| In recent years,the main road in China has entered the "peak period" of comprehensive maintenance.However,the maintenance timing is not appropriate due to the lack of objective understanding and scientific evaluation of the road performance.The decision is unreasonable,and many pavement damage occurs in the early stage.The mountain main road is different from other general trunk roads,which is characterized by high temperature and rainy,complex geology and heavy traffic.It is the great significance to establish an evaluation system suitable for the main road in the region,which is of great significance to improve the quality and efficiency of maintenance work.Based on the investigation of Chongqing trunk road,combined with the pavement condition test data and the maintenance data of previous years,the advantages of GA and BP are complementary to each other,which improves the accuracy of calculation,and establishes a model to study the performance evaluation of asphalt pavement.The main conclusions are as followsFirst,the operation of Chongqing trunk road is analyzed.According to the climatic data and geomorphic characteristics of Chongqing in recent years,Chongqing is divided into four areas,I,II,Ⅲ and Ⅳ.According to the statistical analysis of the traffic inspection points of 37 districts and counties in Chongqing on the traffic vehicles of 55 main highways(154 sections),the main highways in each district are generally overrun,mainly with 2-axis,3-axis and 4-axis models;according to the statistical analysis of disease area in 2018,the main diseases of heavy-duty traffic,medium traffic and light traffic trunk roads are mainly cracks.The main diseases traffic in area I are cracks,and the major diseases in area IV are block cracks.Secondly,the evaluation index of the asphalt pavement of the main road in Chongqing is selected.Based on the statistics and causes of traffic accident mortality in urban and rural areas in Chongqing from 2012 to 2018,combined with the current situation of maintenance of Chongqing trunk road and the characteristics of cracks as the main diseases,the paper selects the pavement failure rate,flatness value and crack rate as the performance evaluation index of Chongqing trunk road asphalt pavement.Based on the investigation of the history and maintenance status of asphalt surface of Chongqing trunk road,combined with the road experts with rich experience in the maintenance and management of local pavement engineering,the paper puts forward the subjective evaluation standards of CRI,PCI and RQI of the road surface condition index of Chongqing trunk road.The model coefficients of CRI,PCI and RQI are fitted and verified by the subjective and objective methods.The results show that the pavement failure rate DR,flatness value IRI and crack rate CR are suitable for the performance evaluation of Chongqing main road asphalt pavement.Thirdly,according to the characteristics of non-linear relationship of evaluation indexes,BP neural network is used to fit the test data to avoid the influence of the evaluation results from the actual situation due to the high or low index of the pavement failure rate,flatness value and crack rate.The weight of BP neural network is optimized by using the key technologies of genetic algorithm gene selection,crossover and variation.The stability and accuracy of BP neural network model are improved.Finally,the example is verified.The model training and 30 trunk road test data of 80 trunk roads in Chongqing are selected.The PQI value is calculated by GA-BP model and BP model by MATLAB tool.The advantages of GA-BP model are obtained by comparing with the current standard method.The GA-BP model is more objective to evaluate the performance of asphalt pavement in Chongqing trunk road. |
PDF Full Text Request