A Game-Based Decision-Making Method for Multi-Ship Collaborative Collision Avoidance Reflecting Risk Attitudes in Open Waters

Jiongjiong Liu,Jinfen Zhang,Zaili Yang,Mingyang Zhang,Wuliu Tian
DOI: https://doi.org/10.1016/j.ocecoaman.2024.107450
IF: 4.295
2024-01-01
Ocean & Coastal Management
Abstract:To accurately reflect risk attitudes towards ship intentions in multi-ship encounters, this paper develops a novel two-stage collaborative collision avoidance decision-making (CADM) model by incorporating intention prediction and real-time decision-making. We acquire prior knowledge of risk attitudes by analyzing Automatic Identification System (AIS) data and further estimate the probability distributions of encountering ship's risk attitude using Bayesian reasoning. By treating collision avoidance procedure as a static game with incomplete information, a predictive model for collision avoidance intentions is developed by taking account into risk attitude probabilities. Real-time decisions are then implemented according to different stages, and a collaborative CADM model is established by a game-decision cycle. Finally, a multi-ship encounter scenario is simulated under all combinations of risk attitudes, and the results are compared with those obtained under complete information. The results demonstrate that the proposed model can formulate avoidance actions that meet safety requirements under all combinations of risk attitudes. Further comparison with complete information proves the effectiveness of the risk attitude probability model, which is conducive to improving the decision-making flexibility and reducing complexity. The research findings enhance the collaborative decision-making, contributing to the development of autonomous navigation in open waters.
What problem does this paper attempt to address?