论文部分内容阅读
研究了多项式基函数神经网络法的结构可靠性计算.当结构的极限状态函数复杂,非线性程度较高,功能函数为隐式时,传统的结构可靠度分析方法计算困难,多项式基函数神经网络法为解决结构可靠性分析提供了一种新方法.基于多项式逼近理论,利用神经网络模拟逼近能力,将多输入多项式作为网络的激励函数,利用激励函数的广义逆矩阵形式计算网络隐层与输出层的连接权值,拟合结构的功能函数.利用可靠度的一阶可靠性方法计算结构的失效概率.通过实例计算,表明了本方法计算精度高,同时公式简单,易于编程,具有通用普遍性.
The structural reliability calculation of polynomial basis function neural network is studied.When the limit state function of the structure is complex, the degree of nonlinearity is high, and the function function is implicit, the traditional structural reliability analysis method is difficult to calculate. The polynomial basis function neural network The method provides a new method to solve the structural reliability analysis.Based on the polynomial approximation theory, the neural network is used to simulate the approximation ability, the multi-input polynomial is used as the excitation function of the network, and the hidden layer and output of the network are calculated by using the generalized inverse matrix of the excitation function The connection weight of layer and the function of fitting structure.The probability of failure of the structure is calculated by the first-order reliability method of reliability.Through the example calculation, it shows that the method has the advantages of high calculation precision, simple formula and easy programming, Sex.