Audio-based Relative Positioning System for Multiple Micro Air Vehicle Systems

Abstract

Abstract-Employing a group of independently controlled flying micro air vehicles (MAVs) for aerial coverage missions, instead of a single flying robot, increases the robustness and efficiency of the missions. Designing a group of MAVs requires addressing new challenges, such as inter-robot collision avoidance and formation control, where individual's knowledge about the relative location of their local group members is essential. A relative positioning system for a MAV needs to satisfy severe constraints in terms of size, weight, processing power, power consumption, three-dimensional coverage and price. In this paper we present an on-board audio based system that is capable of providing individuals with relative positioning information of their neighbouring sound emitting MAVs. We propose a method based on coherence testing among signals of a small onboard microphone array to obtain relative bearing measurements, and a particle filter estimator to fuse these measurements with information about the motion of robots throughout time to obtain the desired relative location estimates. A method based on fractional Fourier transform (FrFT) is used to identify and extract sounds of simultaneous chirping robots in the neighbourhood. Furthermore, we evaluate our proposed method in a real world experiment with three simultaneously flying micro air vehicles.

Design model of microphone arrays for multirotor helicopters

“…α α α i (φ M , p p p) and so on. Selecting the right hand side of (11) as an estimate of the maximum rotor speed, it can be substituted into (9) to evaluate the noise at microphones. Therefore, the cost function of a mic m to optimize can be written as…”

“…The point where the the external force k k k effects is approximated by the farthest point of the microphone array. Figure 5 shows the noise model (9) and the ratio of the estimated maximum rotor speed divided by the hovering speed ω h . It is different from the result in Fig.3 that the noise model shows a convex curve that has its minimum around l ≈ 1.5m, which tells that the microphone array can not be designed arbitrary large in order to attenuate the noise.…”

“…Basiri [8] proposed to utilize artificial tone, whistle, as the target signal to be localized by a fixed-wing UAV, and to process the acoustic signal by a particle filter to fuse estimated sound bearing information derived from Time Difference of Arrival (TDOA) based method. This approach was extended to relative localization of multiple UAVs [9].…”

On-Board Relative Bearing Estimation for Teams of Drones Using Sound

“…This bearing information could be used to achieve tasks such as collision avoidance [27], formation control [28] and cooperative positioning [29]- [31]. Two solutions are presented and separately validated.…”

Audio-based localization for swarms of micro air vehicles

“…An audio-based localization system satisfies the imposed constraints on MAVs and furthermore have the advantage of being independent of illumination, weather conditions, such as fog, smoke and dust, and possible occlusions caused by obstacles or other MAVs. In previous works we showed the success of using sound to locate acoustic targets on the ground [2] and to obtain the relative position between members of a multi-MAV system [9]. Such systems could potentially be exploited to perceive other non-cooperative noise emitting aerial platforms.…”