Research On Dynamicshared Path Planning Of Shared Autonomous Vehicles

Shared Autonomous Vehicles(SAV)are known as a new mode of future transportation to solve chronic problems such as urban traffic congestion,accidents and pollution due to their significant advantages of high efficiency,comfort and flexibility.There are many researches on ride matching algorithms,but the dynamic real-time requirements of SAV in large-scale scenarios cannot be effectively guaranteed.Most of the existing SAV dynamic ride-sharing matching algorithms only consider the path similarity between passengers,and rarely consider satisfying passengers.Individual preference needs to improve the ride experience.Therefore,how to satisfy the SAV dynamic ride sharing service in large-scale scenarios while taking into account the individual preference needs of passengers is an urgent difficulty that needs to be overcome in this paper.To this end,this paper proposes the optimal path planning algorithm for SAV dynamic co-multiplication and the optimal path planning algorithm for SAV dynamic co-multiplication under preference conditions.The details are as follows:In order to meet the real-time requirements of large-scale SAV dynamic carpooling and reduce the computational complexity of the algorithm,this paper proposes an optimal path planning algorithm for SAV dynamic carpooling.Firstly,two dynamic ride-sharing matching methods are proposed.Secondly,when passengers and vehicles are matched,a large number of SAVs that do not meet the requirements are screened out with a small time cost through the pruning technology of the passenger search module and the vehicle screening module.A vehicle sharing matching algorithm based on insertion algorithm is proposed to select the optimal route with the goal of minimum tolerance of detour deviation.The matching process reduces the matching time between passengers and vehicles.In the overall performance analysis of the experiment,the running time of the optimal path planning algorithm and the benchmark algorithm under different parameters was compared and analyzed,and the results showed that the running time of the optimal path planning algorithm was lower than that of the benchmark algorithm.,the influence of different matching methods and different parameter combinations on other operating results is analyzed,thereby verifying the effectiveness and efficiency of the optimal path planning algorithm.In order to meet the individual preference needs of passengers and improve the ride experience,this paper proposes an optimal path planning algorithm for SAV dynamic ride sharing under preference conditions.Firstly,the individual preference of passengers is modeled according to the questionnaire results,and a scoring function model is designed to measure the matching degree between passengers.Secondly,two matching algorithms are proposed when passengers and vehicles are matched.Vehicles that do not meet the requirements are screened out through the rough matching process between passengers and vehicles,and passengers with similar paths and similar interests are preferentially matched to a vehicle.In the overall performance of the experiment,the influence of the two algorithms on the running time under different parameters is compared and analyzed,and the results show that the optimal vehicle carpooling matching algorithm based on branch and bound is better than the optimal vehicle carpooling matching algorithm based on mixed integer programming;Quantitative analysis is made on the factors of "passenger sharing willingness" and "detour deviation tolerance" in the performance of each stage.The results show that when the passenger sharing willingness is 1 and the detour deviation tolerance is 5.37,the total social benefits and matching The success rate is the largest,and the matching effect is the best.