The Design And Implementation Of A Real-time Jobshop Scheduling System

Shop scheduling and management is one of the core functions in manufacturingenterprise production management. Usually, there are many works and machine equipmentin the workshop, each work will be complete processing on different machines accordingby different process route, and these processes will involves large amount of object andinformation to the production scheduling decision. However, in our manufacturing industry,the workshop production process management are often be completed by artificial or halfartificial, so a large number of enterprise have a lot of problem in production management,such as lacking of management means in workshop management and linking bad betweenthe bottom production basic data and ERP system.In order to improve the visibility and controllability of the production process, solvethe existing problem of enterprise workshop management, a real-time dispatching systemfor the Job shop is designed and implemented in this thesis, based on the production relateddata in the manufacturing shop. The goal is to provide certain management methods to helpmanufacturing Job shop achieve the status of zero inventory, no defect, high efficiencyproduction and so forth.This thesis presents the design of the real-time scheduling workshop, and the mainfunctions of the systems including basic data management, production scheduling,information query and statistics and other major modules. In order to make full use ofworkshop real-time data information, effectively achieve shop scheduling function, thisthesis studies the workshop production data analysis base on the method of data mining.This thesis studies from the history data and real-time data accumulated from the Job shopscheduling process through the decision tree method, and regards the useful scheduling knowledge as training data. When the machine gets problem and disturbance appeared,according to the training data to predict the scheduling, realize the online schedulingoptimization.Based on the information above, this thesis proposes a rule-based schedulingalgorithm and with the earliest delivery date, the shortest processing time for index toestablish work, the machine selection strategy. Scheduling results will storage in thedatabase in the form of historical data, so as to modify the scheduling method.A data based Job shop scheduling software system is designed under the platform ofMicrosoft Visual C++6.0environment, and use Microsoft SQL Server2000as thedatabase. Detailed analysis of the main functions of the system implementation processwas given, and some running examples of the elaborate design system is shown at the endof the thesis.