Search

Switched reluctance (SR) motors have a wide range of applications in industries, mainly due to the special properties of this motor. However, because of its dynamical nonlinearities, its control is complex. This paper presents an adaptive intelligent control based on a digital signal processor (TMS320F240) allowing great flexibility for various real time applications to control the speed and the current of SR motors. The Experimental results show good performance of the proposed controller.

The controller, based on differential geometry theory, is designed by transforming the nonlinear system to an equivalent linear system. An analysis of a single-machine infinite-bus power system indicates that the margins of dynamic stability are larger than the margins of conventional linear controllers.

A modeling method based on neuron and RBF neural network for time-varying delay system is presented. Neuron is applied to identify the delay of system; RBF neural network is used to identify other time-varying parameters, for delay has been separated from other parameters, so the structure of RBF neural network is simple, and the training of RBF neural network is quickly. These two neural methods is serial, and training with the same cost function. Because the training time is short, this method can be used in on-line identification of nonlinear delay system

A Smith predictor (SP) neuron based controller for large time-delay system is proposed in this paper. In this combined method, the neuron is used to self-tune the PID controller so the optimal parameters can be obtained, and the SP is going to compensate the large time-delay that is generally associate with different process. The scheme has been simulated on a plant model. Simulation results demonstrate that the controller has enhanced the response speed and robustness