A digital-twin-model-based optimal control system is presented for the steel continuous casting process. The system is designed for the coordinated optimization and dynamic control of secondary cooling and final electromagnetic stirring (FEMS), and involves three related parts. Firstly, a three dimensional real-time heat transfer model is established as the digital twin of the heat transfer process of continuous casting; for high accuracy, it is calibrated offline and calibrated online using measurements of the surface temperatures and shell thicknesses (only offline). Secondly, according to metallurgical rules, cooling and stirring are optimized coordinatively, based on the established digital-twin model and chaos particle swarm optimization algorithm. Thirdly, cooling and stirring are further dynamically controlled for quality stability. Finally, the system is applied in a bloom caster with model errors ≤ ±10 °C and control errors ≤ ±4 °C, which reduces the macro-segregation over grade 1.5 from 11% to 3.3%.
1 January 2011 in Journal of Iron and Steel Research International
Laminar cooling is the key to hot rolling and controlled cooling. It can improve the performance of steel by changing the grain transformation. The key to improve product quality is that the laminar cooling process model can calculate the cooling strategy accurately. The conventional empirical formula calculation method can't guarantee the accuracy of the water cooling heat transfer coefficient, which is the critical parameters of laminar cooling process. Therefore, an experimental study like this is needed. In this paper, the temperature measurement experiment of laminar cooling is carried out by the method of experimental temperature measurement. The experimental design stage uses COMSOL software to simulate the temperature field temperature to determine the diameter and depth of the measurement. Then the laminar cooling experiment is carried out and collects the temperature measurement data under different conditions. The error of the measured temperature field is less than 0.1% under the final experimental conditions. The heat transfer coefficient is obtained by experimental data and input in heat transfer model, the surface temperature is calculated, and the temperature is compared with the experimental data, the relative error is less than 20℃.
1 August 2019 in Journal of Physics: Conference Series
In the established temperature field model of plate hot rolling process, the heat transfer coefficients of each stage are solved by empirical formula, but due to the complexity of the production environment, temperature field models tend to be less accurate. In order to solve this problem, this paper designs a rolling experiment, collects the actual temperature data, and combines with genetic algorithm, simulated annealing algorithm and particle swarm optimization algorithm to correct the heat transfer coefficients of air-cooling zone, and the three algorithms are compared. Finally, the particle swarm algorithm is selected to establish the temperature field model of the hot rolling process. The average error of the final temperature field model calculation is within 14°C.
1 August 2019 in Journal of Physics: Conference Series
The current temperature field model of plate rolling process is generally based on CPU programming, due to the influence of computer performance, the accuracy and speed of the model can not be improved at the same time. To solve this problem, this paper studies the temperature field model with particle swarm optimization to modify the heat transfer coefficients, and the parallel program of this model is written under CUDA architecture, on premise of ensuring the accuracy of the model, the computation speed of the model is improved. To get the best acceleration, this paper analyzes the factors affecting GPU performance, the accuracy and speed of the temperature field model before and after modification are compared under the best acceleration configuration. Final temperature field model calculation error reduced to an average of 15.57 °C, better than the model based on the empirical formula, the computing speed is also improved, which is 1.78 times that of CPU.