We develop a game-theoretic framework in order to estimate market equilibria in the presence of environmental policies at both supra-national and global levels. In a two-stage game, regulators maximize social welfare over their jurisdiction by setting emission charges, whilst airlines compete through frequencies, fares, and fleet choice. Airline decisions include the extent to which costs of the environmental charges are absorbed or passed on to consumers and to what extent the fleet is upgraded and utilized across the network. The equilibria outcomes suggest the presence of multiple potential distortions in aviation markets that could undermine the effectiveness of environmental policies. We apply the model to North American, Western European, and Trans-Atlantic markets and find that the most efficient mechanism is the one of a single regulator able to tax discriminate between regions. Furthermore, we show that an increase in welfare would come at the expense of the environment because the effectiveness of the policy is limited by regulators interacting in a self-interested manner.

Traditional public transportation systems provide efficient mass transportation solutions between designated stations. Despite its advantages, public transportation systems suffer from being restricted to serve a fixed area. This reduces its accessibility considerably, particularly in remote areas. Hub transportation systems overcome this problem by offering shuttle services to serve passengers who face such restrictions. These systems are highly time sensitive, as shuttles are required to return to their station by a due time to ensure that all passengers are served on time. From a decision making perspective, it is essential that the routing decisions made for shuttles are reliable against changes in travel times. To achieve this, we propose a chance constrained programming approach within a column generation framework that guarantees that the final solution remains feasible (with respect to meeting the due time) for at least a given probability of travel time scenarios. We introduce a sampling-based approach to reduce the size of the problem, and incorporate procedures that maintain chance constraint feasibility under this framework. Computational experiments reveal that our approach yields routing decisions that remain immune against travel time uncertainty.

The transportation system is shifting towards a shared mobility ecosystem, emphasizing resource aggregation and coordination among different transport modes to enhance service quality and encourage passengers to shift from private cars to public transit. This study focuses on the coordination between metro and rail networks in a city, to alleviate congestion at transfer hubs and to decrease the travel time of passengers. Given the “tide-like” transfer passengers between rail and metro hubs, we investigate the potential of reserving a number of rolling stocks, strategically allocated to the metro trains with the highest demand to avoid over-congestion. We present a mixed integer linear programming (MILP) model to formulate this problem, with decision variables including rolling stock allocation and coordinated schedules for both metro and rail networks, by considering the time-varying demand of passengers. The objectives aim to minimize the passenger travel/transfer time and operational costs for managers. Due to the computational complexity in solving real-world instances, we analyze the mathematical properties and propose an exact decomposition-based solution algorithm. Our algorithm is similar to a constraint generation method, which reformulates the original problem into a relaxed master problem (corresponding to rolling stock allocation and timetabling) and a series of independent subproblems (corresponding to passenger flows). We provide optimality conditions and prove that the subproblems can be solved by an analytical procedure. According to these properties, we further propose a set of feasibility cuts and we prove that the new cuts provide tighter bounds in comparison with traditional Benders cuts. We test our integrated approach and solution algorithms on real-world instances from the Beijing railway network. The results show that our solution approach can obtain near optimal solutions, while the commercial solvers cannot even return feasible solutions in real-world instances. Compared to the current plan used in practice, our approach by combining rolling stock allocation and timetable coordination can reduce the passenger transfer waiting time by nearly 40\%.

Bike-Sharing Systems provide eco-friendly urban mobility, contributing to the alleviation of traffic congestion and to healthier lifestyles. Efficiently operating such systems and maintaining high customer satisfaction is challenging due to the stochastic nature of trip demand, leading to full or empty stations. Devising effective rebalancing strategies using vehicles to redistribute bikes among stations is therefore of uttermost importance for operators. As a promising alternative to classical mathematical optimization, reinforcement learning is gaining ground to solve sequential decision-making problems. This paper introduces a spatio-temporal reinforcement learning algorithm for the dynamic rebalancing problem with multiple vehicles. We first formulate the problem as a Multi-agent Markov Decision Process in a continuous time framework. This allows for independent and cooperative vehicle rebalancing, eliminating the impractical restriction of time-discretized models where vehicle departures are synchronized. A comprehensive simulator under the first-arrive-first-serve rule is then developed to facilitate the learning process by computing immediate rewards under diverse demand scenarios. To estimate the value function and learn the rebalancing policy, various Deep Q-Network configurations are tested, minimizing the lost demand. Experiments are carried out on various datasets generated from historical data, affected by both temporal and weather factors. The proposed algorithms outperform benchmarks, including a multi-period Mixed-Integer Programming model, in terms of lost demand. Once trained, it yields immediate decisions, making it suitable for real-time applications. Our work offers practical insights for operators and enriches the integration of reinforcement learning into dynamic rebalancing problems, paving the way for more intelligent and robust urban mobility solutions.