Constraint Management For Quadcopter Drones: Reference Governor-Based Approaches
Yudan Liu, Yasaman Pedari, Hamid R. Ossareh · 2022 American Control Conference (ACC) · 2022
This paper presents two constraint management strategies based on Reference Governors (RG) to enforce output and state constraints on nonlinear dynamics of quadcopter drones. The first solution, referred to as the nonlinear decoupled reference governor (NL-DRG), is based on the sequential application of three bisection searches and online numerical simulations to find constraint-aware reference commands for each of the pitch, roll, and yaw channels. While NL-DRG performs well, it relies on online simulations and is thus computationally demanding. For applications where a smaller computational footprint is desired, we propose a second solution, which we refer to as the modified reference governor (M-RG). This solution consists of the sequential application of three scalar reference governors based on linear prediction models, each robustified to account for the worst-case linearization error and coupling behavior. The M-RG can guarantee constraint satisfaction for the quadcopter dynamics while maintaining the highly-attractive computational features of scalar reference governors. However, the performance of M-RG may be more conservative than that for NL-DRG, so a trade-off between computation and performance can be made. Finally, a quantitative comparison between all the methods is presented in terms of performance and computation time.