The Research And Development Of TOC-Based Semi-process Industrial Production Optimizing Decision Support System

With the trend of globalization, the enterprises , especially the manufacturers become to face the continuously increasing press of competition. The efforts to combine with the computer application technology, to enhance the production decision level, to increase the flexibility of production, to shorten the delivery period and to decrease the cost are very much important to the improvement of the enterprises' competitive power.Now, the MRPII/ERP systems, such as SAP, Oracle, Ufsoft, which focus on production management have been widely used in China, and have had good effects on the practical production management. However, the existing software is lack of the decision support functions. The suppositions such as fixed lead-time can't meet the needs of the different kinds of manufacturers.With the background of textile industry, this paper takes production operation as object, the optimizing decision as content, and uses the advanced information technology. It focuses on the research of the establishment of the decision support system that helps the enterprises tosolve the semi-structure and non-structure problems in the production process. At first, it says the history of the production management decision and the decision support system, analyzes the application of the decision support system and the status of the informationization of textile industry, and indicates the essentiality and the significance of this paper. Secondly, it summarizes the basic characteristics of the textile industry— —mixed order-driven and semi-process by the comparison with the general production classification. And identifies the two main bottlenecks of the textile manufacturers by the constraint identification methods and TOC. Thirdly, considering the characteristics of the textile manufacturers it analyzes the production optimizing decision models at three levels and makes some improvements. By the proper transformation of the basic linear program model and goal program model, it discusses the effective method of making the overall production planning according to the mixed order-driven and double-bottleneck; by the combination of the 0-1 integral program model and the transport model, it discourses upon the method of master production planning of distributed factories; by the improving of the scheduling model, it sets forth the scheduling algorithm for the semi-process characteristic. Then it specifies the process of the system analysis and design, ranging from the system aim, decision flow, data relationship, interfaces to software structure, model structure, function design, database design and man-machine dialogue interfacedesign in detail. At last, it utilizes the testing data to simulate the chief functions of the system and evaluates the running statement.The main contribution of this dissertation is improving the production optimizing decision models, optimizing the job scheduling arithmetic, setting forth the systematical assistant decision solution, and developing the prototype system, which based on the TOC and characteristics of mixed order-driven and semi-process. These provide the model-based and man-machine alternant decision support system to all levels of production optimizing decision. And these can also assistant the decisionmakers to make the correct strategy, to organize the production process properly and to increase the economic benefits, especially in comprehensively improving the production decision level and competitive power of the textile manufacturers.