Integrated reliability assessment for Weibull distribution using varied environment data

To improve the stability and accuracy of the reliability assessment, a method of integrated reliability assessment for Weibull distribution was proposed using the varied environment test data. In this method, the location parameter and scale parameter were assumed respectively as a function of relevant environment variables. Then the location parameter function was constructed by the method of radial basis function while the scale parameter function was constructed by the linear model. Thus the influences of environment factors on the reliability of the product were measured quantificationally. With the generalized linear model, the maximum likelihood estimations of the coefficients were obtained, and corresponding arithmetic was put forward. With the estimations, the approximated confidence interval for the reliability function was obtained according to the normal law. The instance analysis shows that this integrated reliability assessment method improves the precision of reliability assessment and is straightforward for engineering application.