Intelligent GPS-based Taxi Passenger Path Trajectory Recommended

With the rapid development of GPS devices, wireless communication technology andwider application of mobile terminals with GPS functions, we are able to more effectivelyfacilitate the tracking of moving objects and collect a variety of behavioral patterns whichtracks the movement of data. As the most familiar characteristics of urban traffic network taxidrivers, they know each time the regional urban road traffic and road conditions and thelength of the road network traffic laws, which can choose a more rational and efficient drivingroute so that they can better arrive quickly the destination. In addition, taxi trajectory datacontains the latitude and longitude, time, speed and other information, are easy to collect,widely distributed, large data characteristics, these data contain a large number of taxi driversdriving experience. In this paper, the track record for effective data analysis, miningexperienced taxi drivers experience in intelligent route planning can guide novice drivers andpassengers of foreign assisted intelligent navigation and also for urban planning andintelligent transportation decision support provide strong support. Therefore, this papercarried out the following tasks:(1) First, in order to analyze taxi mobile behavior patterns and place behavioral law gotpeople at different times in different attributes so we depend on MSRA open track movingobjects datasets to research.Meanwhile, this is also a brief introduction to get these regularknowledge-based temporal data mining technology research trajectory data to track back toprovide some analysis of the knowledge base.(2) We analyze the trajectory of moving objects research data obtained each time thenumber of stay points. Then we will be the number of actual geographic area than the right,and then through the comparison of the results of MSRA been verified feasibility. In addition,we used to get passengers centralized location at different times of temporal clustering-basedapproach, taking into account the concentration of these passengers showing regionalconcentration phenomenon occurs, the paper in the design algorithm K-Means clusteringalgorithm to recommend a taxi around the carrier off locations.(3) Our first question by drawing on classical TSP core idea of the route passenger pointshortest path study, to extract the inter-network model passenger Recommended point shortestpath problem. Second, we use a variable-length encoding mechanism chromosomes,optimized crossover and mutation operations designed components for genetic algorithms tosolve the problem of urban road network and the SP population genetic algorithm based on avariety of planning path planning recommendations, we use probabilistic optimization methodto calculate the shortest path to the exact solution. Then, we also take advantage of BaiduMaps API route search service. Finally, we are driving a taxi in city traffic route network of alarge number of simulation experiments and compare a wide range of genetic algorithms,random genetic algorithm, and the standard genetic algorithm performance real-time traffic inurban road network. Our experimental results show that multi-population genetic algorithmcompared to other algorithms that can more effectively address the taxi driver to optimizeintelligent passenger path.