Abstract: Motivated by the coupling relationship between the machine tools accuracy variation process and parts cutting process, an identification method is developed, which addresses key accuracy characteristics of machine tools for parts machinability evaluation. Hidden Markov model (HMM) is constructed via regarding the accuracy variation process and cutting process as a coupling system. The coupling system state transition matrix consists of the state transition matrix for each cutting characteristics and their importance vector. The state of each cutting characteristics is given capitalizing on the relative error between the measured data and fitted value calculated from improved support vector machine (SVM). Additionally, parts machinability evaluation model is formulated, proposed to get the optimal cutting characteristic state for coupling system. Cutting characteristic importance vector is determined by the parameters estimation of HMM, taking advantage of improved quantum genetic algorithm (QGA). Meanwhile, accuracy characteristics of machine tools are identified through quality function deployment (QFD), based on the fuzzy associated matrix of coupling system and cutting characteristics importance vector. An example is given to verify the feasible and effectiveness of the method.

Key words: Identification of accuracy characteristic, Machinability evaluation, Machine tool