Gas Source Localization Using physics Guided Neural Networks

TL;DR

This paper introduces a physics-guided neural network approach for efficient gas source localization from spatial concentration measurements, reducing reliance on costly simulations and effectively handling noisy data.

Contribution

The paper presents a novel neural network architecture that incorporates physical gas dispersion models to improve source localization accuracy and computational efficiency.

Findings

Accurately localizes gas sources with noisy measurements.

Reduces computational cost by avoiding numerical gas physics simulations.

Demonstrates robustness in real-world scenarios.

Abstract

This work discusses a novel method for estimating the location of a gas source based on spatially distributed concentration measurements taken, e.g., by a mobile robot or flying platform that follows a predefined trajectory to collect samples. The proposed approach uses a Physics-Guided Neural Network to approximate the gas dispersion with the source location as an additional network input. After an initial offline training phase, the neural network can be used to efficiently solve the inverse problem of localizing the gas source based on measurements. The proposed approach allows avoiding rather costly numerical simulations of gas physics needed for solving inverse problems. Our experiments show that the method localizes the source well, even when dealing with measurements affected by noise.