Study on a Differential Evolution Type Neural Network for Intelligent Control of Ultrasonic Motor on a Meal-assistance Robot

Abstract

In this paper, a scheme of intelligent control method using differential evolution (DE) type neural network (NN) for Ultrasonic Motor (USM), which is expected to be applied on a meal-assistance robot. Owing the application of USM in the robot designed with orthogonal structure, excellent control performance in stability and safety with work quietness is expected. To obtain the high accurate control an optimization algorithm of DE was introduced to the NN type PID control for USM. The proposed intelligent method was investigated by the experiments on an existent USM servo system in this paper. The response of the proposed method with converngence of the proposed model was confirmed. With the proposed method, the servo of USM can be expected to use on meal-assistance robots and life support robots in medical and welfare environment.