Analysis On Temperature Prediction Method Of Asphalt Pavements Considering Continuous Temperature Changes

During road service,under the influence of meteorological factors changing continuously over time,the temperature field with continuous changes in time and space will be formed inside asphalt pavement.As the asphalt mixture is a temperature-sensitive material,the asphalt pavement performance will be affected by continuously changing temperature.As the influence of temperature factors is not reasonably considered in the design of pavement structure and the selection of pavement materials,a large number of temperature-related diseases are generated in the actual use of roads.In order to better solve the problem of pavement disease caused by the change of pavement temperature,it is necessary to study the characteristics of pavement temperature distribution and its estimation method.On the basis of measured pavement temperature data,this dissertation improves the prediction method of asphalt pavement temperature from two aspects: statistical analysis method involving environmental factors and numerical analysis method involving pavement material parameters.The main research contents and results are as follows:(1)A large number of measured temperature data and meteorological data were collected in the in-site asphalt pavement temperature monitoring program.According to the collected meteorological data and the definitions of high temperature and low temperature weather in meteorology,the annual temperature was divided into three temperature periods: high temperature period,normal temperature period and low temperature period.The meteorological data and measured temperature data of representative months in different temperature periods were analyzed,and the meteorological variation characteristics and temperature distribution law of asphalt pavement in different temperature periods were obtained.The results show that there are obvious similarities and differences between meteorological characteristics and temperature distribution of asphalt pavement in different temperature periods.(2)Based on the measured data,the applicability of the temperature prediction model established by the traditional statistical analysis method was studied,and the advantages and disadvantages of the prediction model were analyzed.The temperature field prediction model of asphalt pavement was established by analyzing the measured temperature data and meteorological data using neural network method.The results show that compared with the traditional statistical analysis method,the neural network model can better consider the meteorological factors with complex correlation,and the model prediction effect is better,and the BP neural network model has a better evaluation index when it is used to predict the temperature in the high temperature period and the shallow position of the asphalt pavement.The long short term memory neural network model(LSTM)has a good prediction effect on the high temperature period and the low temperature period for each depth of the asphalt pavement.It shows that the neural network model is suitable for predicting the temperature field of asphalt pavement.(3)A numerical analysis model was established based on heat transfer theory and finite element method.The calculated results of the numerical analysis model were verified by measured data,then the pavement material parameters were back-analyzed by genetic algorithm.The results show that the numerical analysis model has a good prediction effect on the temperature at all depths of asphalt pavement in fine weather during all temperature periods of the year,and the radiation absorption rate and thermal conductivity of pavement surface have a significant effect on the temperature field of asphalt pavement.By substituting the material parameters of asphalt pavement obtained by back analysis based on the measured data into the numerical analysis model,the asphalt pavement temperature in different temperature periods can be calculated more accurately.