Optimized torque control of switched reluctance motor at all operational regimes using neural network

Torque in a switched reluctance motor (SRM) at low speed is controlled by PWM chopping of current. The controller provides the appropriate reference current, the phase turn-on, and the turn-off angles based on the torque demand and the motor speed. While the torque at high speed is controlled by the single pulse operation of the current. The controllable parameters at high speeds are the phase turn-os and the turn-off angles. The controller is responsible for providing the appropriate turn-on and the turn-off angles. The complexity in the control of SRM arises from the fact that the SRM operation is highly nonlinear. By design SRM operates in the saturation region in almost all operational points. This results in the non-linearity oft he SRM magnetic field. To find the appropriate control parameters at low as well as at high speeds, an accurate non-linear model of the SRM is, therefore, needed. In this paper a dynamic model of SRM is developed. In order to include the effect of the magnetic non-linearity, static torque and flux linkage data are used in the dynamic model. The static data are generated experimentally. The dynamic model, however, is too detailed and lengthy to be implemented in real time. The optimal control parameters (reference current, phase turn-on, and turn-off angles at low speed and phase turn-on and turn-on angles at high speed), which maximizes torque per ampere, are generated from the dynamic model, off-line, after a complete search. To recreate these control parameters, on-line, artificial neural networks (ANNs) are used. Two separate networks are trained. One is trained with the low speed control parameters for torque control at low speed, while the other is trained with the high speed control parameters for torque control at high speed. The speed at which SRM makes a transition from PWM control to single pulse operation (i.e. low speed to high speed operation), commonly referred to as base speed, is torque (current) dependent. A small table is maintained in the controller to identify the base speed for different torque demands. When the motor exceeds the base speed for a certain torque demand, the controller switches from the low speed neural network to the high speed neural network and vice versa. It is also shown that SRM is capable of producing an extended constant horse power operation with this optimal control. The power factor (the energy ratio) is shown to improve in this extended speed constant horse power range. Simulation and experimental results are presented to demonstrate the effectiveness of the proposed control scheme.