Generative-Model-Based Autonomous Intelligent Unmanned Systems
Zhijun Zhang, Zhentao Wu, Ren Ge · 2023
In order to improve the autonomous decision-making ability and dynamic environment adaptability of unmanned systems, a universal framework of generative intelligent unmanned systems (GIUS) is proposed and designed. The GIUS framework consists of a human-machine interaction module, an environment perceiving module, a task generation and autonomous decision-making module, and a motion planning module. By obtaining the overall task description through human-machine interaction, GIUS can autonomously perceive and predict environment task information through generative models. Based on the understanding of the environment, the unmanned system can achieve task generation and decision-making independently, and realize path generation and motion generation. Under the framework of GIUS, unmanned systems can perform various tasks in complex environments with minimal or no human intervention. Compared to existing unmanned systems, generative intelligent unmanned systems have significant improvements in autonomy, intelligence, and efficiency.