| Title |
Performance Analysis of Vehicle/Cargo Transport Matching Algorithm based on MAML |
| Authors |
김지현(Ji-Hyeon Kim) ; 신다민(Da-Min Shin) ; 이수은(Su-Eun Lee) ; 김형남(Hyoung-Nam Kim) |
| DOI |
https://doi.org/10.5573/ieie.2023.60.9.45 |
| Keywords |
Vehicle/cargo dispatch system; Transportation matching algorithm; MAML |
| Abstract |
In the port logistics environment, it is necessary to consider environmental factors such as fairness/consistency, road conditions, and weather for efficient vehicle/cargo dispatch. These environmental factors have constantly changed and can affect the vehicles transporting cargo and the drivers. Meta-reinforcement learning is a technique that effectively proceeds the optimization process of artificial neural networks through learning and has the advantage of learning models well for new environments or data. In this paper, we propose a vehicle/cargo transportation matching algorithm that applies model-agnostic meta learning (MAML), a meta reinforcement learning algorithm, to efficiently dispatch vehicles in various environments and data changes. Through simulations, we analyze the performance of vehicle and cargo transportation matching algorithms according to various transportation environments. Simulation results show that the proposed algorithm can automatically and quickly distribute vehicles and cargo. |