The design, planning, and implementation of intelligent manufacturing are mainly carried out from the perspectives of meeting the needs of mass customization, improving manufacturing capacity, and innovating business pattern currently. Environmental and social factors should be systematically integrated into the life cycle of intelligent manufacturing. In view of this, a green performance evaluation methodology of intelligent manufacturing driven by digital twin is proposed in this paper. Digital twin framework, which constructs the bidirectional mapping and real-time data interaction between physical entity and digital model, provides the green performance evaluation with a total factor virtual image of the whole life cycle to meet the monitoring and simulation requirements of the evaluation information source and demand. Driven by the digital twin framework, a novel hybrid MCDM model based on fuzzy rough-sets AHP, multistage weight synthesis, and PROMETHEE II is proposed as the methodology for the green performance evaluation of intelligent manufacturing. The model is tested and validated on a study of the green performance evaluation of remote operation and maintenance service project evaluation for an air conditioning enterprise. Testing demonstrates that the proposed hybrid model driven by digital twin can enable a stable and reasonable evaluation result. A sensitivity analysis was carried out by means of 27 scenarios, the results of which showed a high degree of stability.
To address traditional methods’ limitations in neglecting the interaction between random shock loads and progressive degradation, as well as failure correlations, this study proposes a dynamic reliability framework integrating Gamma processes, homogeneous Poisson processes (HPP), and mixed Copula functions. The framework develops a wind turbine gearbox reliability model under shock–degradation coupling while quantifying failure correlations. Gamma processes characterize continuous degradation, with parameters estimated from P-S-N curves. Based on stress–strength interference theory, random shocks within damage thresholds are integrated to form a coupled reliability model. A Gumbel–Clayton–Frank mixed Copula with a multi-layer nested algorithm quantifies failure correlations, with correlation parameters estimated via the RSS principle and genetic algorithms. Validation using a 2 MW gearbox’s planetary gear-stage system covers four scenarios: natural degradation, shock–degradation coupling, and both scenarios with failure correlations. The results show that compared to independent assumptions, the model accelerates reliability decline, increasing failure rates by >37%. Relative to natural degradation-only models, failure rates rise by >60%, validating the model’s effectiveness and alignment with real-world operational conditions.
Mobile notification systems play a major role in a variety of applications to communicate, send alerts and reminders to the users to inform them about news, events or messages. In this paper, we formulate the near-real-time notification decision problem as a Markov Decision Process where we optimize for multiple objectives in the rewards. We propose an end-to-end offline reinforcement learning framework to optimize sequential notification decisions. We address the challenge of offline learning using a Double Deep Q-network method based on Conservative Q-learning that mitigates the distributional shift problem and Q-value overestimation. We illustrate our fully-deployed system and demonstrate the performance and benefits of the proposed approach through both offline and online experiments.
With the widespread application of wind power technology, the detection of abnormalities in wind turbine blades has become a key research area. The use of data from monitoring and data acquisition (SCADA) systems for data-driven fault detection research presents new challenges. This study utilizes short-term SCADA data from wind turbine generators to classify the blade abnormal and normal operational states, thereby introducing a new method called PCABSMMR. This strategy integrates principal component analysis (PCA) and borderline-synthetic minority over-sampling technique (Borderline-SMOTE) for data processing and utilizes an improved multi-dimensional time series classification (MTSC) model. It combines one-dimensional convolution from deep learning with shallow learning’s rigid classifiers. PCA is used for dimensionality reduction, while Borderline-SMOTE expands the samples of minority class fault instances. Comparative analysis with various methods shows that the proposed method has an average F1-score of 0.98, outperforming many state-of-the-art MTSC models across various evaluation metrics.
Accurate estimation of multiaxial fatigue life plays a critical role in maintaining the structural integrity and operational reliability of mechanical components subjected to complex loading conditions. Under non-proportional loading, fatigue life tends to decrease significantly due to the emergence of additional damage mechanisms, such as dislocation accumulation, cyclic hardening, and accelerated propagation of micro-cracks. This study conducts a systematic investigation into the primary factors that influence fatigue behavior under non-proportional loading conditions. A novel damage factor is proposed, which quantifies the additional strengthening effects caused by complex stress and strain interactions. Based on this factor, a new prediction model is developed through the combination of critical plane theory and an energy-based framework. This model captures the influence of non-proportional strengthening on fatigue strength with improved accuracy. Experimental validation is carried out using En8, TC4, and Al7050-T7451 materials under tension and torsion loading conditions. Comparative analysis with three conventional models shows that the proposed method improves the accuracy of predictions and offers a dependable approach for practical engineering applications.
To reduce the significant time and cost associated with wind turbine blade fatigue testing, the applicability of the deep learning model Neural Basis Expansion Analysis (N-BEATS) for modeling the stiffness degradation of wind turbine blades was investigated. First, on the basis of a traditional blade stiffness degradation model, the stiffness data were expanded to meet the data volume requirements of N-BEATS. Second, the basic block structure of N-BEATS was improved (by treating the sequence-to-sequence prediction problem as a nonlinear multivariate regression problem) to meet the specific prediction requirements of this task, and the Pinball Mean Absolute Percentage Error (Pinball-MAPE) loss function was adopted to further reduce bias during the prediction process. Additionally, two data augmentation methods—time series combination and random noise injection—were applied to mitigate the risk of model overfitting and improve prediction accuracy. Experimental results demonstrated that the model can effectively learn underlying patterns in the stiffness data and successfully predict the remaining stiffness.