A5002077422- Metallurgy and Material Forming
- Microstructure and Mechanical Properties of Steels
- Metal Forming Simulation Techniques
- Advanced machining processes and optimization
- Metal Alloys Wear and Properties
- Advanced Surface Polishing Techniques
- Machine Learning and ELM
- Industrial Vision Systems and Defect Detection
- Heat Transfer and Optimization
- Gear and Bearing Dynamics Analysis
- Advanced Algorithms and Applications
- Manufacturing Process and Optimization
- Neural Networks and Applications
- Laser and Thermal Forming Techniques
- Vibration and Dynamic Analysis
- Non-Destructive Testing Techniques
- Advanced Multi-Objective Optimization Algorithms
- Additive Manufacturing and 3D Printing Technologies
- Hydraulic and Pneumatic Systems
- Advanced Sensor and Control Systems
- Infrastructure Maintenance and Monitoring
- Additive Manufacturing Materials and Processes
- Radiative Heat Transfer Studies
- Optical measurement and interference techniques
- Industrial Technology and Control Systems
Northeastern University
2011-2024
Universidad del Noreste
2015-2017
Camber in hot-rolled plates significantly impacts product quality and rolling process stability, making accurate camber prediction crucial. However, it is challenging to measure asymmetric factors impacting real production, hindering the ability of current models predict analyze overall rolled plates. This study proposes a method that combines fluctuations measurable variables with deep learning camber. We developed data analysis platform for plate processing constructed dataset using...
Precision of strip shape is one the important indexes to measure quality steel in hot rolling. Accurate prediction can realise timely adjustment production system, which foundation for ensuring and stable hot-rolled products. Given poor accuracy traditional mechanism model, this paper proposes a stacking ensemble learning crown that consist random forest (RF), extreme gradient boosting (XGBoost), light machine (LightGBM), categorical (CatBoost) linear regression, whose hyperparameters are...
The deformation mechanism is complex in the hot rolling process of clad plates, and head bending a common defect. In this paper, an analytical computational mechanical model metal plate was established by classical elastic mechanics method, relationship between uneven thickness extension warpage obtained. bonding dissimilar bimetallic plates Ti/Steel investigated. On basis, origins defects influence various factors on evolution were revealed. results indicated that forces increased with...
Because of the complexity procedure interface and working conditions, further improvement steel strip quality production efficiency is limited. Realizing optimization product process in multi-process, system-level through intelligent key technology one strategic directions production. (1) Collaborative dynamic scheduling for manufacturing supply chain oriented to customized production, reducing cost raw material purchase operations improving precision service ability; (2) Online monitoring,...
Massive amounts of data are generated during hot strip production in the hot‐rolling metal forming process; resultant dataset is sufficient for model learning steel crown prediction. However, high‐grade nonoriented silicon limited, and rolling process parameters differ, resulting poor Herein, a based on whale optimization algorithm transfer to predict crowns hot‐rolled presented. The composed convolutional linear layers. used optimize hyperparameters obtain an optimal pretraining. then fine...
In article number 2200832, Jingguo Ding and co-workers show that rolling data in information space comes from production lines physical space. The strip shape prediction model is proposed by combining a sparrow search algorithm (SSA) deep multilayer extreme learning machine (DELM). experiments the SSA-DELM has highest accuracy.
The unreasonably high heating process temperature of slab in furnace leads to many defects. To avoid these, a multi‐objective optimization method is proposed for setting based on particle swarm (PSO) algorithm. A 2D model the finite difference scheme, which thickness and width were unequally partitioned, investigated. distance between two neighbouring nodes increased with width, effects more detailed grid surface or side are observed along influence rough core. Then, function setting, from...
Traditional Mizushima automatic plan view pattern control system (MAS) can improve plate rectangular to some extent; as setting curve is not fine, and products yield cannot be further improved, controllable points method for (PVPC) in plate‐rolling process proposed. Prediction model of head metal flow including broadening coefficient deduced, Gaussian mixture introduced, three curves with different shapes are weighted obtain functional expression. To study the influence relationship between...