Search

• Investigate assigning shipping requests to crowdsourcees with time and capacity constraints • Propose a centralized, deep reinforcement learning-based approach • Present new state space representation encompassing spatial-temporal and capacity information • Embed heuristics-guided action choice in DRL to preserve tractability and enhance efficiency • Integrate rule-interposing into DRL to further enhance training and implementation efficiency This paper investigates the problem of assigning shipping requests to ad hoc couriers in the context of crowdsourced urban delivery. The shipping requests are spatially distributed each with a limited time window between the earliest time for pickup and latest time for delivery. The ad hoc couriers, termed crowdsourcees, also have limited time availability and carrying capacity. We propose a new deep reinforcement learning (DRL)-based approach to tackling this assignment problem. A deep Q network (DQN) algorithm is trained which entails two salient features of experience replay and target network that enhance the efficiency, convergence, and stability of DRL training. More importantly, this paper makes three methodological contributions: 1) presenting a comprehensive and novel characterization of crowdshipping system states that encompasses spatial-temporal and capacity information of crowdsourcees and requests; 2) embedding heuristics that leverage information offered by the state representation and are based on intuitive reasonings to guide specific actions to take, to preserve tractability and enhance efficiency of training; and 3) integrating rule-interposing to prevent repeated visiting of the same routes and node sequences during routing improvement, thereby further enhancing the training efficiency by accelerating learning. The computational complexities of the heuristics and the overall DQN training are investigated. The effectiveness of the proposed approach is demonstrated through extensive numerical analysis. The results show the benefits brought by the heuristics-guided action choice, rule-interposing, and having time-related information in the state space in DRL training, the near-optimality of the solutions obtained, and the superiority of the proposed approach over existing methods in terms of solution quality, computation time, and scalability. [ABSTRACT FROM AUTHOR]

• Deal with the schedule-based HSR passenger flow assignment problem for large-scale HSR networks with explicit consideration of continuous time-varying passenger demand and tight capacity constraints. • Establish the assignment model by destination-based arc flow variables and propose the Dantzig-Wolfe decomposition-based reformulation and algorithmic framework to reduce the memory requirement regardless of demand discretization precision in the master problem. • Design an efficient method incorporating the "rooftops" method to solve the sub-problem. This method is an effective technique to tackle the continuous time-varying demand, and its computational complexity improvement for solving sub-problem speeds up the entire solution process. • Prove the convergence of the passenger flow assignment solution with discrete demand would converge to that with continuous demand when the increasing discretization number tends to infinity. • Investigate the computational efficiency and solution quality of the proposed algorithm by three practical cases with different scales. This paper proposes an algorithm for solving the schedule-based passenger assignment problem, with explicit consideration of continuous time-varying demand and tight capacity constraints for large-scale high-speed railway (HSR) networks. We construct the space–time travel network and formulate the assignment model by destination-based arc flow variables, then design a Dantzig-Wolfe (D-W) decomposition-based algorithm to solve the large-scale model of this problem with less memory requirement and computational complexity. The sub-problem in each iteration can be solved as a single-source (destination) minimum cost flow problem without capacity constraint by the "rooftops" method, which provides an effective technique to tackle both continuous and discrete time-varying demand in the assignment problem. It is further demonstrated that the assignment solutions for discrete time-varying demand can converge to that for continuous time-varying demand with the increase of demand discretization number. Three real case studies with different HSR network scales show that the proposed algorithm is efficient in memory and computational complexity, especially when highly accurate solutions for passenger assignment on large-scale HSR networks are needed. [ABSTRACT FROM AUTHOR]