非凸稳态大系统的递阶优化

Hierarchical Optimization Algorithm for Nonconvex Steady State Large Scale System

本文讨论了非凸稳态大系统最优控制的OPBM递阶算法,在此算法的基础上,讨论了通过加入位移因子,对实际系统施加全局反馈或局部反馈,从而克服了模型与实际系统有差异的影响,由于OPBM法应用了加入罚系数项的增广拉格朗日函数,对问题进行凸化,所以它可应用于一类非凸的问题。

This paper presents the coordination algorithm OPBM for steady state systems. On the Basis of this method, the global or local feedback through using a shift vector is introduced. Hence the effect of the error between the model and the real system is reduced. The OPBM method is based on the augmented Lagrangian with penalty function causing the problem to be convex,so as to make it applicable for nonoonvex problems.