神经网络滑模控制在位置伺服系统的应用

针对直流无刷电机伺服系统参数摄动、非线性以及其他不确定因素等问题，提出一种神经网络滑模变结构控制方案。利用一个三层RBF神经网络和滑模变结构控制器实现了无刷电机伺服系统位置的精确控制。该方案综合径向基神经网络和滑模变结构算法的优点，对系统参数变化和外部扰动具有很强的鲁棒性，是一种较理想的智能控制策略。仿真结果表明控制器具有良好的动静态品质，工程实现简单，具有广阔的应用前景。

Application of Sliding Mode Control Based on Neliral Network to Servo System

A neural network sliding mode control scheme for servo system has been proposed, which can overcome the parameter perturbation, non-linearity and uncertainty factors etc. Using a compound controller of a three-layer neural network and sliding mode to realize accurate tracking. This scheme synthesize the advantage of RBF neural network and sliding mode control, and it is a kind of more ideal dynamic and static conditi it has strong adaptability and robustness to load disturbance, parameters change of the system, so intelligence control strategy. The simulation results show that this controller have good quality in ons, and can easily be realized in engineering; therefore it has a vast applying prospect.