We are developing a new hybrid process planning approach that uses a generative procedure to select processes and a variant fixture planning procedure to complete the process plan. This paper describes components of this approach and related work on indexing mechanical designs and estimating manufacturing cycle time during product design. Specifically, we have created data structures that represent features and developed techniques that intelligently retrieve solid models. We have used a neural network to create specialized design similarity measures for fixture planning. We have implemented an algorithm to solve the maximal cutter finding problem for general 2-D milling processes. In addition, we have identified a set of manufacturing system models that can be used to estimate the manufacturing cycle time of a new product.
展开▼