INVESTIGATION OF DIFFERENT INTELLIGENT CONTROL ALGORITHM FOR D-STATCOM

Abstract

The presence of harmonics in power supply network poses a severe power quality problem that results in huge power losses in distribution system, causes interference in the parallel running communication lines and, sometimes, operational failures of electronic equipment. DSTATCOM are employed to feed the required reactive power demand, suppress the current harmonics, minimize the unbalancing in the line currents and reduce the total harmonic distortion (THD) in the load. The voltage source converter (VSC) is the core of a DSTATCOM. The hysteresis current control is an indirect method of controlling the VSC. In this work, Synchronous Reference Frame (SRF) theory with Proportional-Integral controller, Conductance based Fryze Algorithm, Fuzzy logic control technique, different topologies of fuzzy logic controller and Adaptive Neuro-Fuzzy Inference system technique has been implemented for the generation of reference current signals for the controller. This work investigates the effectiveness of the proposed model in mitigation of harmonics currents, feed reactive power demanded by the load and reduce unbalancing due to the different loads in each phase by modelling a 3-φ three-wire D-STATCOM. Simulation results indicate that the proposed D-STATCOM can restrain the unwanted component of current from reaching to source effectively. All topologies are tested for different loads (non-linear load and unbalanced linear load). Performance comparison between the PI controller, PI-like FLC, PI gain scheduled FLC, Hybrid FLC-PI controller and ANFIS controller are made in terms of several performance criteria such as peak overshoot, settling time (ts) and rise time of the DC link voltage. Under all these topologies, THD of the grid side current is well within the IEEE standards and UPF at grid side is maintained.