The path planning of collision avoidance for an unmanned ship navigating in waterways based on an artificial neural network

Abstract: Designing a safe, collision-free navigation route is critical for unmanned ships. This article applies the path planning method to the generation of collision avoidance paths for unmanned ships. Since the path length function is obtained from the distribution points constructed in space, it is necessary to transfer the safe domain of the unmanned ship to the obstacle, treating the unmanned ship as a particle. Then, the constructed artificial neural network (ANN) is applied to compute the collision penalty function for distribution points and obstacles. Furthermore, an evaluation function including the path length function and collision penalty function is designed, and the optimal path is obtained by computing the minimum value of the evaluation function. Meanwhile, the simulated annealing method is introduced to optimize the activation function of the output layer of the ANN to improve its classification performance and suppress the local minima problem. Finally, the application of ANN in ship autonomous dynamic collision avoidance path planning is demonstrated in two types of experiments. Among them, when avoiding static obstacles, the minimum safe passing distance between the two ships reaches 30 m; when avoiding dynamic obstacles (navigating ships), the minimum safe passing distances between the two ships in the head-on situation and the overtaking situation are 378 and 430 m, respectively.

Object Details

• Location: Deutsche Nationalbibliothek Frankfurt am Main
• Extent: Online-Ressource
• Language: Englisch

References and Relationships

• Bibliographic citation: The path planning of collision avoidance for an unmanned ship navigating in waterways based on an artificial neural network ; volume:11 ; number:1 ; year:2022 ; pages:680-692 ; extent:13 ; Nonlinear engineering ; 11, Heft 1 (2022), 680-692 (gesamt 13)

Contributors, Places and Time

• Creator: Wang, Renqiang ; Miao, Keyin ; Li, Qinrong ; Sun, Jianming ; Deng, Hua

Further information

• DOI: 10.1515/nleng-2022-0260
• URN: urn:nbn:de:101:1-2022122213132712415715
• Rights: Open Access; Der Zugriff auf das Objekt ist unbeschränkt möglich.
• Last update: 15.08.2025, 7:37 AM CEST