Research On Key Technologies Of Intelligent Evaluation Of Pavement Driving Quality Based On Smart Phone

Since entering the 21st century,driving had become one of the main ways of daily travel for Chinese residents.With the increasing of the annual driving distance,people have put forward higher requirements for driving quality and driving safety on the travelling.Therefore,China continues to increase the proportion of high-grade roads,strengthen the maintenance of old roads,and maintain the high-level service and high-quality operation of the road network as far as possible,so as to make people’s driving more safe and comfortable.In the face of large-scale road network,the efficient and economical implementation of road traffic quality monitoring helps to supplement the existing detection and evaluation,and improve the level of road management and service.Modern smartphones integrates a variety of sensors,a triaxial acceleration,positioning,speed of data acquisition and transmission and display functions such as friendly interface,has become a feature-rich portable devices.Taking the intelligent monitoring platform of pavement driving quality based on GIS as the application orientation and the smart phone as the mobile intelligent terminal of driving data collection,some key technologies in the intelligent evaluation method of pavement driving quality based on smart phone perception driving data were studied through driving experiment,data correlation analysis,vibration index comparison,vibration threshold determination and other methods.Based on the analysis of the existing literature at home and abroad,aiming at the influence of different combination style of vehicle and smart phone,vehicle speed and pavement roughness on the collected vibration data,an experimental scheme was designed,which comprehensively considered the factors of vehicle,smart phone,fixed position,roughness,vehicle speed and evaluation unit length.Using the smart phone application developed by the research group based on Android environment,the automatic collection of driving data including vehicle speed,triaxial acceleration and so on was realized,and the pavement roughness data of the experimental road section was investigated and measured.On this basis,the reproducibility of driving vibration process characterized by different vibration indexes was analyzed,and the reproducibility of driving vibration process characterized by each index was characterized by the degree of correlation and difference of vibration process.The conversion method of vibration intensity under different vehicle mobile combination conditions was studied.When studying the relationship between driving vibration,vehicle speed and road roughness,the influence of the length of evaluation unit on the prediction of road roughness using the driving data collected by smart phones was analyzed,and the prediction model of kilometer unit was put forward and verified.For the small section evaluation unit,a pavement roughness prediction method based on speed sensitivity parameters was proposed.Then,the evaluation methods of driving quality based on intelligent perception data were studied and analyzed from two aspects of roughness and driving vibration.Finally,a preliminary study on the intelligent application of riding quality evaluation was carried out.According to the above research contents,the the main conclusions could be drawn as follows:1)The performance difference of the two selected smart phones in vibration measurement was small,and the smart phone had good reproducibility in collecting driving vibration data.Vehicle and the position of phone had an impact on the collected driving vibration data,but the vibration data obtained under different working conditions showed the same change law.According to the statistical relationship between the data collected from different vehicle and different phone positions,data standardization could greatly reduce the difference of driving vibration caused by different working conditions.2)The relationship model of"IRI-v-a_1r.MV"was established by multiple regression method,and the prediction accuracy of the model for roughness showed an increasing trend with the increase of the length of evaluation unit.When the IRI was the same,the vehicle vibration increased with the increase of driving speed.When the driving speed was the same,the vehicle vibration increased with the increase of roughness to a certain extent.However,in different IRI intervals,the increase degree of vibration with the increase of speed was different.Therefore,the concept of speed sensitivity（SS）was proposed.Based on the SS,a two-stage prediction method for IRI of smaller evaluation unit was proposed.This method improved the accuracy of the prediction model.However,this method needed to get a lot of traffic information,and its practical application was more complex.3)Vehicle speed was an important factor affecting people’s subjective evaluation of driving vibration.According to whether the influence of vehicle speed was considered or not,the classification range of vector synthetic vibration and the quantitative value of its comfort were determined according to relevant standards.Based on this,three methods were proposed to directly evaluate the riding quality by using the intelligent perception data,namely,the average vibration method,the comfort rate method and the probability comfort index method weighted by the vibration distribution ratio.4)The decision-making method of pavement maintenance was put forward.Firstly,according to the IRI-vibration relationship model established in this paper,the IRI of pavement was predicted by collecting driving vibration,and the pavement maintenance work was carried out according to the requirements of various specifications.Secondly,the driving comfort rate of road section was obtained by collecting driving vibration index with smart phone,and the driving comfort rate of50%was taken as boundary.When the comfort rate is lower than 50%,the pavement maintenance with reasonable strength should be considered.Based on the automatic collection of daily driving data,this paper carried out research on intelligent monitoring of road driving quality.The quantitative and qualitative conclusions and related research results obtained in this paper have certain theoretical and application values,and can provide reference for the links of pavement evaluation,maintenance,management and service.