| With the rapid development of computer technology and Internet of Things technology,people are increasingly relying on location-based services,and spatiotemporal trajectory big data has also increased dramatically.Taxi trajectory data,as an important part of the spatiotemporal trajectory big data,has received widespread attention.Taxi is an important means of transportation in the city,which brings great convenience to residents.At present,for taxi drivers,in the case of no order,they can only choose to wait or search for passengers independently,which will consume a lot of time and fuel,affecting revenue.Due to the lack of reasonable route planning,there is also the problem of traffic jams of popular routes.In this thesis,we study route planning based on spatiotemporal trajectory big data.The main research work of this thesis are as follows.(1)Research on spatiotemporal trajectory big data.We analyze the spatiotemporal distribution of the data and mine optimal profit points based on taxi trajectory data.Firstly,the data is divided into time periods,and then divided according to the geographical location.The purpose is to analyze the change trend of data in different regions in different time periods,which is the basis for route planning.Then we use the DBSCAN clustering algorithm to mine clusters from the divided data.Finally,the K-Means clustering algorithm is used to find the center point of each cluster,which is the optimal profit point.The optimal profit points are the location points,which are the most commonly visited by taxi drivers.If the no-load driver can find these points,attendance can be greatly increased.(2)Plan profitable routes.The data is processed by clustering to form a set of the optimal profit points.We convert the point-to-point route planning problem into a Travelling Salesman Problem(TSP),and propose an Ant Colony Optimization Algorithm based on Optimal Profit Points(ACOOPP).In terms of route selection,we have added influence factors to increase the probability that drivers choose the popular optimal profit points.Secondly,we use the rule of combining global pheromone update and local pheromone update to increase the exploration ability of the algorithm and avoid entering a stagnant state.Finally,we use the real-time driving distance function in Baidu Map API to replace the original distance calculation formula.(3)Introduce the Stackelberg game model.Traffic jams of popular routes often occur in route planning.Driver costs will also increase sharply as congestion time increases.To solve this problem,we introduce the Stackelberg game model.We use traffic guidance by a fixed time interval to reach an equilibrium state from the perspective of two decision makers,traffic leaders and followers.This model maximizes the benefits of traffic leaders and followers,and makes great use of road network resources,which can effectively solve the traffic jams of popular routes.(4)A taxi route planning system is constructed based on the SSM(Spring + Spring MVC + My Batis)framework.The system is developed based on the Baidu Map API,and implements functions such as map display,positioning,retrieval and query of the optimal profit points,route planning,and navigation.In route planning,the Stackelberg game model is introduced in consideration of the traffic jams of popular routes,and the proposed ACO-OPP algorithm is used by the decision makers of the traffic leaders,which maximizes the benefits of taxi drivers and traffic managers. |
