Objective The stress variations during the wear process of an artificial knee joint were studied. Then, a signal acquisition system was designed to capture the vibration signals induced by the wear of knee joint prosthesis. The aim was to provide new technical means for online wear monitoring of the artificial knee joint. Methods To effectively collect vibration signals, the optimal installation position of the vibration sensors was determined by analyzing the dynamic model of the knee joint prosthesis during motion and identifying the main distribution areas of the tibial insert contact stress. The dynamic model of the femoral prosthesis was solved using Lagrangian equations. The torque variation curve of the femoral prosthesis was obtained to validate the effectiveness of finite element analysis. The signals collected by the vibration sensors installed at different positions in the friction wear experiments and the surface morphology in different areas were compared to verify the effectiveness of the acquisition system design and finite element analysis results. Results The stress concentration regions of the tibial pad under four degrees of freedom (flexion, internal and external rotation, anterior-posterior displacement, and up-and-down displacement) were obtained based on a dynamic simulation. A stress concentration was evident in the middle and posterior regions of the tibial pad. A vibration signal with a higher amplitude was collected when the vibration sensor was installed at the rear end of the tibial pad. This aided the vibration feature extraction of the knee joint prosthesis. Conclusions The vibration signal acquisition system designed based on the dynamic simulation analysis effectively collected the vibration signals generated by the artificial knee joint during the wear process. This study provides an important means for evaluating the wear mechanisms of artificial knee joints and monitoring their full-life health status.