Abstract:

The response surface method based on the radial basis function neural nelwork(RBF-NN) was adopted to analyze the fatigue reliability of a loader boom.When the limit state function of structure is complex or highly nonlinear,it is time-consuming or cumbersome to carry out reliability analysis with traditional methods.The RBF-NNbased response surface method provides an effective means for the reliability analysis of the large scale structures.The working process of a loader machine was analyzed with the analytical approach combining the finite element method software ANSYS,the stress and strain histories of the loader boom were calculated by rainflow algorithm,consequently the fatigue life of the structure could be derived by the local stressstrain analysis.The RBF-NN based response method was used to fit the limit state function of the structure as well as its derivatives,so as to calculate the reliability of the structure.The examplified numerical results showed the proposed method is capable of solving the industrial-scale fatigue reliability problems.

Key words: mechanical design, radial basis function, artificial neural network, fatigue reliability, response method, rainflow algorithm