I. Introduction
Autonomous underwater vehicles (AUVs) are qualified for a wide range of subsea missions [1], such as deep surveys of hydrothermal vent fields [2] and underwater communication [3]. Due to the limitation of a single AUV in terms of energy and capacities, multiple AUVs usually cooperatively work to complete complex missions in a large underwater environment [4]. The target missions can be decomposed into multiple subsets and each subset is assigned to one AUV for execution. For example, in monitoring scenarios, a group of AUVs are used to visit particular targets in parallel for quickly investigating the whole area [5]; in tactical edge networks, AUVs work together to achieve a high degree of coordination and communications [6]. To save energy, AUVs are usually launched from movable platforms, e.g., unmanned surface vehicle, on the ocean rather than a coast. The movable surface vehicles and AUVs form a multi-AUV system to fulfill target missions. To build such a system, the path planning of multiple AUVs is one of challenging problems, especially in a large underwater environment with complex bathymetry, irregular obstacles, and risky areas.