Title: Research on inverse model based on ANN and analytic method for induction motor

Authors: Shuo Ding; Qinghui Wu

Addresses: College of Engineering, Bohai University, Jinzhou 121013, China ' College of Engineering, Bohai University, Jinzhou 121013, China

Abstract: The robustness of back propagation neural network-based inverse model is researched by simulation in the paper. Firstly, the dynamic model of induction motor is built by state space theory. Secondly, the corresponding inverse model is got by inverse system theory. However, the analytic inverse model is hardly used in the engineering field by excessively depending on the parameters. Finally, an artificial neural network (ANN)-based inverse model, which synthesises artificial intelligent method and analytic method, is suggested. To accelerate the convergence of ANN and enhance its generalisation ability, the non-linear parts are realised by the analytic expressions and the corresponding results act as the inputs of ANN so that the complex-non-linear mapping relation become a simple-linear mapping and the structure of ANN is simplified. A three-layered feed-forward ANN with 12-10-2 structure is introduced to approach the inverse mode of induction motor drives. Finally, the robustness of ANN-based inverse model is verified by simulation in the case of parameter variance and state disturbance.

Keywords: ANNs; artificial neural networks; inverse models; decoupling control; induction motor drives; simulation; dynamic modelling.

DOI: 10.1504/IJAAC.2011.043617

International Journal of Automation and Control, 2011 Vol.5 No.4, pp.356 - 370

Received: 22 Feb 2011
Accepted: 13 Aug 2011

Published online: 17 Apr 2015 *

Full-text access for editors Full-text access for subscribers Purchase this article Comment on this article