The Research Of Flux Estimation In AC Motor Control

With extensive industrial use of the induction motor, more and more technical requirements for speed control. So higher precision, faster response is the goal of control system. For this reason there had been some methods such as vector control and direct torque control, which two methods performance depends largely on the accuracy of flux estimation.Firstly, in this paper, the classical methods of flux estimation has been analyzed based on vector control and direct torque control. Especially, the flux estimation of direct torque control is analyzed deeply .In order to enhance the performance of AC Drive control, the flux estimation is between current model in low-speed and the voltage model in high-speed .Mixed magnetic chain model resolves the shortcomings of current model and voltage model, and this model can been switched in full range of speed. But model switching has disadvantages of speed of transformation and large torque ripple. Secondly, it analyses the flux estimation based on PI controller, which improves the accuracy of flux estimation and improve the speed performance, but PI controller has a shortcomings of fixed parameters. Finally, this paper presents an improved model of the magnetic chain, using a single neural network PID controller to adjust the stator flux, and set up the corresponding flux estimation model. In this method single neuron network can be self-organization, self-learning, and optimize the PID parameters to overcome the shortcomings of fixed parameters in the traditional PID controller, so improve the accuracy of flux estimation.The results of simulation shows that this method improves the accuracy of flux estimation and testify the feasibility of this method. Finally, this paper studied speed control system of hardware and software design based on DSP.