PFC中约束优先级及可行性问题的混杂处理

A Hybrid Predictive Functional Control for Handling Infeasibility and Constraint Prioritization

针对预测函数控制的约束不可行和优先级问题,基于混杂系统以及混合逻辑动态模型的思想提出了约束不可行及优先级问题的混杂PFC处理方法,讨论了由PFC、混杂系统、松弛变量法所组成的约束不可行及优先级问题的具体实现方法与步骤。在通常PFC跟踪性能指标的基础上结合MLD与松弛变量法,通过将命题逻辑转化为混合逻辑不等式并与系统自身的约束相结合,将问题的可行性与优先级特性引入PFC控制器的求解当中,实现了在约束存在且干扰发生阶跃变化的情况下保证系统输出达到期望设定值的目标。所提出的方法成功应用于典型工业多入多出过程Wood-Berry塔,实现了多变量系统的混杂PFC控制,取得了明显的应用成效,改善了系统的性能指标,获得了较好的跟踪性能。

For constraint infeasibility and prioritization in PFC, hybrid PFC method which based on hybrid system and MLD （ MixedLogical Dynamical） mode is proposed, the implementation and steps which are consist of PFC, hybrid system and relaxation method are discussed. Infeasibility and prioritization features of problem are introduced into the solving of PFC controller, which through the approach that MLD and relaxation are integrated into usual PFC tracking performance, and by the approach that propositional logics are transformed into mixed logic inequalities and are combined with constraints of system, finally the goal that the output of system reach the desired steady states under the circumstances that the constraints are presence and disturbance has a step change. This method is successfully applied into the typical industrial MIMO process Wood-Berry column, hybrid PFC method is implemented for muhivariable system, the better benefits have been acquired, and the performance is improved, ultimately the better tracking performance is obtained.