Research And Implementation Of Road Detection System Based On Mobile Intelligent Terminal

With the rapid development of economy in our country,the number of motor vehicles has grown at an average annual rate of 10%.Automobiles have become an indispensable transportation tool for people.However,due to the large increase in the number of automobiles,the pressure on highways is also increasing,and road pavement damage is inevitable.Meanwhile,the speed of detection and maintenance of roads is far behind the growth rate of automobiles.Road anomalies,such as potholes,not only affect the comfort of the driver and passenger,but also seriously affect the safety of road users and cause major property loss.Aiming at this problem,a road detection system based on mobile intelligent terminal is designed and developed collecting the road condition data through the built-in sensor of the vehicle mobile terminal,transmitting the data to the server,marking the road condition on the map after data processing and classification,and providing early warning service for the driver combined with map navigation function.The main work of the thesis is as follows:(1)By analyzing the basic requirements and application scenarios of road condition detection,a road condition detection system structure based on mobile terminals is designed.It mainly includes three parts,that is data acquisition and processing,road condition classification,and road condition warning.The data acquisition part mainly uses the Raspberry Pi system's built-in acceleration,gyroscopes,and other sensors to collect data for calculating the acceleration,amplitude,offset data,and position data of the vehicle and forming a database of road surface conditions.The classification of road conditions is mainly performed after the server preprocesses the collected data.The road condition warning is mainly to remind the road condition in real time when non-flat road conditions exist in the way forward.(2)In the process of classifying road conditions using sensor data collected by the Raspberry Pi system during the driving of the vehicle,the common classification methods mostly take into consideration the lack of correlation between the feature variables.In view of the accuracy of classification,the road condition detection and classification method based on the Mahalanobis-Taguchi System(MTS)is proposed in this thesis.The similarity of different sample sets is calculated by Mahalanobis distance.This method takes into account the correlation between different feature variables,and the unit of feature variables will not affect the calculation result of Mahalanobis distance,thus effectively classifying road conditions.In the experimental part,the classification method proposed in this thesis is compared with the two classification algorithms of decision tree and support vector machine.The results show that the method has better recognized effect in the classification of road conditions.(3)In view of the potential conflicts between the results of different evidence sources on the same road condition,this thesis adopts the road condition fusion method based on D-S evidence theory.The basic probability assignment function of each evidence source is obtained by using the Mahalanobis distance of each road condition space of the MTS detection model,and calculated according to the weighted combination rule algorithm.The fusion result of the road conditions is determined and taken as the final diagnosis.This method considers the correlation and effectiveness between the evidences and effectively reduces the influence of"bad value" on the data fusion results.Experiments show that this method can perform data fusion and improve system performance compared with other combined algorithms of D-S evidence theory.(4)On the basis of the above research,a road detection system based on mobile intelligent terminals is designed and implemented,including an intelligent gateway based on the Raspberry Pi system and an Android-based navigation APP.The system adopts the C/S architecture,in which the client-side provides the functions of personal information management,personalized settings,navigation and positioning,road condition reminding,driving record management and alarm prompting,and the server-side has the functions of basic information setting and management,data processing,data analysis,data display and message push.