Intelligent Satellite Control for Multi-Target Staring Imaging

Current CMOS chip technology is becoming increasingly advanced, and more remote sensing small satellites are utilizing planar array CMOS cameras to achieve staring imaging of multiple targets. The selection of imaging paths and the accuracy of imaging time control for satellite staring imaging of numerous targets are influenced by the satellite's relative position to the observation targets and its maneuvering capabilities. In this paper, we establish a satellite attitude dynamics model and design an adaptive controller for target staring control. We also present a path selection strategy for multi-target planning, which incorporates the constraints of satellite imaging. Additionally, we propose an intelligent control approach based on deep reinforcement learning to optimize control parameters and imaging paths, enhancing multi-target staring imaging time and stability for optimal benefits. Our solution offers an intelligent method for practical engineering applications involving multi-target staring imaging. Finally, the mathematical simulation results show that this method can improve control stability and best imagine time.