Research On The Evaluation Of Asphalt Pavement Performance Index Based On Big Data Analysis

A scientific and reasonable index system is a prerequisite for performance evaluation and prediction of asphalt pavement.However,the performance index of asphalt pavement in China,especially the diseases condition index(PCI),has a poor practical application effect,which seriously affects the performance evaluation,prediction and maintenance of asphalt pavement.The combination phenomenon of different disease types of asphalt pavement in different use environments(called the disease symbiosis phenomenon in this paper)leads to a great difference in the degree and speed of deterioration of asphalt pavement performance.Due to the limitations of traditional analysis methods and data range,the rule exploration of disease symbiosis in different use environments cannot be realized.Therefore,the research and improvement of the former indicator system are usually carried out for a single disease type and a specific use environment.In order to better reflect the actual use of the pavement performance of the situation,this article is based on the long-term pavement performance(LTPP)data,using big data association rules analysis method,the disease of asphalt pavement under different climate area symbiosis of rule mining,get different diseases of asphalt pavement using environment symbiotic model,and symbiosis model based on different diseases of asphalt pavement using index to evaluate research and the analysis of defects.First,the long-term performance database of asphalt pavement at home and abroad was investigated,and LTPP database was selected as the data source for mining the symbiosis of asphalt pavement diseases.LTPP data were extracted to reconstruct the asphalt pavement performance database,and the data characteristics of asphalt pavement disease data were analyzed.The results showed that the asphalt pavement disease data was a large sparse matrix.Association rule analysis in the big data method is used as the main research method of disease data mining,and the data requirements and algorithm flow of association rules are clarified.Then,based on the data requirements of association rule analysis,the original data of asphalt pavement are distributed and discretized.The LTPP test section was divided into 6 types of different climatic regions by means of factor-cluster analysis.Then,based on the discretization data of pavement diseases in different climate regions,the association rule big data analysis method was used to dig and analyze the symbiosis rules of asphalt pavement diseases in different climate regions,and six types of symbiosis models of asphalt pavement diseases in different climate regions were obtained.Finally,pavement disease data were extracted based on the symbiotic model of asphalt pavement disease in different regions,and PCI scores of pavement distress indicators were calculated according to the index calculation method,and the differences of PCI scores of asphalt pavement distress indicators in different symbiotic models were tested between groups.The results showed that PCI could not effectively reflect the disease symbiosis model of asphalt pavement,especially the disease symbiosis model dominated by transverse cracks.PCI is strongly influenced by area statistics and is not sensitive to diseases type and severity;The main diseases of asphalt pavement are covered by the common statistics and weighting calculation method of each disease.Data analysis in this paper from the LTPP long-term pavement performance database,LTPP contains rich combination of action,structure,basic covers the common load level and structure form,the symbiotic mode difference in the types of these diseases,and therefore based on the LTPP data to establish the index system of evaluation method has certain reference significance and defect analysis,based on the long-term pavement performance data of index system of evaluation and improvement needs further research.