The Job Shop Schedule Problem (JSSP) refers to the ability of an agent to allocate tasks that should be executed in a specified time in a machine from a cluster. The task allocation can be achieved from several methods, however, this report it is explored the ability of the Ant Colony Optimization to generate feasible solutions for several JSSP instances. This proposal models the JSSP as a complete graph since disjunct models can prevent the ACO from exploring all the search space. Several instances of the JSSP were used to evaluate the proposal. Results suggest that the algorithm can reach optimum solutions for easy and harder instances with a selection of parameters.
翻译:工作商店日程问题(JSSP)是指代理商分配任务的能力,这些任务应在某一集体的机器特定时间内执行,任务分配可通过几种方法实现,然而,本报告探讨了 " Ant Colony Optimination " 公司为若干 " JSSP " 公司案例提供可行解决办法的能力,这项提议将JSSP作为完整的图表,因为脱节模式可以防止ACO探索所有搜索空间。JSSP公司的一些实例被用来评估提案。结果显示,算法可以选择参数,为简单和困难的事例找到最佳解决办法。