基于神经元补偿的直线伺服系统全程滑模控制

针对直接驱动的交流永磁直线伺服系统，提出一种基于神经元补偿控制的全程滑模变结构控制策略，该控制策略是使控制系统从开始就处于所设计的滑动模态上并将其保持，解决了以往变结构控制中能达阶段的鲁棒性问题，通过神经元控制的补偿作用，削弱了滑模控制引起的抖振，仿真实验结果表明，该策略对系统参数变化和负载扰动等不确定因素具有很强的鲁棒性，而且提高了直线伺服系统的伺服精度。

Global sliding mode control for linear servo system based on neuron compensation

A global sliding model variable structure control scheme based on artificial neuron compensation for a direct drive AC permanent magnet linear servo system are presented in this paper. The so called global sliding model variable structure control is that the control system states are on the sliding mode from beginning and is holded on it. The robustness of the reaching intervals of the variable structure control system is sovled. The chatting caused by the sliding mode control is eliminated by artificial neuron compensation. Some simulation results show that this control scheme not only has strong robustness to parameter variations and resistance perturbation, but also improves the servo precision of direct drive linear servo system greatly.