| Bridge technical condition assessment plays an important role in bridge operation and maintenance.However,the current bridge technical condition assessment methods are affected by human subjectivity and experience.Machine learning can avoid such problems.Therefore,using machine learning algorithms to analyze bridge technical condition Evaluation has become a trend to solve this problem.In recent years,support vector machine,BP neural network and other single algorithms have been widely used in the research of bridge technical status assessment,but most of them have obvious shortcomings.Therefore,a Stacking fusion model based on logistic regression,random forest,and XGBoost is proposed in this paper,which effectively improves the performance of the technical condition assessment model for concrete beam bridges.The main work carried out in this paper and the research results obtained are as follows:(1)Based on the analysis of relevant literature,bridge inspection reports and the current technical condition assessment specifications of Bridges,six typical diseases such as main beam cracks,pier stripping and exposed reinforcement,expansion joint device damage were selected to build an evaluation index system for machine learning algorithm,and the collected data of 260 concrete beam Bridges were preprocessed.A sample database of machine learning algorithm model is established.(2)Using Python programming software,three kinds of technical condition evaluation models based on single algorithm,namely logistic regression evaluation model,random forest evaluation model and XGBoost evaluation model,are established respectively.The accuracy rate,precision rate,recall rate and F1 score were used to evaluate each model,so as to optimize the parameters of each model.Finally,all the optimized models can evaluate the technical condition of concrete girder Bridges well,and the performance of each model is ranked as XGBoost> Random Forest > logistic regression,but there is still a lot of room for improvement in their performance.(3)The technical status assessment model of concrete beam bridges based on Stacking fusion algorithm is constructed.Three single-algorithm models and the Stacking fusion model are compared for 10 times of experimental analysis.The average values of four evaluation indicators of the Stacking fusion model are the largest,the fluctuations are the smallest,and the performance is more stable.And the relative error between the evaluation result of the Stacking fusion model and the actual value is within the accuracy range of ±4%,which meets the actual engineering needs.(4)The FB bridge and XH bridge were used to verify the real Bridges.The evaluation results of multiple experts,single algorithm model and Stacking fusion model were compared.The evaluation results of Stacking fusion model are closer to the expert results,and the relative error of the evaluation results of the two bridges is within 2%.The results show that the technical condition evaluation model of concrete beam bridge based on Stacking fusion algorithm can deal with complex bridge conditions more effectively than the single algorithm model,and has higher accuracy and generalization ability,which can effectively reduce the working errors of ordinary bridge inspectors. |
