Author : Hasan I. Al-Rubaiey and Rashid H. Al-Rubayi

This paper presents a developed controller for a Static Var Compensator (SVC) System by using an Artificial Neural Networks (ANNs) for compensating unbalanced fluctuating loads and enhancing the efficiency of operating the distribution network namely; source power factor, load voltage profile, total line power losses, and line thermal limit factor. Two types of reactive power compensators are utilized, the Fixed Capacitor (FC), and the Thyristor- Controlled Reactor (TCR) type compensators. The proposed controller designed to reduce and balance the reactive power drifting from the supply bus-bar under many unbalanced load conditions while keeping the harmonic injection to the Point of Common Coupling (PCC) due to the SVC operation quietly low. The first stage of the proposed controller is Gravitational Search Algorithm (GSA). This algorithm determines the optimal thyristor firing angles of TCR that balance the system with a little drafting of the reactive power drawn from the supply and inject minimum harmonics to PCC indicated by Total Harmonic Distortion (THD). The computational speed of finding the optimum TCR’s firing angles is improved by replacing the GSA by a set of online ANNs trained with hundreds of data generated from GSA. The proposed controller has been verified through proper simulation backed by practical Iraqi distribution network (Ghazali – Muhandessen 33 kV feeder). Finally, the study shows that the use of the ANNs is completely a suitable choice for the real-time control, load balancing, and reactive power compensation.

Keywords: Static Var Compensator, Fixed Capacitor etc.

Article published in International Journal of Current Engineering and Technology, Vol.7, No.3 (June-2017)