Research On Mobility Modeling And Mining Method Of Taxi Trajectory Data

As the scale of the city continues to expand,people's quality of life continues to improve,people's travel activities are more and more frequent,and the scope of activities is also growing,and taxis have become the preferred means of transportation for people to travel with their convenient features.Since most taxis are equipped with GPS equipment,the operation of the taxis will generate a large amount of trajectory data.How to mine valuable information from this huge trajectory data has always been a hot issue.Reasonable use of the trajectory data of taxis can improve the operational efficiency of taxis and optimize the traffic environment of the city.In this paper,we will improve the operational efficiency of taxis by analyzing a large number of taxi trajectory data and designing a reasonable taxi operation strategy.We will proceed from the following aspects:(1)We analyze taxi trajectory data from multiple dimensions,explore the impact of taxi drop-off information on the strategy for finding passengers,reasonably define the seeking efficiency and passenger density of the road segments,and integrate the experienced drivers' driving experience to define the weight of each road segment.Finally,A global route recommendation algorithm is proposed based on the weight of each road segment,so that the taxi drivers can find the passengers faster and increase the income of the taxi drivers.(2)By analyzing the taxi trajectory data,we find the unfairness problem in the taxi operation process,and then design a centralized taxi dispatching approach with global fairness.The approach calculates the matching weight between each taxi and passenger in real time,and dynamically adjusts the matching priority of each taxi according to the current operation of the taxi,and then combines the KM algorithm to realize the matching process,thereby ensuring that taxis have certain fairness in the process of centralized dispatching.At the same time,due to the centralized operation strategy of taxis,the cost of taxis is greatly reduced,thereby increasing the income of taxi drivers,reducing passenger waiting time and optimizing traffic efficiency.(3)We improve the operational efficiency of taxis from the perspective of order recommendation.In the taxi operation process,we combine the client and server to achieve the order recommendation.On the client side,we only need to calculate the evaluation function value between the taxi and each order in real time,and then recommend each order to each taxi according to the evaluation function value,but this will lead to the same order recommended to multiple taxis.In this case,in order to avoid taxis snatching orders as much as possible,a further global optimization process is implemented on the server side.Finally,we verify the effectiveness of the above approaches through some experimental simulations.