基于神经网络和改进JC法的预应力混凝土连续梁桥可靠度研究

Reliability Study of Pre-stressed Concrete Continuous Girder Bridges Based on Neural Network and Improved JC Algorithms

由于预应力混凝土连续梁桥极限状态功能函数不能显式表达,导致其可靠度求解困难。本文将BP神经网络与改进JC法引入预应力混凝土连续梁桥可靠度分析领域,首先利用BP神经网络对结构功能函数进行拟合,将高度非线性的极限状态方程显式化,然后采用改进JC法全局搜索验算点并求解可靠度指标,并采用算例来验证该方法的精度和效率。然后以一座预应力混凝土连续梁桥为例,计算结构的可靠指标。计算分析结果表明,BP神经网络和改进JC法能很好地应用于预应力混凝土连续梁桥的可靠度分析,能够弥补传统可靠度分析方法的不足,并且满足工程精度要求。

It is difficult to solve the reliability of the pre-stressed concrete continuous girder bridges because of non-explicit limit state function. The BP neural network and improved JC algorithms are applied into analyzing the reliability of prestressed concrete continuous girder bridges. Firstly,using BP neural network to fit the structural performance function,making the highly nonlinear limit state equation show and then the improved JC method was used to globally search checking points and calculate the reliability index,and the efficiency and accuracy of the method are verified by numerical examples.And a pre-stressed concrete continuous girder bridge reliability index were calculated. The calculation and analysis results showed that the BP neural network and improved JC algorithms compensated the deficiency of the traditional reliability analysis methods,improved the calculation accuracy,provided a new thought and means for the research on the reliability of bridge structure,and well applied to the reliability analysis of pre-stressed concrete continuous girder bridges.