Efficient Paths for UAV-Based Gas Pipeline Leak Detection

Abstract

the study proposes an approach to generating unmanned aerial vehicle (UAV) paths for gas pipeline leak detection applications. The new approach uses a UAV simulation tool to optimize the search for methane leaks in pipelines. The goal is to find optimal UAV paths that maximize the probability of leak detection in the given climatic conditions.

The simulation results show a significant difference in leak detection and gas cloud localization probabilities achieved with the proposed flight path and a pipeline-parallel UAV flight path. Under 0...2 force wind, the UAV path parallel to the pipeline gives sufficient efficiency, but it is a poor strategy in wind forces of 4 points or more. In this case, the proposed approach is more efficient.

We created a simulation model for finding efficient UAV flight paths. The computational experiments confirmed the proposed model, approach and algorithms are sufficiently efficient. The UAV operator is unable to provide or even guess the exact location of a methane leak. Still, preliminary analysis and situation modeling improve UAV utilization. The number of variables considered when generating a flight path can either increase or decrease based on the results of large-scale simulations and field experiments.