Multi-objective Optimization Of Temporary Pier Spacing In Steel Box Girder Jacking Construction Based On ACO

Due to the transport economy’s objective demands,the main steel girders of large-span bridges in urban areas tend to develop extra long.Therefore;temporary piers become an essential auxiliary structure in girder jacking.In the current situation,if the spacing of temporary piers is still determined by previous construction experience,there may be safety hazards and problems,such as high construction costs.Therefore,it is necessary to optimize the spacing of temporary piers according to the characteristics of the current structure.When using conventional optimization algorithms to solve optimization problems based on finite element analysis in engineering structures,they often face the problem of excessive iterative calculations that cannot be calculated or do not converge.At this time,the joint intelligent optimization algorithm to simplify the finite element analysis process is an effective way to solve the problem.This paper uses a span cable-stayed bridge in Ningbo with a total span of 197.2m as a background to transform the temporary pier spacing arrangement problem in steel box girder jacking construction into a multi-objective optimization problem.The following work is mainly accomplished:(1)The steel box girder jacking construction principle and the commonly used jacking construction methods are described.On the one hand,relevant information from home and abroad is collected from sorting out the use and spacing arrangement of temporary piers during steel box girder jacking construction and summarising the current situation of applying intelligent optimization algorithms in bridge construction on the other.(2)The main optimization algorithm is selected according to the commonly used intelligent optimization algorithms combined with the characteristics of the optimisation problem in this paper.A brief discussion of the jacking construction under the original design is given,and a full bridge model is established.By carrying out force analysis on the jacking construction process under the original construction design,a calculation table for the whole jacking process is obtained.Then the basis for selecting optimization variables,state variables,and objective functions is explained,and the maximum jacking span and the corresponding guide beam lengths are determined in conjunction with the optimization range of the maximum jacking span.(3)The penalty parameter C and the kernel parameter σ of support vector regression(SVR)were optimized using the ant colony algorithm(ACO)to improve the prediction accuracy,and four mechanical index prediction models were developed after training the samples.(4)The joint ACO-SVR algorithm and ACO are used to solve the optimization problem in this paper.The optimal solution set is obtained by comparing the solutions in the solution set with different weight ratios of the two objectives to obtain the final optimization solution.In addition,before and after optimization and multi-objective and single-objective optimization are compared separately to verify the superiority and reliability of multi-objective optimization.At the same time,the design control parameters of the optimization problem(different total lengths of the jacking span)are analyzed to investigate the effect of different total jacking spans on the multi-objective optimization results and to summarise the control factors of the optimization results.