發(fā)布時間：2021-04-16 04:12

　　約束單（多）目標(biāo)優(yōu)化問題廣泛存在于科學(xué)、經(jīng)濟(jì)和工程等諸多領(lǐng)域中,具有十分重要的理論意義和應(yīng)用價值。在過去幾十年中,如何有效求解約束單（多）目標(biāo)優(yōu)化問題日益受到相關(guān)學(xué)者的關(guān)注,并且提出了一系列的約束處理方法。由于約束本身的難度,約束間的相互作用,目標(biāo)函數(shù)的難度以及目標(biāo)函數(shù)與約束的相互作用,很難提出一種高效的約束處理方法。目前處理這些問題的算法主要可以為兩種,第一種是使用傳統(tǒng)的數(shù)學(xué)方法來求解約束問題,但通常需要給定問題的梯度信息,這對于有些實際問題,例如離散或目標(biāo)函數(shù)連續(xù)但不可微的約束優(yōu)化問題,是無法求解的。對于求解約束優(yōu)化問題,在不考慮梯度信息的情況下開發(fā)其他方法是非常重要的。進(jìn)化算法受自然的啟發(fā),結(jié)合約束處理方法,在求解約束問題時較于傳統(tǒng)數(shù)學(xué)方法有之獨特的優(yōu)勢。它是一個功能強(qiáng)大的優(yōu)化工具,不需要計算梯度信息,且易于實現(xiàn)。在過去的幾十年里,進(jìn)化算法引起了許多研究者的興趣,并提出了一些求解約束優(yōu)化問題的約束優(yōu)化進(jìn)化算法。然而,約束單（多）目標(biāo)優(yōu)化進(jìn)化算法的研究還沒有得到充分、廣泛的研究。本文對約束單（多）目標(biāo)優(yōu)化問題的算法設(shè)計及其相應(yīng)的實際應(yīng)用進(jìn)行了討論與研究。首先,針對約束單目標(biāo)約束優(yōu)化... 

【文章來源】：廣東工業(yè)大學(xué)廣東省

【文章頁數(shù)】：176 頁

【學(xué)位級別】：博士

【文章目錄】：
ACKNOWLEDGEMENTS
ABSTRACT
摘要
INDEX OF ABBREVIATIONS
CHAPTER 1 Introduction
    1.1 Background
    1.2 Basic Concepts
    1.3 Constrained Single-objective Optimization
    1.4 Constrained Multi-objective Optimization
    1.5 Motivation
    1.6 Contributions
    1.7 Outline of the Thesis
CHAPTER 2 A Novel Constraint-Handling Technique Based on Dynamic Weights forConstrained Optimization Problems
    2.1 Introduction
    2.2 The Proposed Algorithm
        2.2.1 The Proposed Constraint-handling Technique
        2.2.2 The Framework of the Proposed Algorithm
    2.3 Experimental Study
    2.4 Discussion
        2.4.1 The effectiveness of biased dynamic weights
        2.4.2 The Sensitivity of the Proposed Algorithm to a
        2.4.3 The Sensitivity of the Proposed Algorithm to a
    2.5 Conclusion
CHAPTER 3 A Constrained Multi-objective Evolutionary Algorithm Based on BoundarySearch and Archive
    3.1 Introduction
    3.2 The Proposed Algorithm
        3.2.1 Decomposition
        3.2.2 Constraint-handling Method Based on Boundary Search and Archive
        3.2.3 Crossover and Mutation
        3.2.4 The Framework of the CM2M
    3.3 Experiments
        3.3.1 Parameter Settings
        3.3.2 Performance Metrics
        3.3.3 Experimental Results
    3.4 Conclusion
CHAPTER 4 An Evolutionary Algorithm with Directed Weights for ConstrainedMulti-objective Optimization
    4.1 Introduction
    4.2 The Proposed Algorithm
        4.2.1 The Proposed Constraint-handling Technique
        4.2.2 The Proposed Framework
    4.3 Experimental Studies and Discussion
        4.3.1 Parameter Settings
        4.3.2 Experimental results on CFs and CTPs
        4.3.3 Experimental results on two engineering design problems
1 and N₂">        4.3.4 Sensitivity of the proposed algorithm to N₁ and N2

        4.3.5 Sensitivity of the proposed algorithm to λ
    4.4 Conclusion
CHAPTER 5 Handling Multi-objective Optimization Problems with Unbalanced Constraintsand their Effects on Evolutionary Algorithm Performance
    5.1 Introduction
    5.2 The Proposed Test suite
        5.2.1 The Proposed Test Problems
        5.2.2 Illustrative Problem
    5.3 Evolutionary Algorithms for Solving UCMOPs
        5.3.1 Two Related Constraint-handling Techniques
        5.3.2 Characteristics of the UCMOPs and Their Effects on EvolutionaryAlgorithms 615.4 Experimental Studies
    5.4 Experimental Studies
        5.4.1 Parameter Settings
        5.4.2 Performance Metrics
        5.4.3 Numerical Results on UCMOPs
        5.4.4 Results on UCMOPs with different degrees of imbalance
    5.5 Discussion
        5.5.1 Effect of control function k on DW
        5.5.2 Sensitivity of M2M-DW to
    5.6 Conclusion
CHAPTER 6 A Cooperative Evolutionary Framework Based on an Improved Version ofDirected Weights for Constrained Multi-objective Optimization with Deceptive Constraints
    6.1 Introduction
    6.2 The Proposed Test suite
    6.3 The Proposed Framework
        6.3.1 The Pseudo-code of the Proposed Framework
        6.3.2 The First Phase
        6.3.3 The Second Phase
        6.3.4 An Infeasibility Strategy
        6.3.5 The Switching Condition
    6.4 Experiments and Discussion
        6.4.1 Compared Algorithms and Parameter Settings
        6.4.2 Experimental Results on the Compared Algorithms
        6.4.3 A Comparison between Two Frameworks
        6.4.4 A Comparison between M2M-DW and IDW-M2M-CDP
        6.4.5 Investigation of the Tolerance Value r
        6.4.6 Investigation of the Parameter a
        6.4.7 Effectiveness of the Infeasibility Utilization Strategy
        6.4.8 Benefit of the Two Switching Phases
    6.5 Conclusion
Conclusion and Future Work
REFERENCE
List of Published/Submitted Papers
    Published papers
    Papers under review



本文編號：3140715

boshi