Precamber Prediction Method of Continuous Rigid Frame Bridge Based on Heterogeneous Integrated Model
The precamber plays an important role in the monitoring of the construction profile of long-span cantilever bridge,improving the prediction accuracy can ensure that the profile of the construction stage and the finished state of the bridge meets the design requirements as much as possible.In order to obtain better prediction performance,a SSA-BPNN-RF(sparrow search algorithm-back propagation neural network-random forest)heterogeneous integration model based on adaptive set weighting is proposed in this paper.In order to verify the feasibility of this model,the trained model was applied to predict the precamber of a continuous rigid frame bridge in Hunan province,and compared with five prediction models,namely BPNN,RF,BPNN-RF,SSA-BPNN and SSA-RF.The results show that the SSA-BPNN-RF heterogeneous integrated model has the best performance on the average absolute error,root mean square error and fit degree.In addition,BPNN-RF integration and SSA have positive effects on BPNN and RF respectively,further verifying the effectiveness of heterogeneous integration.Therefore,SSA-BPNN-RF heterogeneous integration model has high precision and better adaptability,which has important guiding significance in engineering practice.
万方数据
维普
