Autonomous Underwater Vehicle (AUV) plays an important role in understanding; exploring and exploiting ocean resources; in order to ensure the safety of AUV navigation and improve its working efficiency; it is necessary to conduct path planning research for AUV. Aiming at the path planning problem of AUV in three-dimensional complex environment; an NSGA-II algorithm suitable for global path planning is proposed. Based on the actual situation; a simple and effective three-dimensional seabed environment model is established; and the path length and threat evaluation function are designed according to the principles of economy and safety; and the path points are selected through fast Pareto sorting and congestion distance calculation. Finally; in order to ensure the practicability of the path; the gradient descent method is used to smooth the path. The simulation results show that the improved algorithm can not only safely avoid seabed terrain obstacles in the three-dimensional environment; but also the ability to find the optimal solution and the convergence speed are greatly improved.
With the development of society and economy; the value of the ocean is increasingly being valued by humans. The creation of autonomous underwater vehicle provides a new solution for human to explore and exploit the marine resources (Feng; 2000; Fiorelli; 2006; Xiao; 2014). Autonomous underwater vehicle is a new generation of underwater robots; which has the advantages of good maneuverability; safety; and intelligence; and can complete various underwater tasks. Path planning research is one of the key technologies for autonomous underwater vehicle to complete task.
In the past few decades; many methods have been applied to the path planning research of autonomous underwater vehicle; such as dijkstra algorithm; A* algorithm; grid method; artificial potential field algorithm; fuzzy method and particle swarm optimization algorithm (Zhu; 2019; Zhang; 2019; Pan; 2017; Wang; 2015). Path planning can be divided into two categories: global path planning and local path planning. Global path planning is to find a feasible path in a known environment; usually the optimized path is feasible. However; local path planning is suitable for the unknown environment or the partly unknown environment. In this case; the optimized path is not always available.