Data-driven Parameter Estimation for VTOL UAV Using Opposition-Based Pigeon-Inspired Optimization Algorithm

Abstract

Inertial physical parameter estimation for a Vertical Take-Off and Landing (VTOL) Unmanned Aerial Vehicle (UAV) is helpful to the control and calibration. A novel bio-inspired method, named opposite-based Pigeon-Inspired Optimization (OBPIO) algorithm, is proposed in this paper based on deterministic and dynamic opposition-based learning (OBL) strategies. The deterministic opposition-based Learning strategy is employed in population initialization and offspring generation to accelerate the convergence speed of the globally optimal pigeon. Meanwhile, the dynamic opposition-based learning strategy is introduced to facilitate the central pigeon of the swarm to explore the potential better region. The proposed algorithm is applied to the VTOL UAV system with the data from model tests. On the contrast, the improved OBPIO algorithm showed better performance in convergence speed and overall search ability.