A flexible job shop scheduling system based on multi agents and genetic algorithm is established to overcome job shop automation and optimization problems. The system is made of a management agent,a scheduling agent and machine agents,where the static scheduling is realized by genetic algorithm,while the dynamic scheduling is realized by these coordinative agents. An order is evaluated by the management agent first,and then passed to the scheduling agent with corresponding information as soon as the order is accepted. The order is optimized and decomposed by the scheduling agent with object oriented genetic algorithm,and the result is passed down to the machine agents. The machine agents act according to the information passed down by the scheduling agent,and eliminate the uncertainties in machining process by changing the order dynamically. Illustration examples show that the established system is practical,efficient and advanced.%为解决柔性车间调度自动化及优化问题,建立了基于多Agent及遗传算法的柔性车间调度系统.系统是一个由管理Agent,调度Agent及多个加工单元Agent组成,系统中通过遗传算法实现静态优化调度,而通过Agent之间的协作现实动态调度.加工任务到来时,先经管理Agent评估,接受后打包相应信息传递给调度Agent;调度Agent调用其面向对象遗传算法对任务进行优化分解并传递给各加工单元Agent;加工单元Agent根据调度Agent下达的任务进行加工,同时通过相互协调动态调整加工任务,以消除加工过程中出现的不确定性.示例运行表明所建立的系统可行,并兼有实用性,先进性和有效性.
展开▼