Optimal parking provision for tourist coaches in large-scale bi-modal urban transportation networks

This paper examines the optimal parking provision for tourist coaches in bimodal transportation networks to mitigate traffic congestion and reduce carbon emission. We consider the allocation of parking spaces between tourist coaches and private vehicles, under a network equilibrium model of multi-level granularity that captures the interplay between tourist coaches and private cars as well as the impact of their cruising-for-parking behaviours on traffic congestion. At the zonal level, we characterise how parking space allocation affects cruising-for-parking behaviours and the ensuing traffic congestion. At the link level, we characterise the route choice of travellers influenced by parking space allocations. Overall, the optimal parking space allocation problem is formulated as an optimisation with parking space allocation as decision variables and bi-modal user equilibrium conditions as constraints. To derive the optimal parking space allocation, we utilise the multi-level granularity of our model to develop a tailored sensitivity-based approach for deriving both first-order and second-order gradients of the objective value with respect to decision variables, while avoiding the computation of large-scale matrix inversions. This novel approach enables us to apply the Newton method to determine optimal parking allocations more rapidly, which is not computationally feasible in prior studies, particularly within the context of large-scale transportation networks. Through a realistic case study of Hong Kong, we find that strategic parking space allocation can decrease overall cruising-for-parking traffic by 7.8%, enhance traffic speed by as much as 2.78% within the bimodal transportation network, and reduce CO (Formula presented.) emissions for up to 8% in certain zones.