| Multi-varieties-small-batch production is becoming the dominate mode for medium and small enterprises. However, more kinds of production, tight schedule and frequent switching of the production process operations brought about by the mode, greatly increases the burden of job shop scheduling. At the same time, the perception and understanding of a large number of relevant information caused by complex production scheduling has become more difficult.To solve the above problem, the dissertation mainly talks about the wire harness production workshop of Hefei Derun Electronic Devices Co., Ltd. A hierarchical scheduling policy based on Q learning and simulated annealing genetic algorithm is applied, and the production scheduling information is displayed from different angles through a variety of visualization methods, which provide an effective way of keeping abreast of the production status for corporate executives, increasing their ability to access and deal with effective information for staff. The main research work is listed as following:(1) According to the actual situation of job shop scheduling in Derun, a hierarchical scheduling model is built: for the coarse-grained process before and after, Q learning is used to calculate the latest delivery; for the process schedule inside of process before and after, only considering the work needed to be assigned on that day, SAGA is used to sort by delivery. If the scheduling result can't meet the actual needs of production, manual intervention is allowed to adjust the system scheduling result.(2) Visualization display of production scheduling information. Star diagram is used to display all kinds of information as well as their associated relations; Gantt chart is used to show job shop scheduling results; field simulation is used to reflect real-time production information; statistics, such as bar, line, pie charts, is used to analyze data, thereby find problems.(3) Design and implementation the production scheduling information visualization system. Aiming at the actual production situation in Derun, based on three-tier B/S structure the system can finish automatic planning and manual adjustments, combined with feedback from a variety of visualization methods which display scheduling information. The system achieves good results in production efficiency and transparency management. |