计算机工程与应用 ›› 2011, Vol. 47 ›› Issue (36): 238-241.

• 工程与应用 • 上一篇    下一篇

模糊铁合金配料机会约束模型

袁国强1,2,王小鹏3,姚秋波3   

  1. 1.河北金融学院 基础部,河北 保定 071051
    2.河北省科技金融重点实验室,河北 保定 071051
    3.河北金融学院 管理系,河北 保定 071051
  • 收稿日期:1900-01-01 修回日期:1900-01-01 出版日期:2011-12-21 发布日期:2011-12-21

Fuzzy ferroalloy burdening chance-constrained model

YUAN Guoqiang1,2,WANG Xiaopeng3,YAO Qiubo3   

  1. 1.Fundamental Department,Hebei Finance University,Baoding,Hebei 071051,China
    2.Science and Technology Financial Key Laboratory of Hebei Province,Baoding,Hebei 071051,China
    3.Department of Management,Hebei Finance University,Baoding,Hebei 071051,China
  • Received:1900-01-01 Revised:1900-01-01 Online:2011-12-21 Published:2011-12-21

摘要: 在铁合金配料问题的大量传统研究工作中,多数研究工作均与确定性系统相关。在不确定系统中考虑一类新的带有可信性约束的模糊铁合金配料机会约束模型。由于提出的模糊铁合金配料问题常常包含带有无限支撑的模糊变量参数,因此它是一个很少被直接求解的无穷维优化问题。为了求解这个模糊优化问题,通过逼近方法将模糊铁合金配料机会约束问题转化为一个有限维优化问题。设计一个含有逼近方法、神经网络和遗传算法的混合智能算法求解提出的带有可信性约束的铁合金配料机会约束问题。给出一个数值例子来表明所设计模型和算法的实用性。

关键词: 铁合金配料问题, 可信性约束, 机会约束, 逼近方法, 神经网络, 遗传算法

Abstract: A great deal of conventional research has been done on ferroalloy burdening problem,most of which concerns deterministic systems.This paper considers a new class of fuzzy ferroalloy burdening chance-constrained model with credibility constraint in uncertainty systems.Since the proposed fuzzy ferroalloy burdening problem often includes fuzzy variable parameters with infinite supports,it is infinite-dimensional optimization problem that can rarely be solved directly.In order to solve this fuzzy programming problem,it transforms fuzzy ferroalloy burdening chance-constrained problem into a finite-dimensional optimization problem by approximation approach.It designs a hybrid intelligent algorithm,which combines approximation approach,neural network and genetic algorithm to solve the proposed ferroalloy burdening chance-constrained problem with credibility constraint.A numerical example is given to show the practicality of the designed model and algorithm.

Key words: ferroalloy burdening problem, credibility constraint, chance-constrained, approximation approach, neural network, genetic algorithm