Optimum Switching of TSC-TCR Using GA Trained ANN for Minimum Harmonic Injection

Indian distribution systems are facing various power quality problems such as low power factor, harmonic distortion and reliability. SVC such as TSC-TCR is commonly used for balancing reactive power and correcting power factor. TSC-TCR type of compensators can compensate reactive powers of all the three phases independently by controlling the thyristors used in conjunction with TSC and delta connected Thyristor controlled reactors. TSC-TCR which are installed in distribution lines compensate unbalanced reactive power but while doing so the TCR injects harmonic currents of odd order into the point of common coupling. TSC-TCR types of SVCs have lower cost and moderately complex control strategy as compared to STATCOMs. This paper proposes a new method of controlling the injected harmonics of TSC-TCR system by using genetic algorithm based ANN training. The scheme can be effectively used at the existing TSC-TCR installations to reduce harmonic injections while catering to fluctuating loads.