A5086068186- Metallurgy and Material Forming
- Iterative Learning Control Systems
- Advanced machining processes and optimization
- Microstructure and Mechanical Properties of Steels
- Magnetic Properties and Applications
- Metallurgical Processes and Thermodynamics
- Gear and Bearing Dynamics Analysis
- Power Line Inspection Robots
- Vibration and Dynamic Analysis
- Advanced Algorithms and Applications
- Metal Alloys Wear and Properties
- Machine Fault Diagnosis Techniques
- Viral Infectious Diseases and Gene Expression in Insects
- Industrial Technology and Control Systems
- Elevator Systems and Control
- Industrial Vision Systems and Defect Detection
- Dynamics and Control of Mechanical Systems
- Infrastructure Maintenance and Monitoring
- Advanced Surface Polishing Techniques
- Mineral Processing and Grinding
- Gait Recognition and Analysis
- CCD and CMOS Imaging Sensors
- Metal Forming Simulation Techniques
- Structural Health Monitoring Techniques
- Advanced Control Systems Optimization
Shenyang Jianzhu University
2021-2025
Northeastern University
2018-2024
Real-time insulator defect detection is critical for ensuring the reliability and safety of power transmission systems. However, deploying deep learning models on edge devices presents significant challenges due to limited computational resources strict latency constraints. To address these issues, we propose YOLOLS, a lightweight efficient model derived from YOLOv8n optimized real-time deployment. Specifically, YOLOLS integrates GhostConv generate feature maps through stepwise convolution,...
Abstract Deep learning-based fault diagnosis methods for rolling bearings are widely utilized due to their high accuracy. However, they have limitations under conditions with few samples. To address this problem, a model-data combination driven digital twin model (MDCDT) is proposed in work samples of bearings. The simulation signals generated by different dynamic models and the measured mixed through MDCDT. MDCDT generates virtual bridge gap between simulated combining respective...
Purpose In the cold rolling process, friction coefficient, oil film thickness and other factors vary dramatically with change in speed, which seriously affects strip deviation. This paper aims to improve control precision forecast roll gap model based on CF-PSO-SVM approach process. Design/methodology/approach this paper, a novel forecasting of support vector machine (SVM) optimized by particle swarm optimization compression factor (CF-PSO) is proposed. Based lots online data, models...
Tandem cold rolling is a high-speed, high-efficiency, and complex production process. In some cases, minor interference or disoperation may result in large economic losses. To improve control performance enhance system robustness, an optimal tension thickness method presented this paper. The designed based on the receding horizon (RHC) strategy, which has good tracking strong constraint handling capability. First, state space model constructed to describe of tandem Then, according model, RHC...
Hot strip rolling is a significant process in the fields of manufacturing and processing. In order to further improve control accuracy looper angle, tension gauge hot finishing mill, an innovative looper-gauge integrated scheme developed this paper. Based on inverse linear quadratic (ILQ) theory, proposed designed for system. First, considering numerous interactions between gauge, disturbances from several sources, dynamic 6th state space model established validated. Then, based ILQ theory...
Abstract The walking part of the wall‐building robot is an important component whose health influences robot's overall performance, and bearing a critical but often disregarded component. As result, based on acoustic signals from roadside, this study proposes method fault monitoring parts to ensure more stable operation robot. To begin, due Doppler effect caused by relative displacement between sensor part, collected signal has frequency shift amplitude attenuation, so multiple...
With the expansion of production scale tandem cold rolling and increase complexity, quality monitoring fault detection are particularly important. In this paper, traditional multivariate statistical analysis method deep learning combined to build a model rolling. Firstly, stacked autoencoder (SAE) is used learn reconstruct input samples complete feature extraction data. Secondly, canonical correlation (CCA) establish characteristic variables variables. <tex...