Study On The Prediction And Maintenance Decision Of Urban Bridges In Coastal Area

In recent years,with the progress of urbanization in China,the demand for road traffic has changed,and a series of problems have arisen in the original bridges with low construction standards and untimely maintenance.In this context,it has become an important and urgent task to study the bridge condition prediction and safety maintenance decision-making for public safety and saving public funds.In the coastal area with developed water system and large number of bridges,the bridge diseases show unique characteristics,while the bridge maintenance management mode is random and blind,lacking systematic,preventive and targeted.Therefore,because of the characteristics of the region,it is important to build a bridge degradation prediction and maintenance decision-making model in line with the coastal area.On the basis of the Wenzhou urban bridge maintenance medium and long term planning(2018-2028)project,this paper roundly discusses the characteristics,types and regional distribution of bridge diseases from the actual situation of bridges in this area,and establishes models based on the degradation prediction and maintenance decision-making of urban bridges in coastal areas.The research content and innovation achievements of this paper are as follows:1.This paper analyzes the typical diseases of urban bridges in Wenzhou area,and introduces the formation mechanism of various diseases in detail.The characteristics,types and regional distribution of bridge diseases are systematically discussed to provide modeling indexes for subsequent model establishment;2.Prediction of urban bridge degradation in coastal area.In order to more accurately control and maintain the decision-making period,the following research is focused on the urban bridge degradation prediction plate: firstly,the methods of building the bridge degradation model framework are discussed from durability experiment,Grey Prediction GM model and BP neural network respectively,and the merit and defect of the above three methods are described,which provide reference and suggestions for the follow-up researchers.Secondly,the linear,quadratic,exponential and Fourier regression models are established by combining the bridge degradation with the widely used regression prediction models.Finally,the application of Markov model in bridge condition prediction is introduced in detail.In order to make the result of Markov prediction based on bridge degradation converge,this paper introduces degradation coefficient through formula derivation,so that the state transition matrix is constantly modified in each time period,and a time-dependent state transition matrix is obtained.When different degenerate elements are selected from the target space,the system corresponding to different elements can be more accurately predicted;3.Group level multi-attribute decision-making model of bridge maintenance.Traffic infrastructure is deeply rooted in the society,which is not only subject to technical conditions,but also must meet the characteristics of social and economic development requirements.In order to manage group level bridges and determine the maintenance priority of each bridge,based on the conventional evaluation model,this paper puts the Social Efficiency(SOE)and Structural Efficiency(STE)that affect bridge maintenance into the decision-making scope.Based on fuzzy membership,AHP and penalty variable weight function,this paper proposes a bridge maintenance decision-making model considering group level multi-attribute.Through the case analysis of the group bridge in the coastal area,the fuzzy membership function and penalty variable weight analysis are applied to the indexes of each layer which are difficult to be quantified,and the maintenance priority sequence of each bridge is introduced,which shows the practicability of penalty variable weight model in the maintenance decision of the group bridge.