A new optimal design of ACD-based UPFC supplementary controller for interconnected power systems

Abstract

This study presents a new optimal supplementary neuro-controller design for unified power flow controllers (UPFCs) using wide-area signals. The new design is oriented by the dual heuristic programming (DHP) method, which is a powerful adaptive critic procedure. The proposed controller injects additional signals to the UPFC series inverter to enhance power system stability. A model, action and critic neural networks are also designed for optimizing the neuro-controller. To have an accurate dynamic response of multi-machine power systems, the concepts of two-machine equivalent model (TMEM) and center of inertia (COI) are used to train the adaptive critic design (ACD) neural networks. The ANFIS structure is applied to the suggested DHP technique based on the selected input signals to improve the dynamic performance of the applied DHP controller. The designed optimal controller is applied on a real interconnected power system between Tehran and Khuzestan regions in Iran power grid. The proposed ANFIS-DHP controller performance is compared with various controllers, such as PI-Genetic, PI-Lyapunov and other intelligent approaches. The dynamic responses of the proposed controller are found to be the most effective in increasing system stability and damping out system inter-area oscillations.