Trajectory Optimization of Dynamic Soaring Considering Rigid Body and Actuator Dynamics Using Warmstarting

Achieving sustained autonomous dynamic soaring requires precise trajectory planning and tracking. Typically, simplified point mass and aerodynamic models are utilized for trajectory planning. As dynamic soaring consists of a consecutive sequence of highly agile maneuvers, the significance rotational and actuator dynamics within the planning process to achieve acceptable tracking and energy extraction performance should be investigated. Hence, this work proposes an offline trajectory planner based on optimization of the rigid body equations of motion, considering actuator dynamics and a two-point aerodynamic model. Due to the complexity of the model, convergence without a reasonable initial guess was found to be difficult. Consequently, a warmstarting technique based on closed-loop simulations of dynamic soaring cycles is implemented. Eventually, the optimal trajectories are assessed regarding energy extraction performance. For a given tracking controller, a comparison to trajectories planned based on point mass dynamics revealed that the consideration of rotational and actuator dynamics within the planning process to achieve the aforementioned goals is not of greater significance.     «

Achieving sustained autonomous dynamic soaring requires precise trajectory planning and tracking. Typically, simplified point mass and aerodynamic models are utilized for trajectory planning. As dynamic soaring consists of a consecutive sequence of highly agile maneuvers, the significance rotational and actuator dynamics within the planning process to achieve acceptable tracking and energy extraction performance should be investigated. Hence, this work proposes an offline trajectory planner base...     »