基于贝叶斯融合与仿真的系统剩余寿命预测

针对复杂设备的系统级剩余寿命预测问题,提出一种将贝叶斯理论与仿真相结合的融合预测方法.首先,采用仿真方法将单机级的多源失效信息折合到系统级,并转化为验前分布.然后,结合系统级的现场试验数据,获得多源信息的验后分布,并通过加权融合确定联合验后分布,从而预测系统的剩余寿命.最后,以某系统作为实例进行方法验证.结果 表明,所...

Remaining useful life prediction of system based on Bayesian fusion and simulation

Aiming at the system level remaining useful life (RUL) prediction of complex equipment, a fusion prediction method combining Bayesian theory with simulation is proposed. Firstly, the simulation method is used to convert the multi-source failure information from the single machine level to the system level, and then to the prior distribution. Then, the post test distribution of multi-source information is obtained by combining the field test data at the system level, and the joint post test distribution is determined by weighted fusion, so as to predict the RUL of the system. Finally, a system is taken as an example to verify the method. The results show that the proposed method provides a feasible idea for the system RUL prediction, and can be applied in engineering practice.