Research On Real-time Dispatching Model For Shared Autonomous Vehicles Under Ride-sharing Mode And Its Application

Numerous studies have shown that shared autonomous vehicles(SAVs)ride-sharing can compensate for the poor punctuality,comfort,and coverage of conventional public transportation and help achieve low cost,low energy consumption,and low congestion in travel.However,the analysis of existing shared autonomous vehicles ride-sharing modes shows that most of the existing ride-sharing modes are overlapping ride-sharing modes,which restrict the origin and destination of ridesharing participants to the same route,and due to the large randomness and diversity of travel routes of travelers,the overlapping ride-sharing mode cannot meet the travel demand of travelers,resulting in low efficiency of ride-sharing.Therefore,it is of great practical significance and application value to consider the dispatching of share autonomous vehicles with dynamic detour ride-sharing without limiting the origin and destination of passengers.First,this paper analyzes the ride-sharing mode of shared autonomous vehicles,and then proposes the dynamic detour ride-sharing mode,gives its definition and analyzes its advantages.The dynamic detour ride-sharing mode may cause inconvenience to individuals by picking up other passengers and taking detours,but it can increase the utilization rate of vehicles,reduce travel costs,and satisfy the diversity characteristics of travelers.Secondly,the factors influencing travelers’ dynamic detour ride-sharing preferences are analyzed.Secondly,the factors influencing travelers’ dynamic detour ride-sharing preferences are analyzed.An orthogonal design is used to design the questionnaire and implement the SP willingness survey,and the obtained information is subjected to extreme difference analysis(factor sensitivity analysis)using the orthogonal design method to obtain the important factors influencing travelers’ ride-sharing preferences.Meanwhile,a multi-objective real-time dispatching model is constructed based on mathematical analysis and information gap decision theory with the objectives of minimizing the total travel time cost of passengers and maximizing the ridership ratio.It defines the information of the directed network,passenger request set,and the set of dispatchable vehicles in the system,establishes the objective functions of total travel time cost of passengers and ride-sharing distance ratio,analyzes the constraints of time window,vehicle capacity,and passengers’ ride-sharing preference,and constructs the real-time dispatching model of shared autonomous vehicles in ride-sharing mode with the objectives of minimizing the total travel time cost of passengers and maximizing the ride-sharing distance ratio.Finally,the dynamic ride-sharing dispatching matching algorithm for shared autonomous vehicles is studied.A two-stage ride-sharing dispatching matching algorithm for feasible matching pair acquisition-vehicle dispatching matching optimization is proposed to quickly acquire feasible matching pairs using the greedy insertion method and optimize vehicle matching using an improved Hungarian algorithm.The validity of the model is verified through experimental simulation and arithmetic analysis.To investigate the effect of ride-sharing in shared autonomous vehicles scheduling considering dynamic roundabout ride-sharing mode and overlapping ride-sharing mode(ride-sharing without considering detour).The research results of this paper can be used in the theoretical research and application research of dynamic ride-sharing dispatching of shared autonomous vehicles,providing references and suggestions for the construction,operation and management of the shared autonomous vehicles dispatching system,providing new thinking directions for solving the problems of shared autonomous vehicles ride-sharing scheduling,and having positive significance for alleviating traffic congestion.