Control and optimization of grid-connected inverters for distributed generation using Ziegler–Nichols and Genetic Algorithm

Abstract

The increasing integration of inverter-based distributed generation (DG) into modern power systems has heightened the need for advanced control strategies to maintain power quality and ensure efficient grid operation. This study proposes a control and optimization approach for grid-connected inverters for DG systems using Genetic Algorithms (GA), with performance benchmarked against the conventional Ziegler–Nichols (Z-N) method. A proportional–integral (PI) current control strategy is developed and formulated as a constrained multi-objective minimization problem, where controller gains are optimized using weighted objectives based on the Integral Time Absolute Error (ITAE), to enhance tracking accuracy, dynamic performance, and system stability across varying operating conditions. The system computes reference currents from active and reactive power generated by the DG’s, enhancing dynamic response and adaptability. It is validated under different power injection scenarios in single and dual-inverter DG configurations. The simulation results showed an improvement in the tracking efficiency of $99.502\%$ using GA, surpassing $90.135\%$ achieved with Z-N tuning. ITAE values for the GA-based direct quadrature axis PI controllers are $0.205333$ and 0.151923, with steady-state errors of $0.7$ and$0.8$, and a settling time of$0.00526s$. In contrast, Z-N yielded higher ITAE values (1.76578 and$1.65952$) and a longer settling time of $0.008005s$. FFT analysis shows negligible voltage distortion (0.00 %) and reduced current total harmonic distortion (0.22 % with GA, 1.98 % with Z-N), meeting IEEE 519–2014 and IEC 547 standards. The findings confirm that GA-based optimization significantly improves tracking, control accuracy, and power quality over conventional Z-N tuning, providing a robust solution for inverter control in grid-connected DG systems.