Sensorless control of single switch-based switched reluctance motor drive using neural network

Neural networks (NNs) have proven to be useful in approximating nonlinear systems and in many applications, including motion control. Hitherto, NNs advocated in switched reluctance motor (SRM) control have a large number of neurons in the hidden layer. This has impeded their real-time implementation with DSPs, particularly at high rotational speeds, because of the large number of operations required by the NN controller within a sampling interval. One of the ideal applications of NNs in SRM control is in rotor position estimation using only current and/or voltage signals. Elimination of rotor position sensors is practically mandatory for high-volume, high-speed, and low-cost applications of SRMs, for example, in home appliances such as in vacuum cleaners. In this paper, through simulation and analysis, it is demonstrated that a minimal NN configuration is attainable to implement rotor position estimation in SRM drives. The NN is trained and implemented on an inexpensive DSP microcontroller. NN training data, current, and flux linkage are obtained directly from the system during its operation. Furthermore, the chosen method is implemented on a single-switch-converter-driven SRM with two phases. This configuration of the motor drive is chosen because it is believed that this is the lowest cost variable speed machine system available. Experimental verification of this motor drive system is provided to demonstrate the viability of the proposed approach for the development of low-cost motor drives.