Quantum Inspired GA based Neural Control of Inverted Pendulum

Abstract

This paper deals with comparison of artificial neural network and quantum inspired evolutionary neural network control of an inverted pendulum. First, a properly tuned PID controller was utilized to stabilize the inverted pendulum to generate the training data. Secondly, a feed-forward neural network was trained on the basis of these data. Thirdly, a quantum genetic algorithm optimized neural network was developed. If a disturbance occurs in the system, the controllers counteract this disturbance and balance inverted pendulum. All these three schemes are tested and compared. The results establish that the quantum genetic algorithm neural controller has the best control action.