By taking the joint surface of shafttoolholder and toolholdertool of the vertical CNC milling machine as the study object, combined with the Timoshenko beam theory, the transmission matrix method, receptence coupling substructure analysis (RCSA) and particle swarm optimization (PSO) algorithm, a fixed joint surface parameter identification model was proposed. The study analysis show that calculation of cutting toolholder and tool point frequency response function transmission matrix method is based on Timoshenko beam theory for the model. This model avoids the complex modal analysis error and obtains high precision of calculation. The optimization algorithm is combined with local search and PSO algorithm which considering global PSO powerful search ability and local search algorithm for local search ability, can more easily find the global optimal solution, and realize high precision of identification of joint surface parameters of shafttoolholder and toolholdertool.