Research On Reliability Of Flexible Mechanism Based On Stochastic Extreme Value Response Surface Method

Three kinds of motions consist in flexible mechanism at the same time and coupled interactively when flexible mechanism running.they are deformation motion of flexible components,rigid motion of mechanism,Clearance collision motion.the kinetic characteristics of fiexible mechanism are nonlinear because of the coupled effects?There is ambiguity between failure mode at the same time,in order to improve the precision of reliability analysis,this article with the aid of extreme Response Surface Method(Extremum Response Surface Method,ERSM)agent model of technology,Based on the Stochastic Response Surface Method(SRSM),Genetic Algorithm(GA)and adaptive neuro-fuzzy inference system(ANFIS),an improved Response Surface agent model Method is proposed to analyze the reliability of mechanical systems under clearance coupling between flexible components and moving pairs.It mainly includes:(1)Stochastic extreme value response surface method to analyze the reliability of flexible mechanism.First,SRSM and ERSM are combined to establish a random extreme response surface method(SERSM)model.Then,a kinematic model of the gap coupling between the flexible member and the motion pair is established.Based on this,the crank slider is used,for example,to calculate the movement displacement error in the [0,T] time domain,establish a specific SERSM model,and complete the reliability analysis of the crank slider mechanism under the coupling of flexibility.Clearance of movement pairs.Finally,the reliability calculated by SERSM and the results of MCM and ERSM analysis are compared.The results show that the reliability analysis accuracy of SERSM is higher than ERSM,closer to the calculation accuracy of MCM,and the calculation speed is faster than MCM,which is basically consistent with ERSM.(2)Genetic algorithm-double random extreme value response surface method for reliability analysis of multi-gap flexible mechanism.First,on the basis of SERSM,a Doul stochastic limit extreme value response surface method(DSERSM)model of coupled failure is established,and a genetic algorithm and DSERSM are proposed to optimize the multiple random extreme value extreme surface method of genetic algorithm.(GA-DSERSM);Then,build a virtual prototype model of the aero-engine vector tail nozzle,calculate the [0,T] angular error in the time domain and the deformation of the flexible member,build a specific GA-DSERSM model,and then the tail nozzle performs dynamic Reliability calculations.Finally,the calculated reliability is compared with the MCM and SERSM methods.The results show that GA-DSERSM is closer to MCM calculation accuracy than SERSM in reliability calculation accuracy and greatly improves the calculation speed than SERSM.(3)Fuzzy neural network-multiple random extreme value response surface method for fuzzy reliability analysis of multi-gap flexible mechanism.First,combining ANFIS and MSERSM,an adaptive fuzzy neural network multi-random extreme value response surface method(ANFIS-MSERSM)is proposed.Then,calculate the maximum stress and strain of the component during [0,T] timedomain operation,and calculate the fuzzy reliability of the vector tail nozzle using a specific ANFIS-MSERSM model.Finally,the calculated reliability is compared with the MCM and ERSM methods.The results show that in reliability calculation accuracy,ANFIS-MSERSM is closer to the calculation accuracy of MCM than ERSM,and the calculation speed is even higher than GA-MSERSM.