Novel Adaptive Recurrent Neural Controller based on VSC HVDC Damping Controller to Improve Power System Stability
Subject Areas : International Journal of Smart Electrical Engineering
1 - Faculty Member, Department of Electrical Engineering, Technical and Vocational University (TVU), Tehran, Iran
Keywords: Power system dynamic stability, Recurrent neural network, VSC HVDC, Oscillation Modes Controllability, Supplementary Damping Controller,
Abstract :
The use of high voltage direct current (HVDC) transmission lines in power systems not only increase the capacity of electrical power transmission systems, but also strengthen the stability of the power network. In order to optimize the HVDC influences on voltage-frequency stability, it is necessary to design supplementary controllers in the most optimal path between input-output signals of the whole power system. The supplementary controllers are added to the local control loop of HVDC to improve active-reactive power flow. In this paper, an optimized method based on the controllability concept is proposed for the coupling of the input-output (IO) signals of the power system equipped with voltage source converter (VSC)-based HVDC. Then, the optimal path is used for supplementary damping controller design based on a novel adaptive recurrent neural network (ARNN). The ARNN is trained online Using a new training algorithm. The simulation results, which are carried on using MATLAB software, show the effectiveness of the control strategy to improve the voltage profile and dynamic stability of the power system.