Reliability analysis for system by transmitting, pooling and integrating multi-source data

Reliability analysis of a complex system is vital and challenging. The Bayesian theory is widely used to integrate multiple source data for this problem by decomposing the system into multiple levels. However, the data in the component-level are integrated first and then transmitted to system-level in existing methods. This theory involves extensive computation and is easily affected by the setting of combined weights in the data integration of each level. By contrast, a method is proposed based on multi-source data transmitting first and then pooling together with integration in this paper. Firstly, multi-source data are grouped into lifetime, degradation and other data types and used to determine the corresponding distributions. Next, the data of each data type are separately transmitted to higher level of system. If there are data of identical data type in the higher level, then they are pooled with the transmitted data from the lower level. Finally, all the transmitted data are integrated with the native data in the system-level for reliability analysis of system. An illustrative example shows the application of the proposed method on a system in a satellite. The results, sensitivity analysis, and comparison demonstrate the effectiveness and advantages of the proposed method.