模型与实际有差异时非线性大工业过程递阶优化控制算法

针对实际非线性离散系统与它的模型之间的差异，提出了一种基于模型优化控制问题来求解实际问题最优解的递阶算法，该算法通过上级的关联预测和参数估计与下级的修正的基于模型优化子问题的迭代问题，总可以获得实际非线性离散大系统最优控制．并行计算可以节省计算时间．分析了该算法的收敛性和最优性．仿真例子说明该算法的特色．

Hierarchical Algorithm of Optimal Control for Industrial Dynamic Processes with Model-Reality Differences

According to the differences between model and reality, a hierarchical algorithm of optimal control for nonlinear dynamic systems is presented.The algorithm can give the real optimal solution in spite of the differences of model reality by alternately solving the problem of the interaction predicting and parameters estimation at upper level as well as modified model based subproblems at lower level. The algorithm can greatly save time when each subproblems are solved simultaneously. The analysis of convergence and optimality of the algorithm are also given. Simulation example denotes the characteristic of the algorithm.