含突變機制的粒子群演算法於多目標工程最佳化

A Particle Swarm Algorithm with Mutation for Multi-objective Engineering Optimization

陳志忠(Chih-Chung Chen)

指導教授 ： 史建中

淡江大學/工學院/機械與機電工程學系碩士班/碩士(2011年)

https://doi.org/10.6846/TKU.2011.00991

摘要

本文探討與發展含改良突變機制的粒子群最佳化演算法，除了能避免最終解落入區域極值的可能。本文又探討三種處理限制條件的方法，第一種是改良飛回策略的。第二種是利用量測限制違反量分區的策略。第三種是應用求解多目標最佳化的技術。經由數值例題，檢驗本文所提之三種策略，都能有效及可行的在粒子群最佳化演算法中處理限制條件。
另外，本文應用非支配排序法於粒子群演算法，求解多目標最佳設計的問題。先以無限制的多目標數值例題，探討粒子群演算法求解的可行性，再探討含有限制條件的多目標粒子群演算法。最後，以多目標粒子群最佳化程序，應用於含限制條件的工程設計問題，例如四桿桁架設計，I型樑的截面設計及焊接樑的設計。本研究的含限制之多目標粒子群最佳化程序亦結合有限元素分析軟體ANSYS，再對工程例題進行多目標最佳化的設計，加強本文粒子群最佳化設計具有實用性。 由本文所發展的含限制或不含限制的多目標粒子群最佳化程序，可得到平滑的Parato曲線，亦得到精確的數值。

關鍵字

粒子群最佳化 ； 多目標最佳化 ； 突變 ； 限制條件的處理 ； 有限元素分析

並列摘要

A particles swarm algorithm (PSA) including improving mutation for global optimization is presented in this thesis. The presented global PSA can simplify the solution process. Three constraints handling strategies in such a global PSA are proposed to construct a constrained particles swarm optimization (CPSO). The first strategy is applied the concept of flying back with modification. The second strategy is to modify the measurement technique of constraints violation. The third strategy is the application of the technique of multiobjective optimization in which all constraints are integrated and transformed to an additional objective.
The non-dominated concept is applied for dealing with particles swarm multiobjective optimization (PSMO) problem. The PSMO in the thesis contains those three constraints handling strategies with some illustrative examples and structural design optimization problems. The results shows that all constrained PSO for single or double objective problems are successful to obtain the results and satisfied Pareto front.

並列關鍵字

Particle swarm optimization ； Multiobjective optimization ； Mutation ； Constraint handling ； FEM

參考文獻

[3]Eberhart, R. and Shi, Y., 2000, “Comparing Inertia Weights and Constriction Factors in Particle swarm Optimization,” Proceedings of 2000 Congress on Evolutionary Computation, ,La Jolla, California, USA, July 16- 19, pp. 84-88.

[4]Bergh, F.V.D., and Engelbrecht, A.P., 2006, “A Study of Particle Swarm Optimization Particle Trajectories,” Information Sciences, Vol. 176, No. 8, pp. 937-971.

[6]He, S., Prempain, E., and Wu, Q.H., 2004, “An Improved Particle Swarm Optimizer for Mechanical Design Optimization Problems,” Engineering Optimization, Vol. 36, No. 5, pp. 585-605.

[7]陳孝楷, 2010,粒子群演算法的多及值工程最佳化,淡江大學機械與機電工程學系碩士班,碩士論文.

[8]Li, X., 2003, “A Non-dominated Sorting Particle Swarm Optimizer for Multi-objective Optimization,” in Proc. Genetic and Evolutionary Computation: GECCO 2003, Vol. 2723, pp. 37-48.