Abstract:The precision of the contact measuring systems composed of a robot arm and multiple probes is highly related to the preci-sion of TCP calibration, The center position of each measuring needle relative to the flange needs to be calibrated so that the coordinate position of target points can be obtained and converted to the same base coordinate system accurately. How-ever, traditional TCP calibration methods lead to large random errors and low reliability. To reduce the collaborative cali-bration error of multiple measuring needles in the robotic arm and improve the calibration reliability of the contact meas-urement system, based on classic 4-point TCP calibration method, the external benchmark method was proposed. The con-cept of virtual external reference points is introduced to decrease the errors caused by manual calibration of four points. The center coordinate of the flange is obtained by utilizing the measurement function of the measuring system and the spherical center is calculated by least square method. The terminal TCP parameters of multiple collaborative probes are acquired by constructing a TCP calibration analytical model and using the fitted spherical center as an external reference point. Compared with other TCP calibration methods, the results reveal superior measuring precision performance of the proposed TCP calibration method. Further physical experiments indicates less measuring errors after calibration with an accuracy of within 0.08mm, which identifies key features of high reliability and precision.