Deformation Prediction Analysis Of Cross Sea Bridge Based On Particle Swarm Optimization And Multi Combination Model

The deformation of the sea crossing bridge is influenced by many factors,such as wind force and wave force.In the deformation monitoring work,we should not only monitor the deformation of this kind of bridge in real time,but also study the possible deformation of the bridge.Therefore,this paper selects the deformation point of a certain cross sea bridge to carry out the deformation prediction analysis.The background of the bridge and the data collection work on the deformation of the bridge are introduced.The horizontal and vertical control network are set up to monitor the horizontal and settlement deformation,and the deformation time series data of the bridge are obtained.The settlement sequence of QSC30 points is selected to study the deformation analysis.By means of wavelet denoising,this paper selects the rigrsure function,hard threshold,the rule of adjusting the threshold value according to the noise level estimated by the first layer,decomposes the signal in two layers,and uses Sym7 as the base function to denoise,and then predict.GM(1,1)is used to analyze the sequence,and through the study of GM(1,1),an improved GM(1,1)is proposed.By comparing the prediction results of the model with GM(1,1),it is concluded that the prediction accuracy of the improved model is greatly improved.As the selection of the background value of GM(1,1)is through the way of taking the average value,this is a certain course.The error of the model is increased in degree,so the background value of the grey prediction model should be re selected.The particle swarm optimization algorithm is a optimization algorithm.The algorithm is used to optimize the background value and get the optimal background value,thus improving the precision of the modelBP neural network algorithm is a very strong learning prediction algorithm.The selection of initial weights and thresholds plays a very important role in the quality of prediction.Therefore,the particle swarm optimization algorithm is used to optimize the initial weights and thresholds and improve the prediction accuracy of the BP neural network algorithm.Finally,by comparing the predicted values of the measured settlement time series and the four prediction models,it is concluded that the BP neural network model improved by the PSO algorithm has a higher prediction accuracy.