Yanlin Zhao

The iron and steel industry belongs to the most electricity-intensive branches of manufacturing, and power expenses often account for a large portion of overall production costs. For this reason, reliable short-term forecasts of plant-level electricity consumption are essential for optimizing production planning, avoiding excessive demand charges, and supporting low-carbon operation. This study compares three representative approaches for hourly power-load prediction in a steel enterprise: eXtreme Gradient Boosting (XGBoost), Long Short-Term Memory (LSTM), and Bidirectional LSTM (BiLSTM). Using one year of operating data from an integrated steel plant, we build a dataset of 8,760 hourly records and design a unified preprocessing procedure, including three-sigma outlier detection, time-indexed linear interpolation, and chronological division into training, validation, and test subsets. All three models are trained within a 24-hour sliding-window framework and assessed with RMSE, MAE, and MAPE. The results indicate that each model is capable of depicting the cyclical variation of the load, whereas BiLSTM provides the most accurate predictions, achieving the lowest errors in both absolute and relative terms. XGBoost demonstrates competitive performance and robust trend following. In contrast, the unidirectional LSTM exhibits larger relative errors, particularly during low-load periods. The findings underline the benefits of bidirectional recurrent structures for short-term electricity-consumption forecasting in steel plants and offer guidance for model selection in industrial energy-management practice.

Rolling bearings are essential for modern mechanical equipment and serve in various operational environments. This paper addresses the challenge of vibration data discrepancies in bearings across different operating conditions, which often results in inaccurate fault diagnosis. To tackle this related limitation, a novel lightweight multi-scale attention-based joint adaptive adversarial transfer network, termed MAJATNet, is developed. The proposed network integrates a feature extraction network innovation module with an improved loss function, namely IJA loss. The feature extraction module employs a one-dimensional multi-scale attention residual structure to derive characteristics from monitoring data of source and target domains. IJA loss evaluates the joint distribution discrepancy of high-dimensional features and labels between these domains. IJA loss integrates a joint maximum mean discrepancy (JMMD) loss with a domain adversarial learning loss, which directs the model’s focus toward categorical features while minimizing domain-specific features. The performance and advantages of MAJATNet are demonstrated through cross-domain fault diagnosis experiments using bearing datasets. Experimental results show that the proposed method can significantly improve the accuracy of cross-domain fault diagnosis for bearings.

This study utilizes an improved NSGA-II algorithm to conduct a multi-objective optimization of the hospital outpatient department layout. By simultaneously incorporating patient walking distance, hospital operating costs, patient waiting time, and medical staff work efficiency as optimization objectives, and adopting an adaptive population size adjustment strategy, this paper optimizes the existing outpatient layout in a case study of a three-story outpatient building at Panzhihua Central Hospital. The results show that the new plan reduces patient walking distance by 57.2%, shortens waiting time by 59%, and enhances medical staff collaboration efficiency, while only increasing costs by 5.6%. This demonstrates the effectiveness and feasibility of the improved NSGA-II method in handling complex multi-objective optimization problems for outpatient layouts. The research findings provide a reference for the rational allocation of hospital outpatient resources and the improvement of service quality. Additionally, this paper discusses the applicability and limitations of the study and proposes future research directions, including validating the method’s effectiveness in hospitals of various types and sizes, incorporating dynamic optimization and real-time data, and deeply integrating with hospital information systems.

This paper investigates the layout optimization of multi-layer circular manufacturing cells (MCMC), a topic that has garnered limited attention compared to single-layer circular manufacturing cells (SCMC). With the continuous advancement of global intelligent manufacturing, MCMC has emerged as a viable solution, with several smart factories already implementing this model. Existing literature predominantly utilizes the NSGA II algorithm for SCMC layouts due to their relatively few objectives. However, the layout problem for MCMC encompasses a significantly larger number of objectives, rendering NSGA II inadequate. This study aims to fill this research gap by proposing an innovative approach using NSGA III, specifically designed for complex multi-objective optimization. A multi-dimensional target mathematical model for MCMC is established, facilitating the systematic examination of layout configurations. The methodology employs NSGA III to effectively tackle the challenges posed by MCMC layouts. To validate the effectiveness of NSGA III, empirical data from a smart factory in Zhejiang, China, is utilized. The findings demonstrate that NSGA III significantly outperforms traditional algorithms, yielding superior solutions for MCMC layout problems. This research not only challenges the conventional SCMC layout paradigm but also expands the options available for facility layouts in smart factories. Ultimately, it addresses the pressing engineering needs of smart factory construction and contributes valuable insights to the field of MCMC research, establishing a robust methodology for future studies.

With the rapid development of intelligent manufacturing, the application of virtual reality technology to the optimization of workshop facility layout has become one of the development trends in the manufacturing industry. Virtual reality technology has put forward engineering requirements for real-time solutions to the Workshop Facility Layout Optimization Problem (WFLOP). However, few scholars have researched such solutions. Deep reinforcement learning (DRL) is effective in solving combinatorial optimization problems in real time. The WFLOP is also a combinatorial optimization problem, making it possible for DRL to solve the WFLOP in real time. Therefore, this paper proposes the application of DRL to solve the dual-objective WFLOP. First, this paper constructs a dual-objective WFLOP mathematical model and proposes a novel dual-objective DRL framework. Then, the DRL framework decomposes the WFLOP dual-objective problem into multiple sub-problems and then models each sub-problem. In order to reduce computational workload, a neighborhood parameter transfer strategy is adopted. A chain rule is constructed for the appealed sub-problem, and an improved pointer network is used to solve the bi-objective WFLOP of the sub-problem. Finally, the effectiveness of this method is verified by using the facility layout of a chip production workshop as a case study.

The research on the layout of multi-layer manufacturing cells for smart factories is still in its infancy, but there is an urgent need to address this issue in building smart factories. This paper presents the Manufacturing Cell Integrated Layout (MCIL) Method, which integrates multiple layout forms of multi-layer and single-layer manufacturing cells. The paper develops a mathematical model of the MCIL problem which considers the multi-objective functions of logistics handling, occupied space, cell stability, lost time, and non-logistics relations, as well as the constraints between equipment in the cell and cells. An adaptive RNS-FOA algorithm is proposed to solve the high-dimensional and large-scale characteristics of the MCIL problem based on the research of academics. Lastly, a case demonstrates the outstanding contribution of the mathematical model to the solution of the MCIL problem, while simultaneously validating the efficiency and stability of the RNS-FOA algorithm for solving the MCIL problem.

The article puts forward the new layout methodology of the multi-floor linear cellular manufacturing layout. The proposed equipment layout methodology not just breaks the conventional single-floor linear cellular manufacturing layout but also meets the layout requirements of the intelligent manufacturing workshop for the stereoscopic aisle manufacturing cell. The layout methodology takes into account the least space occupation as well as the shortest total distance of logistics as the objective function, besides considering the limitations that exist between the equipment, different planes, different levels, and so on; also, a mathematical model is put forward. The multi-floor linear cellular manufacturing layout is solved based on the self-adapting multi-objective fruit fly optimization algorithm that refers to an algorithm combining fruit fly optimization algorithm and NSGA-II. Self-adapting multi-objective fruit fly optimization algorithm makes use of the fast nondominated sorting for the multi-target food concentration calculation, together with designing the adaptive olfactory search and visual search, and employing the perturbation operations for flight strategies, aimed at ensuring the population diversity. Simulation cases suggest that self-adapting multi-objective fruit fly optimization algorithm has stronger advantages as compared with multi-objective fruit fly algorithm and elitist non-dominated sorting genetic algorithm (NSGA-II) in the solution of multi-floor linear cellular manufacturing layout problems. The final engineering case application sheds light on the fact that multi-floor linear cellular manufacturing layout saves 57.6% of the area, in addition to 23.7% of space, and 29.2% of the handling distance as compared with single-floor linear cellular manufacturing layout. Accordingly, multi-floor linear cellular manufacturing layout has a specific reference value in the layout of facilities in the intelligent manufacturing plants.