异步电动机的单神经元模糊自适应控制

针对异步电动机矢量控制中PI控制器自适应能力较弱的问题,提出了用单神经元模糊自适应控制器代替转速PI调节器和磁链PI调节器的方法,并基于梯度下降法和单神经元自适应模糊控制原理,将增量式PI控制算法和单神经元自适应学习规则相结合,通过模糊控制修正单神经元的比例系数,提高单神经元模糊自适应控制器参数在线调整能力。为验证单神经元模糊自适应控制器的性能,对系统进行仿真实验。仿真结果表明,采用单神经元模糊自适应控制策略,其控制效果优于传统矢量控制,系统的动态和静态性能比较好,而且控制器对电动机转子电阻变化有较好的鲁棒性。该控制方法具有较好的控制性能,使异步电动机控制系统具有较强的自适应能力。

NeuronAdaptive Controller of Asynchronous Motor Drive Based on Fuzzy Method

Aiming at the weak adaptive ability for the PI controller in the vector control system of asyn- chronous motor, neuron adaptive fuzzy controller is presented instead of the speed regulator and flux regu- lator. In order to improve the on--line adjusting capability of the neuron fuzzy adaptive controller, based on the gradient descent and the neuron fuzzy adaptive principle, combining the incremental PI control algo- rithm with neuron adaptive learning rules, the fuzzy control is used to adjust the proportional coefficient of neuron. To test the performance of the neuron adaptive fuzzy controller, the system using proposed meth- od is simulated. The simulation results show that neuron adaptive fuzzy control strategy leads to better control performance than conventional vector control, which makes the system has good dynamic and static performance. Moreover, the proposed controller has good robustness performance to the change of rotor resistance. The method has a good control performance and the resultant control system of asynchronous motor has a better adaptive ability.