时变大滞后过程的单神经元内模控制

由于常规smith预估器对被控对象参数和结构的变化十分敏感,这对时变大滞后过程的控制极为不利,为此提出了一种新的单神经元内模控制方案,将改进的智能单神经元PID作为Smith预估的主控制器,并证明了该控制系统的内模结构特性。同时,对单神经元自适应PID的学习规则进行了改进,并采用仿人智能思想对神经元的比例系数进行在线自调整。仿真结果表明,这种控制方法具有超调小、抗干扰能力好和鲁棒性强的优点,对时变大滞后过程是行之有效的。

Single Neuron Internal Model Control of Time-various Process with Large Time Delay

Since the traditional Smith compensator is sensitive to the changes of the parameters and structure of the plant,which arc undesirable to time-various processes with large time delay, a novel single neuron internal model controller is proposed. The improved intelligent neuron PID is used as the main controller while Smith estimator is used as compensator.The learning rule of single neuron adaptive PID is improved and the proportional coefficient can be self-tuned online according to human-simulated intelligent technique.The simulation results show that the proposed scheme has the advantages of small overshoot, good anti-disturbance and strong robustness and is effective in controlling the time-various processes with large time delays.