| The reverse design based on prototypes is one of the optimal approaches to creative design from the point of view of reverse engineering. Retrieving the Initial Design Parameters (RIDP) from existing objects is not only a key issue to redesign a product, but also a way to raise reverse engineering to high level. The direct purpose of RIDP is to eliminate the distortion and keep the true state of the material prototypes, is to remove the mistakes of manufacturing, assembling, abrasion, measurement, furthermore to reappear initial design parameters in the case of preventing from spreading mistakes. While its final purpose aims essentially at understanding methods to treat the relation among various factors and looking for correct idea and result of the redesign, therefor improving our design capability. Based on the analysis of actuality in reverse engineering, some research shortages inside and outside our country are pointed out. The theory of RIDP is firstly studied systematically via considering the factors of design and manufacture. The concept, definition, theoretic frame and whole parameter system of RIDP are established. The theoretic basis and realized methods of all RIDP stages are summarized. Based on the above, through analyzing characteristics of the knowledge to be needed in RIDP realization, the knowledge classifying system and representation techniques of RIDP are put forward. Furthermore, the three sub-models combined uniquely with the identifier of object prototypes are promoted. The data layer configuration and design flow of the RIDP are set up. In this dissertation, the basic dimension is sorted scientifically. Some rules to choose parameters in RIDP are abstracted systematically. The design arithmetic related to estimating basic dimension by means of statistics and correcting basic dimension estimation with expert system technique is set up. Based on the research, a neural network model to select fits and tolerances is...
|
PDF Full Text Request