广义预测控制算法改进及其性能仿真研究

Simulation Study on the Performance andImprovement of Generalized Predictive ControlAlgorithm

基于“预测、滚动优化、反馈校正”思想的广义预测控制算法,现己成为一种重要的先进控制策略,并被广泛应用于复杂工业过程控制中.广义预测控制算法可分为显式算法和隐式算法两种,在隐式算法的基础上,为了抑制输出的剧烈振荡,提出了在广义预测控制目标函数中加入输出增量项,为了消弱测量误差、干扰及饱和等因素的影响,采用有平滑滤波作用的输入加权控制增量,并推导了改进算法的控制律.仿真结果表明,改进后的广义预测控制算法无论在跟踪性能、控制精度及鲁棒性上,均优于常规广义预测控制.

The Generalized Predictive Control （GPC） , which based on the thought of ＂prediction, receding horizon optimization, feedbackcorrection＂, now has become an important advanced control strategy, and has been widely used in complex industrial processcontrol. Generalized predictive control algorithm can be divided into explicit algorithm and implicit algorithm. Based on theimplicit algorithm, the method of the output increase item joined in the objective function of GPC is proposed to suppress theacutely vibration of the output, and in order to weaken the measurement error, interference and saturation and other factors, theweighted input control increment with smoothing filter function is adopted, the improvement algorithm control law is given, andthe effect of the improvement algorithm control has been confirmed. Simulation results show that the improved algorithm is betterthan the GPC in track capacity and control precision, and it also puts a check on the disturbance in some sense.