Design and Development of Various Virtual Oscillator Control Techniques for Parallel Inverters in Standalone Microgrid

Abstract

In recent times, renewable energy sources (RES) have been used as potential alternatives to conventional generation systems connected to the grid. The power electronic inverters are the principal media of interface for integrating the RES into the utility grid system. This work is primarily focused on the different virtual oscillator control (VOC) strategies such as Deadzone oscillator (DZO) based VOC, Van der Pol oscillator (VdPO) based VOC, and Andronov-Hopf oscillator (AHO) based VOC for the parallel inverters in a standalone Microgrid (MG). The well-known droop control method for parallel inverters emulates only the droop characteristics of the synchronous machine. Virtual synchronous machine (VSM) is also a familiar control method for parallel inverters. It emulates not only the droop characteristics of the synchronous machine but also the swing equation. Therefore, VSM has a remarkable difference in the dynamic performance of the system compared to the droop control method. The droop and VSM control methods need measurements of both voltage and current. In addition, the calculation of active and reactive power requires low-pass filters (LPFs). Typically, LPFs have low cut-off frequencies that restrict the controller bandwidth. The slow dynamic response, real and reactive powers interaction, sensitive performance with line impedance and non-linear loads, etc., are still evident drawbacks despite the many modified methods to improve the droop and VSM control methods. However, the VOC works on instantaneous feedback signals so that it achieves much faster synchronization and better power-sharing. The idea of the proposed VOC is to control an inverter such that it emulates the behavior of nonlinear oscillators. Compared to droop and VSM techniques, the VOC can achieve faster dynamic response and is easy to implement. It also does not require inner control loops, trigonometric functions, and LPFs. The selection of control parameters in DZO based VOC is difficult and time-consuming by using conventional method. Hence, they are designed by using different optimization techniques such as Particle Swarm Optimization (PSO), Sine Cosine Algorithm (SCA), modified Sine Cosine Algorithm (mSCA), African Vulture Optimization Algorithm (AVOA), and Artificial Jellyfish Search Optimization (AJSO). The MATLAB/Simulink simulations are carried out to examine the performance of the system with the conventional DZO based VOC and the aforesaid optimized DZO based VOC such as VOC-PSO, VOC-SCA, VOC-mSCA, VOC-AVOA, and VOC-AJSO. In comparison to all control methods, the VOC-AJSO is observed to achieve faster synchronization. The effectiveness of the suggested VOC-AJSO control approach is also proved by the experimental results. In the traditional VOC methods such as DZO and VdPO based VOC, there is always the presence of a third-order harmonic in the output voltage, which causes a significant amount of third-order harmonic current in the system. The nonlinear dynamical equations of the oscillator are analyzed, and its nonlinear current source is made simpler in order to develop New-VdPO based VOC for parallel inverters that can effectively get rid of the third-order harmonic component in the oscillator’s output voltage. Finally, an extensive comparison of DZO, VdPO, and a New-VdPO based VOC methods with linear and nonlinear loads is presented. Simulation results of the DZO, VdPO, and the proposed New-VdPO based VOC methods with different loads (resistive, linear RLC, non-linear) are compared and analyzed in detail. The third-order harmonic is dominant in both DZO and VdPO based VOC methods whereas it is very less in the New-VdPO based VOC method. Hardware experimentation is also carried out to analyze the efficacy of the proposed New-VdPO based VOC method for parallel inverters in standalone MG. The results clearly depict that the New-VdPO based VOC strategy is quite efficient in handling the output voltage harmonics. In the DZO, VdPO, and New-VdPO based VOC methods, the output voltage falls as the load increases, which might be problematic for sensitive equipment that demands a constant voltage. This work proposed a fuzzy logic algorithm (FLA) based adaptive VOC methods (DZO-Adaptive, VdPO-Adaptive, and New-VdPO-Adaptive) to eliminate this limitation. The proposed New-VdPO-Adaptive based VOC controller has good voltage regulation and lower total harmonic distortion (THD). The computational and experimental results of the proposed control schemes are presented and compared to other control methods to determine the effectiveness of the proposed adaptive control schemes. The limitations of DZO and VdPO based VOC methods are limit cycle restrictions, stability issues due to initial conditions, and guaranteed synchronization conditions. The AHO based VOC method does not need to consider the limit cycle restriction or the rigorous adequate synchronization requirement and provides improved power quality. The parameter selection of AHO is quite complex because the speed constant and AHO capacitor have a greater range of values. To minimize these limitations, the mSCA based optimization approach is proposed for choosing the AHO parameters. The objective of this is to reduce the settling time of the inverter output voltage during startup and the peak overshoot as a constraint. Finally, the proposed optimized AHO (OAHO) method can give better performance than the other aforesaid control methods in terms of initial response time and THD. Finally, the above-mentioned control techniques are applied to three-phase parallel inverters in standalone MG. When compared to the aforesaid control approaches, the proposed OAHO based VOC method can provide superior performance. The significant works have been carried out in this dissertation: (a) the detailed implementation of various VOC methods (DZO, VdPO, and AHO) for parallel single-phase and three-phase inverters in standalone MG, (b) different optimization techniques are used to improve the existing VOC methods, (c) proposed a FLA based adaptive VOC methods to eliminate the voltage regulation issues in the conventional VOC methods, (d) overall performance of the proposed VOC methods has been evaluated in computational and experimental platforms.