A study of robust ship avoidance avoidance algorithms in an environment with a variety of autonomous vessels

2021 Fiscal Year Final Research Report

• Project/Area Number: 20K14968
• Research Category: Grant-in-Aid for Early-Career Scientists
• Allocation Type: Multi-year Fund
• Review Section: Basic Section 24020:Marine engineering-related
• Research Institution: National Institute of Maritime, Port and Aviation Technology
• Principal Investigator: SATO Keiji 国立研究開発法人海上・港湾・航空技術研究所, その他部局等, 研究員 (90734244)
• Project Period (FY): 2020-04-01 – 2022-03-31
• Keywords: 避航操船 / 多目的最適化 / CORLEGs / OZT

Outline of Final Research Achievements

In this study, the parameters of the ship avoidance algorithm is designed by applying an evolutionary multi-objective optimization method that can obtain a large number of candidate solutions in an environment where there are tradeoffs among multiple objective functions such as economy, convenience, and collision risk.
A multi-objective optimization method, NSGA-II, was applyed to search for a solution to the avoidance algorithm. The obtained parameters was verified through a comparison of ship operation results between a mariner and an automatic ship avoidance algorithm using a ship handling simulator.

Free Research Field

船舶海洋工学

Academic Significance and Societal Importance of the Research Achievements

多くの研究は避航操船アルゴリズムのパラメータは経験的な値を設定するにとどまっている。しかし、避航アルゴリズムのパラメータの解空間は膨大であり、かつアルゴリズムの評価も「目的地点までの距離を短くしたい（経済性・利便性）」と「互いに安全な距離を保持したい（リスク）」はトレードオフの関係にあるため、適切なパラメータ探索手法が求められており、本研究はその解決に資する。