Research On Fuzzy Method Of Reliability Prediction And Distribution For Industrial Robots

With the rapid development of contemporary science and technology,the application of domestic industrial robots is increasing,but compared with developed countries,there is still a certain gap in the reliability of domestic industrial robots.Due to the various types of failures of industrial robots,complex structure and long service life,it is difficult to obtain reliability indicators through probability and statistics methods.Reliability prediction and allocation play a guiding role in all stages of industrial robot production design.The prediction and allocation results can also provide important basis for product optimization and reliability testing,and are of great significance for designing products that meet the specified reliability indicators.Therefore,based on the reliability analysis of industrial robots,this thesis further carries out reliability prediction and allocation.The specific research content is as follows:(1)The reliability analysis of industrial robots is carried out using the failure mode,impact and criticality analysis method based on the fuzzy criticality matrix.After the system is simplified,the fault tree method is used for qualitative analysis.Taking the industrial robot subsystem as the research object,analyzing the structure and function of the system,using the FMEA method,determining the failure mode of the system,and analyzing the impact caused by the failure.Use fuzzy theory to determine the severity level and the interval value of the failure probability level to perform CA analysis,and further obtain the weak links of industrial robots.After simplifying the system,the FTA method is used to qualitatively analyze the industrial robot.(2)A new method of fuzzy reliability prediction of industrial robot system based on weighted geometric average model is proposed.Due to the cumbersome structure,diverse functions,and lack of historical failure data of industrial robots,traditional reliability predictions are not suitable for industrial robot reliability predictions.For this reason,this thesis introduces fuzzy theory to study the method of fuzzy prediction of reliability of industrial robots.Invite multiple experts to score the influencing factors,and calculate the fuzzy number scored by the experts through the aggregate fuzzy number and fuzzy weighted geometric average model,weighting a centroid number,thereby combining the component counting method to achieve the fuzzy reliability prediction of the industrial robot system.(3)The reliability allocation of industrial robot system is based on the fuzzy comprehensive evaluation method of fuzzy Petri net.Through the established fault tree model,it is transformed into a Petri net model.Fuzzy Petri reasoning network is used to determine the complexity and importance of sub-systems.The judgment of the working environment is determined by whether several common harsh working environments affect the failure of the sub-system,and the complexity,importance and working environment are regarded as the three influencing factors of the sub-system.Finally,the fuzzy comprehensive evaluation method is used to pass The membership function of each sub-system is calculated,and the reliability of each sub-system of the industrial robot is allocated.This method does not need to directly determine the weight of each influencing factor,so it can reduce the error caused by the subjectivity of the expert’s understanding of things.