Recommendation Of Hot Spot Areas With High Revenue Based On Taxi GPS

With the rapid development of the economy and the increasing urbanization,the demand for travel services has increased.Taxi,as an important part of urban public transportation,plays an important role in people's daily travel.However,due to the uneven distribution of time and space of taxi passengers,the problem of taxi empty load rate,vicious competition in the taxi industry,traffic congestion and other problems frequently occur,which greatly increases the difficulty of taxi seekers and reduces the operation of taxis.income.With the rapid development of ITS intelligent transportation system technology,more and more scholars at home and abroad have taken taxi homing mode as a research hotspot,but most of the current research results have been studied how to increase the number of rental car passengers,and generally lack of taxi revenue evaluation.Indicators and research on hotspot regional recommendation strategies that comprehensively consider the relationship between taxi supply and demand and traffic capacity in the region.This paper takes the GPS data of 70 million taxis in New York City from January 1,2016 to June 30,2016 as the research object,in order to improve taxi revenue,we divide a day into 24 time periods,divide the space into grid regions,and obtain the spatial and temporal distribution characteristics of taxi passenger source and taxi passenger source income.Two methods are used to recommend taxi hotspot areas with high revenue.The first method uses K-means clustering algorithm to cluster taxi GPS data with average revenue per unit per hour.Then the coordinates of hotspot areas with high revenue per hour are recommended to drivers.The second method takes into account the supply-demand relationship and traffic capacity of taxis in the recommended high-income passengerseeking hotspot area,takes into account the number of passenger trips,passenger income and traffic conditions in the area,and introduces the current regional passenger demand ratio for taxis as the average income index,and then calculates the taxi to passenger-carrying hotspot area according to the average speed of the current regional road network.Travel time and number of orders are taken as the unit capacity index,and a highincome passenger-seeking recommendation strategy based on taxi GPS data is put forward by constructing a high-income passenger-seeking model of maximizing average revenue and unit capacity.This strategy can improve decision support for taxi high-income passenger-seeking.At the same time,it can improve the unbalanced distribution of taxi resources and give full play to taxi-seeking in cities.The complementary role of transportation is of great significance.Finally,the results of K-means clustering algorithm analysis and model solution are compared and verified,and the advantages and disadvantages of the two recommendation methods are analyzed.