| According to the statistical bulletin of the development of the transportation industry,the rapid growth of the number of Bridges has made our country a clear bridge country.However,with the aging of Bridges,the decline of material properties and the continuous development of structural damage and other factors,a large number of bridge diseases become increasingly prominent,leading to bridge maintenance work becomes more and more onerous and urgent.At present,the safe operation status of Bridges can be assessed by carrying out regular technical condition assessment of Bridges.However,most bridge inspection is one bridge,one inspection and one evaluation,and there is no analysis and research on the technical state degradation of the bridge and the next step of inspection maintenance based on a large number of inspection data.Therefore,in order to ensure the operation safety of the bridge,it is necessary to establish a bridge degradation prediction model and a bridge maintenance decision optimization model.Aiming at the above problems,the research content of this paper is as follows:(1)Bridge information statistical analysis and bridge database construction.In order to grasp the overall operation status of the bridge,the bridge inspection data are analyzed statistically from three aspects: basic information,technical status rating and bridge disease.Based on different stages of bridge life cycle,bridge inspection data is divided into bridge basic information module,bridge detection information module and bridge maintenance information module.The bridge database is constructed on Navicat Prem-ium15 database visualization software based on My SQL database language.(2)Build a bridge rating degradation prediction model based on BP neural network.Firstly,the input characteristics of the bridge technical status rating degradation prediction model are obtained based on knowledge and statistics.Secondly,the model input feature data conversion and feature matrix dimension reduction were carried out to determine the input of BP neural network,and a BP neural network bridge rating prediction model containing three hidden layers was established.(3)Build a bridge maintenance decision optimization model based on AHP and CRITIC weight method.Firstly,the maintenance effect of the bridge was quantified from the four perspectives of bridge technical status rating,bridge age,road network importance and design load,and the expression equation of bridge maintenance effect was constructed.The importance score of the bridge to be maintained was obtained by combining the analytic hierarchy process(AHP)and CRITIC weight method.The maintenance fund constraint equation was solved by discrete particle swarm optimization algorithm,and the maintenance decision optimization model was obtained. |
