LIDAR obstacle warning and avoidance system for unmanned aerial vehicle sense-and-avoid

摘要

The demand for reliable obstacle warning and avoidance capabilities to ensure safe low-level flight operations has led to the development of various practical systems suitable for fixed and rotary wing aircraft. State-of-the-art Light Detection and Ranging (LIDAR) technology employing eye-safe laser sources, advanced electro-optics and mechanical beam-steering components delivers the highest angular resolution and accuracy performances in a wide range of operational conditions. LIDAR Obstacle Warning and Avoidance System (LOWAS) is thus becoming a mature technology with several potential applications to manned and unmanned aircraft. This paper addresses specifically its employment in Unmanned Aircraft Systems (UAS) Sense-and-Avoid (SAA). Small-to-medium size Unmanned Aerial Vehicles (UAVs) are particularly targeted since they are very frequently operated in proximity of the ground and the possibility of a collision is further aggravated by the very limited see-and-avoid capabilities of the remote pilot. After a brief description of the system architecture, mathematical models and algorithms for avoidance trajectory generation are provided. Key aspects of the Human Machine Interface and Interaction (HMI2) design for the UAS obstacle avoidance system are also addressed. Additionally, a comprehensive simulation case study of the avoidance trajectory generation algorithms is presented. It is concluded that LOWAS obstacle detection and trajectory optimisation algorithms can ensure a safe avoidance of all classes of obstacles (i.e., wire, extended and point objects) in a wide range of weather and geometric conditions, providing a pathway for possible integration of this technology into future UAS SAA architectures. (C) 2016 Elsevier Masson SAS. All rights reserved.