Study On Rapid Response Technology Of Production Process Under Network Alliance Manufacturing Environment

Market changes that manufacturings face are frequent and unpredictable, and customers' requirements are individualized and diversified, which lead to the shortening of product cycle and the speeding up of product updating. Thus, enterprises should dynamically adjust their manufacturing mode and controlling strategy to make quick response to the changing market so as to succeed in the intense market competition. Network alliance manufacturing is a kind of production mode which is suitable for the medium and small-sized enterprises. It fulfills the manufacturing information integration and resources sharing by means of network; enhaces manufacturing system's agility, flexibility, and ability of meeting market by means of alliance cooperation. Aiming at the production feature of equipment manufacturing, this dissertation studies the rapid response strategies and key technologies of the production process under the network alliance manufacturing environment so as to satisfy customized requirements.1. Rapid response mechanism of the production process is studied, and structure mode of the response system based on network alliance and resources'sharing is established in this research. The improvement of manufacturing system's response property is expected to be achieved from such five levels as enterprise alliance constructing, customer requirements analyzing, production task palnning, production resources reconfigurating, and production schedule executing. And the response system's structure and function model are analyzed and described so as to formulate rapid response architecture of the production process.2. Constructing method of manufacturing enterprise alliance based on service resources matching is put forword. Taking the data model of resource description framework as the basis, manufacturing service resource description model and matching strategy are established, service resources'qualitative property matching and quantitative property matching algorithm are presented, and service resources'searching flow is set up so that a more extensive enterprise alliance, which could promptly achieve other alliance enterprises'support, is formed by using the network manufacturing service sharing plateform.3. On the basis of agent basic structure, multi-agent based structural model of the rapid response system in production process is set up, agent assistance in the transferring from customers'requrements to production requirements is studied; and on the basis of production task planning strategy under the circumstance of indepth study of this assistance mode, annealing evolution algorithm deduced optimal assigning method of production tasks is proposed, and synthetical biding evaluation which takes the tendering result as basis is realizes so that an overall optimal assigning scheme of production tasks is obtained. With the help of reliablity theory, a kind of parallel structure to improve the reliability of muti-agent system is propsed, and the algorithm that can solve the maximized problem of reliability is constructed.4. Resource reconfiguration modeling method and its scheme evaluation indicator system based on polychromatic sets theory are proposed. Mapping relationship between manufacturing resources and system functions are established by the use of element, individual color sets as well as unified color sets. Resource reconfiguration schemes are achieved through searching the contour Boolean matrix. Selecting, processing and evaluating method of the reconfiguration scheme's evaluation indicator are discussed. In which, function redundancy proposed in this dissertation is an important evaluation indicator of the reconfiguration scheme, and its effective calculation method is given in this dissertation. Thereby, production resources rapid organizing of the different production requirements is achieved.5. Production planning and scheduling integrated optimization problem is studied under the multi-order environment. Integer programming models of the individual production and small-lot production are established based on the consideration of product assembly structure constraint, resource capacity constraint, earliness-tardiness penalty, transporting time, and multi-process route problems. The ant colony algorithm with elitist strategy is designed to achieve multiorder-oriented task scheduling, thus planning layer and scheduling layer are unified which can avoid constraints losing in combining process of planning and scheduling. The feasibility of ant colony algorithm to solve the problem and the superiority of arithmetic performance are verified by means of running some practical instances.On the basis of the aforementioned study, a rapid response prototype system of production process is established, and the rapid response system software based on java server pages is developed. The correctness and feasibility of this dissertation proposed theory and method are further verified through partial applications in enterprises.