A neural network controller and a simple circuit of SVPWM technique to increase five-level VSC STATCOM performance during voltage sag and swell

Abstract

The most critical disturbance faced in the electrical distribution systems is power service interruptions due to voltage sag or swell which results in economic losses on the user’s side. To compensate voltage sag or swell, advanced custom power devices are used and one of such devices is the static synchronous compensation (STATCOM). This paper presents the implementation of 5-level voltage source converter (VSC) STATCOM using a neural network (NN) and a simplified space vector pulse width modulation (SVPWM) circuit. The primary objective of the NN controller and SVPWM circuit is to enhance the performance and response time of the STATCOM system, specifically in terms of improving voltage and power factor (PF) when faced with voltage sag or swell. The performance of STATCOM was examined within the context of the IEEE 3-bus system. The investigation focused on two scenarios: a single-line-to-ground fault resulting in voltage sag, and the sudden connection of a capacitive load leading to voltage swell. The findings unequivocally demonstrated the efficacy of the STATCOM with a NN controller in comparison to a conventional controller. The utilization of the NN controller resulted in notable improvements in voltage and PF within a remarkably short time frame of 0.02 seconds.