改进的预测函数控制算法及其应用研究

针对典型的一阶惯性纯滞后2入2出多变量系统，提出了基于Smith 预估补偿思想的分散解耦、集中控制的预测函数控制设计方法：将系统分解成两个具有可测扰动的单入单出子系统，当扰动通道的纯滞后时间大于控制通道的纯滞后时间时，控制器输出可根据系统输入输出的已知信息求出；当2个子系统的扰动通道纯滞后时间均小于控制通道的滞后时间，可以通过求解二元一次方程组求出，从而实现多变量预测函数控制。通过对一典型工业过程的一阶加纯滞后等价描述对象的仿真研究,证明这种算法不仅可以实现静态完全解耦和无静差跟踪，而且能够获得更好的动态解耦品质；同时在各种模型失配情况下有较强的鲁棒性和抗干扰能力。

Improved Predictive Functional Control Algorithm and Its Application

Based on the ideal of Smith predictor, a decentralized decoupling and centralized control design method of predictive functional control(PFC) is proposed for two-input two-output(TITO) first-order plus dead-time process. It decomposes the TITO system into two single-input single-output(SISO) subsystems with measurable disturbance. If the dead-time of disturbance channel is longer than the control channel, the control value can be calculated by the known system information. Otherwise, the PFC control value can be obtained by solving first-order equation group. Simulation results to an industrial process that can be equivalently described by first-order plus dead-time model show that the control algorithm has favorable performances: completely static decoupling, effective dynamic decoupling and zero static error track. Moreover, the control algorithm provides strong robust ability and disturbances rejection under many mismatches between the controlled plant and the predictive model.