Robust Control of Photovoltaic-fed Distribution Static Compensator Using Student Psychology Optimization-Based Least Mean Logarithmic Square Algorithm

The presence of nonlinear and dynamic loads may cause the load current to become severely distorted in the power system. Consequently, harmonic currents penetrate the grid, leading to distortion of the grid current, and the power factor is also affected. Therefore, this paper proposes an improved robust least mean logarithm square (IRLMLS) scheme for controlling the distribution static compensator to improve the power quality of a photovoltaic-fed distribution system. A student psychology-based optimization (SPBO)-tuned artificial neural network method is also implemented to design the maximum power point tracking, which is employed to generate the DC voltage (i.e., the maximum power point (MPP) voltage). This reference DC voltage is the input of the d-axis controller. In this IRLMLS controller, the SPBO algorithm optimizes the d-axis and q-axis PI controller gains. In the current work, the performance of the proposed method has been verified with the SPBO-based \(i{\text{cos}}\emptyset\) technique. A slight DC voltage variation (i.e., 0.22%) occurs while applying the \(i{\text{cos}}\emptyset\) algorithm. Furthermore, the THD values have been reduced due to the implementation of the SPBO-based IRLMLS method. Even the overall system performance is also validated under nonlinear, dynamic, and unbalanced loading conditions.