Uncertainty Finite Element Model Updating Of Long Span Suspension Bridge

Finite element model updating is divided into deterministic updating and uncertainty updating,in which uncertainty updating takes into account the uncertainties of parameters and response,so it is widely used in suspension bridge finite element model updating.Suspension bridges are the most flexible in long-span complex bridge structures,with many structural parameters,and small changes in parameters are likely to cause large changes in the internal forces and dynamic characteristics of the structure,making the structural analysis of suspension bridges difficult to calculate,especially in the updating of suspension bridge finite element models protruding.Therefore,how to improve the efficiency and accuracy of the sampling,regression and optimization process of the uncertainty finite element model updating of suspension bridge is worthy studying deeply at this stage..The uncertainty updating of the traditional suspension bridge finite element model cannot reflect the influence of the parameters to be corrected on the state of the bridge,and there are problems of large calculation and low solution efficiency,and the updating needs to focus on the construction of the alternative model and the solution of the objective function.Three alternative optimization models are constructed by uniform design and Latin hypercube sampling.Three intelligent optimization algorithms,bacterial foraging algorithm,genetic algorithm,and particle swarm algorithm,solve the objective function to find the optimal solution efficiently and quickly.In view of the uncertain updating of the finite element model of the suspension bridge,the following aspects of research are mainly carried out:(1)Introduce the precise solution method of the main cable forming bridge shape and three intelligent optimization algorithms,and combine the two to form the solution scheme of the intelligent optimization algorithm for the shape finding of the segmented catenary main cable.Taking the cable with a span of 660 m as an example,the bridge completion status test is carried out.The results show that the solution result of this scheme satisfies the bridge completion status.(2)Summarizing the uncertain correction process of the long-span suspension bridge and the difference with the general bridge type,the uniform design and Latin hypercube sampling method that belong to the U-shaped design are analyzed and introduced,and three intelligent optimization algorithms are used in the calculation example.The results show that the mean square error of the samples constructed by the two experimental design methods can reach the order of 10-5,which can well replace the original model;the three intelligent optimization algorithms can meet the requirements in terms of solving accuracy,and the range of the obtained interval is in line with the real Interval.(3)Taking a large-span suspension bridge as the research object,the finite element model is uncertainly updated according to the measured data collected by the health monitoring system.Two test design methods are used to construct the sample,and two bridge state inspection methods are used to check the bridge state corresponding to the sample points.The test results show that all sample points meet the bridge state,and both test design methods can be applied to uncertainty finite element model updating of long span suspension bridge;in addition,the three intelligent optimization algorithms are applied to solve the objective function,and the corresponding parameter range has also done well.