A human-like collision avoidance method for autonomous ship with attention-based deep reinforcement learning

Jiang, Lingling; An, Lanxuan; Zhang, Xinyu; Wang, Chengbo; Wang, Xinjian

doi:10.1016/j.oceaneng.2022.112378

Cited by 37 publications

(5 citation statements)

References 16 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Lucas [21] constructed a safe encounter distance model by analyzing the principles of radar target parameter calculation and ship collision risk assessment. Jiang [22] proposed a new method for autonomous ship collision avoidance based on attention based deep reinforcement learning (ADRL). This study focuses on open water areas, using a comprehensive 𝑆 mathematical model.…”

Section: ( )mentioning

confidence: 99%

Research on Multi vessel Collision Avoidance Decision Considering China's Inland River Collision Avoidance Rules

Xiao,

Li,

Jiang

et al. 2024

Preprint

View full text Add to dashboard Cite

In order to quickly develop a multi ship collaborative collision avoidance plan in multi ship encounter scenarios, a multi ship collision avoidance decision model is proposed in accordance with the requirements of the International Maritime Collision Avoidance Regulations, taking into account the collision avoidance operations in multi ship encounter scenarios. Firstly, the collision risk between ships is determined by the closest encounter distance (DCPA) and closest encounter time (TCPA). Secondly, a decision objective function is established based on rules, degree of ship deviation, and ship risk. Finally, the prediction mechanism is used to calculate the minimum value of the objective function and complete the collision avoidance decision. The simulation results show that in the scenario of multiple ship encounters, the collision avoidance model can achieve safe collision avoidance with a smaller steering frequency and a shorter col-lision avoidance time. The proposed multi ship collision avoidance model can provide reference for collision avoidance decision-making in multi ship encounter scenarios, which helps to im-prove the collision avoidance efficiency in multi ship collision avoidance

show abstract

Section: ( )mentioning

confidence: 99%

Research on Multi vessel Collision Avoidance Decision Considering China's Inland River Collision Avoidance Rules

Xiao,

Li,

Jiang

et al. 2024

Preprint

View full text Add to dashboard Cite

show abstract

“…In recent years, intellectualization has become the trend of ship development, and the development of artificial intelligence, big data and other emerging technologies has fuelled the research progress of intelligent ships. Intelligent ships have made great progress in the fields of path planning [1] , autonomous collision avoidance [2] , and automatic berthing [3] , etc., but the research on anchoring safety and early warning of the risk of dragging anchor for intelligent ships is still in its infancy. Anchoring is one of the routine operations of a ship, and the risk caused by intelligent ships that failed to detect and control anchoring in time can't be ignored.…”

Section: Introductionmentioning

confidence: 99%

Risk early warning method of ship anchor dragging based on simulation experiments

Chen,

Xi,

Jiang

2024

Third International Conference on Intelligent Traffic Systems and Smart City (ITSSC 2023)

View full text Add to dashboard Cite

In response to the issue of limited timeliness in the current monitoring of ship anchor dragging, a method of early warning of ship anchor dragging risk utilising fuzzy set theory was proposed. Based on sailing simulator, ship anchoring simulation experiments were conducted to collect trajectory data of ship anchoring under various environmental conditions, and the characteristics of the trajectory of normal anchored ships were analyzed. During a yawing motion, the yawing width, ground speed and heading change rate of were established as early warning indicators for ship anchor dragging risk. Triangular fuzzy numbers and the distribution of early warning indicators obtained from simulation experiments were combined to determine the membership functions of each indicator, and the entropy weight method was used to determine indicator weights. Then ship anchor dragging risk was comprehensively evaluated and the anchor dragging probability was quantified. Warning levels were set according to actual conditions, and the warnings will be alarmed if the anchor dragging probability reaches the corresponding warning level. The results indicate that during the yawing motion of anchored ships, larger yawing width, higher ground speed, and smaller ship heading change rate correspond to higher anchor dragging probability. Finally, sailing simulators were utilized for simulation experiments ,and the feasibility of the proposed method was validated. The method exhibits practical operability and demonstrates favorable timeliness in early warning capabilities.

show abstract

“…Based on the experience and lessons learned, a high-level architecture of collaborative collision-avoidance protocol is presented. Jiang et al [31] proposed a decision model based on deep reinforcement learning (DRL) and considering the attention distribution mechanism of ship pilots, in which assessment of collision risk and planning of ship movement was mainly considered. Rothmund et al [32] proposed a method using dynamic Bayesian networks to infer the intention of collision-avoidance actions of other ships in open waters, further improving the safety of results.…”

Section: Introductionmentioning

confidence: 99%

A Construction Method of a Sequential Decision Chain for Unmanned-Ship Autonomous Collision Avoidance Based on Human-Like Thinking

Wang,

Wang

et al. 2023

JMSE

View full text Add to dashboard Cite

As one of the key technologies restricting the development of intelligent ships, autonomous collision avoidance has attracted the attention of many scholars all over the world. Existing research on collision-avoidance behavior focuses more on collision risk assessment and local path-planning methods for studies on the human-like sequential logic of the whole collision-avoidance process, as well as the decision-making process of various stages. Further in-depth thinking is needed urgently. Based on this, a construction method of a human-like sequential decision chain for the autonomous collision avoidance of unmanned ships is proposed through the construction of a collision-avoidance rule base and strategy set, efficient data access based on the Knowledge Graph concept, global collision risk assessment considering sequential decision process, and the construction of a complete collision-avoidance logic process to simulate the decision-making process of humans in complex multi-ship encounters in open waters. For multi-ship encounter scenarios, considering the sequential decision-making process of collision avoidance, a method was proposed to divide the collision risk of the target ship into direct collision risk and potential collision risk. The validity and reliability of the constructed sequential decision chain are verified by simulation experimental results. The results show that the method is effective for collision avoidance (especially multi-ship collision avoidance) in open waters and can provide a theoretical basis and technical support with good interpretability for the decision-making process of an unmanned ship’s autonomous collision avoidance.

show abstract

A human-like collision avoidance method for autonomous ship with attention-based deep reinforcement learning

Cited by 37 publications

References 16 publications

Research on Multi vessel Collision Avoidance Decision Considering China's Inland River Collision Avoidance Rules

Research on Multi vessel Collision Avoidance Decision Considering China's Inland River Collision Avoidance Rules

Risk early warning method of ship anchor dragging based on simulation experiments

A Construction Method of a Sequential Decision Chain for Unmanned-Ship Autonomous Collision Avoidance Based on Human-Like Thinking

Contact Info

Product

Resources

About