神经网络鲁棒控制器的设计及应用研究

Design and Application of Neural Network and Robust Controller

针对变频调速系统中交流电机模型的不确定性和非线性等特点,提出一种神经网络鲁棒控制方法。这种控制方法采用已知模型与神经网络在线逼近非线性相结合来共同辩识系统的实际模型,同时通过引入鲁棒因子将神经网络控制与传统控制方法有机结合起来。仿真研究表明它能明显提高系统性能,具有实际应用价值。

A Controller of neural network and robust is presented considering uncertainty and non-linear of ac induction motor model. This control method adopts neural network and presented model to identify system real model. At the same time system combine tradition control method with neural network control method by leading into robust factor. Simulation proves that the controller can improve performance of system. It has a high value of practical application.