基于差分型思维进化算法的受限广义预测控制

摘要

针对广义预测控制(GPC)的滚动优化对受限控制量求解的复杂性,笔者提出一种基于差分型思维进化算法的受限广义预测控制方法(DMEAGPC).用差分型思维进化算法处理带约束的非线性优化问题,以此作为滚动优化策略,求得最优控制律.并将该算法应用于电厂再热汽温系统,仿真结果验证了该算法的有效性和可靠性.%This paper presented a Generalized Predictive Controller based on the Differential Mind Evolutionary Algorithm (DMEAGPC). The Differential Mind Evolutionary Algorithm (DMEA) realized rolling optimization of GPC and the optimal △U was searched in the constrained ranges. DMEAGPC improved the real-time and stability of the system and lowered calculation complexities. When DMEAGPC was used to control boiler reheat steam temperature system,the simulation results showed its effectiveness and feasibility.