Jiankui Chen
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Jiankui Chen has a total of 107 co-authors. Here are the top co-authors by number of shared publications:
- Zhouping Yin
Flexible Electronics Research Center, State Key Laboratory of Intelligent Manufacturing Equipment and Technology, School of Mechanical Science and Engineering, Huazhong University of Science and Technology, Wuhan, China
24 shared publications Zhouping Yin
Institute for Advanced Electronics Manufacturing, State Key Laboratory of Intelligent Manufacturing Equipment and Technology, School of Mechanical Science and Engineering, Huazhong University of Science and Technology, Wuhan, Hubei 430074, PR China
20 shared publications- Wei Chen
Huazhong University of Science and Technology
14 shared publications - Hua Yang
School of Mechanical Science and Engineering, State Key Laboratory of Intelligent Manufacturing Equipment and Technology, Huazhong University of Science and Technology, Wuhan, China
8 shared publications Ziwei Zhao
7 shared publicationsWei Tang
6 shared publications
Latest Publications
- Open Access
- 29 October 2026
- Published by MDPI AG in Micromachines
- Vol. 17  (11) , 1245
Deformation Characteristics of Flexible Sheets Across Discontinuous Air-Floating Supports
Air-floating transport provides a noncontact solution for handling large flexible sheets in precision manufacturing. However, variations in the spatial distribution of air-film support during transport can induce localized deformation, particularly when the sheet crosses gaps between adjacent support regions. This study investigates the deformation characteristics of thin flexible sheets during air-floating transport using a coupled model of the sheet and supporting air film. The evolution of sheet deformation is analyzed at representative transport positions. A relative displacement measure is introduced to quantify local edge sagging. The effects of guide gap, air supply conditions represented by the equivalent parameter Q, initial installation gap, and negative pressure are systematically examined. The results show that deformation is particularly pronounced near the sheet edges during transitions between support regions. Among the investigated parameters, guide gap has the strongest influence on edge sagging, while the effects of the equivalent supply parameter Q and negative pressure are relatively limited within the examined ranges. These findings provide guidance for configuring air-floating transport systems to reduce undesirable sheet deformation.
- 29 October 2026
- Published by MDPI AG in Micromachines
- Vol. 17  (11) , 1245
Open AccessDeformation Characteristics of Flexible Sheets Across Discontinuous Air-Floating Supports
Air-floating transport provides a noncontact solution for handling large flexible sheets in precision manufacturing. However, variations in the spatial distribution of air-film support during transport can induce localized deformation, particularly when the sheet crosses gaps between adjacent support regions. This study investigates the deformation characteristics of thin flexible sheets during air-floating transport using a coupled model of the sheet and supporting air film. The evolution of sheet deformation is analyzed at representative transport positions. A relative displacement measure is introduced to quantify local edge sagging. The effects of guide gap, air supply conditions represented by the equivalent parameter Q, initial installation gap, and negative pressure are systematically examined. The results show that deformation is particularly pronounced near the sheet edges during transitions between support regions. Among the investigated parameters, guide gap has the strongest influence on edge sagging, while the effects of the equivalent supply parameter Q and negative pressure are relatively limited within the examined ranges. These findings provide guidance for configuring air-floating transport systems to reduce undesirable sheet deformation.
- Open Access
- 21 September 2026
- Published by Elsevier in SSRN Electronic Journal
Adaptive Digital Twin System for Inkjet Manufacturing with Small-Sample Model Updating
Integrating equipment data with computational models supports decision-making in industry, but reliable decisions require models that remain valid as operating conditions change. When process responses deviate from the nominal relationship, rebuilding the model for each new condition requires substantial additional measurements, making efficient model updating with limited online data a critical challenge. Digital twin technology provides a promising paradigm for addressing this problem by integrating online model updating within a physical–digital loop. This study develops an adaptive digital twin system with small-sample model updating and validates it in industrial inkjet manufacturing. In this application, droplet volume is treated as the process response and driving-waveform parameters as decision variables. A digital twin-driven similarity-blended residual adaptation (DT-SBRA) method is proposed as the model-updating mechanism, using only a few measurements from the current condition to preserve the nominal mapping while correcting condition-dependent prediction deviations. The updated model is then reused for virtual process evaluation and waveform-parameter optimization. Validation under actual drift conditions showed that droplet-volume prediction accuracy exceeded 96% and target-volume control accuracy exceeded 95% over the range of 3.0–5.0 pL. Data-transmission latency remained below 50 ms, and defect-free deposition on organic light-emitting diode (OLED) substrates further demonstrated practical applicability. The results indicate that integrating small-sample model updating into the physical–digital loop can sustain model-based process evaluation and parameter feedback under operating-condition drift while reducing additional data acquisition and physical experimentation.
- 21 September 2026
- Published by Elsevier in SSRN Electronic Journal
Open AccessAdaptive Digital Twin System for Inkjet Manufacturing with Small-Sample Model Updating
Integrating equipment data with computational models supports decision-making in industry, but reliable decisions require models that remain valid as operating conditions change. When process responses deviate from the nominal relationship, rebuilding the model for each new condition requires substantial additional measurements, making efficient model updating with limited online data a critical challenge. Digital twin technology provides a promising paradigm for addressing this problem by integrating online model updating within a physical–digital loop. This study develops an adaptive digital twin system with small-sample model updating and validates it in industrial inkjet manufacturing. In this application, droplet volume is treated as the process response and driving-waveform parameters as decision variables. A digital twin-driven similarity-blended residual adaptation (DT-SBRA) method is proposed as the model-updating mechanism, using only a few measurements from the current condition to preserve the nominal mapping while correcting condition-dependent prediction deviations. The updated model is then reused for virtual process evaluation and waveform-parameter optimization. Validation under actual drift conditions showed that droplet-volume prediction accuracy exceeded 96% and target-volume control accuracy exceeded 95% over the range of 3.0–5.0 pL. Data-transmission latency remained below 50 ms, and defect-free deposition on organic light-emitting diode (OLED) substrates further demonstrated practical applicability. The results indicate that integrating small-sample model updating into the physical–digital loop can sustain model-based process evaluation and parameter feedback under operating-condition drift while reducing additional data acquisition and physical experimentation.
- Open Access
- 8 September 2026
- Published by Elsevier in SSRN Electronic Journal
Dry-Film-Based Planar Flow-Channel Piezoelectric Inkjet Printing Chip for High-Frequency, Small-Droplet Printing
Piezoelectric inkjet printheads are essential microelectromechanical systems (MEMS) for high-precision additive manufacturing and are continuously advancing toward smaller droplets, higher operating frequencies, and denser nozzle integration. However, conventional silicon-based printheads typically require complex fabrication processes involving wafer thinning, deep reactive ion etching, and multilayer wafer bonding, which impose significant challenges in process complexity, structural reliability, and manufacturing scalability. Here, we present a MEMS piezoelectric inkjet printing chip featuring a planar flow-channel architecture enabled by dry-film technology. By replacing conventional three-dimensional silicon micromachining and multilayer bonding with dry-film lamination and standard photolithographic patterning, the proposed approach enables the fabrication of integrated planar channels and nozzle arrays through a simplified low-temperature process (<150 °C). Multiphysics simulations coupling piezoelectric actuation and ink flow were employed to optimize the actuator and flow-channel structures, and the optimized design was experimentally validated through MEMS fabrication and jetting characterization. The fabricated chips exhibit excellent structural uniformity and stable droplet ejection at frequencies up to 30 kHz, achieving a minimum controllable droplet volume of 2.5 pL with a coefficient of variation below 7%. These results demonstrate that the proposed dry-film-based planar architecture provides a scalable and manufacturable strategy for high-frequency, small-droplet MEMS piezoelectric inkjet printheads.
- 8 September 2026
- Published by Elsevier in SSRN Electronic Journal
Open AccessDry-Film-Based Planar Flow-Channel Piezoelectric Inkjet Printing Chip for High-Frequency, Small-Droplet Printing
Piezoelectric inkjet printheads are essential microelectromechanical systems (MEMS) for high-precision additive manufacturing and are continuously advancing toward smaller droplets, higher operating frequencies, and denser nozzle integration. However, conventional silicon-based printheads typically require complex fabrication processes involving wafer thinning, deep reactive ion etching, and multilayer wafer bonding, which impose significant challenges in process complexity, structural reliability, and manufacturing scalability. Here, we present a MEMS piezoelectric inkjet printing chip featuring a planar flow-channel architecture enabled by dry-film technology. By replacing conventional three-dimensional silicon micromachining and multilayer bonding with dry-film lamination and standard photolithographic patterning, the proposed approach enables the fabrication of integrated planar channels and nozzle arrays through a simplified low-temperature process (<150 °C). Multiphysics simulations coupling piezoelectric actuation and ink flow were employed to optimize the actuator and flow-channel structures, and the optimized design was experimentally validated through MEMS fabrication and jetting characterization. The fabricated chips exhibit excellent structural uniformity and stable droplet ejection at frequencies up to 30 kHz, achieving a minimum controllable droplet volume of 2.5 pL with a coefficient of variation below 7%. These results demonstrate that the proposed dry-film-based planar architecture provides a scalable and manufacturable strategy for high-frequency, small-droplet MEMS piezoelectric inkjet printheads.
- 25 June 2026
- Published by Trans Tech Publications, Ltd. in Key Engineering Materials
- Vol. 1059, 27-31
Study on Film Morphology of Liquid Droplet Evaporation in Confined Space
Organic light-emitting diode (OLED) technology is particularly favored for such devices due to its broad color spectrum, design flexibility, low power consumption, and suitability for miniaturization. While conventional OLED fabrication predominantly relies on vapor deposition, inkjet printing has recently emerged as a promising alternative for large-area OLED manufacturing because of its high material utilization and capability for scalable patterning. In this process, light-emitting materials are deposited in liquid form onto the substrate and subsequently dried to form a uniform thin film. However, several challenges remain. In particular, the solvent evaporation rate must be precisely controlled to minimize droplet shrinkage during drying, which can compromise film uniformity and adhesion. Moreover, external parameters such as temperature and humidity significantly influence the drying dynamics, and the choice of solvents and polymers plays a critical role in achieving the desired film quality. To address these challenges, we propose optimizing the ambient pressure to achieve improved curing morphology.
- 25 June 2026
- Published by Trans Tech Publications, Ltd. in Key Engineering Materials
- Vol. 1059, 27-31
Study on Film Morphology of Liquid Droplet Evaporation in Confined Space
Organic light-emitting diode (OLED) technology is particularly favored for such devices due to its broad color spectrum, design flexibility, low power consumption, and suitability for miniaturization. While conventional OLED fabrication predominantly relies on vapor deposition, inkjet printing has recently emerged as a promising alternative for large-area OLED manufacturing because of its high material utilization and capability for scalable patterning. In this process, light-emitting materials are deposited in liquid form onto the substrate and subsequently dried to form a uniform thin film. However, several challenges remain. In particular, the solvent evaporation rate must be precisely controlled to minimize droplet shrinkage during drying, which can compromise film uniformity and adhesion. Moreover, external parameters such as temperature and humidity significantly influence the drying dynamics, and the choice of solvents and polymers plays a critical role in achieving the desired film quality. To address these challenges, we propose optimizing the ambient pressure to achieve improved curing morphology.
- 24 June 2026
- Published by Trans Tech Publications, Ltd. in Advanced Engineering Forum
- Vol. 59, 17-22
Electrohydrodynamic Printing Droplet Volume Control Using Supervised Learning
Electrohydrodynamic inkjet printing technology can generate femtoliter-scale droplets, which provides significant advantages in additive manufacturing. With these advantages, electrohydrodynamic inkjet printing technology shows broad application prospects in repairing micro/nano-scale complex structures in flexible electronic devices and high-resolution displays. During the repair process, precise control of printed droplet volume is required according to target volume requirements. However, due to the complexity of the printing process, traditional theoretical and simulation methods face challenges in achieving effective volume control. This paper proposes a supervised learning-based electrohydrodynamic droplet volume control method. The algorithm innovatively establishes new strategy samples through historical datasets, which include the deviation between current droplet volume and target volume, current process parameters, and changes in process parameters for the next iteration. Based on a feedforward control strategy, we employ a multilayer perceptron (MLP) supervised algorithm to achieve printing parameter recommendation, significantly improving printing efficiency. Volume control experiments conducted on the established electrohydrodynamic printing platform show that the standard volume filling rate can reach 98%, and the control can be completed within a single control cycle.
- 24 June 2026
- Published by Trans Tech Publications, Ltd. in Advanced Engineering Forum
- Vol. 59, 17-22
Electrohydrodynamic Printing Droplet Volume Control Using Supervised Learning
Electrohydrodynamic inkjet printing technology can generate femtoliter-scale droplets, which provides significant advantages in additive manufacturing. With these advantages, electrohydrodynamic inkjet printing technology shows broad application prospects in repairing micro/nano-scale complex structures in flexible electronic devices and high-resolution displays. During the repair process, precise control of printed droplet volume is required according to target volume requirements. However, due to the complexity of the printing process, traditional theoretical and simulation methods face challenges in achieving effective volume control. This paper proposes a supervised learning-based electrohydrodynamic droplet volume control method. The algorithm innovatively establishes new strategy samples through historical datasets, which include the deviation between current droplet volume and target volume, current process parameters, and changes in process parameters for the next iteration. Based on a feedforward control strategy, we employ a multilayer perceptron (MLP) supervised algorithm to achieve printing parameter recommendation, significantly improving printing efficiency. Volume control experiments conducted on the established electrohydrodynamic printing platform show that the standard volume filling rate can reach 98%, and the control can be completed within a single control cycle.
- 24 June 2026
- Published by Trans Tech Publications, Ltd. in Advanced Engineering Forum
- Vol. 59, 9-16
Three-Dimensional Geometrically Exact Dynamic Modeling of Electrohydrodynamic Direct-Writing under a Non-Uniform Electric Field
Electrohydrodynamic direct-writing (EDW) is widely applied in the field of micro-nano manufacturing due to its advantage of rapid deposition of line structure. However, due to the difficulty of bending-torsion coupled modeling for EDW with large deformation, the mechanism and effective control strategies of EDW still unclear. This study firstly establishes a three-dimensional geometrically exact model for EDW using Euler angles. It contains differential equations for the mass conservation, the jet geometry, the jet kinematics, the jet dynamics, and the electric field and charge. Euler angles are used to exactly describe the jet’s geometric variation. A theoretical electric field description for non-uniform electric field is introduced. The universality of the model is validated by reducing the governing equations system to a two-dimensional steady-state situation. The geometrically exact model established in this study is useful for revealing the mechanism of EDW and exploring its control strategies.
- 24 June 2026
- Published by Trans Tech Publications, Ltd. in Advanced Engineering Forum
- Vol. 59, 9-16
Three-Dimensional Geometrically Exact Dynamic Modeling of Electrohydrodynamic Direct-Writing under a Non-Uniform Electric Field
Electrohydrodynamic direct-writing (EDW) is widely applied in the field of micro-nano manufacturing due to its advantage of rapid deposition of line structure. However, due to the difficulty of bending-torsion coupled modeling for EDW with large deformation, the mechanism and effective control strategies of EDW still unclear. This study firstly establishes a three-dimensional geometrically exact model for EDW using Euler angles. It contains differential equations for the mass conservation, the jet geometry, the jet kinematics, the jet dynamics, and the electric field and charge. Euler angles are used to exactly describe the jet’s geometric variation. A theoretical electric field description for non-uniform electric field is introduced. The universality of the model is validated by reducing the governing equations system to a two-dimensional steady-state situation. The geometrically exact model established in this study is useful for revealing the mechanism of EDW and exploring its control strategies.
- Open Access
- 11 June 2026
- Published by Trans Tech Publications, Ltd. in Engineering Innovations
- Vol. 19, 101-106
Bayesian Strategy for the Volume Accuracy of Ink Droplets Continuous Estimation during Printing Based on Prior Knowledge with Earth Mover’s Distance
Monitoring the volume accuracy of ink droplets during printing is important for OLED panel production. This article uses data from ink droplet observation volumes before printing to build a membership distribution histogram and intuitively describes the volume status of multiple nozzles. A histogram in conjunction with the Earth Mover’s Distance (EMD) indicator is used to design prior knowledge, which describes a subjective estimate of ink droplet volume distribution based on experimental data. And using the likelihood function, combining the sampled data of the ink droplets observation volume, obtains the volume accuracy estimate of the ink droplets during printing. This article proposes a Bayesian strategy for the volume accuracy of ink droplet continuous estimation during printing based on prior knowledge with EMD. This strategy can randomly inspect any number of samples of nozzles in the production process and get a reasonable estimate of volume after each sampling observation, which can realize continuous estimation and adjustment for the volume accuracy of ink droplets during printing.
- 11 June 2026
- Published by Trans Tech Publications, Ltd. in Engineering Innovations
- Vol. 19, 101-106
Open AccessBayesian Strategy for the Volume Accuracy of Ink Droplets Continuous Estimation during Printing Based on Prior Knowledge with Earth Mover’s Distance
Monitoring the volume accuracy of ink droplets during printing is important for OLED panel production. This article uses data from ink droplet observation volumes before printing to build a membership distribution histogram and intuitively describes the volume status of multiple nozzles. A histogram in conjunction with the Earth Mover’s Distance (EMD) indicator is used to design prior knowledge, which describes a subjective estimate of ink droplet volume distribution based on experimental data. And using the likelihood function, combining the sampled data of the ink droplets observation volume, obtains the volume accuracy estimate of the ink droplets during printing. This article proposes a Bayesian strategy for the volume accuracy of ink droplet continuous estimation during printing based on prior knowledge with EMD. This strategy can randomly inspect any number of samples of nozzles in the production process and get a reasonable estimate of volume after each sampling observation, which can realize continuous estimation and adjustment for the volume accuracy of ink droplets during printing.
- Open Access
- 11 June 2026
- Published by Trans Tech Publications, Ltd. in Engineering Innovations
- Vol. 19, 107-113
The Printing of Highly Uniform Display Device by Graph Convolutional Neural Network-Based Manufacturing
The field of print display manufacturing still faces the challenge of uneven volume printing. This paper proposes an integer programming framework tailored to the printing process for volume optimisation, alongside a mimetic solving algorithm based on a graph convolutional neural network. These methods enable the accurate resolution of models and achieve volume uniformity with an error margin of less than 2%, effectively addressing inconsistencies in pixel pit printing. Simulation and experimental results validate the efficacy of the approach.
- 11 June 2026
- Published by Trans Tech Publications, Ltd. in Engineering Innovations
- Vol. 19, 107-113
Open AccessThe Printing of Highly Uniform Display Device by Graph Convolutional Neural Network-Based Manufacturing
The field of print display manufacturing still faces the challenge of uneven volume printing. This paper proposes an integer programming framework tailored to the printing process for volume optimisation, alongside a mimetic solving algorithm based on a graph convolutional neural network. These methods enable the accurate resolution of models and achieve volume uniformity with an error margin of less than 2%, effectively addressing inconsistencies in pixel pit printing. Simulation and experimental results validate the efficacy of the approach.
- 5 March 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Automation Science and Engineering
- Vol. 23  (15455955) , 5686-5699
Biomimetic Visual Perception Network for Industrial Image Anomaly Detection
Anomaly detection in the industrial manufacturing process is important for controlling product quality. In real-world scenarios, anomalies manifest in a number of diverse and unforeseen ways and can be categorized into two main types: structural anomalies, such as scratches and breakages, and logical anomalies, such as missing components and dislocations. Current mainstream methods tend to focus on knowledge generalization but are inadequate for detailed and logical detection tasks. To address this issue, this study proposes Biomimetic Visual Perception Network (BVPN), a biomimetic design-based novel paradigm that simulates biological vision and human logical discrimination with detailed distillation dependencies and semantic component information. BVPN introduces three convolution-based networks to extract information from different regions, imitating biological central, para-central, and peripheral vision, and facilitates the integration of disparate information through the process of knowledge distillation. On the basis, BVPN develops two modules based on the student-teacher network: Focus Module, for local detailed detection, and Peripheral Module, for global context detection. Moreover, BVPN proposes the Logical Module, a clustering-based segmentation branch, assessing logical anomalies by the number and color information of segmentation components. Through a fused structure highly aligning with the perception mode of human system, BVPN achieves excellent performance in the detection of structural and logical anomalies. Extensive experiments with mainstream anomaly detection datasets and real-world inkjet printing dataset demonstrate that BVPN outperforms state-of-the-art competitors in accuracy. Note to Practitioners—This work addresses the challenge of unsupervised anomaly detection in complicated industrial scenarios. Due to the inconsistency between structural and logical anomalies, current mainstream methods are inadequate for general anomaly detection tasks. The presented BVPN is a robust and generalized detection method, which optimizes the effective receptive field of the model from the perspective of biomimetic visual system. BVPN demonstrates substantial improvements in both structural and logical defects, providing robust detection of surface anomalies across multiple industrial domains, from fabrics to metal components and electronic assemblies. Furthermore, it also achieves practical computational complexity and inference speed, which has been applied in the real-world inkjet printing inspection equipment.
- 5 March 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Automation Science and Engineering
- Vol. 23  (15455955) , 5686-5699
Biomimetic Visual Perception Network for Industrial Image Anomaly Detection
Anomaly detection in the industrial manufacturing process is important for controlling product quality. In real-world scenarios, anomalies manifest in a number of diverse and unforeseen ways and can be categorized into two main types: structural anomalies, such as scratches and breakages, and logical anomalies, such as missing components and dislocations. Current mainstream methods tend to focus on knowledge generalization but are inadequate for detailed and logical detection tasks. To address this issue, this study proposes Biomimetic Visual Perception Network (BVPN), a biomimetic design-based novel paradigm that simulates biological vision and human logical discrimination with detailed distillation dependencies and semantic component information. BVPN introduces three convolution-based networks to extract information from different regions, imitating biological central, para-central, and peripheral vision, and facilitates the integration of disparate information through the process of knowledge distillation. On the basis, BVPN develops two modules based on the student-teacher network: Focus Module, for local detailed detection, and Peripheral Module, for global context detection. Moreover, BVPN proposes the Logical Module, a clustering-based segmentation branch, assessing logical anomalies by the number and color information of segmentation components. Through a fused structure highly aligning with the perception mode of human system, BVPN achieves excellent performance in the detection of structural and logical anomalies. Extensive experiments with mainstream anomaly detection datasets and real-world inkjet printing dataset demonstrate that BVPN outperforms state-of-the-art competitors in accuracy. Note to Practitioners—This work addresses the challenge of unsupervised anomaly detection in complicated industrial scenarios. Due to the inconsistency between structural and logical anomalies, current mainstream methods are inadequate for general anomaly detection tasks. The presented BVPN is a robust and generalized detection method, which optimizes the effective receptive field of the model from the perspective of biomimetic visual system. BVPN demonstrates substantial improvements in both structural and logical defects, providing robust detection of surface anomalies across multiple industrial domains, from fabrics to metal components and electronic assemblies. Furthermore, it also achieves practical computational complexity and inference speed, which has been applied in the real-world inkjet printing inspection equipment.