The forging press plays a crucial and indispensable role in industries such as metallurgy, automotive, and aerospace. Its stable and efficient operation directly impacts the productivity and product quality of the entire production line. Therefore, it is particularly important to carry out efficient fault diagnosis for forging presses. In this paper, a method combining Convolutional Neural Network (CNN), Principal Component Analysis (PCA) and Transmission Support Vector Machine (TSVM) is proposed, which is called Multi-modal Attention-based CNN with PCA and TSVM (MACNN-PTSVM). This method effectively integrates multi-sensor information fusion technology, attention mechanism, CNN, PCA technology and TSVM, fully utilizing data collected from various sensors of the forging press. The constructed network model takes smooth monitoring signal data from the die forging press as input samples. Fault features are extracted through CNN, and PCA is utilized for feature fusion. Applications of this model in forging press fault diagnosis demonstrate significant advantages in fault classification performance. Specifically, its semi-supervised learning technology exhibits higher generalization ability and diagnostic accuracy when dealing with unlabeled data.

Rotating machine fault diagnosis plays a vital role in reducing maintenance costs and preventing accidents. Machine learning (ML) methods and Internet of things (IoT) technologies have been recently introduced into machine fault diagnosis and have generated inspiring results. An ML model with more trainable parameters can typically generate a higher fault diagnostic accuracy. However, the IoT nodes have limited computation and storage resources. How to design an ML model with high accuracy and computational efficiency is still a difficulty and challenge. This work develops an enhanced sparse filtering (ESF) method for mining and fusing the features of the machine signals for fault diagnosis. First, a dimension reduction algorithm is utilized for obtaining the principal components of the vibration signals that are hindered by noises. The distinct features of the principal components are then exploited by using sparse filtering (SF). To reduce the overfitting of the SF model, the L_1/2 norm is applied to regularize the objective function. Finally, the obtained features are combined as the inputs of a softmax classifier for machine fault pattern recognition. The effectiveness, superiority, and robustness of the proposed ESF method are validated by the simulated signals and the practical bearing and motor fault signals compared with the other conventional methods. The lightweight and intelligent ESF algorithm is also deployed onto an edge computing node to realize online motor fault diagnosis. The designed model and the proposed method show great potential in highly accurate and efficient rotation machine fault diagnosis.

Generally, traditional uncertainty design optimization (UDO) methods are based on probability density distribution function or fuzzy membership function. In this situation, a large amount of uncertain information is necessary to construct the UDO model accurately. While, the interval UDO methods require less design information. Only the upper and lower bounds of interval uncertainty are utilized to construct the optimization model. In this study, to enhance the efficiency and accuracy of UBDO considering interval uncertainty, a reliability-based multidisciplinary design optimization (RBMDO) strategy using the point-infilled Kriging model is proposed. In the given method, a double-nested RBMDO model considering interval uncertainty is established. The collaborative optimization is utilized to deal with coupling relationships among complex systems. Then, the point-infilled Kriging response surface strategy is introduced to approximate the RBMDO model. The procedure of the interval multidisciplinary collaborative optimization method based on the Kriging model is discussed. Two examples are given to illustrate the application of the proposed method.

The mapping of digital twins to physical entities suffers from a lack of real-time capability, which affects the twin’s ability to reflect the actual status of the equipment accurately. In manufacturing, latency in critical information, such as industrial equipment parameters, product processing status, and tool usage conditions, can lead to discrepancies between the information displayed on human-machine interfaces and the actual conditions, increasing the risk of incorrect decisions. To address such issues, this paper focuses on the synchronization between digital twins and physical entities in industrial production processes. This paper designs a measurement method for overall time latency and node-level latency in the digital twin process, applies it to a production unit at a forging enterprise, analyzes key factors affecting the real-time performance of digital twins, and discusses methods to improve real-time synchronization between virtual and physical systems. The proposed latency measurement method is universally applicable. It enables researchers to conduct measurements using low-cost tools. Simultaneously, this method can rapidly quantify both types of latency between virtual and physical systems: end-to-end latency and node-level latency. It provides a reliable real-time evaluation approach for digital twins during engineering design and application stages.

The deterministic topology optimization (TO) has become one of the better-known optimization approaches in the engineering design of structures. However, its approximation errors have placed several restrictions on designers as it does not consider inherent uncertainties such as loads, geometrical dimensions, and materials in structures. To remedy this issue, the TO methods were incorporated with reliable design with highly martials involved under uncertainties for structural/mechanical problems. In the current reliability-based topology optimization (RBTO) procedures under multi-source uncertainties, dynamic mean value (DMV) was applied for evaluating the probabilistic constraints. For this aim, a sequential TO and reliability analysis (STORA) was proposed for application in various TO methods. The TO loop coupled bisection method as inverse topology layout applied in reliability loop that the DMV was utilized for evaluating the probabilistic constraint using sufficient descent condition. In the current RBTO model based on STORA, different TO methods named moving iso-surface threshold (MIST), evolutionary structural optimization (ESO), Level-set, and solid isotropic material with penalization (SIMP) are discussed for various problems with continuous design domains. The comparative results are discussed for different TO methods basis bisection approach for TO and RBTO solutions. The results demonstrated that the stable results provided by DMV and different optimal shapes between inverse TO-based bisection and RBTO solutions are captured. The proposed RBTO method using STORA provides safe optimum shapes for almost TO methods with uncertainties in material properties and loads.