Mingrong Dong
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Mingrong Dong has a total of 41 co-authors. Here are the top co-authors by number of shared publications:
- Yan Zhu
Kunming University of Science and Technology
5 shared publications - Tao Shen
Kunming University of Science and Technology
3 shared publications - Yixuan Yang
Faculty of Information Engineering and Automation, Kunming University of Science and Technology, Kunming 650500, China
3 shared publications Baoguo Zhang
Northwest Institute of Nuclear Technology, Xi’an, China
2 shared publicationsBin He
Northwest Institute of Nuclear Technology, Xi’an, China
2 shared publicationsDeyu Sun
Northwest Institute of Nuclear Technology, Xi’an, China
2 shared publications
Latest Publications
- Open Access
Laser-Scattering-Image-Based Intelligent Sensing for Dust Concentration Prediction in Top-Blown Furnace Applications with ICPO-CKGPR
Top-blown furnaces are important equipment in non-ferrous metal smelting processes, where dust concentration variations provide useful information for process monitoring and environmental management. However, conventional dust measurement methods may suffer from limited measurement efficiency and insufficient adaptability under complex operating conditions. To address these challenges, this study proposes a laser-scattering-image-based intelligent sensing framework with an improved crested porcupine optimizer-based combined-kernel Gaussian process regression model, termed ICPO-CKGPR, for dust concentration prediction under simulated top-blown furnace conditions. First, a laboratory-scale laser scattering image acquisition platform was established to collect scattering images under different dust concentration conditions, and 5000 timestamp-matched image–concentration sample pairs were obtained using a particulate matter sensor as the reference measurement source. Subsequently, a multi-domain feature representation strategy integrating grayscale-domain, frequency-domain, and spatial-domain information was developed to enhance the characterization capability of dust scattering images under multi-scale variations and disturbances. Furthermore, a combined-kernel Gaussian process regression model consisting of radial basis function, Matérn, and rational quadratic kernels was constructed to improve the modeling capability for complex nonlinear dust concentration variations. Meanwhile, the crested porcupine optimizer was improved by incorporating good-point-set initialization and periodic oscillation mutation strategies for adaptive hyperparameter optimization. Experimental results on the independent test set demonstrate that the proposed ICPO-CKGPR model achieves an RMSE of 3.086 mg/m3, an MAE of 1.784 mg/m3, and an R2 of 0.992 against the sensor-derived reference measurements. Compared with the benchmark models, the proposed method shows closer agreement with the sensor-derived PM10 reference measurements, indicating its potential for laser-scattering-image-based PM10 estimation under the investigated laboratory conditions.
Open AccessLaser-Scattering-Image-Based Intelligent Sensing for Dust Concentration Prediction in Top-Blown Furnace Applications with ICPO-CKGPR
Top-blown furnaces are important equipment in non-ferrous metal smelting processes, where dust concentration variations provide useful information for process monitoring and environmental management. However, conventional dust measurement methods may suffer from limited measurement efficiency and insufficient adaptability under complex operating conditions. To address these challenges, this study proposes a laser-scattering-image-based intelligent sensing framework with an improved crested porcupine optimizer-based combined-kernel Gaussian process regression model, termed ICPO-CKGPR, for dust concentration prediction under simulated top-blown furnace conditions. First, a laboratory-scale laser scattering image acquisition platform was established to collect scattering images under different dust concentration conditions, and 5000 timestamp-matched image–concentration sample pairs were obtained using a particulate matter sensor as the reference measurement source. Subsequently, a multi-domain feature representation strategy integrating grayscale-domain, frequency-domain, and spatial-domain information was developed to enhance the characterization capability of dust scattering images under multi-scale variations and disturbances. Furthermore, a combined-kernel Gaussian process regression model consisting of radial basis function, Matérn, and rational quadratic kernels was constructed to improve the modeling capability for complex nonlinear dust concentration variations. Meanwhile, the crested porcupine optimizer was improved by incorporating good-point-set initialization and periodic oscillation mutation strategies for adaptive hyperparameter optimization. Experimental results on the independent test set demonstrate that the proposed ICPO-CKGPR model achieves an RMSE of 3.086 mg/m3, an MAE of 1.784 mg/m3, and an R2 of 0.992 against the sensor-derived reference measurements. Compared with the benchmark models, the proposed method shows closer agreement with the sensor-derived PM10 reference measurements, indicating its potential for laser-scattering-image-based PM10 estimation under the investigated laboratory conditions.
- Open Access
RSSM-Based Virtual Sensing and Sensorless Closed-Loop Control for a Multi-Temperature-Zone Continuous Crystallizer
Precise temperature control is crucial for maintaining product quality and optimizing energy efficiency in multi-zone continuous crystallizers. However, such industrial processes typically exhibit complex nonlinear dynamics and strong coupling effects. More critically, physical constraints often prevent sensor installation, rendering temperatures in key regions unobservable and challenging traditional closed-loop control strategies. To address partial observability and model uncertainty, this paper proposes a Model-Based Reinforcement Learning (MBRL) framework utilizing solely offline historical data. The core innovation lies in developing a Recursive State Space Model (RSSM) that serves not only as a high-fidelity digital twin but, more critically, is deployed as a real-time “virtual sensor” to infer unobservable system states. This virtual sensing capability provides precise state estimates for downstream policy optimization. Additionally, a multi-objective reward function is designed to balance tracking error, stability, and control cost. Experimental results demonstrate that the proposed virtual sensor exhibits exceptional long-term stability, maintaining high fidelity and effectively suppressing error accumulation during long-term multi-step autoregressive predictions. Consequently, the trained agent outperforms traditional Proportional-Integral-Derivative (PID) and Model Predictive Control (MPC) controllers, achieving over 67% improvement in temperature tracking accuracy while reducing control action costs by more than 93%, indicating smoother system operation and enhanced energy efficiency.
Open AccessRSSM-Based Virtual Sensing and Sensorless Closed-Loop Control for a Multi-Temperature-Zone Continuous Crystallizer
Precise temperature control is crucial for maintaining product quality and optimizing energy efficiency in multi-zone continuous crystallizers. However, such industrial processes typically exhibit complex nonlinear dynamics and strong coupling effects. More critically, physical constraints often prevent sensor installation, rendering temperatures in key regions unobservable and challenging traditional closed-loop control strategies. To address partial observability and model uncertainty, this paper proposes a Model-Based Reinforcement Learning (MBRL) framework utilizing solely offline historical data. The core innovation lies in developing a Recursive State Space Model (RSSM) that serves not only as a high-fidelity digital twin but, more critically, is deployed as a real-time “virtual sensor” to infer unobservable system states. This virtual sensing capability provides precise state estimates for downstream policy optimization. Additionally, a multi-objective reward function is designed to balance tracking error, stability, and control cost. Experimental results demonstrate that the proposed virtual sensor exhibits exceptional long-term stability, maintaining high fidelity and effectively suppressing error accumulation during long-term multi-step autoregressive predictions. Consequently, the trained agent outperforms traditional Proportional-Integral-Derivative (PID) and Model Predictive Control (MPC) controllers, achieving over 67% improvement in temperature tracking accuracy while reducing control action costs by more than 93%, indicating smoother system operation and enhanced energy efficiency.
- Open Access
- 17 October 2025
- Published by MDPI AG in Electronics
- Vol. 14 (20) , 4092
Enhanced Heart Sound Detection via Multi-Scale Feature Extraction and Attention Mechanism Using Pitch-Shifting Data Augmentation
Cardiovascular diseases pose a major global health threat, making early automated detection through heart sound analysis crucial for their prevention. However, existing deep learning-based heart sound detection methods have shortcomings in feature extraction, and current attention mechanisms perform inadequately in capturing key heart sound features. To address this, we first introduce a Multi-Scale Feature Extraction Network composed of Multi-Scale Inverted Residual (MIR) modules and Dynamically Gated Convolution (DGC) modules to extract heart sound features effectively. The MIR module can efficiently extract multi-scale heart sound features, and the DGC module enhances the network’s representation ability by capturing feature interrelationships and dynamically adjusting information flow. Subsequently, a Multi-Scale Attention Prediction Network is designed for heart sound feature classification, which includes a multi-scale attention (MSA) module. The MSA module effectively captures subtle pathological features of heart sound signals through multi-scale feature extraction and cross-scale feature interaction. Additionally, pitch-shifting techniques are applied in the preprocessing stage to enhance the model’s generalization ability, and multiple feature extraction techniques are used for initial feature extraction of heart sounds. Evaluated via five-fold cross-validation, the model achieved accuracies of 98.89% and 98.86% on the PhysioNet/CinC 2016 and 2022 datasets, respectively, demonstrating superior performance and strong potential for clinical application.
- 17 October 2025
- Published by MDPI AG in Electronics
- Vol. 14 (20) , 4092
Open AccessEnhanced Heart Sound Detection via Multi-Scale Feature Extraction and Attention Mechanism Using Pitch-Shifting Data Augmentation
Cardiovascular diseases pose a major global health threat, making early automated detection through heart sound analysis crucial for their prevention. However, existing deep learning-based heart sound detection methods have shortcomings in feature extraction, and current attention mechanisms perform inadequately in capturing key heart sound features. To address this, we first introduce a Multi-Scale Feature Extraction Network composed of Multi-Scale Inverted Residual (MIR) modules and Dynamically Gated Convolution (DGC) modules to extract heart sound features effectively. The MIR module can efficiently extract multi-scale heart sound features, and the DGC module enhances the network’s representation ability by capturing feature interrelationships and dynamically adjusting information flow. Subsequently, a Multi-Scale Attention Prediction Network is designed for heart sound feature classification, which includes a multi-scale attention (MSA) module. The MSA module effectively captures subtle pathological features of heart sound signals through multi-scale feature extraction and cross-scale feature interaction. Additionally, pitch-shifting techniques are applied in the preprocessing stage to enhance the model’s generalization ability, and multiple feature extraction techniques are used for initial feature extraction of heart sounds. Evaluated via five-fold cross-validation, the model achieved accuracies of 98.89% and 98.86% on the PhysioNet/CinC 2016 and 2022 datasets, respectively, demonstrating superior performance and strong potential for clinical application.
- 1 July 2025
- Published by IOP Publishing in Physica Scripta
- Vol. 100 (7) , 075999
Tailoring the phase structure and optical properties of Cu2ZnSnS4 nanocrystals through co-substitution of Ag and Fe
Straightforward two-step heating method, employing oleylamine as the reaction solvent, was utilized to incorporate Ag and Fe into Cu2ZnSnS4 (CZTS) nanoparticles, resulting in the synthesis of (Cu1-xAgx)2FeyZn1-ySnS4 (CAFZTS, where x = 0, 0.2 and y = 0, 0.05, 0.1) nanoparticles under atmospheric conditions. The produced samples underwent comprehensive characterization through techniques such as XRD, Raman spectroscopy, SEM, EDS, TEM, XPS, and UV–Vis–NIR absorption spectroscopy. The findings revealed that the initial CZTS nanoparticles exhibited a wurtzite phase structure. However, with the introduction of Ag and Fe, a gradual transition from the wurtzite phase to the kesterite phase was observed in the nanoparticles’ structure. Notably, when the Ag substitution ratio (x) reached 0.2 and the Fe substitution ratio (y) was set at either 0 or 0.05, the nanoparticles displayed coexistence of both wurtzite and kesterite phases. Conversely, at an Ag substitution ratio (x) of 0.2 and an Fe substitution ratio (y) of 0.1, the nanoparticles exclusively exhibited a kesterite phase. The band gap of the original CZTS nanoparticles was identified as 1.47eV. This value increased to 1.58eV with an Ag substitution ratio (x) of 0.2. Conversely, with Fe substitution ratios (y) of 0.05 and 0.1, the band gap decreased from 1.30 eV to 1.17 eV, respectively. These results demonstrate that the incorporation of Ag and Fe not only alters the mineral phase of CZTS nanoparticles but also provides a means to adjust the band gap.
- 1 July 2025
- Published by IOP Publishing in Physica Scripta
- Vol. 100 (7) , 075999
Tailoring the phase structure and optical properties of Cu2ZnSnS4 nanocrystals through co-substitution of Ag and Fe
Straightforward two-step heating method, employing oleylamine as the reaction solvent, was utilized to incorporate Ag and Fe into Cu2ZnSnS4 (CZTS) nanoparticles, resulting in the synthesis of (Cu1-xAgx)2FeyZn1-ySnS4 (CAFZTS, where x = 0, 0.2 and y = 0, 0.05, 0.1) nanoparticles under atmospheric conditions. The produced samples underwent comprehensive characterization through techniques such as XRD, Raman spectroscopy, SEM, EDS, TEM, XPS, and UV–Vis–NIR absorption spectroscopy. The findings revealed that the initial CZTS nanoparticles exhibited a wurtzite phase structure. However, with the introduction of Ag and Fe, a gradual transition from the wurtzite phase to the kesterite phase was observed in the nanoparticles’ structure. Notably, when the Ag substitution ratio (x) reached 0.2 and the Fe substitution ratio (y) was set at either 0 or 0.05, the nanoparticles displayed coexistence of both wurtzite and kesterite phases. Conversely, at an Ag substitution ratio (x) of 0.2 and an Fe substitution ratio (y) of 0.1, the nanoparticles exclusively exhibited a kesterite phase. The band gap of the original CZTS nanoparticles was identified as 1.47eV. This value increased to 1.58eV with an Ag substitution ratio (x) of 0.2. Conversely, with Fe substitution ratios (y) of 0.05 and 0.1, the band gap decreased from 1.30 eV to 1.17 eV, respectively. These results demonstrate that the incorporation of Ag and Fe not only alters the mineral phase of CZTS nanoparticles but also provides a means to adjust the band gap.
- 24 January 2025
- Published by Springer Nature in Applied Physics A
- Vol. 131 (2) , 1-9
Optical and structural properties of Fe- and Ge-substituted CZTS thin films produced by alternating sputtering deposition
- 24 January 2025
- Published by Springer Nature in Applied Physics A
- Vol. 131 (2) , 1-9
Optical and structural properties of Fe- and Ge-substituted CZTS thin films produced by alternating sputtering deposition
- 18 December 2024
- Published by Springer Nature in The Journal of Supercomputing
- Vol. 81 (1)
Research on the improvement of domain generalization by the fusion of invariant features and sharpness-aware minimization
- 18 December 2024
- Published by Springer Nature in The Journal of Supercomputing
- Vol. 81 (1)
Research on the improvement of domain generalization by the fusion of invariant features and sharpness-aware minimization
- Open Access
- 6 August 2024
- Published by MDPI AG in Remote Sensing
- Vol. 16 (16) , 2877
Implicit Sharpness-Aware Minimization for Domain Generalization
Domain generalization (DG) aims to learn knowledge from multiple related domains to achieve a robust generalization performance in unseen target domains, which is an effective approach to mitigate domain shift in remote sensing image classification. Although the sharpness-aware minimization (SAM) method enhances DG capability and improves remote sensing image classification performance by promoting the convergence of the loss minimum to a flatter loss surface, the perturbation loss (maximum loss within the neighborhood of a local minimum) of SAM fails to accurately measure the true sharpness of the loss landscape. Furthermore, its variants often overlook gradient conflicts, thereby limiting further improvement in DG performance. In this paper, we introduce implicit sharpness-aware minimization (ISAM), a novel method that addresses the deficiencies of SAM and mitigates gradient conflicts. Specifically, we demonstrate that the discrepancy in training loss during gradient ascent or descent serves as an equivalent measure of the dominant eigenvalue of the Hessian matrix. This discrepancy provides a reliable measure for sharpness. ISAM effectively reduces sharpness and mitigates potential conflicts between gradients by implicitly minimizing the discrepancy between training losses while ensuring a sufficiently low minimum through minimizing perturbation loss. Extensive experiments and analyses demonstrate that ISAM significantly enhances the model’s generalization ability on remote sensing and DG datasets, outperforming existing state-of-the-art methods.
- 6 August 2024
- Published by MDPI AG in Remote Sensing
- Vol. 16 (16) , 2877
Open AccessImplicit Sharpness-Aware Minimization for Domain Generalization
Domain generalization (DG) aims to learn knowledge from multiple related domains to achieve a robust generalization performance in unseen target domains, which is an effective approach to mitigate domain shift in remote sensing image classification. Although the sharpness-aware minimization (SAM) method enhances DG capability and improves remote sensing image classification performance by promoting the convergence of the loss minimum to a flatter loss surface, the perturbation loss (maximum loss within the neighborhood of a local minimum) of SAM fails to accurately measure the true sharpness of the loss landscape. Furthermore, its variants often overlook gradient conflicts, thereby limiting further improvement in DG performance. In this paper, we introduce implicit sharpness-aware minimization (ISAM), a novel method that addresses the deficiencies of SAM and mitigates gradient conflicts. Specifically, we demonstrate that the discrepancy in training loss during gradient ascent or descent serves as an equivalent measure of the dominant eigenvalue of the Hessian matrix. This discrepancy provides a reliable measure for sharpness. ISAM effectively reduces sharpness and mitigates potential conflicts between gradients by implicitly minimizing the discrepancy between training losses while ensuring a sufficiently low minimum through minimizing perturbation loss. Extensive experiments and analyses demonstrate that ISAM significantly enhances the model’s generalization ability on remote sensing and DG datasets, outperforming existing state-of-the-art methods.
- Open Access
- 22 March 2024
- Published by Public Library of Science (PLoS) in PLOS ONE
- Vol. 19 (3) , e0300832
A high-intensity low-frequency acoustic generator based on the Helmholtz resonator and airflow modulator
The high-intensity low-frequency acoustic sources have essential applications in acoustic biological effects research, airport bird repelling, and boiler ash removal. However, generating high-intensity low-frequency acoustic waves in open space is difficult. In this paper, a low-frequency acoustic generator with a resonant cavity used to enhance the acoustic intensity in open space was developed, which is an aerodynamic acoustic generator to radiates a high-intensity acoustic wave of 52Hz. Some experiments were carried out to measure this generator’s internal flow field and radiated acoustic field characteristics, including the propagation characteristics at 100m. The experimental results show that the resonant enhancement effect is presented near the predetermined resonance frequency, and the enhanced value is about 4dB. The acoustic intensity for 52Hz at 1m position is 124dB. By combining the Helmholtz resonator with the airflow modulator, the airflow resonance in the resonator enhances the air pressure pulsation inside the chamber and increases the disturbance of acoustic radiation to the air. So as to improve the sound intensity and radiation efficiency in the low-frequency range.
- 22 March 2024
- Published by Public Library of Science (PLoS) in PLOS ONE
- Vol. 19 (3) , e0300832
Open AccessA high-intensity low-frequency acoustic generator based on the Helmholtz resonator and airflow modulator
The high-intensity low-frequency acoustic sources have essential applications in acoustic biological effects research, airport bird repelling, and boiler ash removal. However, generating high-intensity low-frequency acoustic waves in open space is difficult. In this paper, a low-frequency acoustic generator with a resonant cavity used to enhance the acoustic intensity in open space was developed, which is an aerodynamic acoustic generator to radiates a high-intensity acoustic wave of 52Hz. Some experiments were carried out to measure this generator’s internal flow field and radiated acoustic field characteristics, including the propagation characteristics at 100m. The experimental results show that the resonant enhancement effect is presented near the predetermined resonance frequency, and the enhanced value is about 4dB. The acoustic intensity for 52Hz at 1m position is 124dB. By combining the Helmholtz resonator with the airflow modulator, the airflow resonance in the resonator enhances the air pressure pulsation inside the chamber and increases the disturbance of acoustic radiation to the air. So as to improve the sound intensity and radiation efficiency in the low-frequency range.
- Open Access
- 14 December 2023
- Published by SAGE Publications in Journal of Low Frequency Noise, Vibration and Active Control
- Vol. 43 (2) , 906-919
Theoretical research on the application of impedance calculation in the design of high-intensity low-frequency acoustic generator
The impedance matching is essential in producing high-intensity acoustic waves. This paper studies the impedance calculation methods of modulators and double-chamber Helmholtz resonators based on fluid mechanics, aeroacoustics, and electro-acoustic analogy principles. The results: ① The impedance expression of the resonators is derived. ② The impedance variation law of the chamber volume, the stator size, the air supply pressure, the air supply flow, and the working frequency is calculated. ③ The impedance-matching design scheme of the two-chamber Helmholtz low frequency and high acoustic intensity generator is given. Based on the experimental results of the high acoustic intensity generating device that the author has manufactured in the closed space, the correctness of the impedance calculation and matching design scheme in this paper is verified. This method has guiding significance for the design of high-intensity acoustic generating devices in confined space.
- 14 December 2023
- Published by SAGE Publications in Journal of Low Frequency Noise, Vibration and Active Control
- Vol. 43 (2) , 906-919
Open AccessTheoretical research on the application of impedance calculation in the design of high-intensity low-frequency acoustic generator
The impedance matching is essential in producing high-intensity acoustic waves. This paper studies the impedance calculation methods of modulators and double-chamber Helmholtz resonators based on fluid mechanics, aeroacoustics, and electro-acoustic analogy principles. The results: ① The impedance expression of the resonators is derived. ② The impedance variation law of the chamber volume, the stator size, the air supply pressure, the air supply flow, and the working frequency is calculated. ③ The impedance-matching design scheme of the two-chamber Helmholtz low frequency and high acoustic intensity generator is given. Based on the experimental results of the high acoustic intensity generating device that the author has manufactured in the closed space, the correctness of the impedance calculation and matching design scheme in this paper is verified. This method has guiding significance for the design of high-intensity acoustic generating devices in confined space.
- Open Access
- 7 December 2023
- Published by MDPI AG in Micromachines
- Vol. 14 (12) , 2211
Cubic AgBiS2 Powder Prepared Using a Facile Reflux Method for Photocatalytic Degradation of Dyes
The ternary chalcogenide AgBiS2 has attracted widespread attention in the field of photovoltaic and photoelectric devices due to its excellent properties. In this study, AgBiS2 powders with an average diameter of 200 nm were prepared via a simple and convenient reflux method from silver acetate, bismuth nitrate pentahydrate, and n-dodecyl mercaptan. The adjustment of the ratios of Ag:Bi:S raw materials and of the reaction temperatures were carried out to investigate the significance of the synthesis conditions toward the composition of the as-synthesized AgBiS2. The results of XRD indicated that the powders synthesized at a ratio of 1.05:1:2.1 and a synthesis temperature of 225 °C have the lowest bismuth content and the highest purity. The synthesized AgBiS2 crystallizes in a rock salt type structure with the cubic space group. Scanning and transmission electron microscopy, thermogravimetric analysis, ultraviolet–visible–near-infrared spectra, and photocatalytic degradation performance were employed to characterize the as-synthesized samples. The results demonstrated that AgBiS2 powders display thermal stability; strong absorption in the ultraviolet, visible, and partial infrared regions; and an optical bandgap of 0.98 eV. The obtained AgBiS2 powders also have a good degradation effect on the methylene blue solution with a degradation efficiency of 58.61% and a rate constant of 0.0034 min−1, indicating that it is an efficient strategy for sewage degradation to reduce water pollution.- 7 December 2023
- Published by MDPI AG in Micromachines
- Vol. 14 (12) , 2211
Open AccessCubic AgBiS2 Powder Prepared Using a Facile Reflux Method for Photocatalytic Degradation of Dyes
The ternary chalcogenide AgBiS2 has attracted widespread attention in the field of photovoltaic and photoelectric devices due to its excellent properties. In this study, AgBiS2 powders with an average diameter of 200 nm were prepared via a simple and convenient reflux method from silver acetate, bismuth nitrate pentahydrate, and n-dodecyl mercaptan. The adjustment of the ratios of Ag:Bi:S raw materials and of the reaction temperatures were carried out to investigate the significance of the synthesis conditions toward the composition of the as-synthesized AgBiS2. The results of XRD indicated that the powders synthesized at a ratio of 1.05:1:2.1 and a synthesis temperature of 225 °C have the lowest bismuth content and the highest purity. The synthesized AgBiS2 crystallizes in a rock salt type structure with the cubic space group. Scanning and transmission electron microscopy, thermogravimetric analysis, ultraviolet–visible–near-infrared spectra, and photocatalytic degradation performance were employed to characterize the as-synthesized samples. The results demonstrated that AgBiS2 powders display thermal stability; strong absorption in the ultraviolet, visible, and partial infrared regions; and an optical bandgap of 0.98 eV. The obtained AgBiS2 powders also have a good degradation effect on the methylene blue solution with a degradation efficiency of 58.61% and a rate constant of 0.0034 min−1, indicating that it is an efficient strategy for sewage degradation to reduce water pollution.