Zhiqiang Geng

Zhiqiang Geng


Research Focus & Metrics

Academic Career Length

1Year (2026-Present)

(Calculated once Scilit publication data is synced: current year − year of first publication + 1)

Representative Publications

View All Publications
8 October 2026 in Minerals
Verified
Journal Article
Open access
PDF file
CWChen Wei
Chen Wei
YLYuxin Liao
Yuxin Liao
HYHao Yan
Hao Yan
JYJian Yao
Jian Yao
JXJie Xu
Jie Xu
ZGZhiqiang Geng
Zhiqiang Geng
The conventional attenuation-ratio feature R remains affected by particle thickness, material heterogeneity, and beam hardening in practical dual-energy X-ray transmission (DE-XRT) measurements. This study proposes an ore-sorting method based on multi-feature fusion of dual-energy X-ray transmission responses. An aluminium step wedge was used to establish a reference relationship between R and high-energy attenuation αH. The deviation from this relationship was defined as the residual feature Rc. Particle-level statistics of R, αH, and Rc were combined in a linear support vector machine (SVM) classifier. For lead–zinc ore, the SVM was compared with a conventional double-threshold R-value rule. In batch-grouped out-of-fold evaluation, the SVM increased combined Pb–Zn recovery from 90.99% to 97.42% and reduced the combined Pb–Zn grade of the tailings from 2.101% to 0.357%. In an independent 216-particle sorting-and-assay test, the SVM recovered 98.78% of Pb–Zn at 22.03% mass rejection, compared with 90.72% recovery at 26.05% rejection for the conventional R-value rule.

13 November 2026 in Minerals
Verified
Journal Article
Open access
PDF file
WCWen Chen
Wen Chen
YLYu-Xin Liao
Yu-Xin Liao
ZGZhiqiang Geng
Zhiqiang Geng
Dual-energy X-ray transmission sorting of copper ores is often constrained by strong coupling between conventional features and particle thickness. To address this, we propose two decoupled features: R3, an improved R-value feature, and Zeff from basis material decomposition, which are designed to mitigate thickness interference and improve high-Z mineral identification. Copper and lead–zinc ore particles from the Yinshan Mine (Dexing, China), with particle thickness ranges of 8–60 mm for the copper ore and 8–25 mm for the lead–zinc ore, were tested on a laboratory DE-XRT sorter. A support vector machine regression model was trained on 314 phyllite copper ore particles (8–60 mm) to predict a weighted comprehensive grade (Cu+Fe+S), using R-value, R3, Zeff, and w3–w2 as inputs. The method showed favorable performance for porphyry copper ores (particle thickness: 8–60 mm): feed grades of 0.068%–0.109% Cu were upgraded to concentrates of 0.127%–0.244% Cu, with waste grades of 0.030%–0.051% and recoveries of 61.69%–79.67%. For phyllite copper ores (particle thickness: 8–60 mm), enrichment was observed through yield adjustment, with recoveries of 33.38%–71.69%. A plausible reason for the poor performance of the intermediate-grade Cu phyllite ore (Cu: 0.098%) is a combination of two factors: the chalcopyrite grain size (~0.1 mm) being below the detector resolution (0.8 mm), which causes mineral signals to be swamped by gangue, and the fundamental limitation of dual-energy XRT in distinguishing Cu from Fe. For lead–zinc ore (4.744% Pb+Zn feed), the sorter rejected 21.99% as waste while achieving 98.87% Pb+Zn recovery, with only 0.245% Pb+Zn in the waste. These results indicate that the proposed multi-feature fusion and SVM regression strategy provides a feasible approach for the preconcentration of low-grade copper ores and waste rejection for lead–zinc ores.

28 September 2026 in Minerals
Verified
Journal Article
Open access
PDF file
JXJie Xu
Jie Xu
YLYuxin Liao
Yuxin Liao
HYHao Yan
Hao Yan
JYJian Yao
Jian Yao
CWChen Wei
Chen Wei
ZGZhiqiang Geng
Zhiqiang Geng
Sensor-based ore sorting can discard barren rock before grinding, but dual-energy X-ray transmission (DE-XRT) grade models are often judged by block-wise error that poorly reflects concentrate and tailings mass balance at the plant cut-off. Process-aware sorting knowledge (PSK) encodes attenuation, sorting tendency, and grade-band text from patch statistics and split rules; captions are aligned via a frozen contrastive language–image pre-training (CLIP) encoder during training, while inference remains Vision-only. Stream outcomes are reported as metal recovery, mass yield (mass fraction to concentrate), concentrate and tailings grades, and enrichment ratio, with the Process Fidelity Index (PFI) summarizing oracle-aligned fidelity at cut-off τ. On a fixed PbZn holdout with full assay labels, PSK raises the PFI from 0.716 to 0.830 and lowers mean absolute error (MAE) from 4.59% to 3.77% versus Vision-only training; categorical and physics-feature controls support added value from CLIP alignment. Backbone ablation shows positive PSK gains on ResNet-18, MobileNet v3 small, and ShuffleNet v2_×0.5 (ΔPFI = +0.114/+0.118/+0.089), with lightweight Vision PFI near MobileNet and below ResNet-18. Field ablation shows that attenuation and sorting-tendency fields drive most stream gains, while the grade-band field trades a higher PFI for worse tailings grade when removed.