Unmanned surface vessel (USV) autonomous navigation on the open sea involving real-time path planning and collision avoidance is still one of the essential problems to ensure the USV’s safe and efficient navigation. Especially in a congested and uncertain marine traffic environment, not only will static obstacles be taken into account but other target vessels in motion should also be considered. Also, the general requirement of the Convention on the International Regulations for Preventing Collisions at Sea (COLREGs) should be satisfied. Hence, an improved A-star algorithm for USV path planning and improved dynamic window approach (IDWA) for collision avoidance were proposed. First, considering the requirement of COLREGs, the velocity search space was filtered again, and the quantity of USV trajectories was reduced. Then, the improved A-star algorithm was introduced to let the USV avoid static obstacles and reach its destination without trapping in local optimization. Moreover the Deep Q-network method was utilized to train weight coefficients of the IDWA objective function. Thereby, the improved algorithm-generated path during the process of collision avoidance was more reasonable and safer. To verify feasibility of the proposed path-planning algorithm, a comparison experiment with the traditional DWA method was carried out. The results showed that whether it was for a single USV to a single target or for multiple USVs to multiple targets, path planning, the proposed method, could work effectively to avoid obstacles safely and reach the destination quickly. The improved algorithm will be expected to provide a reference for USV path planning and collision avoidance as well as contribute to the implementation of autonomous ship navigation.

The congestion of waterways can easily lead to traffic hazards. Moreover, according to the data, the majority of sea collisions are caused by human error and the failure to comply with the Convention on the International Regulation for the preventing Collision at Sea (COLREGs). To avoid this situation, ship automatic collision avoidance has become one of the most important research issues in the field of marine engineering. In this study, an efficient method is proposed to solve multi-ship collision avoidance problems based on the multi-agent reinforcement learning (MARL) algorithm. Firstly, the COLREGs and ship maneuverability are considered for achieving multi-ship collision avoidance. Subsequently, the Optimal Reciprocal Collision Avoidance (ORCA) algorithm is utilized to detect and reduce the risk of collision. Ships can operate at the safe velocity computed by the ORCA algorithm to avoid collisions. Finally, the Nomoto three-degrees-of-freedom (3-DOF) model is used to simulate the maneuvers of ships. According to the above information and algorithms, this study designs and improves the state space, action space and reward function. For validating the effectiveness of the method, this study designs various simulation scenarios with thorough performance evaluations. The simulation results indicate that the proposed method is flexible and scalable in solving multi-ship collision avoidance, complying with COLREGs in various scenarios.

The increasing congestion of maritime traffic leads to frequent maritime traffic accidents. Unmanned search and rescue equipment provides a guarantee for rapid rescue operations. In marine accidents, the waiting time is short and the rescue risk factor is high. Unmanned surface vessel (USV) can transport rescue equipment to the accident site in the first time and carry out searching and rescuing, which will greatly improve rescue efficiency. With the development of technology, USV can autonomously conduct area exploration to assist in the discovery of survivors. This paper proposes a complex area exploration of multiple unmanned surface vessels based on the Proximal Policy Optimization algorithm (PPO). The PPO algorithm is used to improve training efficiency, achieve 100% search coverage in complex water surface areas, and make searching and rescuing operations safe, simple and efficient. Experimental simulations show that the algorithm can realize 100% area exploration, have high search efficiency, and consumes less energy. This is a new intelligent ship application that can complete 100% exploration without human involvement.