基于约束最小方差基准的模型预测控制性能评价方法

考虑到实际工业过程中存在各种约束,基于模型预测最小方差控制器的设计原理,提出了一种带约束模型预测控制(MPC)性能评价方法。最小方差控制器的最优预测输出采用控制增量进行预测,目标函数采用最优预测输出和控制增量加权的二次型形式,通过求解二次规划(QP)问题获取最优控制律。该方法不仅考虑了控制输入和输出约束,而且考虑了控制增量约束,因此能够更真实地反映模型预测控制系统的性能。在Wood-Berry二元精馏塔上的仿真研究验证了该方法的有效性。

Performance Assessment of Model Predictive Control based on the Constrained Minimum Variance Benchmark

Considering various constraints in the practical industry process,the performance assessment of MPC with constraints was presented based on the design principle of the model-based predictive minimum variance controller.The control increments are used to predict the optimal predictive outputs of the minimum variance controller.The objective optimization function employs quadratic form of optimal predicted outputs and control increments weighted,and then it obtains the optimal control law through solving the quadratic programming(QP) problem.This approach not only takes the constraints of control inputs and outputs into consideration but also covers the factor of the constraints of the control increments,which made it reflect the performance of the model predictive control system more authentically.The simulation example on the Wood-Berry binary distillation column illustrated the validity of the method.