Residual life estimation fusing life data and expert information

A method fusing life data and expert information was proposed to estimate residual life of some kinds of single machine on satellite platform, which solved the problem that performance degradation information of these products could not be detected real-time and the life data collected were almost no-failure data. Firstly, expert information about reliability was transformed into prior distributions of parameters. Then life data were fused with expert information to determine posterior distribution of parameters. When ranges of parameters were determined, the residual life probability density distribution could be obtained. Besides, this paper took GPS receivers on the satellite platform as example, residual life estimation was given based on the method proposed in this paper. Lastly, a simulation method and a comparison with other methods were provided to prove the presented method effectiveness.