| For large-scale structures,high safety factor is generally adopted in design,so its reliability is high.At this time,Monte Carlo method is more expensive to calculate its reliability index.And the failure equation of large-scale structure is usually implicit,so it is difficult to accurately fit the limit state equation by traditional response surface method.In view of the above problems,this paper proposes a structural reliability method based on zero residual fit and line sampling.By dividing the total space into several subspaces,using multi response surface to realize zero residual fitting of sample points,the accurate limit state equation can be obtained,and then the fitted failure equation can be rotated to a new coordinate system by using the coordinate system rotation technology,and then the reliability analysis of the failure equation can be carried out by using the line sampling method.The main contents are as follows:(1)Firstly,the zero residual fitting method and the line sampling method based on coordinate rotation proposed in this paper are described in theory,and the realization process is given.(2)Through fitting and comparing the failure equations of the existing literature,and calculating the failure probability of the examples of the existing literature,and combining Monte Carlo,subset simulation and other methods to verify the accuracy and efficiency of this method.The results show that this method has high accuracy in fitting failure equation and calculating failure probability.(3)Using the reliability analysis method proposed in this paper,the structural failure equations of the long-span cable steel truss structure and the long-span suspension bridge structure with high-dimensional random variables are fitted respectively,and the line sampling method based on coordinate rotation is used for sampling calculation,and then their system failure probability is obtained,which verifies the application of the failure equation fitting method and reliability calculation method in this paper The feasibility of reliability analysis of large structures. |