if the original signal has a highest frequency of ω, which requires a sampling frequency of 2ω radians, then it now has a highest frequency of ω/2 radians. It can now be sampled at a frequency of ω radians thus discarding half the samples with no loss of information. This decimation by 2 halves the time resolution as the entire signal is now represented by only half the number of samples. Thus, while the half band low pass filtering removes half of the frequencies and thus halves the resolution, the decimation by 2 doubles the scale. With this approach, the time resolution becomes arbitrarily good at high frequencies, while the frequency resolution becomes arbitrarily good at low frequencies. The time-frequency plane is thus resolved as shown in figure 1.1(d) of Chapter 1. The filtering and decimation process is continued until the desired level is reached. The maximum number of levels depends on the length of the signal. The DWT of the original signal is then obtained by concatenating all the coefficients, a[n] and d[n], starting from the last level of decomposition. Fig. 4.10 Three-level wavelet reconstruction tree. Figure 2.3 shows the reconstruction of the original signal from the wavelet coefficients. Basically, the reconstruction is the reverse process of decomposition. The approximation and detail coefficients at every level are up sampled by two, passed through the low pass and high pass synthesis filters and then added. This process is continued through the same number of levels as in the decomposition process to obtained the original signal. The Mallat algorithm works equally well if the analysis filters, G and H , are exchanged with the synthesis filters, G . 0 Dept of EEE 0 11 KMCET - DVK CLASSIFICATION OF POWER QUALITY PROBLEMS USING WAVELET BASED ARTIFICIAL NEURAL NETWORK| 39 4.6.3.2 Conditions for Perfect Reconstruction In most Wavelet Transform applications, it is required that the original signal be synthesized from the wavelet coefficients. To achieve perfect reconstruction the analysis and synthesis filters have to satisfy certain conditions. Let G0(z) and G1(z) be the low pass analysis and synthesis filters, respectively and H0(z) and H1(z) the high pass analysis and synthesis filters respectively. Then the filters have to satisfy the following two conditions as given in [4] : G0 (-z) G1 (z) + H0 (-z). H1 (z) = 0 G0 (z) G1 (z) + H0 (z). H1 (z) = 2z -d The first condition implies that the reconstruction is aliasing-free and the second condition implies that the amplitude distortion has amplitude of one. It can be observed that the perfect reconstruction condition does not change if we switch the analysis and synthesis filters. There are a number of filters which satisfy these conditions. But not all of them give accurate Wavelet Transforms, especially when the filter coefficients are quantized. The accuracy of the Wavelet Transform can be determined after reconstruction by calculating the Signal to Noise Ratio (SNR) of the signal. Some applications like pattern recognition do not need reconstruction, and in such applications, the above conditions need not apply. 4.6 Applications : There is a wide range of applications for Wavelet Transforms. They are applied in different fields ranging from signal processing to biometrics, and the list is still growing. One of the prominent applications is in the FBI fingerprint compression standard. Wavelet Transforms are used to compress the fingerprint pictures for storage in their data bank. The previously chosen Discrete Cosine Transform (DCT) did not perform well at high compression ratios. It produced severe blocking effects which made it impossible to follow the ridge lines in the fingerprints after reconstruction. This did not happen with Wavelet Transform due to its property of retaining the details present in the data. In DWT, the most prominent information in the signal appears in high amplitudes and the less prominent information appears in very low amplitudes. Data compression can Dept of EEE KMCET - DVK CLASSIFICATION OF POWER QUALITY PROBLEMS USING WAVELET BASED ARTIFICIAL NEURAL NETWORK| 40 be achieved by discarding these low amplitudes. The wavelet transforms enables high compression ratios with good quality of reconstruction. At present, the application of wavelets for image compression is one the hottest areas of research. Recently, the Wavelet Transforms have been chosen for the JPEG 2000 compression standard. Fig.4.11 Signal processing application using Wavelet Transform. Figure 4.11 shows the general steps followed in a signal processing application. Processing may involve compression, encoding, denoising etc. The processed signal is either stored or transmitted. For most compression applications, processing involves quantization and entropy coding to yield a compressed image. During this process, all the wavelet coefficients that are below a chosen threshold are discarded. These discarded coefficients are replaced with zeros during reconstruction at the other end. To reconstruct the signal, the entropy coding is decoded, then quantized and then finally Inverse Wavelet Transformed. Wavelets also have numerous applications in digital communications. Orthogonal Frequency Division Multiplexing (OFDM) is one of them. Wavelets are used in biomedical imaging. For example, the ECG signals, measured from the heart, are analyzed using wavelets or compressed for storage. The popularity of Wavelet Transform is growing because of its ability to reduce distortion in the reconstructed signal while retaining all the significant features present in the signal. 4.7 Summary This chapter has explained brief discussion about the wavelet analysis, this analysis shows the comparison between the sine wave and wavelet wave form. Different types of wavelet families and their related waveforms are also studied. The applications of wavelet analysis is also studied in brief. Dept of EEE KMCET - DVK CLASSIFICATION OF POWER QUALITY PROBLEMS USING WAVELET BASED ARTIFICIAL NEURAL NETWORK| 41 CHAPTER 5 TEST SYSTEM AND SIMULATION RESULTS 5.1 Introduction to MATLAB The MATLAB is a high performance language for technical computing integrates computation, visualization, and programming in an easy –to- use environment where problems and solutions are expressed in familiar mathematical notation. MATLAB software can provide solution for you in technical computing, some of the common applications of MATLAB and the types of add-on, application-specific solutions that are available in MATLAB toolbox. The MATLAB documentation is organized into these main topics Graphics: Tools and techniques for plotting graph annotation, printing and programming with handle graphics. Desktop tools and development environment: start up and shutdown, the desktop and other tools that help you use and become more productive with MATLAB. Mathematics: Mathematical operations Data analysis: Data analysis, including data fitting, Fourier analysis, and time series tools. Programming fundamentals: the MATLAB language and how to develop MATLAB applications MATLAB classes and Object-Oriented programming: Designing and implementing MATLAB classes. 3-D visualization: Visualizing surface and volume data, transparency, and viewing and lighting techniques Creating graphical user interfaces: GUI-building tools and how to write callback functions External interface: MEX-files, the MATLAB engine, and interfacing to sun Microsystems java software, COM, and the serial port Dept of EEE KMCET - DVK CLASSIFICATION OF POWER QUALITY PROBLEMS USING WAVELET BASED ARTIFICIAL NEURAL NETWORK| 42 5.2 POWER SYSTEM NETWORK In this work, we consider single source feeding to 0.8 MW and 0.4 MVAR load through a transmission line of 200 Km length at 5 kV. Here the test system have been simulated for normal loading condition, sag, swell and interruption. The MATLAB test system shown in fig.5. Fig.5.1network used for creating various PQ problems 5.2.1 SPECIFICATION OF TEST SYSTEM Rating of the source voltage is 5 k.v and 50 Hertz The length of the transmission line is 100 K.M The ratings of the transmission line parameters Resistance is 2.568 ohm Inductance is 20 Mh Capacitance is 0.86 Micro farad Total length of the transmission system used is 200 K.M. The rating of RL load connected to this transmission line is 0.8 M. and 0.4 Mvar The one by one added loads for this transmission line are 0.34 M.W,0.16 Mvar Dept of EEE KMCET - DVK CLASSIFICATION OF POWER QUALITY PROBLEMS USING WAVELET BASED ARTIFICIAL NEURAL NETWORK| 43 0.18 M.W, 0.06 Mvar 0.3 M.W, 0.14 Mvar Parallel RC load is 0.06 M.W, 0.6 Mvar 0.36 M.W, 0.17 Mvar 0.38 M.W, 0.19 Mvar Breaker resistance 0.01 Ohm Snumbber resistance 1M Ohm These are the specifications used to simulate the test system using MATLAB SIMULINK. SIMULATION The above network had simulated in MATLAB using simulink blocks with the given specifications mentioned above. Case 1: For the normal condition simulated load voltage and currents are presented. Fig .5.2 normal voltage Dept of EEE KMCET - DVK CLASSIFICATION OF POWER QUALITY PROBLEMS USING WAVELET BASED ARTIFICIAL NEURAL NETWORK| 44 Case 2: After applying the load at breaker 1 the obtained distorted signal (Sag in voltage magnitude) is shown below Fig 5.3 Sag voltage To get the voltage sag by adding a RL Load of 0.3 MW& 0.16 Mvar at breaker 1. For sag magnitude levels varies from 10% to 90% with different duration of time with a base value of voltage (1 p.u). Case 3: After applying the parallel RC load (0.06 MW, 0.6 Mvar) at breaker 6 the obtained load voltage waveform is shown in fig 5.4 Fig.5.4 voltage for swell Dept of EEE KMCET - DVK CLASSIFICATION OF POWER QUALITY PROBLEMS USING WAVELET BASED ARTIFICIAL NEURAL NETWORK| 45 For voltage swell magnitude of PQ disturbance varies from 110% to 180% with different duration of time. Case4: To make the load end ded short circuit by closing the breaker 8, the obtained load voltage waveform is shown in Fig5.5. . Fig 5.5 voltage for interruption Because of shot circuit at load the load voltage becomes zero. So we obtained the interruption during the time 0.2 to 0.6 seconds. 5.3 wavelet decomposition In the pre-processing voltage signals are captured during a power quality problem. Due to the absence of field results different power quality problems have been simulated in the MATLAB/SIMULINK environment. The PQ data thus obtained matches the field data. Subsequently, using the Wavelet toolbox of the MATLAB, multi signal decomposition on the captured signals is performed. Depending on the selected resolution levels, the PQ signal is decomposed into a number of wavelet levels. If the resolution level is defined as n, after decomposing the signal, there will be one level of approximation coefficients with n level of detail coefficients. The information contained in the coefficients of details and approximations is very useful for further processing. 5.3.1 Procedure of power quality problem Various possible PQ disturbances viz. sag, swell, and interruption are classified using Wavelet analysis. Network is shown in Fig.5.1 is used for simulation as well as for visualizing disturbances in MATLAB/SIMULINK. After obtaining the disturbances data with Dept of EEE KMCET - DVK CLASSIFICATION OF POWER QUALITY PROBLEMS USING WAVELET BASED ARTIFICIAL NEURAL NETWORK| 46 wide range of parameters variation, MSD wavelet analysis is applied on the PQ signal for extracting the useful information relating the event. Thereafter, Parseval’s energy theorem is used to determine the energy of the PQ signal corresponding to each detailed level. Parseval theorem relates the energy of the distorted signal to the energy in each of the expansion components and their wavelet coefficients. This means that energy of the PQ signal can be partitioned in terms of the expansion coefficients as per equation 5.1. 2 N 2 N f ( n) = c j (n) + d j (n) n 1 n 1 j N 2 (5.1) j 1 n 1 Where, f(n)is the time domain signal under analysis N is the total number of samples of the signal N f ( n) 2 =total energy of the signal n 1 2 N c n 1 j (n) =total energy connected in the level ‘j’ of the approximated version of the signal j 2 N d j 1 n 1 j ( n) is the total energy connected in the detailed version of the signal from levels 1 to j In the present work the signal is decomposed in 6 levels using db-3. The average energies, Ej up to sixth level of detailed signals are determined. These six energy values contain the information of energy distribution of particular PQ signal. Thereafter normalization of these energy values is carried out. Normalization is done to extract the useful information of the PQ problem in a better way for classification. Further this enhances the information embedded in lower energy levels. The normalized Shannon entropy will be used. The definition of normalized Shannon entropy is as follows EJ k E jk log E jk Dept of EEE 5.2 KMCET - DVK CLASSIFICATION OF POWER QUALITY PROBLEMS USING WAVELET BASED ARTIFICIAL NEURAL NETWORK| 47 Where Ejk is the wavelet energy spectrum at scale j and instant k and it as follows is defined EJK DJ (k ) 2 5.3 Equations (5.2) and (5.3) are used for normalization of the extracted energy values. where j has values from 1, 2, ……6 Ej= Energy at each level In order to make the information easy to handle scaling of the energy values is done by scaling factor ‘ k ’ so as to make the system efficient and consuming less memory in computation. 5.3.2 Wavelet tool box main menu Fig 5.6 wavelet toolbox main menu 5.3.3 Decomposition levels for each power quality disturbance The simulation results of load voltage signal have been loaded to wavelet menu and decomposed using db-3 with 6 levels. The decomposed signals are presented in fig.5.7 . Dept of EEE KMCET - DVK CLASSIFICATION OF POWER QUALITY PROBLEMS USING WAVELET BASED ARTIFICIAL NEURAL NETWORK| 48 Fig 5.7 decomposition levels for normal voltage The simulation results of load voltage signal have been loaded to wavelet menu and decomposed using db-3 with 6 levels. The decomposed signals are presented in Fig5.8 Fig5.8 decomposition levels for sag voltage The simulation results of load voltage signal have been loaded to wavelet menu and decomposed using db-3 with 6 levels. The decomposed signals are presented in fig 5.9 Dept of EEE KMCET - DVK CLASSIFICATION OF POWER QUALITY PROBLEMS USING WAVELET BASED ARTIFICIAL NEURAL NETWORK| 49 Fig5.9 decomposition levels for voltage swell The simulation results of load voltage signal have been loaded to wavelet menu and decomposed using db-3 with 6 levels. The decomposed signals are presented in fig 5.10 Fig 5.10 decomposition levels for voltage interruption 5.3.4 Wavelet entropy table for different power quality disturbances From the results of wavelet decomposed values, we find the entropy using shaman formula and normalized. Finally finding energy levels of each signal and based on energy values, made the classifying the various power quality problems. Dept of EEE KMCET - DVK CLASSIFICATION OF POWER QUALITY PROBLEMS USING WAVELET BASED ARTIFICIAL NEURAL NETWORK| 50 Condition o e a. load=0.6MW 32.7628 114.3192 b. load=0.7MW 26.9059 88.5833 c. load=0.8MW 15.1958 41.3481 d. load=0.9MW 24.8938 80.0242 e. load=1MW 18.1092 52.4517 1.normal 2.sag a.0.2-0.4 17.6076 b.0.2-0.6 15.9491 50.5044 44.1535 c.0.2-0.8 17.1694 48.4581 d. 2 breakers 17.228 e. 3 breakers 14.2696 f.5 breakers 12.8830 49.0337 37.9306 32.9297 3.swell a.0.2-0.4 25.8427 84.0410 b. 0.2-0.6 27.8535 92.3534 c. 0.2-0.8 35.1693 125.5135 a.0.2-0.4 14.6258 39.2514 b. 0.2-0.6 10.5068 24.7122 c. 0.2-0.8 7.5918 15.2129 4. interruption Table 5.1 wavelet entropy for different power quality problems Dept of EEE KMCET - DVK CLASSIFICATION OF POWER QUALITY PROBLEMS USING WAVELET BASED ARTIFICIAL NEURAL NETWORK| 51 6. CONCLUSION In this work, the literature work carried out for various power quality disturbances for classification of power quality problems. We carried out the MATLAB SIMULINK, simulation and observed the power quality problems. Further analysis of power quality problems using wavelet analysis, by calculating the entropy and energy levels for various faults. Based on entropy and energy values it is to discriminate various power quality problems. Multi resolution signal decomposition using wavelet transformation is used as a basis for feature vector for obtaining the training and testing data for ANN. The integration of wavelet transform with ANN classifies various power quality problems with over 90% accuracy. This is the extension for this thesis. Dept of EEE KMCET - DVK CLASSIFICATION OF POWER QUALITY PROBLEMS USING WAVELET BASED ARTIFICIAL NEURAL NETWORK| 52 7. REFERENCES [1] M. H. Bollen, Understanding Power Quality Problems, IEEE Press,2001 [2] A.K. Chandel, G. Guleria, and R. Chandel Classification of Power Quality Problems Using Wavelet Based Artificial Neural Network, 978-1-4244-1904-3/08/$25.00 ©2008 IEEE [3] Resende, J.W., Chaves, M.L.R., Penna, C.Universidade Federal de Uberlandia (MG)-Brazil,” Identification of power quality disturbances using theMATLAB wavelet transform toolbox”. [4] Sudipta Nath, Arindam Dey and Abhijit Chakrabarti., “Detection of Power Quality disturbances using Wavelet Transform”, proceedings of worldacademy of science, engineering and technology volume 37 January 2009 ISSN 2070-3740 PWASET VOLUME [5] F. Truchetet, O. Laligant Le2i, Université de Bourgogne, France, [email protected], “Wavelets in industrial applications: a review” [6] Nguyen Huu Phuc, Truong Quoc Khanh, Nguyen Nhan Bon* “DISCRETE WAVELETS TRANSFORM TECHNIQUE APPLICATION IN IDENTIFICATION OF POWER QUALITY DISTURBANCES” Department of Electrical & Electronics Engineering, HCMC Univ.of Technology, Vietnam *Department of Electrical Engineering, HCMC Univ. of Technical Education, Vietnam [7] CHIEN-HSING LEE, YAW-JUEN WANG, AND WEN-LIANG HUANG, “ A Literature Survey of Wavelets in Power Engineering Applications” , Proc. Natl. Sci. Counc. ROC(A) Vol. 24, No. 4, 2000. pp. 249-258 [8] S. El Safty, and A. El-Zonkoly. “Applying Wavelet Entropy Principle in Fault Classification” , proceeding world economy of science, engineering and technology volume 30 July 2008 ISSN 1307-6884 [9] M.Sushama, G. Tulasi Ram Das,” DETECTION AND CLASSIFICATION OF VOLTAGE SWELLS USING ADAPTIVE DECOMPOSITION & WAVELET TRANSFORMS” Journal of Theoretical and Applied Information Technology. [10] Sudipta Nath “Power Quality Assessment by Wavelet Transform Analysis TIG Research Journal, Vol. 1, No. 2, September 2008 Dept of EEE KMCET - DVK CLASSIFICATION OF POWER QUALITY PROBLEMS USING WAVELET BASED ARTIFICIAL NEURAL NETWORK| 53 [11] Thomas E. Grebe, PE Electrotek Concepts, Inc, “ CAPACITOR SWITCHING AND ITS IMPACT ON POWER QUALITY” Prepared on Request of Cigré 36.05/CIRED 2 (Voltage Quality) [12] C. Benachaiba and B. Ferdi , “Power Quality Improvement Using DVR “American Journal of Applied Sciences 6 (3): 396-400, 2009 ISSN 1546-9239 © 2009 Science Publications [13] F. Choong*, M. B. I. Reaz, “ Implementation of Power Quality Disturbance Classifier in FPGA Employing Wavelet Transform, ANN and Fuzzy Logic SETIT 2005” 3rd International Conference: Sciences of Electronic, Technologies of Information and Telecommunications March 27-31, 2005 – TUNISIA [14] J.-P. ANTOINE, “ WAVELET ANALYSIS AND SOME OF ITS APPLICATIONS IN PHYSICS “ Institut de Physique Th´eorique, Universit´e Catholique de Louvain Dept of EEE KMCET - DVK