累积法在随机删失数据可靠性评估中的应用

Application of Cumulative Sum Method in Censored Data's Reliability Assessment

在可靠性数据分析中引入累积法,结合平均秩次法解决随机删失数据下的参数估计问题。以双参数威布尔分布为例,设计了一种对比分析累积法和最小二乘法进行参数估计时估算效果优劣性的蒙特卡洛仿真方法。仿真结果表明,对于不同删失比、不同样本量下的随机删失数据,累积法的计算误差比最小二乘法更小,其结果更加接近真实值。

Combined with mean rank order method,cumulative sum method is introduced to data analysis of reliability to solve the parameter estimation problem in censored data. With double parameters Weibull distribution as an example, a Monte Carlo simulation method is designed to compare the cumulative sum method with least squares estimate in parameter estimation. The result shows that the cumulative sum method＇s calculation error is smaller than the least squares estimate＇s. The cumulative sum method＇s results are more close to the real value.