共因故障数据的极大似然分析方法

A Maximum Likelihood Method for Common Cause Failures Data

系统存在共因故障时,故障数据处理方法与传统的可靠性数据处理方法存在区别。在Marshall-Olkin的指数分布和相似模型条件下,根据极大似然方法,提出了共因故障数据的快速处理方法,所提出的方法,由于根据极大似然函数的导数变化,求解极大似然方程,可以高精度的确定系统中若干单元同时故障的故障率,为工程中共因故障数据分析,提供了实用方法。

When the system has the common cause failures, the testing data analysis method differed with traditional reliability data analysis method. Based on the homogeneous MarshaU-Olkin model, using maximum likelihood method, the testing data analysis method is presented for common cause failures. The maximum likelihood equation is computed by the sign change of the equation derivative, which testing data parameter is quickly estimated in high precision. Failure rate of specific failure units can be easily calculated by the proposed method and easily used in engineering.