Due to the confined span of parking location choice for human-driven vehicles (HVs), the spatiotemporal imbalance in parking space utilization has always been a challenging problem in many major cities, leading to a substantial waste of precious land resources. However, in the era of autonomous vehicles (AVs) with parking autonomy, there are more alternative parking options available as AV can park at more distant locations away from the trip maker’s destination. This paper aims to investigate the morning commute problem with consideration of AV commuters’ distant parking choices in a many-to-one network. Parking sharing scheme is explicitly considered, wherein AV travellers can choose to park at city centre at public parking facilities and then lease their own parking space out, park at home, or park at a shared parking space. We first examine how AV commuters in different residential clusters prioritize their parking location choices, as well as their willingness to share the vacant parking space. Then, we investigate the travellers’ trip timing decisions and determine the resultant equilibrium travel pattern. Without other dynamic managing schemes, the model results indicate that appropriate CBD parking supply, together with differentiated parking charges/subsidy, can reduce the total queueing congestion significantly, yet at the cost of higher total travel cost.

由于人用车辆（HV）的停车位选择范围有限，因此在许多主要城市中，停车位利用率的时空不平衡一直是一个具有挑战性的问题，导致大量宝贵土地资源的浪费。但是，在具有停车自主权的自动驾驶汽车（AV）时代，有更多可供选择的停车选择，因为AV可以停在旅行社目的地以外的较远位置。本文旨在在多对一网络中考虑AV通勤者的远距离停车选择来研究早晨通勤问题。明确考虑了共享停车方案，AV旅行者可以选择在市中心的公共停车设施停车，然后将自己的停车位出租，在家停车或在共享停车位停车。我们首先研究不同居住区中的AV通勤者如何优先选择停车位置，以及他们共享空置停车位的意愿。然后，我们调查旅行者的旅行时机决策，并确定最终的平衡旅行方式。在没有其他动态管理方案的情况下，模型结果表明，适当的CBD停车供应以及有区别的停车收费/补贴可以显着减少总排队拥堵，但代价是总出行成本较高。