工程中经常需要用少量现场故障数据来评估设备的可靠寿命。针对现场故障数据很少,因而难以判断寿命分布类型的情况,提出了采用威布尔分布方法近似表示设备寿命分布。在分析统计量选择和经验系数蒙特卡罗仿真模型基础上,提出了保守估计可靠寿命的方法。仿真结果表明:当样本数N为1,仿真次数n为5000,形状参数m为2~5时,经验系数的蒙特卡罗仿真解具有较高精度,且误差为0.04%。该方法为小样本条件下可靠寿命评估提供了一种工程实用方法。

Reliable life evaluation of equipment based on a few of failure data is frequent in engineering. Since there are not many failure data, it is difficult to analyze exact life distribution type. Due to this, Weibull distribution was adopted to denote the life distribution of equipment approximately. Based on selection of statistics and experience coefficient with Monte Carlo simulation model, conservative estimation method for reliable life was advanced. The simulation results showed that it has a high accuracy and the error is 0.04% when the sample N, simulation times n equal to 1 and 5000 respectively in the field of shape parameter m valuing between 2 and 5. The method provides an engineering method for reliable life assessment with small samples.