An Evaluation and Prediction Method of Reliability Growth under Interval Censoring

Abstract

The invention relates to a reliability increase evaluation and prediction method under the condition of interval censoring. The method comprises steps that an order statistics model following power-law distribution is constructed; an initial value of a reliability increase process parameter is determined according to priori knowledge of the equipment, on the basis of the order statistics model following power-law distribution, random interpolation of the missed fault time is carried out; on the condition that the missed fault time is completely filled, an expectation-maximization method is employed, the parameter of the reliability increase process is updated iteratively; the average fault-free time interval at present is evaluated, and the future fault generation time is predicted. According to the method, through integrating the observed data, failure time distribution can be iteratively estimated, random interpolation of the missed fault time can be carried out on the basis, determination of system reliability increase rules can be realized on the basis of an integral data set, evaluation of the present average fault-free time interval and prediction of the future fault generation time are realized, inconvenience in statistics analysis is made up, and deviation caused data deletion under the non-gaussian hypothesis can be effectively reduced.

Description

technical field [0001] The invention belongs to the field of reliability engineering, and in particular relates to an evaluation and prediction method of reliability growth under the condition of interval censoring. Background technique [0002] A complex device or system will inevitably and continuously expose various defects and failures in the initial stage of being put into formal use. These defects are usually closely related to the design scheme, manufacturing process, operation method, maintenance technology and management level of the product. Aiming at these soft defects, adopting the strategy of "operation-exposure-improvement-operation" to take timely improvement measures can continuously improve the reliability of equipment, which is the process of equipment reliability growth. In this process, it is necessary to use the "reliability growth evaluation method" to judge whether the reliability of the equipment is increasing and to quantitatively evaluate and predi...

AI Technical Summary

Problems solved by technology

[0005] In view of the above-mentioned difficulties and challenges, the purpose of the present invention is to provide a method for evaluating and predicting reliability growth in the case of interval censoring, which can fuse the observed data to iteratively estimate the failure time distribution and randomly interpolate accordingly Find out the missing failure time, and then realize the determination of the system reliability growth law on the complete data set, and use the generalized Gamma distribution to realize the evaluation of the current failure interval time and the prediction of future failure occurrence time

Images