Dynamic Modeling and Control for a Collision-Resilient Tensegrity Aerial Vehicle

The tensegrity aerial vehicle dexterously integrates the tensegrity structure and drones, embracing transitions between ground locomotion and aerial motions while maintaining resilience against collisions. However, there still lacks analytically tractable dynamic models and theoretically provable control methods to attain efficient maneuvering performance. Hence, in this article, a comprehensive dynamic model of the tensegrity aerial vehicle is first derived by employing the Euler-Lagrange formalism. In accordance with the construction manner of the robot prototype, we initially formulate the interconnection relationship between the two subsystems, with which their dynamics are merged into a coherent framework. In addition, the environmental interactions, including ground contact and friction effects, are also taken into account to refine the dynamic model's accuracy. Then model-based controllers are designed for the resulting underactuated system, utilizing second-order sliding mode control techniques. Finally, the precision of the derived model and the effectiveness of the tailored control scheme are validated through simulations and experiments conducted on the tensegrity aerial vehicle.