一种自适应PSD伺服控制器的研究

本文针对常规PID伺服控制器设计时依赖精确数学模型同负载、模型参数的不确定性和非线性因素影响精确模型获得的矛盾，设计并分析了单神经元自适应PSD控制器。利用神经元的自学习、自组织能力，根据被控对象的参数变化情况进行权值的在线调整，加入了增益的自适应学习算法，使得调节更方便。同时，对PSD算法快速性方面的不足，采用复合算法进行了一定的改进。仿真表明，所设计的自适应PSD伺服控制器在一直线电机伺服系统中获得了满意的结果。

Research on an Adaptive PSD Servo-Controller

Considering the big influence of load disturbance, model parameter uncertainty and nonlinear factors,a single-neuron adaptive PSD controller is designed and studied. Based on the self-learning and self-organizing ability of the neuron, the weight of the plant is adjusted online.Furthermore, the algorithm of self-regulating the gain is given.Meanwhile,the speed of the PSD algorithm is improved. The simulation results show that the adaptive PSD servo-controller has a good performance.