Research On The Order Promising Decision-Making For The Discrete Manufacturing System

The order promising decision is responsible for responding the incoming orders' specifications, such as quoted quantity of the order items, due date, material preference and production route preference. Within the current complicated manufacturing environment, how to quickly respond the customers, reduce lead time, improve on time delivery and optimize enterprise resource by order promising function becomes an increasingly urgent issue to all companies. This dissertation mainly focus on the information system architecture of implementing order promising, and researches the order promising decision approaches of three kinds of specific discrete manufacturing systems.Since order promising is tightly relative to the other functions of the order fulfillment process, the dissertation investigates the characteristics of the order promising of discrete manufacturing systems from the view of point of the order fulfillment process and even the supply chain management, and then presents the requirement attributes of the order promising system. Subsequently, the service contract meta-model is refined and extended on the basis of the reference model of service oriented architecture (SOA), and formalized with UML to contract the service, and then the SOA meta-model is constructed. At last, an integrated framework for order fulfillment system is presented based on the SOA meta-mode to implement the order promising and fulfillment process.For the order promising decision of job shop manufacturing, a decision approach is developed based on allocated available to promise (AATP) and capacity to promise (CTP) in this dissertation, called C-AATP. The promise using the C-AATP is based on capacity constraints because it introduces the forecasted demands and maps the lower level resources (raw material and production resources) to the higher level resources (products) by simultaneously considering constraints such as company's object, material supply, production resource and market demands. Therefore, the C-AATP can overcome some lacks of quick response customer's order and global optimality yielded by CTP. In addition, the C-AATP also advances the two serial steps processing method in AATP into a one step processing method. Consequently, the C-AATP guarantees a reliable and optimal order promise while responding customers' order quickly.The cycle time of the products of semiconductor production system, which is characteristic of multi-reentrance, is long and fluctuates dramatically, and the fact that the production system is frequently impacted by the arriving of rush orders has increased the variability of the cycle time. All these lead to make a reliable promise tougher. The difficulties of order promising in semiconductor production system are under investigated in this dissertation. After analyzing the issues of existing order promising approach, we develop an order promising decision approach with capacity reserved in advance for multi-priority orders. It is an effective solution to the frequent rush orders arriving on the system. In addition, we also develop a kind of algorithm based on queuing theory to accurately predict the production cycle time to make the promise more reliable.The products of the project manufacturing vary so greatly in specification and have so complicated structure that the probability of producing the same product is small. It's almost necessary to design and to prepare tools and technology again from start for each product, so the production has same characteristics of project, and content of the order promising decision is to make a preliminary production plan, then to calculate the cycle time of product for determining a reliable due date. We apply the theory of project scheduling management to support the order promising decision in this dissertation. An order promising decision approach based on the theory of critical chain project management is developed. And then for the theory of critical chain project management sizes buffer without considering attributes of a project, we proposed a revised method with attributes dependent for sizing buffer. The revised method brings the order promising approach with shorter cycle time and higher on time delivery.Finally, a prototype of order promising system was constructed based on integrated framework of order fulfillment system.