Research on Particle Swarm Optimization algorithm incorporating Entropy Weight Method for UAV gas plume tracking subtask

Abstract

Unlike ground robot that performs Gas Source Localization task (GSL) in a two-dimensional plane, UAV can be applied to various three-dimensional space scenarios due to their flexibility. However, due to the interference of wind speed and wind direction, the gas diffusion distribution is uneven, resulting in different confusion degrees in the gas distribution. Based on this, this paper proposes a Particle Swarm optimization algorithm (PSO) incorporating Entropy Weight method (EWM) for UAV gas plume tracking strategy, which referred to as EWM-PSO algorithm below, and the Gaussian gas plume model in three-dimensional space is used to simulate the interference between wind speed and wind direction. As for this algorithm, EWM is used to calculate the entropy value and comprehensive score at each position, PSO uses the comprehensive score as its fitness value, which will guide the UAV move towards the position with high concentration value, i.e., the position of the gas plume source. After setting up two groups of comparative experiments, the results verify that this algorithm has high universality and accuracy.