Neural Network–Assisted Harmonic Reduction and Power Quality Enhancement in Vsc-Hvdc Links

Authors: Asst. Prof Anil Choubey

Abstract: Long distance AC transmission is often subjected to certain problems which limit the transmission capability. HVDC is a better option for transmission of power over long distances. Power is being transmitted between two generating stations via dc link. The control of power flow in DC link can be achieved through control of current or voltage. For minimization of loss considerations, it is important to maintain constant voltage in link and adjust current to meet required power. In this project, a HVDC system is designed to control the power flow between two converter stations with conventional PI controller and Artificial Neural Networks. For rectifier side current control is used for inverter side both current and extinction angle control is implemented. The error signal is passed through a VSC+HVDC and Artificial Neural Networks controller, which produces the necessary firing angle order. The firing circuit uses this information to generate the equidistant pulses for the valves in the converter station. Here Artificial Neural Networks is designed for both rectifier and inverter control and its performance is compared with conventional PI controller.