Track Prediction Method of Unmanned Platform Based on Reinforcement Learning and Improved Potential Field Method

In the modern battlefield environment, unmanned platforms and unmanned clusters are playing an increasingly important role, and the analysis and prediction of unmanned platform behavior and command and control have become a hot research topic. In this paper, the track parameters of the unmanned platform in a continuous time period are used as sample data, and the historical track of the unmanned platform is fitted by the function fitting method. Based on the static obstacle information and dynamic obstacle information in the application scene, the corresponding repulsion field is constructed. Based on the empirical playback training network model, the trained network model is finally used to predict the track of the unmanned platform in the future. Simulation experiments show that this method can significantly improve the training efficiency of the unmanned platform track prediction model and the accuracy of the prediction results.