Jiayang Dai

Regenerative aluminum melting serves as a core process in recycled aluminum production. In the regenerative aluminum melting process, the furnace temperature is a key variable which affects product performance and energy costs. The extreme in-furnace temperature necessitates sensors equipped with protective jackets, which increases measurement costs and severely compromises real-time monitoring capability. Accordingly, accurate furnace temperature prediction is highly valuable for regenerative aluminum melting. In regenerative aluminum melting furnaces, periodic burner nozzle commutation and frequent material charging and discharging lead to complex and time-varying operating conditions, posing considerable challenges to high-precision furnace temperature prediction. To address these issues, a condition-adaptive multi-scale convolution-enhanced Transformer (CA-MC-Transformer) model is proposed for furnace temperature prediction. Firstly, an agglomerative hierarchical clustering algorithm based on the weighted dynamic time warping (WDTW) distance is designed to perform unsupervised clustering on historical process data, thereby extracting physically interpretable prior labels for macroscopic operating conditions. Secondly, multi-scale dilated causal convolutions are utilized to capture local dynamic features at diverse temporal resolutions. A soft attention mechanism is further introduced to dynamically assign fusion weights to condition embeddings and local features, enabling condition-adaptive feature reconstruction. Finally, the fused adaptive features are fed into an encoder-only Transformer network to capture the global long-range temporal dependencies and achieve accurate furnace temperature prediction. Comparative experiments conducted on real operational datasets from an aluminum plant verify that the proposed method effectively eliminates the inherent tracking lag of conventional deep learning models, and substantially improves prediction accuracy and anti-noise robustness under complex and variable operating conditions.

In the process of regenerative aluminum smelting, the temperature of the furnace needs to be maintained between 700 and 850 by adjusting the setting parameters of the smelting furnace. The setting parameters are usually adjusted by manual work, and inaccuracies in manual operation can lead to wasted energy as well as unstable temperatures. Energy consumption and temperature stability are two conflicting objectives, which are difficult to find optimal parameters for the aluminum smelting process. In this paper, an improved multi-objective artificial vulture algorithm (IMOAVOA) is developed to solve a multi-objective problem of energy consumption and temperature deviations in the regenerative aluminum smelting process. The dynamic switching–elimination mechanism based on crowding distance is proposed to maintain the archive, which enhances the diversity of solutions by dynamically switching the operation space for deleting redundant solutions in the archive and dynamically deleting the solution with the smallest crowding distance in the operation space. The multi-directional leader selection mechanism is developed to select better leaders. To improve the convergence of the algorithm, the bounce strategy is introduced in the IMOAVOA. The effectiveness of the proposed algorithm is verified by UF1-UF10, kursawe, Viennet2, Viennet3, ZDT1-ZDT6, DTLZ4, and DTLZ6 test functions with several multi-objective algorithms. The experimental results indicate that IMOAVOA outperforms the original algorithm and three other multi-objective algorithms in terms of the algorithm convergence, the Pareto front coverage, and the solution diversity. Finally, the proposed algorithm is tested in an application case of regenerative aluminum smelting process. The results show that the optimal parameters for the aluminum smelting process using the proposed algorithm can reduce the consumption while meeting the objective of furnace temperature.

In the regenerative aluminum smelting process, the furnace temperature is critical for the quality and energy consumption of the product. However, the process requires protective sensors, making real-time furnace temperature measurement costly, while the strong nonlinearity and distribution drift of the process data affect furnace temperature prediction. To handle these issues, a multi-step prediction model for furnace temperature that incorporates reversible instance normalization (RevIN), convolutional neural network (CNN), and Transformer is proposed. First, the self-attention mechanism of the Transformer is combined with CNN to extract global and local information in the furnace temperature data, thus addressing the strong nonlinear characteristics of the furnace temperature. Second, RevIN with learnable affine transformation is utilized to address the distribution drift in the furnace temperature data. Third, the temporal correlation of the prediction model is enhanced by a time-coding method. The experimental results show that the proposed model demonstrates higher prediction accuracy for furnace temperature at different prediction steps in the regenerative aluminum smelting process compared to other models.

The grinding process is an essential part of iron and steel smelting. Due to the changeable ore source and the complex dynamic characteristics of the process, the grinding particle size often cannot meet the requirements, resulting in an increase in process energy and material consumption, as well as reworking. In response to the above problems, the following goal was achieved through this study: a collaborative optimization problem for minimization of the grinding particle size error, process water, amountand electricity consumption has been established, which yields the grinding particle size in real time through the use of a neural network. In order to improve the optimization performance, an enhanced Harris hawks optimization (EHHO) algorithm has been developed. EHHO strengthens the team cooperation between individuals in the algorithm initialization phase and location update phase, as well as introduces an attenuation factor in the search phase to improve the stability and accuracy of the algorithm. The effectiveness of the proposed method is demonstrated through the use of some benchmark functions and an industrial grinding process. The results show that the enhanced Harris hawks optimization algorithm has better optimization accuracy and generalizability; furthermore, it can be effectively applied to the collaborative optimization of grinding process.

In the process of recycled aluminum smelting, timely measurement of the temperature of the smelting furnace is very important for the aluminum yield and quality. However, it is sometimes difficult or costly to measure the temperature in a timely manner due to the high temperature and pressure environment in the furnace. To tackle this problem, a soft sensor modeling framework which combines an operating condition classification and a prediction model based on locally sample-weighted long short-term memory (LSTM) neural network is proposed. In the operating condition classification, a hybrid of dynamic time warping (DTW) based fuzzy c-means and convolutional neural network is used to cluster the training samples and to classify the query samples. In the prediction model, the dynamic time warping and locally sample-weighted technique are introduced to LSTM to solve time-varying and strong nonlinear problems of the process. By adopting the method of classifying the operating conditions of the query samples before temperature prediction, the prediction time can be effectively reduced and the prediction accuracy can be maintained. The results of the experiment show that the proposed method can meet the prediction accuracy and time efficiency requirements of the regenerative aluminum smelting furnace.

In this paper, a double locally weighted extreme learning machine model based on a moving window is developed to realize process modeling. To improve model performances, an improved sparrow-searching algorithm is proposed to optimize the parameters of the proposed model. The effectiveness of the proposed model and algorithm are verified by data from a hematite grinding process. The experimental results show that the proposed algorithm can significantly improve the global search ability and convergence speed in the parameter optimization of the proposed model. The proposed model can correctly estimate the grinding particle size which is expected to be applied to other complex industries.

Iron removal by goethite is a key procedure in zinc hydrometallurgy. Due to its complex chemical reaction mechanism, multiple reactors cascade, and high uncertainty, it is difficult to optimize and control in industrial process. In this paper, a distributed optimization control method which is based on a novel hybrid model of the iron removal process is proposed. By combining the mechanism model and a data-driven oxygen mass transfer coefficient model, a hybrid model is first established. Then, to overcome the influence of the former reactor on the latter reactor, the ratio of the status in each subsystem to the set point is taken as a new status, and a distributed optimization control problem is constructed. Considering the high dimensionality of this problem, it is necessary to reconstruct it by Virtual Motion Camouflage (VMC), so that the optimal control problem is transformed into a nonlinear constrained optimal trajectory planning problem. And a Legendre pseudo-spectral method is used to solve the problem accurately to obtain the optimal trajectory of the ion concentration. Finally, simulation results show that the proposed method can effectively reflect the industrial process, and track the fluctuation of inlet ion concentrations with a nice real-time performance.