Speed Identification of Ultrasonic Motors Based on Evolutionary Elman Network

A learning algorithm for dynamic recurrent Elman neural networks is proposed based on an improved adaptive genetic algorithm. The proposed algorithm performs the evolution of network structure, weights, initial inputs of the context units and self-feedback coefficient of the modified Elman network together. Two dynamic identification algorithms for nonlinear systems are constructed successively based on the proposed algorithm to perform the speed identification for ultrasonic motors. Numerical results show that the proposed algorithms not only realize the fully automatic optimization design for the dynamic recursive neural network, but also improve the precision of convergence for model identification.