基于BP神经网络的腐蚀管道可靠性分析

利用BP人工神经网络算法建立基于BP神经网络腐蚀管道失效预测模型。通过BP神经网络拟合极限状态方程,借助神经网络的函数映射关系产生大量的极限状态函数值,作为下一步的分析数据。采用蒙特卡洛法随机抽样的思路,对大范围的数据进行概率分析,通过概率分析得到极限状态函数值的均值和标准差,求得腐蚀管道可靠性指标,解决了腐蚀管道的可靠性分析问题。

Reliability Analysis of Corroded Pipelines Based Upon BP Neural Network

BP artificial neural network algorithm was applied to establish the failure prediction model of corroded pipelines based on BP artificial neural network. With artificial neural network fitting limit state equation, a large amount of limit state function values generated from artificial neural network function mapping relationship are used as analytical data. The random sampling by Monte Carol is applied, and the probability of a wide range of statistic data is analyzed. The reliability index of corroded pipelines is obtained through the probability analysis to get mean value and standard deviation of the state functions of limit, which has solved the reliability analysis problem for corroded pipelines.