Stability Enhancement of PV Powered Microgrid using Levenberg-Marquardt Algorithm Based Intelligent Controller Under Grid-connected Mode
An effective and robust controller is designed using Levenberg-Marquardt (LM) algorithm-based Artificial Neural Network (ANN) for the solar Photo-Voltaic (PV) based distributed generation units for stabilizing the grid-connected microgrid (MG) under load changes and irradiance variations. A test system comprising of two PV units and one diesel generator unit connected to the utility grid is modelled and considered for the controller design in MATLAB/Simulink environment. PV generated power is injected into the grid through voltage source converter (VSC) regulated by using the proposed ANN controller. Based on the grid voltage and available PV generation, the ANN controller regulates the inverter current by setting the reference voltage vector to synthesize gating pulses for the inverter. The robustness of the controller design is analysed and validated through time-domain simulations by subjecting it to extreme operating conditions. The controller performance is evaluated by Integral Square Error (ISE) and Integral Time Absolute Error (ITAE) for the test system. The results are compared with conventional PI and PID controllers to prove the superior performing ANN controller.