Classifying grains using behaviour-informed machine learning

Laudari, Sudip; Marks, Benjy; Rognon, Pierre

doi:10.1038/s41598-022-18250-4

Cited by 5 publications

(2 citation statements)

References 23 publications

(22 reference statements)

Supporting

Mentioning

Contrasting

Order By: Relevance

“…More generally, the flow dynamics of bi-disperse mixtures was found to depend on their size ratio [17,18]. Finally, we evidenced in our recent studies that contrast in physico-mechanical properties of particles including size, density, Young's modulus and cohesion, could be detected from their trajectories and acceleration using machine learning classifier [19,20].…”

Section: Introductionmentioning

confidence: 56%

Using tracer particle kinematics to predict particle size in rotating drums

Laudari,

Marks,

Rognon

2024

Preprint

Self Cite

View full text Add to dashboard Cite

Comminution is an energy intensive process. In ball mills, it is achieved by rotating a drum in which large metal balls crush ore particles. In-situ monitoring of particle size would be of considerable interest to optimize their operation. However, there is no established solution to measure particle size in such a harsh mechanical environment. We show here that the acceleration of the grinding media, which can be monitored using embedded accelerometers, can be used to sense the particle size during operation. In DEM simulations, we find that a machine learning classifier is able to detect the size of small "ore" particles solely based on the knowledge of the acceleration of larger grinding media particles. Results show that this kinematic sensing is effective over a wide range of particle size ratios, mixture ratio and mill charge. Beyond their practical interest in mineral processing, these results point out that the kinematics of large particles is affected by the size of the smaller particles, an observation which can help advance rheological models for bi-disperse granular flows.

show abstract

Section: Introductionmentioning

confidence: 56%

Using tracer particle kinematics to predict particle size in rotating drums

Laudari,

Marks,

Rognon

2024

Preprint

Self Cite

View full text Add to dashboard Cite

show abstract

“…Recently some review articles have published in the same area [1], [2], [3], [4], [5], [6], [7], [8], [9], [10], [11], [12], [13], [14], [15], [16], [17], [18], [19], [20] creating an exchange platform for cost effective engineering solutions in sensor based ore sorting. More beneficiation and advanced techniques for minerals like copper, coal, and diamond have also been proposed in [21], [22], [23], [24], [25], [26], [27], [28], [29], [30], [31], [32], [33], [34], [35], and [36], but none of these approaches intently reviews a generalized approach to sensor based ore sorting through the application of the interaction of electromagnetic waves with matter.…”

Section: Introductionmentioning

confidence: 99%

Sensor-Based Ore Sorting—A Review of Current Use of Electromagnetic Spectrum in Sorting

et al. 2022

View full text Add to dashboard Cite

Sensor-based sorting has had a wide range of industrial use in automating and speeding up the process which requires substances or objects to be segregated from each other. The high demand for goods including raw materials, food, minerals, and waste recycling has increased the pressure for high-speed sorting. The first part of this paper presents a comprehensive theoretical and practical survey and comparison of current sorting methods relying on the use of the electromagnetic spectrum. Sorting methods are classified among other things, which portion of the EM spectrum is it, background noise rejection capacity, sample size limitations, sample chemistry limitations, sample surface cleanliness, spatial resolution capacity, spectral resolution, and feed rate limitations. The analysis focuses on color or visible light sorting, gamma-ray sorting, infra-red sorting, x-ray transmission-based sorting and x-ray fluorescence sorting, coupling the findings to the classification criteria outlined. We see a need to define a universal sorting scheme that will in general be applied to most sorting tasks. To do this, the final part of this paper re-looks at the x-ray transmission and x-ray fluorescence sorting scheme in line with the established limitations and proposes a dual x-ray transmission and fluorescence method to mitigate the challenge affecting the different schemes.

show abstract

Nut Image Enhancement: Effective Noise Removal with Enhanced Selective Median Filters (ESMF) Methodology

Saranya,

Durga

2024

Innovative Computing and Communications

View full text Add to dashboard Cite

Classifying grains using behaviour-informed machine learning

Cited by 5 publications

References 23 publications

Using tracer particle kinematics to predict particle size in rotating drums

Using tracer particle kinematics to predict particle size in rotating drums

Sensor-Based Ore Sorting—A Review of Current Use of Electromagnetic Spectrum in Sorting

Nut Image Enhancement: Effective Noise Removal with Enhanced Selective Median Filters (ESMF) Methodology

Contact Info

Product

Resources

About