论文标题
超级池：将私人旅行汇集到高占用的类似于运输的诱人共享游乐设施
Hyper-pool: pooling private trips into high-occupancy transit-like attractive shared rides
论文作者
论文摘要
我们提出了Hyper-Pool，这是一种分析，离线，公用事业驱动的乘坐算法，以将单独的跳闸请求汇总到有吸引力的高占用范围中。我们偏离了乘车式的Exmas算法，在该算法中，单一游乐设施被集中在有吸引力的挨家挨户的骑行中，并提出了两种新颖的需求方算法，以进一步汇总个人需求，以更紧凑。首先，我们生成了停留的游乐设施，单个拾取和下降点为所有旅行者最佳。其次，我们再次将这种游乐设施捆绑在一起，导致超紧凑的骑行足以与公共交通运营相似。我们提出了一个自下而上的框架，在该框架中，确定的游乐设施的合并度逐渐增加，从而确保了随后的聚合水平的吸引力。我们的Hyper-pool方法为给定的旅行需求输出了每次服务变体的一组有吸引力的合并乘车。该算法可公开可用且可再现。它适用于实际尺寸的需求数据集，并为利用乘车潜力的限制打开了新的机会。在我们的阿姆斯特丹案例研究中，我们设法将超过220位旅行者列入了40次超级乘车的平均入住5.8 pax/vear。
We propose Hyper-pool, an analytical, offline, utility-driven ride-pooling algorithm to aggregate individual trip requests into attractive shared rides of high-occupancy. We depart from our ride-pooling ExMAS algorithm where single rides are pooled into attractive door-to-door rides and propose two novel demand-side algorithms for further aggregating individual demand towards more compact pooling. First, we generate stop-to-stop rides, with a single pick up and drop off points optimal for all the travellers. Second, we bundle such rides again, resulting with hyper-pooled rides compact enough to resemble public transport operations. We propose a bottom-up framework where the pooling degree of identified rides is gradually increased, thereby ensuring attractiveness at subsequent aggregation levels. Our Hyper-pool method outputs the set of attractive pooled rides per service variant for a given travel demand. The algorithms are publicly available and reproducible. It is applicable for real-size demand datasets and opens new opportunities for exploiting the limits of ride-pooling potential. In our Amsterdam case-study we managed to pool over 220 travellers into 40 hyper-pooled rides of average occupancy 5.8 pax/veh.
