Research On Methodology And Techniques For Cost Decision And Control In Product Design Process

On the basis of in-depth research of product cost and related techniques, methodology and techniques for cost decision and cost control in product design phase are discussed detailed against the shortcomings of current cost decision and cost reduction research. The contributions are as follows:Firstly, the precision of cost estimation in early product design phase is studied. An weighted eulerian-distance based model is proposed to evaluate the similarity of estimating sample and product to be estimated with poor information and samples, then the precision of this cost estimation type is improved effectively by handling priority of the similarity calculated by weighted eulerian-distance. A concept named "cost clustering" is put forward for improving the estimation precision by clustering analysis estimation samples in adaptive design and variant design process.Secondly, main factors which affect precision of cost estimation in product detailed design phase are discussed, then the idea of combining several single-feature together to form CAPP oriented feature combinations is proposed, and its effects to machining process are also analyzed. Cost estimation process with feature combinations improves the accuracy of process reasoning and precision of cost estimation compared to single-feature reasoning. A CAD/CAPP integration cost estimation system for machining part is presented based on the feature combinations method.Thirdly, knowledge-based solving process for cost reduction in product design phase is studied. A multi-agent based solving process for cost reduction in product design phase is presented on the basis of comparisons of current three solving patterns of cost reduction, and the structures of several kinds of sub agents are built. Also solving flow using multi-agent based model are presented.Fourthly, the representation of cost reduction knowledge is studied. The architecture of cost reduction knowledge is built on the basis of classifications for multi-agent based solving process. The concept of "one another nestification of knowledge" is presented to represent the coupling relations between different kinds of knowledge. Knowledge base for cost reduction is constructed with knowledgemaintenance and management functions.Fifthly, the key techniques to implement multi-agent based reasoning system for cost reduction are discussed, which are made up by the product cost model, strategy of task decomposition and allocation, cooprating resolving process based on nestification of knowledge and reasoning mechanism of sub agents, then a software system for cost reduction reasoning which supports product design procedure is constructed. A couple of samples show the system is effective to reduce product's cost.Finally, necessity to integrate cost estimation and cost reduction in whole product design process is discussed. The framework of cost decision for product design process is proposed, the framework can be divided into product cost modeling, cost estimation & analysis and cost reduction decision.