An ANN based X-PC target controller for speed control of permanent magnet brushless DC motor

In this paper an artificial neural net (ANN) is suggested for implementation as a reference commutation signal generator for speed control of permanent magnet brushless DC motor in X-PC target: domain in overloaded condition. The ANN is trained in the continuous time and then it is transformed to discrete time. The reason for this transformation is that the X-PC target accepts only discrete models. The ANN is initially optimised by obtaining continuous time domain data by experimentation. Since the actual system is an embedded system hence the ANN is transformed to discrete model and then it is interfaced to the transfer function model of PMBLDC motor to analyse its performance. After optimisation the system is targeted for an X-PC target by discretising the model for a specified sampling rate based on the performance of the interface card. The system is tested in real time and is working as designed. The system is responding to the overload conditions in an almost linearised manner which is not the normal characteristics of the PMBLDC motor