A HMM Based Method to Select Solutions for Function Modules

A systematic method is developed to select the optimal combination of solutions for each function module existing in a pre-defined Product Function Model (PFM). Based on the definition of three fundamental modes for constructing a PFM, any PFM generated from customer requirements could be transformed into a uniform series- connection format. Then, Hidden Markov Model (HMM) is introduced to model the process of selecting solutions for each function module, and all the parameters in HMM can be calculated from the existing available information of these solutions. Furthermore, In order to avoid generating an exponential number of instantiations, the Viterbi algorithm is proposed to seek the optimal combination of solution during the conceptual design phase of a new product.