Structural reliability analysis based on neural network and finite element method

A new approach for structural reliability analysis with implicit performance function is presented by combining the neural network, finite element method (FEM) and first-order reliability method (FORM). BP network is applied to approximate the implicit performance function, which is more flexible and adaptive than the polynomial function used in response surface method. The first derivatives of the performance function with respect to random variables can be obtained by the successful trained BP network, which plays an important role in calculating the reliability index. The training samples come from the numerical results of FEM. Two examples are given to demonstrate the validity of the method. The method can be applied to the structural reliability analysis with implicit performance function.