| With the popularization of GPS devices,more and more track big data are generated.In order to improve the comprehensive transportation service level and speed up the construction of a transportation power,it is particularly important to use the processing method of big data to effectively process,analyze and store the track data of taxis.Based on the GPS track data of taxis in Shanghai,this paper analyzes several factors that affect the income of drivers.Firstly,the GPS positioning data is extracted and cleaned to obtain the data set of effective travel.MATLAB based big data processing and storage method is used to process the track big data.For the problems of latitude and longitude anomaly and temperature drift in GPS data,several methods are proposed to filter the anomaly points.Clustering method can be used to accurately identify the abnormal points with missing tracks,and fuzzy c-means clustering can be used to process a large number of redundant data obtained from sampling,so as to reduce the size of invalid data and verify the accuracy of clustering results.Aiming at the shortcomings of the existing map matching method,which is too complex and limited,a map matching method based on ellipse is proposed,which USES the road width and high GPS sampling accuracy to effectively match the GPS data points to the correct road.According to the processed data,the correlation analysis of taxi drivers’ income was carried out: the OD point of the taxi was extracted,so as to obtain the passenger carrying track and time period of the taxi in operation.Then,the shift changing behavior of the taxi was detected to distinguish the number of drivers driving the taxi every day,and the taxi income was quantified to analyze the average hourly income of the drivers at different times of the day.Statistical analysis was conducted on taxi travel frequency,taxi driving distance and taxi occupancy rate to determine whether it is related to taxi drivers’ income,and Suggestions were given to improve taxi drivers’ income and relieve taxi drivers’ pressure.Visualization for taxi drivers of passenger behavior,and puts forward the mobile entropy,swing radius,number of passenger transport,grid,quantity of heat,POI distribution density five indicators for users,were analyzed,and the validation of analytical results are accurate,using the Analytic hierarchy process to analyze the decision-making problems affecting the taxi driver,get a proper strategy plan,provides reference for the taxi driver.The results of this paper can help taxi drivers to make decisions,so as to increase their income. |
