Adaptive Load Optimization and Precision Control Scheme for Vertical Landing Rockets with Sparse Sensing Data

Reusable rockets face fundamental challenges regarding recovery safety, and high-altitude winds affect the aerodynamic loads, flight attitudes, and trajectories of rockets during reentry and aerodynamic deceleration. Difficulties remain in the accurate real-time assessment of wind conditions at the predicted landing site, and existing methods are limited in their real-time wind estimation capabilities and performance, requiring new measurement and additional computational resources. This study proposes a methodology for estimating and predicting wind fields in real-time using deep learning. The proposed methodology uses rocket attitude angles and apparent acceleration variations as inputs to train a specific deep neural network. Additionally, a non-recursive, simplified, high-order sliding mode control (HOSM) technique is proposed to

resolve the complex attitude control of reentry and landing of reusable launch vehicles (RLV) during recovery. This control technique is designed to achieve finite-time convergence of online wind disturbance compensation. The proposed methodology also provides a comprehensive dynamic model of the attitude control of the RLV during recovery. Based on the dynamic model, homogeneity theory is used to design a non-recursive homogeneous high-order sliding mode controller to ensure finite-time tracking control for the recovery of RLVs, effectively damping chattering and guaranteeing optimal control.

Message: Hi FE, I need help with “Adaptive Load Optimization and Precision Control Scheme for Vertical Landing Rockets with Sparse Sensing Data” in “Aerospace Engineering”