Lecture 3 Chapter 3 Multiobjective Programming 相关管理科学英文版.pptVIP

Lecture Lecture Management Science:Operations Research(OR) Chapter 3 Multiobjective Programming * * Lecture Multiobjective Programming Multiple objectives: house, size vs costs, car: economy vs comfort, economic policy: inflation vs unemployment: tradeoffs between objectives. MCDM: multicriteria decision making problems, MADM: multi-attribute decision making MOLP: multiobjective (linear) programming. * Lecture * Multiobjective (linear) Programming Here, MOLP. Observation: the concept of optimality collapses. Optimal w.r.t. what??? MOLP has two main classes: Vector optimization goal programming. * Lecture * Vector Optimization Max z1 = 3x1 + x2 Max z2 = ?2x1 + x2 s.t. ?x1 + x2 ≤ 3 (I) x2 ≤ 4 (II) x1 + x2 ≤ 6 (III) x1 ≤ 5 (IV) x1, x2 ≥ 0 * Lecture * * Lecture * Gradient z iso-profit line through x with “+++” indicating improving objective values “---” showing deteriorating objective values Generating compromises by introducing new issues * Lecture * * Lecture * Unions Management $ Management · · C+ + Working conditions Unions Wage Efficient Frontier A point is nondominated, noninferior, efficient, or pareto-optimal, if we cannot move into its improvement cone without leaving the feasible set. Collection of all nondominated points = efficient frontier. Note: interior points can never be efficient, as long as all objectives are linear. * Lecture * Efficient Frontier: A-B-C-D (Note: The nondominated set is connected. ) O and E is not nondominated * Lecture * Lecture Lecture * *