3.3.1. Model Structure

Figure 5 shows the model structure of unmanned ships path planning based on the DDPG algorithm. The model mainly includes AC algorithm, Environment (Ship Action Controller and Ship Navigation Information Fusion Module), and Experience Replay Memory. Among them, the model obtains the environmental information and the ship's state data through the Ship Navigation Information Fusion Module, and it serves as the input state of the AC algorithm. The optimal ship action strategy is output, which satisfies the ship's maximum cumulative return during the learning process, by randomly extracting data from the experience buffer pool for repeated training and learning. Finally, unmanned ships can avoid obstacles and reach the destinations with the help of the Ship Action Controller.

**Figure 5.** Structure of DDPG algorithm based unmanned ships path planning model.
