Random parameter structure reliability evaluation method based on point collocation algorithm

Abstract

The invention provides a random parameter structure reliability evaluation method based on a point collocation algorithm and relates to the technical field of engineering structure reliability analysis. The method includes steps: for an engineering structure design problem having multiple independent random variables, building a structure performance function containing random parameters; building an equivalent weak form of integration of the structure performance function to acquire a unitary substitution equation of a real structure performance function; determining structure performance function statistics including average value, variance and similar expressions of high-order original moment and central moment; performing point collocation in a random variable definition domain, and determining a point collocation scheme; determining structure failure probability to evaluate structure reliability. The method takes balance between calculation accuracy and calculation efficiency of the engineering structure reliability problem into comprehensive consideration, and the structure failure probability is given out on the basis of a unitary equivalent integration weak form of the structure performance function and a point collocation method, so that model errors caused by normal assumption of random response are reduced.

Description

technical field [0001] The invention relates to the technical field of engineering structure reliability analysis, in particular to a random parameter structure reliability evaluation method based on collocation algorithm. Background technique [0002] Engineering structure design needs to ensure that the structure completes the predetermined function within the specified time and under the given load. However, there are often a large number of uncertainties in the design process and actual operating environment, which can be roughly divided into physical uncertainties, statistical uncertainties and model uncertainties. Physical uncertainties are caused by variability in parameters such as loads, materials, geometry, and boundary conditions. Statistical uncertainty is caused by the estimated deviation of the distribution parameters due to insufficient sample size, and is often caused by the deviation between the true distribution form and the estimated distribution form. M...

AI Technical Summary

Problems solved by technology

The theoretical basis of the Monte Carlo simulation method is the large number theorem in probability theory, which has the advantage of not being limited by the scope of application, but its disadvantage is that it needs to consume a huge calculation cost. Although various improved sampling methods have appeared, their It is still rarely used in actual engineering calculations, and is generally used to test the accuracy of some newly proposed methods

Therefore, in the current engineering structure reliability analysis methods, there is a lack of a method with high calculation accuracy and affordable calculation amount.

Images