Power Quality Improvement using Hopfield Neural Network in Grid Distribution System

The objective of this study revolves around the power quality enhancement in grid connected distribution systems through Hopfield Neural Network (HNN) algorithm. The distribution system is linked to a non-linear load resulting in the generation of harmonics in the grid current. Distribution Static Synchronous Compensator (DSTATCOM) is implemented to mitigate harmonics. DSTATCOM contains IGBTs, that require firing signals to operate. The control algorithm provides the gate pulse by comparing grid and reference grid current; and aims at removing grid current harmonics for the improvement of system’s performance. A comparative analysis of the HNN algorithm is executed with Self Tuning Filter and Synchronous Reference Frame Theory where metrics such as harmonic content, settling time and oscillations are taken under consideration. The distribution system and different control algorithms incorporated by the DSTATCOM have been simulated using MATLAB 2018b Simulink software.