A Supervised Hybrid Algorithm based DSTATCOM to Cater to Dynamic Load Changes

Abstract

Renewable energies like Photo Voltaic (PV)/ solar power are abundantly available and waiting to be harnessed. Being weak system renewable energy based power systems require reactive power generation close to load to unburden the source. Study reveals fixed tuned Proportional & Integral (PI) controller based DSTATCOM may not be able to provide satisfactory voltage regulation with wide load changes. DSTATCOM generally requires tuning of PI controllers by utility engineers during installation. This process is mostly trial and error approach. It is necessary to re-tune the DSTATCOM controller when there is change in operating condition. To ensure automatic control action irrespective of load conditions, soft-computing technique is implemented in the DSTATCOM. A supervised hybrid algorithm named Neuro- Fuzzy controller is adopted to ensure automatic adaptation of the controller parameters during changing load conditions. The paper presents a Synchronous Reference Frame theory (SRF) based DSTATCOM on MATLAB platform and uses a Neuro- Fuzzy inference system to get a better response in terms of dynamic voltage profile and Total Harmonic Distortion (THD) as compared to simple PI Controller.