一种基于知识的多目标优化方法及其应用  

Knowledge Based Multiobjective Optimization Approach and Its Application

作　　者：王劭伯[1] 吕勇哉[1] 

机构地区：[1]浙江大学

出　　处：《控制与决策》1989年第1期1-5,共5页Control and Decision

摘　　要：本文以非线性规划逐步线性化法KBSLP为基础,进一步开发了基于知识的非线性多目标优化方法KBMOP.KBMOP法是在凸性假设下,用知识库辅助权系数的选择,将多目标问题化为单目标问题,再用高效率的KBSLP法求解,通过知识库实现对满意解的判定。本文最后给出一个计算实例,说明这种方法的应用.This paper presents a development of knowledge based approach to multiobjective nonlinear optimization (KBMQP) based on knowledge based successive linearization method in nonlinear programming (KBSLP) proposed, by the authors earlier. Under the assumption of system convex, the prohem proposed KBMOP may convert a multiobjective problem into a singleobjective through the knowledge based aided selection of the relevant weighting coefficients, such that the effective KBSLP can further be used to solve the problem under study. The satisfactory solution could then be judged via the knowledge base. An illustrative numerical example is given to describe the application of the proposed method.

关 键 词：KBSLP KBMOP 知识库 运筹学 

分 类 号：TP11[自动化与计算机技术—控制理论与控制工程]