為能提高傳統土木營建產業之競爭能力,並推廣節能減碳與永續發展的概念,本研究主要針對橋梁基礎中之樁基礎與淺基礎進行最佳化設計之研究。首先摘要回顧各種多點隨機搜尋之最佳化方法,以及其混合最佳化技術,並針對目前樁、淺基礎最佳化設計之相關文獻進行剖析探討。依此本研究建立起嚴謹且詳實之基礎設計分析模式,並整合各混合最佳化技術至實數編碼遺傳演算法、和聲搜尋法、差分演化法、粒子群法與螞蟻群法之中,並提出以最佳化技術對各演算法中最佳化參數研究工作進行分析探討。透過七個實際設計案例之分析結果顯示,於全部所率定出之最佳化參數中,自訂參數模式下以和聲搜尋法(HS/M)表現最為良好,而自適應參數模式下則以差分演化法(DE/A)表現最為優良。且此兩演算法之率定參數於各種案例搭配不同年度單價之情況下,皆能有效找出全域最佳解,且與最佳解差距也較其他演算法為低。若從可較原設計解節省之工程造價來看,本研究之各最佳化方法平均節省幅度可達48%,顯示已能滿足工程低價化設計與經濟性之需求。
A concept for the optimization of pile and shallow foundations using Hybrid Real-coded Genetic Algorithm (HRGA), Hybrid Harmony Search (HHS), Hybrid Differential Evolution (HDE), Hybrid Particle Swarm Optimization (HPSO), and Hybrid Ant Colony Optimization (HACO) is introduced. The design variables of pile foundation are seven independent variables, including pile diameter, pile length, pile spacing of x and y direction, pile number of x and y direction, and thickness of pile cap and the design variables of shallow foundation include size and depth of shallow foundation. Implementations of six test functions and seven paradigms of bored piled design case are discussed and compared, including their global optimum solutions. The conclusions show that the performance of the proposed algorithm HHS/M and HDE/A are better than other algorithms; furthermore, the result which saving costs from original design case by all algorithms are around 48% can manifest the application of least cost design efficiently as well.