本研究提出一整合限制規劃及和聲搜尋之啟發式演算法求解資源可使用成本 問題。此演算法績效將使用由Mohring 所提出之測試例[1]以及Yamashita at al. 修 正自PSPLIB 之數值例[2]進行測試。前者為一由16 項作業、4 種資源之橋梁建造 專案，根據所給定之15 組資源可使用成本組合案例進行測試，測試結果發現：當 專案完工限制設為64 至69 天時，案例3/1/1/3 及案例3/3/1/3 可以找到新的最佳 解。後者則又可分為30 項、60 項、90 項、120 項作業四大類型，皆需使用4 種資 源。總共進行720 組專案案例測試，測試結果發現：本研究所提出的整合式演算 法所求得的總資源使用成本上界平均改善率優於野草最佳化演算法(Invasive Weed Optimization Algorithm)及人工免疫系統演算法(Artificial Immune System Algorithm) 之求解結果。

We developed a hybrid heuristic procedure of constraint programming and harmony searching to solve a resource availability cost problem (RACP). Performances of the proposed solution algorithm on testing instances for the RACP is evaluated, which are proposed by Mohring [1] and modified by Yamashita at al. [2] from PSPLIB dataset. The former is a bridge construction project consisting of 16 activities and four resource types. Given 15 combinations of unit availability for these four resource types, new best known solutions of this project were found for the Instance 3/1/1/3 and Instance 3/3/1/3 when the time limit of project completion sets as 64 to 69. The latter are RACPs for 30, 60, 90, and 120 activities and 4 resource types. 720 different project situations were tested in total. The average percentage of improvement of the upper bound achieved by the proposed algorithm is superior compared to the invasive weed optimization algorithm and artificial immune system algorithm.