Cycling is a popular and sustainable mode of transportation. However, few studies have examined experimental and modeling studies on bicycle flow dynamics. In particular, for wide roads, the characteristics of high-density regimes in the fundamental diagram have not been well addressed, and the emergence of stop-and-go waves has not been fully understood. This study experimentally investigated bicycle flow dynamics on wide roads, using two types of 3-m-wide track. Different riding behaviors, namely, free riding, following, and overtaking, were analyzed. The bicycle flow dynamics were found to be essentially the same on the two tracks: (i) The bicycle flow rate remained nearly constant across a wide range of densities, in marked contrast to the single-file bicycle flow, which exhibited a unimodal fundamental diagram. By studying the weight density of the radial and lateral locations of cyclists, we argue that this behavior arises from the formation of more lanes with increase in global density. The newly formed lanes prevented the flow rate from decreasing. (ii) When the density exceeded 0.5 bicycles/m², the flow rate began to decrease, and stop-and-go traffic emerged. Based on these behavioral observations, we propose an improved heuristic-based model to simulate bicycle flow on roads of different radii and explicitly account for the centrifugal effect of bicycles. The calibration and validation results demonstrate that the proposed model can reproduce the traffic dynamics of bicycle flow.

骑自行车是一种流行且可持续的交通方式。然而，很少有研究检查关于自行车流动动力学的实验和模型研究。特别是对于宽阔的道路，基本图中的高密度区域特征尚未得到很好的解决，“走走停停”浪潮的出现还没有被完全理解。这项研究使用两种类型的3米宽的轨道，对宽阔道路上的自行车流动动力学进行了实验研究。分析了不同的骑行行为，即自由骑行，跟随和超车。发现在两个轨道上自行车流量动力学基本相同：（i）自行车流量在各种密度下几乎保持恒定，这与单档自行车流量形成了单峰基本图形成鲜明对比。 。通过研究骑自行车者的径向和横向位置的重量密度，我们认为这种行为是由于随着全球密度的增加而形成了更多的车道。新形成的通道防止流速降低。（ii）当密度超过0.5辆/米时^{如图2所示}，流量开始下降，出现走走停停的交通。基于这些行为观察，我们提出一种改进的基于启发式的模型，以模拟不同半径的道路上的自行车流量，并明确考虑自行车的离心作用。校准和验证结果表明，该模型可以重现自行车流量的动态变化。