论文部分内容阅读
因试验成本原因,仿真试验将成为一个重要的先验信息来源。通过对正态分布进行研究,本文提出以一个分布函数用于描述仿真试验能力。在对仿真试验能力量化描述的基础上,推导出了成败型仿真试验验后分布的计算方法,可用于有限次和无限次仿真试验验后概率分布计算,该方法有效地解决了仿真大样本淹没现场小样本核心问题。经分析,多种试验手段之间为相互独立关系,试验结果应与试验手段实施先后无关,并根据贝叶斯理论,推导出了多源数据融合算法。
Due to the cost of testing, simulation tests will become an important source of prior information. By studying the normal distribution, this paper proposes a distribution function to describe the simulation test capability. Based on the quantitative description of the simulation test ability, the calculation method of the post-test distribution of the success or failure simulation test is deduced, which can be used to calculate the probability distribution after the finite and infinite simulation test. This method effectively solves the problem of large sample flooding On-site small sample core issues. After analysis, a variety of test means are independent of each other, the test results should have nothing to do with the experimental means, and according to Bayesian theory, derived multi-source data fusion algorithm.