The Research And Realization Of The Rapid Design And Cost Estimating Of Complex Technical Product

The main requirements of the customer in product can be summarized as quality, cost and delivery term. With the development of science and technology, it is not hard for enterprises to produce the products of high quality. However, on the premise of guaranteeing the quality, it is becoming more and more difficult to provide for customer the product of low price within the given time. In the light of this case, the relevant theories and key technologies directing at complex machinery product is systematically, deeply studied based on the actual situation of enterprise. The application of BP neural network in the cost estimating is analyzed. Finally the relevant software platform is developed.The mapping between function and structure and the technology of Case-Based Reasoning are adopted to realize the product rapid design. The design method based on function is studied. The function model of product is established through analyzing and decomposing the function. Then corresponding solution of the structure is searched through the mapping from function to structure for every end function cell. Successively, the holistic solution of the structure is obtained using the relationship of subordination between the function cell and that belonging to it. Finally, the optimal solution is accomplished after screening, evaluation and decision through the synthesized method of morphology.Variant design of complex technical product means the effective usage of the existing resource of the enterprise. The investigation on design activity of the enterprise shows that up to 70% of the actual activity belongs to variant design. At the same time, the main problem facing variant design is how to effectively express and retrieve the product. In view of this condition, the principle, the procedure and the state of the art are briefly introduced. Combining the advantages of two product modeling methods, one is based on tabular layout of article characteristics, the other is object-oriented, the product model of multi-view multi-layer model is proposed, which is convenient for the reusage of product structures and function and can express procedure and layer of the product as well. The computing formula for different kinds of attribute similarity is briefly discussed, due to variety of attribute type of product case. Based on this, the similarity between the historical cases in database and the current design requirements is calculated, thus, the case retrieval is completed, the optimal solution is obtained. Finally, the feature analyzing and information extraction are carried on, the new case is saved into case base. While the rapid design is completed, it also enriches the case base.The information of the product is gradually clear during the process, from scheme design to detail design. If we can estimate the cost in every phase, it not only conducts and evaluates the designer's task, but also can provide the basis for the quotation of prices and decision of the leaders. The relevant theories and the application in the aspect of cost estimate according to the structure tree assisted by BP artificial neural network are analyzed.According to the theories and technique above, the software platform for product rapid design and cost estimating is developed. The design tool is applied to shot blasting machine design, which proved to be efficient.