【摘要】 求解有独立到达时间和完工时间的单机调度问题目前主要采用启发式算法。为研究仿生类算法的应用可行性,采用蚁群算法,以提前/拖期的总惩罚量达到最小为目标。将邻位工作交换法(AP I)用于局部搜索以提高解的质量,并对每一个解的相邻工作间隔时间进行优化调整。用90个测试算例,将蚁群算法求解结果与分支定界法和禁忌搜索法的结果进行比较。结果表明,蚁群算法与分支定界法和禁忌搜索法的结果相当。更多还原

【Abstract】 Single machine scheduling problems with distinct ready times and due dates are usually solved using heuristic algorithms.An ant colony optimization(ACO algorithm was used to minimize the total weighted early and tardy penalties and to analyze the applicability of bionic algorithms.The adjacent pairwise interchange(API) method was used for a local search to improve the solution quality with idle times optimally inserted into each solution.The results of 90 tests made using the ACO algorithm compare well with results of the Branch & Bound method and the Tabu search method.更多还原