Bridge deck elevation fitting method based on Bayesian-Kriging model

Abstract

The invention relates to the technical field of bridge deck elevation measurement in bridge engineering, and discloses a bridge deck elevation fitting method based on a Bayesian-Kriging model. The method comprises the steps that S1 a Bayesian-Kriging fitting model is established; S2 test optimization of an elevation test point sample is carried out; S3 a Bayesian-Kriging prediction model is established; and S4 elevation fitting evaluation is carried out on the whole bridge deck. The model has the advantages that Bayes and Kriging are combined: based on non-information super-priority, multi-layer priori constraints are imposed on the basis function coefficients and related parameters of a Kriging model; an EM algorithm is used for the solving of the basis function coefficients and the maximum posteriori estimation of the related parameters; the Kriging model is improved, and the Bayesian-Kriging model is established; the adaptability and robustness of the model are enhanced; and the test optimization design of an elevation measurement sample based on a number theory method is carried out.

Description

technical field [0001] The invention relates to the technical field of bridge deck elevation measurement in bridge engineering, in particular to a bridge deck elevation fitting method based on a Bayesian-Kriging model. Background technique [0002] During the use of the bridge, the elevation of the bridge deck will change (settlement or upward deflection) under the action of internal factors (such as shrinkage and creep, material aging, etc.) and external factors (such as traffic load, temperature effect, etc.). When detecting the deflection or elevation of the bridge deck, it is impossible to measure every point on the entire bridge deck, and only a small number of local points (such as mid-span, 4-span, 8-span, etc.) can be selected for measurement. In fact, due to the complex influence of internal and external factors of the bridge, the elevation distribution of the bridge deck is unevenly distributed, and those local elevation point data information cannot effectively de...

AI Technical Summary

Problems solved by technology

In addition, there is an important problem with these models that the structural forms of these methods are relatively rigid and not self-adaptive. What kind of parameter expressions should be selected for different problems, and different parameter expressions have great influence on the fitting calculation results. How much of an impact is unclear

[0042] 2. The nearest neighbor point method and the spline function interpolation method are essentially local interpolation fitting techniques. Among them, the nearest neighbor point method is greatly affected by the sample points (existing test points), and only considers the distance factor, and does not affect other spatial factors. and certain regularities inherent in variables without much consideration

They also have the same problems as the above methods: the form is relatively rigid and not adaptive, and it is impossible to effectively know the influence of parameters on the fitting model

[0043] 3. The BP neural network method belongs to the category of machine learning, but the neural network does not have a strict mathematical foundation, and there are the following problems: it needs to pre-set the structure of the neural network or continuously explore during the training process, so that this method is not suitable for " Excessive reliance on the user's prior knowledge and experience leads to difficult structure selection; BP neural network may fall into local minima, resulting in local extremum problems; BP neural network has a large demand for training data and requires more test data However, in the actual bridge deck fitting test, the data obtained from the test is relatively limited, and the problem of "under-learning" is likely to occur at this time

Images