Aerospace and Electronic Systems Magazine September 2016 - 40

Feature Article:

DOI. No. 10.1109/MAES.2016.150166

An Advanced Sense and Collision Avoidance Strategy
for Unmanned Aerial Vehicles in Landing Phase
Yu Fu, Youmin Zhang, Concordia University, Montreal, Quebec, Canada
Xiang Yu, Hunan University, Changsha, Hunan, China

INTRODUCTION
Unmanned Aerial Vehicles (UAVs) become promising in different civilian and military applications due to their lower cost and
greater flexibility in comparison to manned aircraft. However, the
growing diversity of flight makes UAVs vulnerable to mid-air collisions. To ensure the safety for manned aircraft, human pilots are
responsible for detecting intruders in airspace and performing appropriate maneuvers to avoid collisions [1]. Unlike manned aircraft with human pilots involved, UAVs must be equipped with
Sense and Avoid (SAA) systems to guarantee the flight safety [2].
Thus, SAA systems play an important role in merging UAVs into
the National Airspace System. As illustrated in Figure 1, an SAA
scheme is composed of four units, which are sensing, conflict detection, collision avoidance, and flight controller, respectively.
The sensing technologies to provide traffic information of the
surrounding environment for UAVs rely on existing infrastructures, such as cooperative sensors (Traffic Alert and Collision
Avoidance System, Automatic Dependent Surveillance Broadcast, Airborne Collision Avoidance System [3], [4], and non-cooperative sensors (e.g., radar systems, cameras, infrared sensors)
[5]-[9]. Based on the sensed data, the conflict detection function
extracts useful information and projects the states into the future to
determine whether a potential conflict will occur or not [10]. When
a potential collision is predicted to take place, the parameters of
the conflict (position of predicted collision and time of possible
conflicts) are transferred to the collision avoidance unit. This unit
Authors' current addresses: Y. Fu, Y. M. Zhang, Concordia
University, Department of Mechanical and Industrial, Engineering, 1455 Maisonneuve Blvd. W., Montreal, Quebec
H3G 1M8 Canada; X. Yu, College of Mechanical and Vehicle
Engineering, Lushan South Road, Changsha, Hunan,
410082 China; Y. M. Zhang was also with the Shaanxi Key
Laboratory of Complex System Control and Intelligent
Information Processing, Xi'an University of Technology, #5
Jinhua South Road, Xi'an, Shaanxi, 710048 China. Corresponding author is Y. M. Zhang at E-mail: (youmin.zhang@
concordia.ca).
Manuscript received July 31, 2015, revised February 6, 2016,
and April 11, 2016, and ready for publication April 14, 2016.
Review handled by G. Fasano.
0018-9251/16/$26.00 © 2016 IEEE
40

is applied to generate trajectories in order to resolve multiple collisions without incurring too much cost.
A large amount of heuristic collision avoidance approaches
have been proposed with application to UAVs in the literature.
However, there are several challenges in the existing literature: 1)
an investigation by the Air Safety Foundation reveals that 45% of
collisions occur in the traffic pattern, and more than 67% of these
collisions occur during landing when aircraft are at the final landing stage or over the runway [11]. Moreover, many civilian UAV
applications are performed at low altitude airspace, increasing the
possibility of conflicting with other aircraft during a landing period. To ensure a safe and successful landing, it is desirable that
the SAA systems perform necessary maneuvers to avoid potential
collisions with intruder aircraft; 2) one of the most practical factors
in real flights is that the dynamic constraints have to be respected.
In practice, there are allowable bounds related to the control inputs
and system states (e.g., control surface deflections, pitch rate). To
guarantee a safe flight, the produced trajectory should follow the
rules without violating these limits; and 3) traditional heuristic algorithms are easily trapped into a local minimum after updating the
solution. The generated trajectory may not be optimal due to the
constraints of the algorithms.
Motivated by the above-mentioned factors, this study develops an optimal path through a Biogeography-Based Optimization
(BBO) approach to avoid multiple threats in the UAV landing period. Compared with the heuristic algorithms, BBO provides a high
convergence rate and derives an optimal strategy within a short
spectrum of time without trapping into a local minimum. Additionally, to consider the dynamic constraints during the collision avoidance application, the differential flatness algorithm is developed
to smoothen the planned trajectory. Furthermore, by integrating a
passivity based control technique with singular perturbation ideas,
an energy-based controller is designed to ensure that the UAV can
follow the planned path to complete the collision avoidance. The
main contributions of this article include: 1) considering the safety
requirement of UAV landing, which is identified as the most complex and challenging period during UAV flight and the SAA mission, BBO is employed to produce an optimal solution for avoiding
multiple collisions; 2) with respect to the path planner, the UAV
constraints are accounted for such that the generated trajectory is
feasible for the UAV; and 3) the control design derived from energy function is exploited for path following.

IEEE A&E SYSTEMS MAGAZINE

SEPTEMBER 2016

Table of Contents for the Digital Edition of Aerospace and Electronic Systems Magazine September 2016