Jixie qiangdu (Jan 2020)
RELIABILITY ANALYSIS OF RIGHT CENSORED WEIBULL DISTRIBUTION BASED ON GENETIC ALGORITHM
Abstract
In order to improve the reliability of the atmospheric probes of the B737 fleet,this paper takes the fault data of the probes as research object to analyze their reliability. Due to the incompleteness of fault data,this paper first employs the distribution fitting evaluation method to determine that the fault data is in accordance with the right censoring three-parameter Weibull distribution,and then proposes the maximum likelihood method to estimate the parameter of right censoring threeparameter Weibull distribution. According to the definitions of Weibull distribution and maximum likelihood,this paper converts the problem into a two-parameter equation to work out the optimal solution to the problem. Moreover,in light of genetic algorithm,this paper manages to work out the shape parameter β = 2. 2614,the size parameter δ = 5317. 186 and the location parameter θ =13253. 36. Finally,with the aid of KS test,this paper tries to verify the degree of model fitting,proving that it is appropriate to employ the right censoring three-parameter Weibull distribution model. On the basis of the above analyses,this paper predicts the lifespan and the maintenance interval of the probes and offers corresponding suggestions for maintenance. For the airspeed heads with complete maintenance records and limited use,their maintenance and management should be carried out within 10216 flight hours.
