Research On Evaluation And Predication Of Asphalt Pavement Roughness

In China’s high-grade highway construction,asphalt pavement has become the main form of pavement structure because of its excellent performance.Therefore,ensuring and improving the performance of asphalt pavement is not only related to the economic benefits of highway management,but also whether it can be used to provide the public with a stable,comfortable and safe driving environment.In the evaluation of the performance of asphalt pavement,the driving quality,driving safety and comfort of the pavement are all related to the roughness of the pavement,and the roughness of the pavement also reflects the overall condition of the asphalt pavement.However,because the current pavement roughness evaluation indicators and prediction models have some drawbacks,the highway maintenance department cannot accurately grasp the overall situation of the pavement.Therefore,it is of great significance to adopt more effective means to evaluate and predict the pavement smoothness.1.Based on a detailed analysis and summary of the deficiencies of the existing flatness evaluation indicators,a weighted longitudinal proflie evaluation method was proposed to evaluate the asphalt pavement roughness,and the weighted longitudinal proflie evaluation indicators were calculated and deduced,thus establishing the weighted longitudinal proflie evaluation indicators Mathematical model;the weighted proflie evaluation criteria and index evaluation intervals for roughness were determined;the scientific and rationality of the weighted profile evaluation method was verified by evaluating the actual project roughness.2.Based on the theoretical research on the influencing factors of asphalt pavement roughness,combined with the actual situation of the project,the most important influencing factors of asphalt pavement roughness were determined,respectively:pavement disease,traffic load,pavement structural factors;further The variables involved were analyzed by variable clustering,and finally the damage rate DR,the pavement structure strength index PSSI,the annual average daily traffic volume AADT,and the time t were used as the input variables of the roughness prediction model.3.Based on the analysis and summary of the shortcomings of the current mainstream roughness prediction model,it is proposed to use the mixed effect model as the roughness prediction model;during the construction of the model,the fixed effect is used to achieve the overall roughness development trend of all road sections,and the random effect is used to Realize the difference of roughness between road sections;the mixed effect model can predict and analyze multi-dimensional data.When analyzing the roughness data type of panel,the time variable is used to reflect the roughness in the longitudinal dimension over time the trend of evolution is to reflect the specific observations of the roughness data at a specific moment on the cross section by introducing covariates in the model.Through verification,the mixed-effect model makes full use of various additional information contained in the roughness data,which makes the prediction of roughness more accurate,and can provide more accurate basis for the maintenance decision of the highway maintenance department.