Speed Control of Switched Reluctance Motor Using Artificial Neural Network Controller

Abstract

Abstract. Switched reluctance motors (SRMs) have an intrinsic simplicity and low cost that makes them well suited to many applications. However, the motor has doubly salient structure and highly non-uniform torque and magnetization characteristic. Since it was hard to determine the accurate mathematical model of (SRM) .The Artificial Neural Networks (ANNs) solve the problem of nonlinearity of SRM drive. It ensures fast, accurate, less overshoot and high precision dynamic response with perfect steady state performance. In the simulation analysis, this paper tests the (SRM) motor adopting two different control modes at starting process under full load torque with a reference speed of 2000 rpm, and the load disturbance under full load torque with a reference speed 2250 rpm. Simulation results show that speed control is better using (ANN) controller than using the (PID) controller. Matlab/Simulink tool is used for the dynamic simulation study.Keywords: Switched relcutance motor (SRM), proportional plus integrator plus diffrrential controler (PID), artificial neural network controller (ANN). IntroductionSwitched reluctance motors (SRMs) and drives have received growing interest recently. Their simple structure and low cost of manufacture make (SRMs) potentially attractive for various general purposes adjustable speed applications [1]. However, the motor has many drawbacks due to the doubly salient structure and the extremely comprehensive mathematical modeling which is complicated due to its highly nonlinear torque output and magnetization characteristic [2][3]. This due to the fact that (SRMs) primarily operate in magnetic saturation and hence the developed torque is a non-linear function of rotor position and stator current [4].The non-linearity of (SRM) uses the general feedback theory or the conventional PID controller, but it is impossible to obtain good control performance. It required applying the advanced control strategy to the SR drive to improve the system performance.

Energy Saving and Economic Analysis of Switched Reluctance Motor in Agricultural Applications

“…Different methods of controllers are implemented in different power system problems including the optimization algorithms [15][16][17][18]. In the implementation stage, many speed control techniques of drive systems such as PI control, sliding mode control and artificial neural network control applied in SRM are suggested [19][20][21]. However, PI based control method is commonly adopted due to the advantages such as the easiness of the implementation of control strategy, precise control in the transient and steady state period of operation etc.…”

Artificial Neural Network-Based Control of Switched Reluctance Motor for Torque Ripple Reduction

“…SRM nonlinearity characteristics are trained by neural networks and then the current graph for ripple reduction is obtained. ANN is used as an intelligent controller [18,19]. Torque ripple reduction is done through PI and fuzzy logic controller [20].…”

Degradation assessment and trend prediction of rolling bearing based on MD-ESN

“…In previous studies the authors have shown that SOM (Self-organizing Maps), GMM (Gaussian Mixture Model), etc., can identify the different working conditions of rolling bearings [3,4]. Meanwhile, traditional ANN (Artificial Neutral Network), BP (Back Propagation) and Elman models have been widely used in the field of intelligent prediction of rotating machinery [5][6][7]. But some of the methods mentioned above are complex and time consuming, such as GMM, meanwhile, some are too simple to get approving results, like BP, etc.…”