The Study And Application Of Hall For Workshop Of Metasynthetic Engineering (HWME) For SBA System

Simulation Based Acquisition (SBA) is a kind of new idea of acquisition progressively adopted by Department of Defense (DOD) and Industry in recent years, which represents the direction of weapon system acquisition modes. While, Hall for Workshop of Metasynthetic Engineering(HWME) is an enable technology to impel the development of acquisition from traditional mode to SBA mode. Thus, take the project supported by the National Defence pre-research Foundation as background, this dissertation mainly develops an inland SBA framework and the corresponding prototype system, and explores a sub-system of prototype system, which is HWME. As follows in detail:In this paper, by analyzing the SBA architecture proposed by the Joint SBA Task Force and using it as a guide, the inland traditional sequential acquisition process is studied, the shortfalls within which are identified, and then the SBA Engineering architecture is established based on Concurrent Engineering theory. SBA process is a spiral or iterative process rather than a waterfall process that leads to the development and fielding of a new weapon system. Following, according to the characteristics and requirements of SBA Engineering, a framework of SBA is developed, which describes the support platforms/tools and application systems etc. the SBA system should include; and an architecture of prototype SBA system which can direct implementation is put forward, which including collaborative platform of management and control, Hall for Workshop of Metasynthetic Engineering, collaborative simulation platform, collaborative PDM, ERP, Computer-Aided Logistic Support and distributed resource database system etc. Next, the key technologies piloting the implement of the prototype system are analyzed.To deal with complicated decision-making problems solved difficultly in rule decision-making methods and system engineering methods, such as problems of demonstration of weapon and equipment, the Meta-synthetic Decision Support System (MSDSS, i.e. Hall for Workshop of Metasynthetic Engineering for complicated decision-making problems) is proposed. By analyzing characteristics and existing problems in current DSS, the requirements of MSDSS are put forward according to the Meta-synthetic process for complicated decision-making. Furthermore, the openarchitecture of MSDSS catering to requirements based on Browser/Server is established; the components and functions of the system are discussed. And the static design of the system is conducted using method of Object-oriented modeling. Also, the key technologies such as model synthesizing, simulation synthesizing, information synthesizing, opinion synthesizing and knowledge synthesizing are introduced.To model synthesizing, two approaches to distributed model management and automatic composition are developed. One method designs the model as agent, so, a multi-agent model system with star structure is built, which consists of one management agent and amount of model agents. Then the structures of two kinds of agents are designed, and their functions and actions are analyzed. Furthermore, a backward chaining algorithm is addressed to compose the existing model agents in a dynamic and automatic way, which realizes the following functions: model can be managed in a distributed way and calling remotely, model can be reused, customers needn't know how models are implemented, and composing several models to solve complicated problems manually is unnecessary etc. Another method takes models as Web services, describes Web service with OWL-S, and using semantic web technology to finding and matching models, and combining semantic web technology and an improved backward chaining algorithm to compose the models semantically and automatically, which also implements the functions that multi-agent model system does, while, it is more flexible. Realizing the function of distributed model management and automatic composition makes the machine system of HWME may employ its quantitative meta-synthesis to maximum advantage, and also makes the experts may employ their experience and wisdom to a more perfect advantage to solve complicated problems, with no necessary to master too much knowledge about model.During the course of discussing, a problem should be considered is that how to aggregate experts' distributed thinking to reach consensus. Thus, a new consensus method on basis of linguistic assessment information is proposed, which considers not only using nature language to evaluate is more approaching to fact and more favored by experts, but also high efficiency feature of computer in dealing with numeric problems. Within the method, a new algorithm for aggregation of several triangle fuzzy numbers is proposed, according to it, the opinion which has high consensus degree with other opinions will be more reflected in group opinion, while the opinion which has little consensus degree with other opinions will occupy a subordinate position in group opinion. Also, the algorithm may integrate the weights of experts, which will make the...