A probabilistic risk approach for the collision detection of multi-ships under spatiotemporal movement uncertainty

• Published in: Reliability engineering & system safety Vol. 215; p. 107772
• Main Authors: Xin, Xuri, Liu, Kezhong, Yang, Zaili, Zhang, Jinfen, Wu, Xiaolie
• Format: Journal Article
• Language: English
• Published: Barking Elsevier Ltd 01.11.2021; Elsevier BV
• Subjects: AIS data; Algorithms; Collision avoidance; Maritime safety; Monte Carlo simulation; Multi-ship spatiotemporal movements; Probabilistic conflict detection; Reliability engineering; Risk analysis; Risk reduction; Ship collision risk; Ship motion; Ships; Trajectories; Uncertainty; Uncertainty time-dependent analysis; AIS data; Multi-ship spatiotemporal movements; Maritime safety; Uncertainty time-dependent analysis; Probabilistic conflict detection; Ship collision risk
• ISSN: 0951-8320; 1879-0836
• DOI: 10.1016/j.ress.2021.107772

•Proposed a novel probabilistic conflict detection approach to quantify the conflict criticality of multi-ships in dynamic and uncertain situations.•Develop a two-stage MC simulation algorithm to efficiently and accurately compute the ship conflict probability.•Introduce an AIS data-driven mechanism to extract the vessel spatiotemporal uncertainty patterns to support a robust estimate of the collision risk.•Demonstrate the performance of the new conflict detection approach through the real AIS-based trajectory data in port. It is vital to analyse ship collision risk for preventing collisions and improving safety at sea. The state-of-the-art of ship collision risk analysis focuses on encountering conflict between ship pairs, subject to a strong assumption of the ships having no/little spatiotemporal motion uncertainty. This paper proposes a probabilistic conflict detection approach to estimate potential collision risk of various multi-vessel encounters, in which the spatiotemporal-dependent patterns of ship motions are newly taken into account through quantifying the trajectory uncertainty distributions using AIS data. The estimation accuracy and efficiency are assured by employing a two-stage Monte Carlo simulation algorithm, which provides the quantitative bounds on the approximation accuracy and allows for a fast estimation of conflict criticality. Several real experiments are conducted using the AIS-based trajectory data in Ningbo-Zhoushan Port to demonstrate the feasibility and superiority of the proposed new approach. The results show that it enables the effective detection of collision risk timely and reliably in a complicated dynamic situation. They therefore provide valuable insights on ship collision risk prediction as well as the formulation of risk mitigation measures.