To solve the global path planning for autonomous underwater vehicles (AUV) in complex underwater environment with static seabed terrain obstacles and irregular ocean currents, the moth-flame optimization (MFO) algorithm was conducted to get the optimum energy comsumption path based on obstacle avoidance. First, a set of initial paths that meet the requirements of collision avoidance as the initial moth population were randomly generated after the search space was rasterized. Then, the moth position was updated using the MFO algorithm, and autonomous path planning for AUV was realized . At last, cruise energy consumption model for AUV is combined static current distribution model as a cost function to obtainthe final path through iterative searching. A series of simulation experiments were carried out to verify the effectiveness and superiority of the method. The experimental results show that the proposed algorithm has obvious advantages over the traditional ant colony algorithm in global autonomous path planning for AUV. The resulting path is smooth, and it shows excellent global convergence ability in different environments.