Dynamic shared autonomous vehicles dispatching algorithm

Numerous studies have shown that shared autonomous vehicle (SAV) ride-sharing models are beneficial for achieving low cost, low energy consumption, and low congestion for travel. However, an analysis of existing shared autonomous vehicle ride-sharing modes shows that existing modes suffer from inefficient ride-sharing. To address this problem, in this study, a dynamic detour ride-sharing mode (DDRM)as a new ride-sharing mode is proposed, and a dynamic SAV ride-sharing model for with the multi-objective of maximizing passengers' ride-sharing sharing distance ratio and minimizing their total travel time based on this new mode is formulated. According to the constraints of the model, a new passenger ride-sharing intelligence algorithm that determines passenger requests-analysis passenger matches-optimizes matching is designed, which solve the dynamic ride-sharing matching model with Hungarian algorithm in this study. In simulation results, the algorithm framework is able to provide quality dynamic ride-sharing solutions in real time, and the DDRM showed a much higher ride-sharing rate (the ration of the number of passengers who participant in ride-sharing to the total number of passenger) and more significant ridership benefits compared to the non- detour ride-sharing mode. The dynamic detour ride-sharing method of shared autonomous vehicle studied in this study can achieve the global optimum of receiving the most passengers using the least number of SAVs.