A Comparative Study of TDLAS-based UAV Plume Tracing Algorithms

In this paper, we proposed a method that can be applied to solve the problem of plume tracing, in which a TDLAS gas sensor was installed on an unmanned rotorcraft to perform conical rotation scanning (CRS) to search for the plume, and compared with the Spiral-Surge algorithm. The TDLAS was installed on a three-axis aerial gimbal and it can maintain an angle to the vertical direction. While the drone is powered on and flying, by adjusting the yaw of the gimbal, the TDLAS continuously rotates and scans the three-dimensional space below the UAV. According to the time series of concentration line integrals of the odor detected in real-time during the flight, the UAV plans the path for finding the plume and tracing the plume, gradually approach the odor source and finally estimate its location. In the comparison on the CPTS-3D simulation platform, the detection direction of the TDLAS was a constant angle to the vertical direction in the CRS algorithm, whereas in the spiral-surge algorithm, the detection direction of TDLAS kept perpendicular to the ground. The simulation results show that, compared to the Spiral-Surge algorithm, the CRS algorithm proposed in this paper can find the plume in less time and has a higher success rate, and takes less average time to complete the whole odor source search process.