The Statistical Research Of The Bridge Vehicle Loads Probability Model Based On WIM Data

Nowadays,with the increasing number of aging bridges in the world and the increasingly serious phenomenon of overloading,the evaluation of bridges has become a very important topic in the bridge research.Therefore,making analysis of the current situation about vehicle flow and establishing a practical vehicle load model are of great significance to the follow-up research of the reliability analysis of the bridges.Based on data of WIM(Weigh-in-Motion System)of expressway bridge,the present situation of vehicle load is analyzed.Based on that,the probability model of vehicle load is studied.The thesis mainly includes four parts.Firstly,the research background and significance,research status of vehicle load are outlinedSecondly,the relevant theories and methods are introduced.The main contents include the common probability distribution of vehicle load,the parameter estimation and non-parametric estimation of probability density function,and the inspection and evaluation of the model.Thirdly,the present situation of vehicle load in Nanxi Yangtze River Bridge is analyzed.On the basis of introducing the WIM system,data collection and pretreatment,the current situation of vehicle load is analyzed from five aspects of vehicle types,lanes,vehicle speeds,vehicle weights and vehicle distancesFinally,different estimation methods are used to estimate the probability density function of weight variable,and they are compared with each other based on Cross Validation.Three estimation methods of NLSF,MLE and Kernel Estimation are applied to estimate the probability density function of vehicle load.Cross validation is used for verification and comparison.The results shows that:on the basis of Cross Validation,the fitting results of the three methods are great,and they can be applied to vehicle load data fitting.In the specific application,the estimation method should be chosen according to the principle of simplicity and convenience.