Quantitative analysis of heavy metals in soilviahierarchical deep neural networks with X-ray fluorescence spectroscopy

Abstract

Potentially toxic metals in soils are accurately predicted using hierarchical depth neural networks by X-ray fluorescence spectroscopy.

Quantitative analysis of potentially toxic elements in soil by XRF based on efficient reinforcement learning and sparse partial least squares

“…To obtain a suitable count rate and characteristic peak resolution, the XRF spectra of soil samples were collected using the spectrometer at 45 kV tube voltage, 26 μA current, and 90 s measurement time. 34 Due to the inherent heterogeneity and complexity of soil samples, spectra collected from the same samples will have varying degrees of fluctuating behaviour. To reduce the error caused by such fluctuation, we scanned each sample three times and took the average value as the final XRF spectral data for analysis.…”

A deep spectral prediction network to quantitatively determine heavy metal elements in soil by X-ray fluorescence

“…9 Moreover, XRF has the ability of real-time detection and can be used for in situ or ex situ measurements. Due to its promising applications, XRF technology has been widely utilized in soil analysis, 10 geological exploration 11 and quality monitoring. 12 However, the elemental composition of soils is extremely complex.…”