基于贝叶斯统计的驱动桥壳可靠性设计

描述了基于贝叶斯统计的汽车驱动桥壳可靠性设计，回顾并分析了驱动桥壳的受力及传统可靠性设计。在传统方法的基础上，为了实现给定可靠度求出驱动桥壳内径，对贝叶斯统计后验期望方法进行了简化整理，运用符号换元积分和Newton—Raphson迭代法得出了数值解(收敛速度极快)，并通过例子予以验证。本方法克服了传统方法的保守性，使设计更合理。本方法特别适于经典统计方法不适应的个性化和特殊化设计，对设计方案的频繁变更及贫乏的实验和现场数据都能得出比传统方法更精确的解。

Reliable design of vehicular rear axle on the base of bayes statistics

A kind of reliable design of vehicular rear axle on the base of bayes statistics was proposed and the force of rear axle was given and reliability of bayes was analyzed . For achieving diamond of rear axle by given reliable value, post-check expectation on the base of bayes statistics was simplified and ordered. Using the integral of changing elements with symbols and methods of Newton , we can get more accumulative outcome of diamond of rear axle and reliability than traditional methods. This method is very useful for personal and special character design on which traditional statistics method is out of fitness. Bayes method can get more accumulate outcome than traditional method in the condition of changing design and less test-data.