Doubly-Fed Generation System Based on Neural Network Inverse Control

Neural network inverse control is applied to doubly-fed generation system, and the mathematical model of inverse system is derived from the power control model of the doubly-fed induction generator. Through the proper selection of input and output signals of inverse control system and the use of neural network inverse control algorithm, the system is decomposed into two single-variable linear subsystems of active power and reactive power. With the comprehensive approach of linear system, the two closed-loop subsystems are designed separately which consist of PI controllers. Finally the simulation model is built and run. Simulation results show that doubly-fed generation system with neural network inverse control has good performance, for it can not only conveniently control active power but also provide reactive power for power grid independently.