Bridge Finite Element Model Modification Based On Quantum Genetic Algorithm And Response Surface

Accurate finite element model is the premise of health monitoring,static and dynamic response reanalysis,structural optimization,damage identification,etc.However,due to construction error,environmental effect,material parameter simplification and idealization of structure boundary,there are some differences between the finite element model based on the design drawings and the actual structure of the bridge,which is difficult to ensure the reliability of structure analysis.Therefore,the initial finite element model needs to be modified.There are many shortcomings in traditional finite element model updating methods,such as inaccurate parameter selection based on sensitivity,too much calculation cost caused by repeated calls of finite element software,etc.In order to reduce the workload of model modification and improve the accuracy of parameter optimization,this paper deeply studies the bridge finite element model modification based on quantum genetic algorithm and response surface method.Firstly,the basic principles and implementation steps of the finite element model modification based on response surface method are described in detail.In order to improve the local nonlinear fitting ability of response surface and ensure the good fitting of all sample points,a new method to improve the response surface is proposed:taking quadratic polynomial as the initial response surface,using the student residual after the initial fitting to find the abnormal points,and introducing Gaussian radial basis interpolation function at the abnormal points to reconstruct the response surface.The response surface fitting results of numerical examples and engineering examples show that the overall fitting effect is improved to a certain extent and the local nonlinear fitting ability is improved obviously,which verifies the correctness of the improved response surface method.Secondly,the basic characteristics of quantum computation and the mathematical principles and algorithm implementation steps of quantum genetic algorithm and double chain quantum genetic algorithm are introduced in detail.In view of the shortcomings of the shiftstep function of double chain quantum genetic algorithm,a method of improving the shiftstep function by using antisinusoidal function is proposed.The test results of the two-dimensional Ackley’s function show that the search accuracy and robustness of the improved double chain quantum genetic algorithm are improved,which verifies the feasibility of the improvement of the shiftstep function.Finally,taking Shending River Bridge as an example,combined with the measured data from the static and dynamic load tests,the initial finite element model is modified step-by-step based on the improved double chain quantum genetic algorithm and improved response surface.After static correction,the maximum deflection error decreased from 15.74% to 1.37%.Using the modified elastic modulus to recalculate the vertical third-order frequency,the frequency error has been significantly reduced,which proves that the modified elastic modulus is an effective structural parameter.After dynamic correction,the maximum frequency error decreased from 8.36% to1.15%.The final results show that the modified model can accurately reflect the actual working conditions of the bridge,and can be used as the benchmark finite element model for subsequent structural analysis,which verifies the feasibility of the finite element model modification based on the quantum genetic algorithm and response surface method.