基于混合粒子群算法的线性预测PID控制

针对传统PID控制对大时滞过程控制效果不佳的问题,提出了一种基于线性预测的PID控制算法,用系统过去的几个输出值预测系统未来的输出值,用此预测值与期望设定值所得的偏差作为PID控制输入,再依PID控制律来设定控制器的输出,从而使被延迟了的被控量超前反映到控制器,使控制器提前动作,实现事先调节,从而减少超调量和调节时间,消除时滞对系统稳定性影响。利用混合粒子群算法对模型系数和系统参数进行了优化和在线调整。实例仿真结果表明,将线性预测与传统PID控制相结合并经混合粒子群算法优化的控制方法远好于单一的PID控制方法,而且具有操作简单、容易程序实现等特点,有效地提高了系统的控制品质。

Linear Prediction PID Control Based on Hybrid Particle Swarm Algorithm

In view of poor control of traditional PID over large time lag process,this paper presents a PID control algorithm based on linear prediction,whereby future output values of the system are predicted by means of its past output values,the deviation between the predicted value and the expected set value is used as PID input,and the controller output is set according to the PID control law.In this way,the delayed control quantity is reflected to the controller in advance so that the controller may be actuated in advance to realize prior adjustment,thus reducing overshooting and adjustment time and eliminating influence of time lag on the system stability. Optimization and online adjustment of the model coefficients and system parameters are implemented by using the hybrid particle swarm algorithm.The case simulation results show that the combination of linear prediction and traditional PID and optimization through hybrid particle swarm algorithm is far better than PID alone,and this control algorithm,characterized by simple operation and easy programming,effectively improves the control quality of the system.