维修性小子样验证中先验信息的融合方法

研究了在Bayes小子样验证方法中多源先验信息的融合问题。由于维修时间取值具有跳跃性大、范围较宽的特点,采用随机加权法将先验信息表示为先验分布。分别提出了基于均值和基于样本量的先验分布权重因子,在此基础上,建立了基于质量因子计算先验分布权重的模型。通过计算综合先验分布与假设的正态分布的概率密度误差,对其融合后为正态分布的假设进行了检验,同时也提出了准确计算分布模型参数的方法。通过某舰船装备维修性先验信息融合的实例表明了方法的有效性及准确性。

Prior Information Fusion Method in Maintainability Small Sample Test

The multiple source prior information fusion in Bayesian small sample test method was studied. Because the maintenance time values are much bouncing and have a wide range, the random weighting method was used to convert a piece of prior information into a prior distribution. The prior distribution weight factors based on mean and based on sample number were put forward respectively. Then the weight model of prior distributions based on the quality factor was established. By calculating the probability density errors between the integrated prior distribution and the hypothetic normal distribution, the hypothesis that the fused integrated prior distribution was normal distribution was tested, and the method to accurately calculate the distribution parameters was also put forward. The analysis by an example of some ship equipment maintainability prior information conversion and fusion shows the method is effective and accurate.