Towards an Integrated Framework for Fugitive Methane Source Localization and Monitoring in a Wireless Sensor Network

Gas leaks are the main cause of industrial fires and accidents. These cause countless fatalities, equipment damage, and other severe environmental effects. In this paper, we provide a framework for the monitoring and detection of methane leakage using a diffusion model based on the gas diffusion theory. Given that centralized Least Square methods are not efficient and robust as they require the gathering and processing of large-scale measurements on a central node. We propose a detection technique which makes use of the distributed (Non-linear) least squares method to overcome this problem. Then, a network of connected methane sensors is used to detect gas leaks. In order to estimate the parameters of the diffusive model for the gas leakage on each sensor node, a distributed recursive estimator of the consensus plus an innovation type technique is used. The characteristics being estimated include the gas source's distance, which will be effectively triangulated to determine the source's precise location. The targeted location is subsequently estimated using a location dispersed algorithm-based LS.