Design and implementation of an Inverse Neural Network Controller applied To VSC Converter for active and reactive Power Flow, based on regions of work

Abstract

Voltage Source Converter (VSC) usually used in High Voltage Direct Current (HVDC) systems, where a VSC can be used as inverter or rectifier. VSC systems allow the independent control of active or reactive power flow using different techniques. VSC systems present nonlinear behaviors, multiple inputs and multiple outputs, therefore nonlinear controllers can be used to obtain an adequate behavior. Inverse Neural Control is an alternative of an intelligent control since a mathematical model of the system is not required for designing controllers. Additionally, Inverse Neural Control can easily manage uncertainties and nonlinear behaviors typically presented in VSC systems. In this paper are presented the design, simulation and implementation of an Inverse Neural Control applied to the control of active and reactive power flow in a VSC system. Initially, is presented the simulation of the controller, where is evaluated the behavior of the system using a MIMO controller for the control of two parameters in the same time. Subsequently, the implementation of the controller is done and the e obtained is presented. Finally, a modular Inverse Neural Network Control is proposed to overcome the drawbacks presented in the behavior of the system when it was controlled in real implementation.