Reliability Analysis And Optimization Of Industrial Robot Drivers With Multiple Failure Modes

Today,industrial robots are the most representative equipment in high-end manufacturing industry,which combines many technologies such as mechanical engineering,electronic engineering,sensor,automatic control,artificial intelligence.Although China is the world's largest market for industrial robots,there is a considerable gap between China and the West in the field of robot research and development.Especially,the market for advanced industrial robots is dominated by Western countries.In order to break the monopoly of advanced industrial robots in Western countries,it is imperative to strengthen the research on the reliability of our industrial robots.As the core component of industrial robots,As the core part of industrial robot,the driver has a significant impact on the system reliability of industrial robot.In this thesis,the reliability of electrical driver is studied considering the failure dependence.The main work can be drawn as follows:(1)This thesis analyses and elaborates the types,characteristics and structure of industrial robot drivers.And divides the driver into modules according to the structure and function,builds the reliability model of the driver,and draws the reliability block diagram.Based on the literature and engineering statistics,the failure modes are sorted out and the causes are found.By evaluating the Risk Priority Number of each failure mode,the main failure modes of the driver are determined,which provides the basis for reliability analysis.(2)There must be correlation between these failure modes of the driver,because a large part of the variables that affect the failure are same or related.This thesis uses the MRGP surrogate model method to estimate the failure probabilities of the driver system and its main failure modes at 4000 hours.And by comparing the accuracy and efficiency of AK-MCS,AK-SYS and MRGP methods to solve the same failure dependent problems,the advantages of MRGP surrogate model method in dealing with failure dependent problems are proved.(3)Combining MCS-and variance-based sensitivity index with MRGP method,the reliability sensitivity of the driver is analyzed.The input random variables that affect the reliability of the driver are sorted according to the contribution of uncertainty to the failure probability of the driver system and its main failure modes.The random variables that contribute a lot are found and suggestions are made to improve the reliability of the driver.(4)Aiming at fatigue fracture of PTH,the objective function and constraint conditions for optimizing the lifetime of plating through holes are established.Particle Swarm Optimization and Genetic Algorithm are used to optimize the reliability of plating through holes in the driver to extend the lifetime of plating through holes and improve the reliability of driver.