基于贝叶斯方法的兰利法测试引信解除保险距离数据处理

针对兰利法测试引信解除保险距离试验使用的试验样本量较少与评价误差较大的矛盾，提出应用贝叶斯方法处理兰利法测试引信解除保险距离数据。该方法建立贝叶斯数学模型，明确先验分布选择和后验分布形式，利用试验数据确定先验分布超参数，再根据先验信息求解后验分布。仿真和试验结果表明：新方法能够改进兰利法本身对方差精度的影响，使方差精度提高20%；减小了估计引信解除保险距离95%点、5%点的误差，将误差控制在5%以内；能够对试验条件造成试验结果不一致的情况进行修正，获得了修正均值和方差的关系；贝叶斯方法处理兰利法测试引信解除保险距离试验数据能够修正兰利法的方差偏差，也能减弱试验条件对结果的影响，在引信解除保险距离试验中具有应用价值。

Data Processing of Test of Fuze Arming Distance Measured Using Langlie Method Based on Bayesian Method

For the contradiction between smaller sample size and large evaluation error in measuring the fuze arming distance by Langlie method, Bayesian method is used to process the data of fuze arming distance measured by Langlie method. A Bayesian mathematical model is established, and the prior distribution choice and the posterior distribution form are clarified. The prior data hyperparameters are determined from the experimental data, and the posterior distribution is solved from the prior information. The simulated and experimental results show that the influence of the Langlie method on accuracy of variance can be improved, and the accuracy of variance can be increased by 20%. The error of 95% and 5% points of the fuze arming distance that can be reduced is controlled within 5%. The inconsistent test results that are caused by test conditions are corrected to reduce the influence of condition errors on the mean and variance, and obtain the relationship between modified mean and variance. Bayesian method can modify the variance of Langlie method, and also weaken the influence of the test conditions on the results.