基于神经元和直接转矩控制技术的高性能交流调速系统研究

针对传统的异步电动机直接转矩控制系统中,电机在低速时存在较大的转矩脉动问题,提出了一种新的解决方法.在传统直接转矩控制系统的基础上,应用神经网络自适应控制能力,采用神经元控制器取代传统PlD(proportion integration differentiation)控制器,并利用Matlab/simulink软件进行仿真.仿真实验结果表明该方法能够有效地解决直接转矩控制系统转矩脉动问题,在低速时系统依然具有较好的动静态性能.

Research of high-performance system for AC drive based on a single neuron controller and DTC technology

The large torque ripples exist in induction motors based on conventional direct torque control system (DTCS), especially at a low speed. On the basis of conventional DTCS, this paper provides a new control strategy that can effectively reduce the torque ripples. The adaptive ability of neuron network was used in this strategy. The proportion integration differentiation (PID) controller in the conventional DTCS was replaced by a neuron controller, and the proposed strategy was simulated by Matlab/simulink. The experimental results illustrate that the torque response is improved, and the torque ripples are effectively decreased. The system has a better dynamic and steady performance in low-speed range.