Precision Sensitivity Optimization for Parallel Robotic System Based on Multi-Source Preventive Maintenance

This paper presents a multi-source preventive maintenance (MPM) based precision sensitivity optimization method for parallel robotic systems. Taking a parallel mechanism as an example, the input error can be divided into finite mutually independent uncertain error sources, and the probabilistic error model is established by the Monte Carlo stochastic method. Considering that several times of maintenance can lead to an increase of the accuracy degradation factor, a generalized hierarchy maintenance yield model is established, where the cost of preventive maintenance and failure replacement maintenance are distinguished between the independent error source and the overall error source. The nonlinear relationship between the maintenance yield per unit cost and influencing factors, including the error threshold and the maintenance times is obtained, which allows to develop an optimal maintenance strategy. Preventive maintenance is introduced to avoid unintended failure of the mechanism while extending the lifetime. The high-precision parallel mechanism has demonstrated promising applications in medical, industrial and other fields, and its economic benefits can be effectively improved by incorporating the MPM method. The experiment of manufacturing human lung models using a parallel 3D printing device demonstrates that the MPM method can improve the long-term precision and reliability of the parallel device, and the optimized human lung contour tracking precision can be improved by up to 55.56%.