چكيده به لاتين
In this research, using intelligent optimal sliding-mode controller, speed control and torque ripple reduction have been performed for switch-reluctance motor in two different structures. For the linear model of this motor, a first-order sliding-mode controller is designed. Then, by defining the cost function as a derivation of the sliding surface, the sliding-mode control problem is transformed into an optimization problem in the search space of the control signal. By online solving the constrained optimization problem, the control signal has no chattering and satisfies the sliding condition as well. To solve the real-time optimization problem, projection recurrent neural network is used. Therefore, the optimization problem must be transformed as applicable to this network. Using such networks enables implementation of the control algorithm with minimal computational burdens. To reduce the torque ripple, two different structures have been proposed: using the inner current loop and preventing current changes in the speed controller loop. In the first method, the cascaded control loops reduce the torque ripple and the optimal sliding-mode controller is designed based on the system mechanical model, which provides the desired set-point for the inner loop. In the second method, using only one control loop, the speed control, and the torque ripple reduction are obtained. In this method, the fluctuations of the current that cause the torque ripple, are taken as a term in the sliding surface of the speed controller. If the sliding condition is satisfied, then the sliding surface will converge to its desired level. This desired set-point includes reducing the current changes and consequently reducing the torque ripple. In simulation and experiments, the proposed methods are compared with the existing methods in the literature. The results showed that the proposed method is effective in most experimental cases.
