替換數位資訊法處理粒子群最佳化的限制

Handling Constraints in Particle Swarm Optimization Using Digital Information Substitution

林博然(Po-Jan Lin)

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(2013年)

https://doi.org/10.6846/TKU.2013.01202

摘要

基本粒子群演算法即為一種不含處理限制條件的仿生最佳化演算法。在近代粒子群演算的文獻中，以間接的方式處理限制函數居多。本文以仿生的觀念，發展替換數位資訊法處理粒子群演算中的限制條件處理問題，做法是藉由基因演化法則，將不可行個體與可行個體經由逐位元的資訊結合交換運算，使得不可行個體逐漸趨向且相似於一最佳可行個體。此法是直接處理限制函數，顯示穩健效果。另外本研究亦修改非支配排序策略，使其合理化。將限制條件違反量轉換為另一設計目標，與原目標函數使用非支配排序法求解，收斂於另一目標值為零。本文敘述整體最佳化解題程序，由數值題目驗證兩種限制處理策略，可有效且穩健的處理粒子群最佳化演算法含限制條件的問題。微致動器是微機電系統的一種動力源，大致可分為電熱式、壓電式、靜電式及電磁式。電熱式微制動器具有輸出力大，製作容易等優點。此微致動器若應用於夾取人體細胞，夾持點的溫度需限制在35℃以內，本研究設計電熱式微致動的結構，以最大化挾持點的位移量為目標，其中結合有限元素軟體ANSYS的分析，使用本文之粒子群演算法進行有限制條件的最佳化設計，可得到最佳的結果。本文之限制粒子群最佳化程序，可解一般含限制的工程設計問題。

關鍵字

粒子群最佳化 ； 限制條件的處理 ； 有限元素分析 ； 最佳結構設計 ； 工程設計

並列摘要

Partical Swarm Algorithm (PSA) naturally is a unconstrained biocomputational optimization. In recent publications, several indirect constraints treatments were proposed named constrained partical swarm optimization (CPSO).The present thesis simulates birds flock communication to develop the information substitution strategy in treating constraints. This approach can gradually convert the infeasible individual becomes the feasible individual by refering the information of globally effective individual. The presenting strategy can be categorized to a direct method of constraints handling. Another approach presented here in is a modification of non-dominating strategy. The idea is that all constraints are transformed to an additional single objective, then non-dominating sorting technique can be applied. At the end, pareto front will converge to zero value of additional objective. Illustrative numerical examples shows that the proposed two approaches of constraint handling are effective and reliable for engineering optimization. Microthermal actuator is broadly used in MEMS, it has large power output and is easy to manufacture. It can be adapted for grasping human cell, however the 35°C is the high temperature limit on the grasping point. This thesis provides the optimum structural design by using CPSO proposed in the thesis with the aid finite of element analysis ANSYS.

並列關鍵字

partical swarm optimization ； constraint handling ； optimum structural design ； engineering design

參考文獻

[3] L.Wang, Q.Kang, H.Xiao, and Q.Wu, “A Modified Adaptive Particle Swarm Optimization Algorithm,” Industrial Technology ,IEEE, pp. 209-214, 2005.

[4] R.C.Eberhart, and Y,Shi, “Comparing Inertia Weights and Constriction Factors in Particle swarm Optimization,” Evolutionary Computation,IEEE, vol 1, 84-88, 2000.

[5] Y. Shi, and R.C. Eberhart, .“Empirical Study Of Particle Swarm Optimization,” Evolutionary Computation, IEEE ,vol. 3,1945-1950,1999.

[6] R. C. Eberhart and Y. Shi, “Particle Swarm Optimization: Development, Applications and Resources,” Evolutionary Computation,IEEE, vol. 1, pp. 81-86, 2001.

[7] A.Ratnaweera, S.K Halgamuge, and H.C. Watson, “Self-Organizing Hierarchical Particle Swarm Optimizer with Time-Varying Acceleration Coefficients,” Evolutionary Computation, IEEE,vol.8