Implementation of distribution static compensator for power quality enhancement using learning vector quantisation

Abstract

An implementation of a three-phase distribution static compensator (DSTATCOM) using learning vector quantisation (LVQ) control algorithm is presented for compensation of loads in a distribution system. The DSTATCOM is used for harmonics elimination, load balancing and reactive power compensation with self-supporting dc bus in distribution system. The LVQ-based control algorithm is used for extraction of fundamental weighted values of active and reactive power components of load currents which are required for estimation of reference supply currents. It is based on a double supervisory system where each weighted vector is modified by input weights. The input weights are trained outputs of the first stage learning weights. It is needed for accurate estimation of compensation currents from the voltage source converter (VSC) used in DSTATCOM. A prototype of the VSC-based DSTATCOM is developed using a digital signal processor to validate the proposed control approach and its performance is studied under a steady state and dynamic conditions. The performance of the DSTATCOM is found satisfactory with the proposed control algorithm under linear and non-linear loads which are commonly used in the distribution system. The dc bus voltage of the VSC of DSTATCOM has a negligible effect on varying loads.