直流伺服系统中单神经元PSD应用的改进

Improvement of the Application of Single-neuron PSD in the Direct Current Servo-control System

常规PID控制具有结构简单、稳定性好、可靠性高等优点,在调速系统中被广泛应用。但常规PID控制的设计需依靠数学模型,负载、模型参数的变化及非线性因素等影响常规PID的精确调节。单神经元PSD控制器利用神经元的自学习、自组织能力,根据被控对象的变化情况对控制器的权值进行在线调整,达到在线调整PID参数的目的,并且采用无需对象模型的控制算法构成了自适应控制。与常规的PID调节器相比,具有更好的鲁棒性。同时,通过对PSD控制器的改进,在直流伺服系统的仿真应用中得到了较理想的结果。

The conventional PID regulator has been widely applied in speed control system, with its simple structure, good stability and high reliability. But the control of a conventional one depends heavily on the precise adjustment of all the factors including the variation of the mathematical model, the load, the model parameter, as well as the nonlinear factors, which affect the conventional PID. The singleneuron PSD controller has achieved the goal of online adjustment of PID parameter through making online adjustment on the weight of the controller according to the variation of the controlled plant, based on the selflearning and selforganizing ability of the PSD controller. Further, by using the control algorithm independent of the objective model, the adaptive control has been established to be more robust, comparing with the conventional one. In the mean time, a perfect performance has been obtained in the simulation of the direct current servocontrol system through the improvement of the PSD controller.