* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project
Download Enhanced Virtual Synchronous Generator Control for Parallel
Opto-isolator wikipedia , lookup
Three-phase electric power wikipedia , lookup
Power factor wikipedia , lookup
Distributed control system wikipedia , lookup
Stray voltage wikipedia , lookup
Control theory wikipedia , lookup
Solar micro-inverter wikipedia , lookup
Electrification wikipedia , lookup
Electrical substation wikipedia , lookup
Resilient control systems wikipedia , lookup
Audio power wikipedia , lookup
History of electric power transmission wikipedia , lookup
Pulse-width modulation wikipedia , lookup
Electric power system wikipedia , lookup
Distributed generation wikipedia , lookup
Power over Ethernet wikipedia , lookup
Variable-frequency drive wikipedia , lookup
Power inverter wikipedia , lookup
Amtrak's 25 Hz traction power system wikipedia , lookup
Buck converter wikipedia , lookup
Control system wikipedia , lookup
Power engineering wikipedia , lookup
Voltage optimisation wikipedia , lookup
Alternating current wikipedia , lookup
University of Kurdistan Dept. of Electrical and Computer Engineering Smart/Micro Grid Research Center smgrc.uok.ac.ir Enhanced Virtual Synchronous Generator Control for Parallel Inverters in Microgrids Liu J, Miura Y, Bevrani H, Ise T Published (to be published) in: IEEE Transactions on Smart Grid (Expected) publication date: 2016 Citation format for published version: Liu J, Miura Y, Bevrani H, Ise T (2016) Enhanced virtual synchronous generator control for parallel inverters in microgrids. IEEE Transaction on Smart Grids. DOI: 10.1109/TSG.2016.2521405 Copyright policies: Download and print one copy of this material for the purpose of private study or research is permitted. Permission to further distributing the material for advertising or promotional purposes or use it for any profitmaking activity or commercial gain, must be obtained from the main publisher. If you believe that this document breaches copyright please contact us at [email protected] providing details, and we will remove access to the work immediately and investigate your claim. Copyright © Smart/Micro Grid Research Center, 2016 This article has been accepted for inclusion in a future issue of this journal. Content is final as presented, with the exception of pagination. IEEE TRANSACTIONS ON SMART GRID 1 Enhanced Virtual Synchronous Generator Control for Parallel Inverters in Microgrids Jia Liu, Student Member, IEEE, Yushi Miura, Member, IEEE, Hassan Bevrani, Senior Member, IEEE, and Toshifumi Ise, Member, IEEE Abstract—Virtual synchronous generator (VSG) control is a promising communication-less control method in a microgrid for its inertia support feature. However, active power oscillation and improper transient active power sharing are observed when basic VSG control is applied. Moreover, the problem of reactive power sharing error, inherited from conventional droop control, should also be addressed to obtain desirable stable state performance. In this paper, an enhanced VSG control is proposed, with which oscillation damping and proper transient active power sharing are achieved by adjusting the virtual stator reactance based on state-space analyses. Furthermore, communication-less accurate reactive power sharing is achieved based on inversed voltage droop control feature (V–Q droop control) and common ac bus voltage estimation. Simulation and experimental results verify the improvement introduced by the proposed enhanced VSG control strategy. Index Terms—DC-AC power converters, distributed power generation, droop control, microgrids, power control, power system dynamics, power system modeling, reactive power control, state-space methods, virtual synchronous generator. N OMENCLATURE Sbase P0 , Q0 Pin Pout , Qout Pload Qref ω0 , E0 ωm ωg θm E Vout , Vbus Iout Vpwm , θpwm J, D Power rating Set value of active and reactive power Virtual shaft power Measured output active and reactive power Load active power Reference value for reactive power control Nominal angular frequency and voltage Virtual rotor angular frequency Output voltage angular frequency Virtual rotor phase angle Virtual internal electromotive force Inverter output and common ac bus voltage Inverter output current Voltage and phase reference of PWM inverter Virtual inertia and virtual damping factor Manuscript received September 9, 2015; revised December 2, 2015; accepted January 20, 2016. Paper no. TSG-01106-2015. J. Liu, Y. Miura, and T. Ise are with the Division of Electrical, Electronic, and Information Engineering, Osaka University, Osaka 565-0871, Japan (e-mail: [email protected]). H. Bevrani is with the Department of Electrical and Computer Engineering, University of Kurdistan, Sanandaj 66177-15175, Iran. Color versions of one or more of the figures in this paper are available online at http://ieeexplore.ieee.org. Digital Object Identifier 10.1109/TSG.2016.2521405 M∗ Inertia constant ω–P and V–Q droop coefficient kp , kq R, L, X, Z Resistance, inductance, reactance and impedance δ Power angle K Synchronizing power coefficient Gain and time constant of reactive power Kpq , Tiq control Time constant of reactive power 1st order filter Tfq V̂ ∗ Estimation error in V̂bus Subscripts/Superscripts * i f line ls α, β Per unit value (i =1, 2) ith distributed generator Inverter output LC filter Distribution line Virtual stator α-axis and β-axis components. I. I NTRODUCTION ECENT years, inverter-interfaced distributed generators (DGs) with renewable energy sources (RES), e.g., photovoltaics and wind turbines, have been developed to solve energy crisis and environmental issues. To facilitate the integration of DGs in distribution system, the concept of microgrid is proposed [1]. The control strategies of microgrids are preferred to be in a communication-less manner because of its decentralized feature. Although in a hierarchical microgrid control structure, communication is required for the secondary and tertiary control, it is still recommended to realize the basic functions of a microgrid in the primary control level without communication [2], [3]. Droop control is a widely adopted communication-less control method in a microgrid. By drooping the frequency against the active power (P–ω droop) and the output voltage against reactive power (Q–V droop), load sharing between DGs can be performed in an autonomic manner, which is similar to the power sharing between parallel synchronous generators (SGs) [4], [5]. In some references [6]–[8], it is proposed that P–V and Q–ω droop controls are more suitable for low voltage (LV) microgrid in the light of the resistive line impedance feature. Meanwhile, the P–ω and Q–V droop controls are still valid in LV microgrid by adding inductive virtual impedance [2], [3], [9]. However, like most of DG control methods, a conventional droop control provides barely any inertia support for R c 2016 IEEE. Personal use is permitted, but republication/redistribution requires IEEE permission. 1949-3053 See http://www.ieee.org/publications_standards/publications/rights/index.html for more information. This article has been accepted for inclusion in a future issue of this journal. Content is final as presented, with the exception of pagination. 2 the microgrid, thus a droop-control-based microgrid is usually inertia-less and sensitive to fault. To provide inertia support for the system, control methods to emulate virtual inertia are proposed in recent literatures, such as virtual synchronous generator (VSG) [10]–[18], virtual synchronous machine [19], and synchronverter [20], [21]. Although their name and control scheme differ from each other, the principles are similar in the aspect that all of them mimic the transient characteristics of SG by emulating its fundamental swing equation. For simpler explication, all of these methods are called VSG control in this paper. A comprehensive survey on VSGs and the existing topologies are given in [17]. Besides, a unique method to provide virtual inertia by modifying the droop coefficient in droop control is presented in [22]. To share the load in parallel operation, droop characteristics are also emulated in some VSG control schemes [12], [14], [20], [21]. In this case, as it is demonstrated in [23] and [24], VSG control inherits the advantages of droop control, and outperforms the latter in terms of transient frequency stability owing to its lower df /dt rate. Therefore, VSG control can be considered as a potential upgrade for the communication-less control method of a microgrid. However, when VSG control is applied in microgrids, several problems have been noticed, such as oscillation in active power during a disturbance, inappropriate transient active power sharing during loading transition and errors in reactive power sharing. Active power oscillation during a disturbance is introduced by the well-known feature of the swing equation, thus it is an inherent feature for a real SG as well as a VSG. It is not a critical problem for SGs because they usually have considerable overload capabilities, but the overload capabilities of inverter-interfaced DGs are not high enough to ride though a large oscillation. However, this oscillation can be damped by properly increasing the damping ratio [15] or using alternating moment of inertia [16]. Using smaller inertia may also lead to reduced oscillation [18]; however, it is not encouraged because providing a large amount of virtual inertia is an advantage that distinguishes VSG from other control methods. In this paper, a novel method for oscillation damping is proposed based on increasing the virtual stator reactance. Due to the oscillatory feature of VSG, inappropriate transient active power sharing during loading transition may also cause oscillation, which is avoidable if the swing equation and output impedance are designed properly, as it is analyzed in this paper. Sharing transient loads between SG and DG is addressed in [25], but theoretical analysis is not provided. The inaccurate reactive power sharing is a well-known problem in conventional Q–V droop control, and the same problem is reported in active power sharing of P–V droop control. In Q–V or P–V droop controls, output voltage is regulated according to reactive/active power sharing, but the output voltage of each DG is not equal due to unequal line voltage drop. This problem has received considerable attention in the literature, and many control strategies are proposed to address this issue [26]–[40]. A comprehensive solution is to eliminate IEEE TRANSACTIONS ON SMART GRID the mismatch of DG output impedance [26], [27]; however, this method cannot guarantee accurate reactive power sharing if active power is not shared according to the power rating ratio. An approach based on line voltage drop compensation is proposed in [28]. However, a grid-connected mode operation is required for the evaluation of line parameters, which is not feasible for an isolated microgrid. Other communication-less approaches, e.g., Q–dV/ dt droop control [29], adaptive voltage droop [30], and virtual capacitor control [31] are also proposed. However, the reactive power sharing errors cannot be completely eliminated by these methods, as it is demonstrated in respective experimental results. In some approaches, communication is used to improve reactive power sharing accuracy, such as secondary control signals from MGCC [32]–[37], master-slave communication [38], and communication between DGs [15]. However, as accurate reactive power sharing is a basic function of a microgrid, it is always preferred to solve this problem in a communicationless manner considering the probable communication fault. In this paper, a communication-less approach is proposed based on inversed voltage droop control (V–Q droop control) and common ac bus voltage estimation. By applying the proposed method, reactive power sharing is immune to line impedance mismatch and active power sharing change. The idea to use ac bus voltage as a common reference shares some similarities with the approaches presented in [21], [39], and [40]. However, in these works, measured bus voltage is used directly; while, in microgrid applications, it may not be feasible if DGs are not installed in the proximity of the ac bus. In this paper, bus voltage is estimated based on the available local measurement, thus there should be no installation difficulty in the field applications. The rest of this paper is organized as follows. In Section II, a brief description of the basic VSG control is presented. In Section III, a state-space model of islanded microgrid using VSG control is built, and the principle of proposed oscillation damping method is derived from eigenvalue analysis of this model. Proper parameter design for appropriate transient load sharing based on poles-zeros cancellation is also discussed based on the same model. In Section IV, the cause of reactive power sharing errors is discussed and a novel accurate reactive power sharing method is proposed. The enhanced VSG control strategy is presented in Section V. Simulation and experiment results are shown in Sections VI and VII, respectively. Finally, conclusions are given in Section VIII. II. BASIC VSG C ONTROL S CHEME Fig. 1 shows the structure of a DG using the basic VSG control [14], [41]. The primary source of the DG could be photovoltaic panels, fuel cells, a gas engine or other distributed energy resources (DERs). The energy storage is designed for emulating the kinetic energy stored in rotating mass of a SG, in order to supply or absorb insufficient/surplus power generated by the primary source in transient state [13]. As this paper focuses on the control scheme of the inverter, the design and This article has been accepted for inclusion in a future issue of this journal. Content is final as presented, with the exception of pagination. LIU et al.: ENHANCED VSG CONTROL FOR PARALLEL INVERTERS IN MICROGRIDS Fig. 2. 3 Structure of a microgrid in islanded mode. by the LC filter is not included in Qout . Therefore, no specific inertial process is required for the reactive power PI controller. In a microgrid, in order to share the active and reactive power according to the ratings of DGs without communication, kp∗ = (kp ω0 )/Sbase , kq∗ = (kq E0 )/Sbase , P∗0 = P0 /Sbase and Q∗0 = Q0 /Sbase should be designed equally for each DG in default [2]. In this paper, to simplify the explication for the case of different power ratings, per unit values are calculated based on respective power ratings of DGs. Fig. 1. Block diagram of (a) the basic VSG control, (b) the “Governor Model” block and (c) the “Q Droop” block. III. A NALYSES OF T RANSIENT ACTIVE P OWER P ERFORMANCE A. Closed-Loop State-Space Model control of the primary source and energy storage are beyond the scope of this paper. In the block “Swing Equation Function” in Fig. 1(a), ωm is solved from the swing equation (1) by an iterative method. Pin − Pout = Jωm dωm + D ωm − ωg dt (1) The block “Governor Model” in Fig. 1(a) is a ω–P droop controller as shown in Fig. 1(b). In some previous studies [12]–[14], a first order lag unit is used to emulate the mechanical delay in the governor of a real SG. However, in this paper, this delay is removed, because it degrades the dynamic performance of DG, as it is discussed in [23]. The block “Q Droop” in Fig. 1(a) is a V–Q droop controller as shown in Fig. 1(c), which differs from the conventional Q–V droop controller in the reversed input and output. It is noteworthy that inner current or voltage loop is not adopted in this control scheme, in order to make the filter inductor Lf contribute to the output impedance and be considered as the stator inductance of the VSG. This stator inductance results in more inductive output impedance, which is especially important for active and reactive power decoupling in a low voltage microgrid in which line resistance is dominant. Nevertheless, output voltage is still regulated indirectly by the V–Q droop controller and the PI controller of reactive power. In order to diminish the influence from ripples in measured output power, a 20Hz 1st order low-pass filter is applied for Qout as shown in Fig. 1(a). As the output current is measured after the LC filter stage, the reactive power consumed In the present work, an islanded microgrid which consists of two DGs using VSG control is studied, as it is shown in Fig. 2. The DGs are connected to a common ac bus via a distribution line, to supply the loads in the microgrid. Note that the capacitor of the DG output LC filter in Fig. 1 is neglected, as its susceptance is usually negligible at fundamental frequency. In order to understand the reasons of active power oscillation and to find proper solutions, a state-space model for the closed-loop active power control of the microgrid shown in Fig. 2 can be obtained as given in (2)–(9), of which the deduction process is shown in [23]. To simplify the model and focus on the specific eigenvalues causing oscillation, the reactive power part is not included in this model and the line resistance is neglected in inductive output impedance point of view. It is shown in [23] that these simplifications do not affect the precision of the model. ẋ = Ax+Bw (2) y = Cx+Dw where T w = Pload P0_1 P0_2 T y = ωm1 ωm2 Pout1 Pout2 ⎡ ⎤ D1 ωm1 + Pload ⎢ ⎥ J1 ω0 (K1 + K2 ) ⎢ ⎥ D2 ⎢ ⎥ Pload ⎥ x = ⎢ ωm2 + ⎢ ⎥ J2 ω0 (K1 + K2 ) ⎣ ⎦ 1 δ1 − Pload K1 + K2 (3) (4) (5) This article has been accepted for inclusion in a future issue of this journal. Content is final as presented, with the exception of pagination. 4 IEEE TRANSACTIONS ON SMART GRID ⎤ 1 0 0 ⎢0 1 0 ⎥ ⎥ (8) C=⎢ ⎣ 0 0 K1 ⎦ 0 0 −K1 ⎡ ⎤ D1 − 0 0 ⎢ J1 ω0 (K1 +K2 ) ⎥ ⎢ ⎥ D2 ⎢− ⎥ 0 0 ⎢ J ω (K +K ) ⎥ ⎥ 2 0 1 2 (9) D=⎢ ⎢ ⎥ K1 ⎢ ⎥ 0 0 ⎢ ⎥ K1 +K2 ⎣ ⎦ K2 0 0 K1 +K2 and A and B are shown in (6) and (7), respectively, at the bottom of the page. Here, Ki = (Ei Vbus cosδi )/Xi , and Xi ≈ Xf i + Xline i . Analyses of transient active power control performance in the following parts of present section are based on this model, as it describes the transient performance of variables in y after a given disturbance w. ⎡ B. Oscillation Damping It is a known conclusion in the control theory that the poles of transfer function of Yj (s)/Wk (s) are available in the eigenvalues of A, for any j, k. Therefore, the studies on eigenvalues of A should give some clues to damping methods for oscillations in Pout i . The loci of eigenvalues of A with a variation of D1 or X1 are shown in Fig. 3. Nominal parameters are listed in Table I, in which Mi∗ = (Ji ω02 )/Sbase i , D∗i = (Di ω0 )/Sbase i , and Xi∗ = (Xi Sbase i )/E02 . In the eigenvalue loci plots, radial dash lines indicate damping ratio ζ , and circle dash lines indicate natural frequency ωn . As it is shown in Fig. 3, damping ratio of the complex-conjugate eigenvalues increases if the damping factor Di and/or the output reactance impedance Xi are increased. It should be pointed out that increasing Xi causes a decrease in damped natural frequency ωd , which is indicated by the distance between eigenvalue and the real axis. This may result in longer settling time compared to the method of increasing Di . However, the approach of increasing output reactance has other merits as follows. 1) The state-space model is obtained under the assumption that the output impedance of DGs is inductive. ⎡ Fig. 3. Eigenvalue loci with a variation of (a) D∗1 (2−4 ×17 pu ∼ 25 ×17 pu) or (b) X1∗ (2−4 × 0.7 pu ∼ 25 × 0.7 pu). TABLE I S TATE -S PACE M ODEL PARAMETERS This assumption is less valid if Xi is small, especially in a LV microgrid in which the line impedances are mainly resistive [42]. If this assumption is not valid, the active power and reactive power control cannot be decoupled correctly and the system may become more oscillatory and even unstable. 2) To share transient active power properly, output reactance of each DG should be designed equally in per unit value, as it is discussed in the next part of this section. Therefore, the problem of oscillation and that of transient active power sharing can be solved simultaneously by proper stator reactance design. 3) Moreover, the influence of output reactance mismatch on transient active power sharing becomes smaller if output reactance of DGs is increased, owing to decreased relative errors. C. Transient Active Power Sharing The response of output active power of DGs during a loading transition can be calculated from transfer functions Pout1 (s)/Pload (s) and Pout2 (s)/Pload (s), which are the elements G31 (s) and G41 (s) of the matrix G(s) kp1 D1 K2 D1 K2 K1 ⎤ − − ⎢ J1 ω0 (K1 + K2 ) J1 ω0 J1 ω0 (K1 + K2 ) J1 ω0 ⎥ ⎢ kp2 D2 K1 K1 ⎥ D2 K1 ⎢ ⎥ − − A=⎢ ⎥ ⎢ J2 ω0 (K1 + K2 ) J2 ω0 (K1 + K2 ) J2 ω0 J2 ω0 ⎥ ⎣ ⎦ K2 K2 − 0 K1 + K2 K1 + K2 ⎡ D2 K2 D1 kp1 K1 D1 D2 K2 − − + 2 2 1 + 2 2 ⎢ 2 2 2 J ω + K (K ) J ω J J ω + K J ω + K (K ) (K ) 1 0 1 2 ⎢ 1 2 0 1 2 2 1 0 (K1 + K2 ) 1 0 1 ⎢ 2K D D2 kp2 D1 D2 K1 K1 1 ⎢ 1 B = ⎢− − + + 2 2 2 + 2 2 2 2 2 ⎢ J2 ω0 J2 ω0 (K1 + K2 ) J1 J2 ω0 (K1 + K2 ) J2 ω0 (K1 + K2 ) J2 ω0 (K1 + K2 ) ⎢ ⎣ D1 K2 D2 K2 − + J1 ω0 (K1 + K2 )2 J2 ω0 (K1 + K2 )2 − (6) 1 J1 ω0 0 0 ⎤ 0 ⎥ ⎥ ⎥ 1 ⎥ ⎥ J2 ω0 ⎥ ⎥ ⎦ 0 (7) This article has been accepted for inclusion in a future issue of this journal. Content is final as presented, with the exception of pagination. LIU et al.: ENHANCED VSG CONTROL FOR PARALLEL INVERTERS IN MICROGRIDS 5 Fig. 4. Poles and zeros of (a) Pout1 (s)/Pload (s) and (b) Pout2 (s)/Pload (s). Left column: variation of X1∗ (2−4 × 0.7 pu ∼ 25 × 0.7 pu); middle column: variation of M1∗ (1.3−4 × 8 s ∼1.35 ×8 s); right column: variation of D∗1 (2−4 × 17 pu ∼ 25 × 17 pu). shown in (10), respectively. G(s) = C(sI − A)−1 B + D. (10) The loci of poles and zeros of Pout1 (s)/Pload (s) and Pout2 (s)/Pload (s) are shown in Fig. 4. It is shown that if the per unit value of Xi∗ , Mi∗ , and D∗i of each DG are set equally, all poles are cancelled by zeros. This cancellation implies a desirable step change of Pout1 and Pout2 directly to their respective steady-state values without any oscillation during a loading transition. Equal values of Mi∗ and D∗i are not difficult to realize since they are virtual parameters that can be easily changed in the VSG control program. As for Xi∗ , a control method to adjust stator reactance is presented in Section V. IV. I MPROVEMENT OF R EACTIVE P OWER S HARING Fig. 5 shows the principles of ω–P and V–Q droop controls in the “Governor Model” and “Q Droop” blocks shown in Fig. 1 for the case of Sbase1 : Sbase2 = 2 : 1. As discussed in Section II, kp∗ , kq∗ , P∗0 and Q∗0 are designed equally. Based on the predefined linear droop characteristic, the desired power sharing Pin1 : Pin2 = 2 : 1 can be obtained because the governor input is ωm , and ωm1 = ωm2 is guaranteed in steady state. Following the same principle, to share the reactive power according to the power rating ratio, an equal voltage reference is required. However, for the V–Q droop in basic VSG control shown in Fig. 1(c), the voltage reference is the inverter output voltage, which may be a different value for each DG even in steady state due to the line voltage drop. As most of previous studies are based on Q–V droop, in which the output voltage Vout i should be regulated based on measured reactive power Qout i , the basic idea to address this problem is to equalize Vout i Fig. 5. Principles of ω–P and V–Q droop control. by equalizing the output impedance [26], [27], or to compensate the line voltage drop [28]. Both methods need great effort in design process and complex computations in DG control law, whereas the resulted reactive power sharing is still influenced by active power sharing. As the voltage does not need to be controlled directly in a V–Q droop control scheme shown in Fig. 1(a), the reference voltage can be chosen other than inverter output voltage. If the common ac bus voltage Vbus is used instead of inverter output voltage Vout i , equal reactive power reference value Qref 1 = Qref 2 can be guaranteed, as it is illustrated in Fig. 5. Therefore, accurate reactive power sharing Qout1 = Qout2 should be obtained through the using of reactive power PI controller. Moreover, unlike output voltage, bus voltage is not influenced by line voltage drop, which is determined by both active and reactive power. Therefore, reactive power sharing according to the bus voltage is independent from active power. In some previous researches [21], [39], [40], direct bus voltage measurement is suggested. However, in the field applications, it is difficult to measure Vbus directly, as DGs may be installed far away from the common ac bus, and the utilization of communication is not preferred for reliability reason. Therefore, a bus voltage estimation method using local measurement is proposed in next section. This article has been accepted for inclusion in a future issue of this journal. Content is final as presented, with the exception of pagination. 6 IEEE TRANSACTIONS ON SMART GRID Fig. 6. Block diagram of (a) the proposed enhanced VSG control, (b) the “Stator Reactance Adjuster” block and (c) the “Vbus Estimator” block. V. P ROPOSED E NHANCED VSG C ONTROL S CHEME The proposed enhanced VSG control scheme is shown in Fig. 6. Compared to the basic VSG control, two major modifications are made, i.e., the stator reactance adjuster and the bus voltage estimator, as shown in Figs. 6(b) and 6(c), respectively. The function of stator reactance adjuster is to adjust the output reactance of the DG freely. It is operating as a virtual impedance controller. The virtual stator inductor is realized by multiplying output current by the virtual stator inductor in stationary frame. It will be more accuracy if inductor current through Lf is used. However, this increases the number of current sensors, which is not necessary. As the current flowing into Cf at fundamental frequency is less than few percent of the inductor current, using output current instead of inductor current does not affect the performance of the control scheme. Based on the given analyses in Section III and according to (11), tuning of virtual stator inductor Lls is suggested to set total output reactance Xi∗ for both DGs in same large per unit value. This approach increases active power damping ratio and shares transient load without oscillation. The target value is proposed to be 0.7 pu because it is a typical value for the total direct-axis transient reactance X d of a real SG. Xi∗ = Sbase i ωm i (Lls i + Lf i + Lline i )/E02 = 0.7 pu. (11) The Lf i and Zline i (Rline i + jLline i ) are considered as known parameters in this paper. As the scale of microgrid is usually small, the line distance is easily to be measured or fed by the planner. Even if it is not the case, several online measurement or intelligent tuning methods for Zline i are available in [42] and [43]. With the proposed design of stator reactance adjustment, oscillation in a VSG-control-based microgrid should be almost eliminated during a loading transition in islanded mode. Particularly, transition from grid-connected mode to islanded mode can also be considered as a loading transition; therefore, the oscillation during an islanding event should also be eliminated with the proposed control strategy, as it is proved by simulation results in next section. As for other disturbances in islanded mode, e.g., change of active power set value of DG(s), connection/disconnection of DG(s), etc., oscillation cannot be eliminated, but can still be damped by the increased total output reactance. The principle of bus voltage estimator in Fig. 6(c) is similar to that of stator reactance adjuster in Fig. 6(b). By calculating the line voltage drop in stationary frame using measured output current and line impedance data, the bus voltage can be estimated from the difference of output voltage and calculated line voltage drop. Since the RMS value of estimated bus voltage V̂bus for each DG should be approximately equal, as it is discussed in last section, accurate reactive power sharing can be obtained by using estimated bus voltages as the input references of “Q Droop” instead of respective output voltages of DGs. Although the principle of presented bus voltage estimator is not new, the idea of using this estimator to realize communication-less accurate reactive power sharing can be considered as a contribution in the present work. However, if there is an estimation error in V̂bus , it will cause ∗ ∗ + V̂ ∗ = Vbus a reactive power sharing error. Supposing V̂bus1 1 ∗ ∗ + V̂ ∗ , and V̂bus2 = Vbus 2 Q∗out1 − Q∗out2 = −kq∗ (V̂1∗ − V̂2∗ ). (12) That is to say, the reactive power sharing error caused by estimation errors is determined by the V–Q droop gain kq∗ . The design of kq∗ is a well-known trade-off between voltage deviation and reactive power control accuracy. Considering the probable ripples in the measured RMS value of V̂bus , kq∗ is recommended to be 5 pu for the present example. It should be pointed out that the increased output reactance by adding the virtual stator inductor Lls causes a decrease in the reactive control plant gain, as shown in Fig. 7. Therefore, to obtain a same bandwidth of 20 Hz for the reactive power control loop, the gain of PI controller should be increased to compensate the decreased plant gain, as illustrated in Fig. 8. The parameters used to plot Fig. 8 are related to DG1, which are shown in Fig. 9 and Table II. The 20Hz bandwidth is relatively low compared to control methods working on instantaneous value; however, it is fast enough to track the reactive power and regulate the output voltage as it is demonstrated in the simulation and experimental results. This article has been accepted for inclusion in a future issue of this journal. Content is final as presented, with the exception of pagination. LIU et al.: ENHANCED VSG CONTROL FOR PARALLEL INVERTERS IN MICROGRIDS 7 TABLE II S IMULATION PARAMETERS Fig. 7. Small-signal model of reactive power control loop. TABLE III S IMULATION S EQUENCE Fig. 8. Bode plot of reactive power control loop. Fig. 9. Simulation circuit. VI. S IMULATION R ESULTS Simulations are executed in PSCAD/EMTDC environment to verify the effectiveness of the proposed enhanced VSG control scheme. A microgrid shown in Fig. 9 is studied. As it is shown in Fig. 9, impedances of output filters and lines of each DG differ in per unit values. Other main parameters are listed in Table II, and the sequence of simulation is shown in Table III. Events of islanding from grid, loading transition, and intentional active power sharing change are simulated at 21 s, 24 s, and 27 s, respectively. The simulation results are shown in Fig. 10. As it is illustrated in Fig. 10(a), when the microgrid is islanded at 21 s, and when load 2 is connected at 24 s, oscillation can be observed in active power when the basic VSG control is applied for both DGs. This oscillation is almost eliminated by applying the proposed enhanced VSG control shown in Fig. 10(b). As the disturbance at 27 s is caused by change of active power set value of DG1, which is not a loading transition, active power oscillation cannot be eliminated in this case. However, the proposed enhanced VSG control increases the damping ratio; therefore, the overshoots in Fig. 10(b) are smaller than that in Fig. 10(a). Meanwhile, the oscillation periods become longer, because the damped natural frequencies are decreased as it is discussed in Section III-B. Note that the rate of change of frequency remains the same in all cases, which suggests that the proposed enhanced VSG control has no influence on the inertia support feature of VSG control. Moreover, in the case of the basic VSG control, reactive power is not shared properly in islanded mode, and is not controlled at set value in grid-connected mode, due to the voltage drop through the line impedance, as shown in Fig. 10(a). Besides, reactive power control is not independent from active power control, as a change of set value of active power at 27 s also causes a change of reactive power sharing. These problems are all solved in the enhanced VSG control, as it is shown in Fig. 10(b). It is also noteworthy that the steady-state deviations of DG voltage and bus voltage become smaller when the enhanced VSG control is applied. Fig. 11 illustrates the dynamic performance of reactive power and voltage during the loading transition at 24 s. Although the virtual internal emf E1 becomes much higher in the proposed enhanced VSG control owing to the voltage drop on virtual stator inductance Lls , the maximum voltage sag of PWM inverter reference Vpwm1 and output voltage Vout1 are kept within the same level as the basic VSG control. This implies that the voltage drop on Lls is compensated well by the reactive power PI controller. Besides, although in the enhanced VSG control, the voltage becomes slightly oscillatory, the reactive power oscillation converges within 0.1 s. This article has been accepted for inclusion in a future issue of this journal. Content is final as presented, with the exception of pagination. 8 IEEE TRANSACTIONS ON SMART GRID Fig. 10. Simulation results of active power and frequency in the left column and reactive power and voltage in the right column when both DGs are controlled by (a) the basic VSG control, (b) the proposed enhanced VSG control. Fig. 12. Setup of experiment system. TABLE IV E XPERIMENT S EQUENCE Fig. 11. Zoom-in simulation results of reactive power and voltage of DG1 at 24 s. (a) The basic VSG control; (b) the proposed enhanced VSG control. VII. E XPERIMENTAL R ESULTS Experiments are executed in an islanding microgrid, of which the circuit is the same as that of simulation shown in Fig. 9, except that instead of dc sources, ac supply rectified by diode bridges is used to imitate the dc output of DGs, and the breaker BK3 is opened. The setup of experiment system is shown in Fig. 12 and experiment sequence is shown in Table IV. Control Parameters are the same as those listed in Table II, and the experimental results are shown in Fig. 13. Experimental results verify again the effectiveness of the proposed enhanced VSG control. First, the oscillation due to loading transition at 0.5 s is eliminated, and the oscillation due to change of set value of acitive power at 3.0 s is damped. It implies that the proposed enhanced VSG control This article has been accepted for inclusion in a future issue of this journal. Content is final as presented, with the exception of pagination. LIU et al.: ENHANCED VSG CONTROL FOR PARALLEL INVERTERS IN MICROGRIDS 9 Fig. 13. Experimental results of active power and frequency in the left column and reactive power and voltage in the right column when both DGs are controlled by (a) the basic VSG control, (b) the proposed enhanced VSG control. is able to track the loading transition rapidly and accurately without oscillation; meanwhile, the inertia support of the basic VSG control is kept. Even when an oscillation occurs, the overshoot is suppressed owing to increased system damping. Furthermore, by applying the enhanced VSG control, reactive power is shared according to power rating ratio, and is immune to active power sharing change and line impedance mismatch in per unit values. Although ripples in RMS value of output voltage can be observed due to a slight load unbalance, the reactive power is controlled well when the enhanced VSG control is applied. VIII. C ONCLUSION In this paper, an enhanced VSG control is proposed as a novel communication-less control method in a microgrid. A stator reactance adjuster is developed based on state-space analyses, in order to increase the active power damping and to properly share transient active power. A novel communicationless reactive power control strategy based on inversed voltage droop control (V–Q droop control) and common ac bus voltage estimation is also proposed to achieve accurate reactive power sharing, which is immune to active power sharing change and line impedance mismatch. Simulation and experimental results demonstrated that the proposed enhanced VSG control achieves desirable transient and steady-state performances, and keeps the inertia support feature of VSG control. As a result, the proposed enhanced VSG control is a preferable choice for the control system of DGs in microgrids. R EFERENCES [1] R. H. Lasseter, “Microgrids,” in Proc. IEEE Power Eng. Soc. Winter Meeting, New York, NY, USA, 2002, pp. 305–308. [2] J. M. Guerrero, J. C. Vasquez, J. Matas, L. G. de Vicuña, and M. Castilla, “Hierarchical control of droop-controlled AC and DC microgrids— A general approach toward standardization,” IEEE Trans. Ind. Electron., vol. 58, no. 1, pp. 158–172, Jan. 2011. [3] A. Bidram and A. Davoudi, “Hierarchical structure of microgrids control system,” IEEE Trans. Smart Grid, vol. 3, no. 4, pp. 1963–1976, Dec. 2012. [4] M. C. Chandorkar, D. M. Divan, and R. Adapa, “Control of parallel connected inverters in standalone AC supply systems,” IEEE Trans. Ind. Appl., vol. 29, no. 1, pp. 136–143, Jan./Feb. 1993. [5] H. Bevrani, M. Watanabe, and Y. Mitani, Power System Monitoring and Control. Hoboken, NJ, USA: Wiley, 2014. [6] J. M. Guerrero, J. Matas, L. G. De Vicuña, M. Castilla, and J. Miret, “Decentralized control for parallel operation of distributed generation inverters using resistive output impedance,” IEEE Trans. Ind. Electron., vol. 54, no. 2, pp. 994–1004, Apr. 2007. [7] T. L. Vandoorn, B. Meersman, L. Degroote, B. Renders, and L. Vandevelde, “A control strategy for islanded microgrids with DC-link voltage control,” IEEE Trans. Power Del., vol. 26, no. 2, pp. 703–713, Apr. 2011. [8] T. L. Vandoorn, B. Meersman, J. D. M. De Kooning, and L. Vandevelde, “Analogy between conventional grid control and islanded microgrid control based on a global DC-link voltage droop,” IEEE Trans. Power Del., vol. 27, no. 3, pp. 1405–1414, Jul. 2012. [9] J. C. Vasquez, J. M. Guerrero, M. Savaghebi, J. Eloy-Garcia, and R. Teodorescu, “Modeling, analysis, and design of stationary-referenceframe droop-controlled parallel three-phase voltage source inverters,” IEEE Trans. Ind. Electron., vol. 60, no. 4, pp. 1271–1280, Apr. 2013. [10] J. Driesen and K. Visscher “Virtual synchronous generators,” in Proc. IEEE Power Energy Soc. Gen. Meeting—Convers. Del. Elect. Energy 21st Century, Pittsburgh, PA, USA, 2008, pp. 1–3. [11] L. M. A. Torres, L. A. C. Lopes, T. L. A. Moran, and C. J. R. Espinoza, “Self-tuning virtual synchronous machine: A control strategy for energy storage systems to support dynamic frequency control,” IEEE Trans. Energy Convers., vol. 29, no. 4, pp. 833–840, Dec. 2014. [12] Y. Hirase et al., “Virtual synchronous generator control with double decoupled synchronous reference frame for single-phase inverter,” IEEJ J. Ind. Appl., vol. 4, no. 3, pp. 143–151, May 2015. [13] K. Sakimoto, Y. Miura, and T. Ise, “Stabilization of a power system with a distributed generator by a virtual synchronous generator function,” in Proc. 8th IEEE Int. Conf. Power Electron. ECCE Asia, Shilla Jeju, Korea, 2011, pp. 1498–1505. [14] T. Shintai, Y. Miura, and T. Ise, “Reactive power control for load sharing with virtual synchronous generator control,” in Proc. 7th Int. Power Electron. Motion Control Conf., Harbin, China, 2012, pp. 846–853. This article has been accepted for inclusion in a future issue of this journal. Content is final as presented, with the exception of pagination. 10 [15] T. Shintai, Y. Miura, and T. Ise, “Oscillation damping of a distributed generator using a virtual synchronous generator,” IEEE Trans. Power Del., vol. 29, no. 2, pp. 668–676, Apr. 2014. [16] J. Alipoor, Y. Miura, and T. Ise, “Power system stabilization using virtual synchronous generator with alternating moment of inertia,” IEEE J. Emerg. Sel. Topics Power Electron., vol. 3, no. 2, pp. 451–458, Jun. 2015. [17] H. Bevrani, T. Ise, and Y. Miura, “Virtual synchronous generators: A survey and new perspectives,” Int. J. Elect. Power Energy Syst., vol. 54, no. 1, pp. 244–254, Jan. 2014. [18] C. Cheng, Z. Zeng, H. Yang, and R. Zhao, “Wireless parallel control of three-phase inverters based on virtual synchronous generator theory,” in Proc. Int. Conf. Elect. Mach. Syst. (ICEMS), Busan, Korea, 2013, pp. 162–166. [19] Y. Chen, R. Hesse, D. Turschner, and H.-P. Beck, “Improving the grid power quality using virtual synchronous machines,” in Proc. Int. Conf. Power Eng. Energy Elect. Drives, Málaga, Spain, 2011, pp. 1–6. [20] Q.-C. Zhong and G. Weiss, “Synchronverters: Inverters that mimic synchronous generators,” IEEE Trans. Ind. Electron., vol. 58, no. 4, pp. 1259–1267, Apr. 2011. [21] Q.-C. Zhong, P.-L. Nguyen, Z. Ma, and W. Sheng, “Self-synchronized synchronverters: Inverters without a dedicated synchronization unit,” IEEE Trans. Power Electron., vol. 29, no. 2, pp. 617–630, Feb. 2014. [22] N. Soni, S. Doolla, and M. C. Chandorkar, “Improvement of transient response in microgrids using virtual inertia,” IEEE Trans. Power Del., vol. 28, no. 3, pp. 1830–1838, Jul. 2013. [23] J. Liu, Y. Miura, and T. Ise, “Comparison of dynamic characteristics between virtual synchronous generator and droop control in inverterbased distributed generators,” IEEE Trans. Power Electron., vol. 31, no. 5, pp. 3600–3611, May 2016. [24] M. Guan et al., “Synchronous generator emulation control strategy for voltage source converter (VSC) stations,” IEEE Trans. Power Syst., vol. 30, no. 6, pp. 3093–3101, Nov. 2015. [25] A. D. Paquette, M. J. Reno, R. G. Harley, and D. M. Divan, “Sharing transient loads: Causes of unequal transient load sharing in islanded microgrid operation,” IEEE Ind. Appl. Mag., vol. 20, no. 2, pp. 23–34, Mar./Apr. 2014. [26] J. He, Y. W. Li, J. M. Guerrero, F. Blaabjerg, and J. C. Vasquez, “An islanding microgrid power sharing approach using enhanced virtual impedance control scheme,” IEEE Trans. Power Electron., vol. 28, no. 11, pp. 5272–5282, Nov. 2013. [27] Y. Zhu, F. Zhou, F. Wang, B. Liu, and Y. Zhao, “A wireless load sharing strategy for islanded microgrid based on feeder current sensing,” IEEE Trans. Power Electron., vol. 30, no. 12, pp. 6706–6719, Dec. 2015. [28] Y. W. Li and C.-N. Kao, “An accurate power control strategy for power-electronics-interfaced distributed generation units operating in a low-voltage multibus microgrid,” IEEE Trans. Power Electron., vol. 24, no. 12, pp. 2977–2988, Dec. 2009. [29] C.-T. Lee, C.-C. Chu, and P.-T. Cheng, “A new droop control method for the autonomous operation of distributed energy resource interface converters,” IEEE Trans. Power Electron., vol. 28, no. 4, pp. 1980–1993, Apr. 2013. [30] E. Rokrok and M. E. H. Golshan, “Adaptive voltage droop scheme for voltage source converters in an islanded multibus microgrid,” IET Gener. Transm. Distrib., vol. 4, no. 5, pp. 562–578, May 2010. [31] H. Xu et al., “A reactive power sharing method based on virtual capacitor in islanding microgrid,” in Proc. Int. Power Electron. Conf., Hiroshima, Japan, 2014, pp. 567–572. [32] J. He and Y. W. Li, “An enhanced microgrid load demand sharing strategy,” IEEE Trans. Power Electron., vol. 27, no. 9, pp. 3984–3995, Sep. 2012. [33] J. He, Y. W. Li, and F. Blaabjerg, “An enhanced islanding microgrid reactive power, imbalance power, and harmonic power sharing scheme,” IEEE Trans. Power Electron., vol. 30, no. 6, pp. 3389–3401, Jun. 2015. [34] H. Mahmood, D. Michaelson, and J. Jiang, “Accurate reactive power sharing in an islanded microgrid using adaptive virtual impedances,” IEEE Trans. Power Electron., vol. 30, no. 3, pp. 1605–1617, Mar. 2015. [35] H. Han, Y. Liu, Y. Sun, M. Su, and J. M. Guerrero, “An improved droop control strategy for reactive power sharing in islanded microgrid,” IEEE Trans. Power Electron., vol. 30, no. 6, pp. 3133–3141, Jun. 2015. [36] Q. Shafiee, J. M. Guerrero, and J. C. Vasquez, “Distributed secondary control for islanded microgrids—A novel approach,” IEEE Trans. Power Electron., vol. 29, no. 2, pp. 1018–1031, Feb. 2014. [37] A. Micallef, M. Apap, C. Spiteri-Staines, J. M. Guerrero, and J. C. Vasquez, “Reactive power sharing and voltage harmonic distortion compensation of droop controlled single phase islanded microgrids,” IEEE Trans. Smart Grid, vol. 5, no. 3, pp. 1149–1158, May 2014. IEEE TRANSACTIONS ON SMART GRID [38] Y. Zhang and H. Ma, “Theoretical and experimental investigation of networked control for parallel operation of inverters,” IEEE Trans. Ind. Electron., vol. 59, no. 4, pp. 1961–1970, Apr. 2012. [39] Q.-C. Zhong, “Robust droop controller for accurate proportional load sharing among inverters operated in parallel,” IEEE Trans. Ind. Electron., vol. 60, no. 4, pp. 1281–1290, Apr. 2013. [40] C. K. Sao and P. W. Lehn, “Autonomous load sharing of voltage source converters,” IEEE Trans. Power Del., vol. 20, no. 2, pp. 1009–1016, Apr. 2005. [41] H. Bevrani, Robust Power System Frequency Control, 2nd ed. New York, NY, USA: Springer, 2014. [42] H. Bevrani and S. Shokoohi, “An intelligent droop control for simultaneous voltage and frequency regulation in islanded microgrids,” IEEE Trans. Smart Grid, vol. 4, no. 3, pp. 1505–1513, Sep. 2013. [43] L. Asiminoaei, R. Teodorescu, F. Blaabjerg, and U. Borup, “A digital controlled PV-inverter with grid impedance estimation for ENS detection,” IEEE Trans. Power Electron., vol. 20, no. 6, pp. 1480–1490, Nov. 2005. Jia Liu (S’15) was born in Xuzhou, China, in 1986. He received the B.Eng. and M.Eng. degrees in electrical engineering from Xi’an Jiaotong University, Xi’an, China, in 2008 and 2011, respectively, and the Engineering degree in mechanical systems from the University of Technology of Troyes, Troyes, France, in 2011. He is currently pursuing the Ph.D. degree with Osaka University, Osaka, Japan. He was with Delta Electronics (Jiangsu) Ltd., Nanjing, China, from 2011 to 2012. His research interests include power converter control, microgrids, modular multilevel converters, and motor control. Yushi Miura (M’06) received the Doctoral degree in electrical and electronic engineering from the Tokyo Institute of Technology, Tokyo, Japan, in 1995. From 1995 to 2004, he joined the Japan Atomic Energy Research Institute as a Researcher, and developed power supplies and superconducting coils for nuclear fusion reactors. Since 2004, he has been an Associate Professor of the Division of Electrical, Electronic, and Information Engineering, Osaka University, Osaka, Japan. His areas of research involve applications of power electronics for power systems. Hassan Bevrani (S’90–M’04–SM’08) received the Ph.D. degree in electrical engineering from Osaka University, Osaka, Japan, in 2004. From 2004 to 2014, he has worked as a Postdoctoral Fellow, a Senior Research Fellow, a Visiting Professor, and a Professor with Kumamoto University, Japan; the Queensland University of Technology, Australia; the Kyushu Institute of Technology, Japan; Ecole Centrale de Lille, France; and Osaka University. He is currently working as a Professor with the University of Kurdistan, Iran. He has authored five books, ten book chapters, and over 200 journal/conference papers. His current research interests include smart grid control, robust power system monitoring and control, and microgrid dynamics and control. Toshifumi Ise (M’86) was born in 1957. He received the B.Eng., M.Eng., and Doctor of Engineering degrees in electrical engineering from Osaka University, Osaka, Japan, in 1980, 1982, and 1986, respectively. From 1986 to 1990, he was at the Nara National College of Technology, Nara, Japan. Since 1990, he has been with the Faculty of Engineering, Osaka University, where he is currently a Professor of the Division of Electrical, Electronic, and Information Engineering, Graduate School of Engineering. His research interests are in the areas of power electronics and applied superconductivity for power systems, including superconducting magnetic energy storage and future power systems including many distributed generations.