Multi-AUV (Multi-autonomous underwater vehicle) formation control is one of the research hotspots of marine resources exploration technology, and it is also the key technology to realize the large-scale marine resources exploration. It is of great research value and practical significance to effectively control multiple AUV formation to carry out underwater exploration orderly. Aiming at the demand of marine resources exploration, the deep-sea resource exploration problem of AUV formation is deeply explored and studied in this paper. The AUV formation strategy and formation control method are proposed respectively for the large-scale ocean exploration mission and the local high-precision exploration mission which are the subdivided from in deep-sea resource exploration. Main research contents of this paper are follows: According to the characteristics of the marine resource exploration, the problem of underwater exploration can be subdivided into two problems: large-scale marine exploration and local high-precision seabed exploration. And some key in factors like, AUV carrier, sensor, communication, navigation, and environment in the marine resource detection system of AUV formation are modeled. In addition, combined with the process and system decision frame of AUV formation marine resource exploration, the ideas of AUV formation exploration is expounded. Focusing on the characteristics of large-scale ocean exploration, this paper proposes a multiple AUVs formation strategy. The system states of the AUV formation are divided into initial states, scanning states, detection states, formation recombination and termination states. After receiving the command, the AUV formation enters the scanning state. The AUV formation adjusts the distance between adjacent AUV according to the characteristics of the target area and the operating environment. then increases the coverage area during the AUV parallel scanning. When the target area is found, the AUV formation enters the detection state. According to the target information and the AUV status information, the USV decides an AUV follows the detection path along the petal-shaped to carry out high-precision detection for the target area. The AUV will continue to work together after the AUV completes its detection mission. Since the AUV formation needs the USV to provide the USBL positioning support, the USV needs to track the AUV formation in time, and there is a designed tracking control law for the AUV formation. And the feasibility and effectiveness of the proposed formation strategy and tracking control law are verified by simulation. Taking the constraints of AUV formation in actual operating environment into consideration, the height between AUV and seabed should be maintained in vertical direction except for keeping the formation shape in the horizontal plane for getting a better detection effect. The proposed AUV formation control method in this paper decouples formation control into horizontal plane and vertical plane respectively. In the horizontal plane, the formation control law based on leading-following and feedback linearization method is proposed and proved by the lyapunov stability theory. In the vertical plane, the least square method is used to fit the estimation of the sea bottom topographic height data. At the same time, the corresponding obstacle avoidance strategies are proposed in case of the possible extreme terrain. Apart for that, the formation optimization strategy is proposed for the lawnmower path of AUV formation. Finally, the effectiveness of the proposed AUV formation control method and strategy are simulated and verified.