Based on the finite element method,a low cycle fatigue model,

RBF neural network technique and reliability theory, the reliability sensitivity design for one turbine disc with low cycle fatigue failure were studied in detail.First, the explicit relational expression among the design parameters and the failure life were given in term of RBF neural network. On the basis of the trained neural network,the MPPPM can be used to begin the sensitivity design. Hence,the introduction to RBF neural network solves the explicit expression of the limit state function which is difficult to obtain in engineering, and the method of sensitivity which expresses the influence the parameters posed on the reliability in quantity has a strong significance.

将有限元方法、低周疲劳模型、径向基函数神经网络技术与可靠性理论相结合,给出了涡轮盘低周疲劳可靠性灵敏度设计方法。应用径向基函数神经网络拟合得到随机设计变量与失效寿命之间的函数关系式,根据随机摄动法与可靠性灵敏度技术进行灵敏度设计。径向基函数神经网络技术的引入克服了工程实际无法给出极限状态函数显式的问题,而灵敏度技术使各变量对可靠度的影响得以定量表达,具有很好的指导意义。

可靠性, 灵敏度, 径向基函数神经网络, 最大可能点摄动法, 蒙特卡洛法