An extended social force model for the dynamics of electric bicycles in isolated non-motorized lanes

This study proposes a model for simulating the dynamics of electric bicycle flow in isolated non-motorized lanes. The model is an extension of the conventional social force model with an additional rotational degree of freedom that describes the steering angle of electric bicycles. In this model, two corrective relaxation terms are included in the driving force for the rotational degree of freedom: one on the steering angle and another on the angular velocity. Stability analysis shows that an appropriate amount of correction on the angular velocity helps to avoid oscillation and overshoots of the turning angle. The proposed model is validated both macroscopically and microscopically, and the agreement of the simulated fundamental diagram with that in the literature demonstrates the rationality of the model in characterizing the collective motions of electric bicycles. Comparison of simulated and observed trajectories confirms that the proposed model could reproduce electric bicycle motions with a certain accuracy at the microscopic level. Finally, predictions based on the validated model are provided. According to the simulation results, congestion starts to occur when the inlet flow rate of the lane reaches 1500–1750 ebike/h/m. For a mixed configuration, the maximum allowed flow rate in the lane is 2350–2400 ebike/h/m, and this value depends on the proportion of fast electric bicycles and the riding style in the flow.