Indirect torque control of switched reluctance motors using iterative learning control

This paper proposes the use of iterative learning control (ILC) in designing a torque controller for switched reluctance motors (SRMs). The demanded motor torque is first distributed among the phases using a torque-sharing function. Following that, the phase torque references are converted to phase current references by a torque-to-current converter and the inner current control loop tracks the phase current references. SRM torque is a highly nonlinear and coupled function of rotor position and phase current. Hence, the phase current references for a given demanded torque can not be obtained analytically. Assumption of linear magnetization characteristics results in an invertible torque function. However, the nominal phase current references obtained using this torque function will lead to some torque error as motor enters into magnetic saturation. For a constant demanded torque, the error in the phase current references will be periodic with rotor position. Hence, we propose to use ILC to add a compensation current to the nominal phase current references so that torque error is eliminated. Similarly, current tracking for the nonlinear and time-varying system is achieved by combining a simple P-type feedback controller with an ILC controller. The proposed scheme uses ILC to augment conventional feedback techniques and hence, has better dynamic performance than a scheme using only ILC. Experimental results of the proposed scheme for an 8/6 pole, 1-hp SRM show very good average as well as instantaneous torque control.