本論文中提出兩種以群體智慧(swarm intelligence)為基礎之連續性結構拓樸(continuum structural topology)與桁架結構(truss structure)最佳化設計方法。一種為模仿真實螞蟻行為的蟻群演算法(Ant Colony algorithm),而另一種則為模仿鳥群覓食行為的粒子群最佳化演算法(Particle Swarm Optimization algorithm)。 在連續性結構拓樸最佳化設計方面,利用蟻群演算法及二進制粒子群最佳化演算法(binary Particle Swarm Optimization algorithm),找尋多值域結構拓樸問題的最佳解。並以四個常用的範例,印證所提出最佳化方法的確有效。 至於桁架結構最佳化設計方面,同時考慮到桁架設計問題的拓樸(topology)、尺寸(sizing)、與外型(shaping)最佳化設計。本論文提出兩階段螞蟻演算法,包括蟻群演算法及API(after "apicails" in Pachycondyla apicails)演算法,以及兩階段粒子群最佳化演算法,包括二進制粒子群最佳化演算法及吸引擴散粒子群最佳化演算法,搜尋最佳桁架結構。首先利用蟻群演算法及二進制粒子群最佳化演算法搜尋最佳桁架拓樸,再利用API演算法及吸引擴散粒子群最佳化演算法最佳化桁架尺寸及外型。並以多個常見的桁架設計問題,印證所提出的最佳化方法。所得到的結果也指出,所提出的演算法略優於其他文獻中所得到的結果。
In this dissertation, two novel approaches to swarm intelligence-based methodology for optimal design of continuum structural topology and truss structure are presented. One is the ant colony algorithm mimicking the behavior of real ant colonies, and the other is the particle swarm optimization algorithm mimicking the social behavior of bird flocking. In terms of optimal design of structure topology, ant colony algorithm and binary particle swarm optimization algorithm were implemented for finding optimal solutions to multi-model structural problems. Four well-studies benchmark examples in continuum structural topology optimization problems were used to evaluate the proposed approach. The results indicate the effectiveness of the proposed algorithm. And, in terms of optimal design of truss structure, truss structure optimization considering topology, sizing, and shaping simultaneously. A two-stage ant algorithm, consisting of the ant colony algorithm and API(after "apicails" in Pachycondyla apicails) algorithm and a two-stage particle swarm optimization algorithm, consisting of the binary particle swarm optimization and the attractive and repulsive particle swarm optimization were proposed in this thesis for finding optimal truss structure. First, ant colony algorithm and binary particle swarm optimization were used to optimize the topology of truss, and then API algorithm and attractive and repulsive particle swarm optimization ware used to optimize the size and shape of truss. To confirm the effectiveness of the proposed method, several well-know truss optimization problem were used to evaluate the proposed approach. The results indicated that the proposed algorithm have better performance than those reported in the literature.