Diverse Working Conditions Research Articles

Transient Sand Scour Dynamics Induced by Pulsed Submerged Water Jets: Simulation Analysis

Water jet scouring technology is extensively applied in marine engineering, harbor maintenance, river training, and various other fields, showcasing a broad spectrum of potential applications. However, achieving a comprehensive understanding of the transient sand scouring characteristics of water jets remains challenging due to the inherent complexity of the coupled flow structure involving submerged jets and environmental fluids, along with the intricate dynamics of two-phase flow. This study, rooted in numerical simulation and experimental validation, introduces pulse characteristics into a submerged jet. A thorough investigation is conducted to explore the transient sand scouring characteristics and sand transport laws of the submerged jet under diverse working conditions. The results of this study revealed that the main reason for the asymmetry of the sand pit morphology is not the non-uniform distribution of sand grains, but more likely caused by turbulence effects. Simultaneously, within the initial 0.25 s of the pulse cycle, suspended sediment resulting from the pulsed jet in the preceding cycle gradually transports to the dune and its surrounding areas. Subsequently, from 0.25 s to 0.5 s, sediment on both sides of the pit’s bottom undergoes movement and amalgamation with the sediment that remained unsettled during the previous cycle. The findings reveal that higher jet velocities significantly enhance sediment suspension, migration, and redeposition, leading to deeper erosion and the rapid formation of the sand pit’s outline within 2 s. Additionally, the jet velocity and the impact distance are identified as critical factors influencing erosion depth and sediment dynamics. These insights advance the understanding of erosion mechanisms driven by pulsed jets, highlighting their impact on sediment transport processes. The research findings provide important guidance for dredging and ocean engineering fields and offer a theoretical basis for improving the understanding of submerged jet scouring mechanisms.

Fog Computing Enabled Adaptive Prognosis of Cutting Tool Remaining Life through Multi-source Data

Abstract Cutting tool remaining life during machining processes is essential for achieving sustainable production. Traditional prognosis methods are often crippled by the inability to adapt to diverse working conditions across the machining process lifecycle. This paper introduces a fog computing-enabled adaptive prognosis framework utilising multi-source data to address these challenges effectively. The key innovations include: (1) The proposed system integrates power and vibration data collected from LGMazak VTC-16A and IRON MAN QM200 machines. A standardised data fusion method combines multi-source data to enhance robustness and accuracy. (2) The transformer model is employed to improve prognosis accuracy of cutting tool remaining life, best accuracy of 98.24% and an average accuracy of 97.63% are achieved. (3) Finite Element Analysis (FEA) is incorporated to validate the model's predictions to validate reliability of deep learning model. (4) The fog computing optimisation mechanism based on the bees algorithm, which shows fitness value of 0.92 and convergence within 15 iterations. The proposed method reduces total data volume in cloud by 54.12%, prediction time by 33.64% and time complexity in the cloud layer by 4.62%. The effectiveness of fog computing in enhancing the operational efficiency and reliability of manufacturing systems is validated through advanced data integration and deep learning techniques.

Maternal employment characteristics as a structural social determinant of breastfeeding duration

Abstract Background The European Region exhibits the world’s lowest exclusive breastfeeding rate at 6 months. Addressing work-related breastfeeding challenges is crucial, particularly for women navigating precarious work situations, compounding their adversity. This scoping review surveys research on maternal employment characteristics facilitating breastfeeding continuation post-return to work in the European Region. Methods Studies from 2013 to 2023 were gathered from Scopus, PubMed, and PsycInfo. Inclusion criteria comprised English or French quantitative/qualitative studies exploring links between maternal employment characteristics and breastfeeding status, duration, or experience. Participants were employed mothers of healthy children sustaining breastfeeding post-resumption of work. Key determinants included work-related factors contributing to socially diverse working conditions, encompassing type of employment, working conditions, and work environment. Geographic scope included World Health Organization European Region countries. Results Among 693 hits, 13 studies met criteria. Eight focused on work and breastfeeding, others explored broader breastfeeding determinants. Represented countries: Spain (n = 4), France (n = 4), UK (n = 2), Ireland (n = 2), Netherlands (n = 1). Findings highlighted methodological heterogeneity, indicating a lack of conceptual framework linking work, breastfeeding, and social health inequalities. Nonetheless, self-employment, non-manual professions with time flexibility, workplace breastfeeding facilities, coworker support, and breastfeeding workplace policies emerged as pivotal factors supporting working mothers’ breastfeeding. Conclusions Backing breastfeeding choices for working mothers is crucial amid adversities faced by mothers and children. These results underscore the need for targeted workplace interventions, including time flexibility, breastfeeding facilities, and promotion of breastfeeding-friendly policies. Key messages • Self-employment, non-manual occupations, time flexibility, breastfeeding facilities, support of co-workers and breastfeeding workplace policies promote continuation of breastfeeding. • Policy directives targeting low skilled or precarious jobs are needed to reduce social health inequalities broadly, and in particular, in relation to breastfeeding practices.

Extending cutting tool remaining life through deep learning and laser shock peening remanufacturing techniques

Predicting and extending the remaining life of cutting tools during machining processes is crucial for sustainable manufacturing. Traditional prognosis methods often struggle to adapt to diverse working conditions throughout the machining process lifecycle. This paper introduced a comprehensive framework that effectively addressed the challenges by integrating multi-source data and using deep learning techniques. The system integrated power and vibration data collected from LGMazak VTC-16A and IRON MAN QM200 machines with the following innovations: (1) A standardized data fusion method was developed to integrate multi-source data sources, the hybrid graph convolutional network (GCN) with attention mechanisms was employed to improve the prognosis accuracy of cutting tool remaining life, best accuracy of 98.56% and average accuracy of 97.71% were achieved. (2) The optimization of laser shock peening (LSP) remanufacturing parameters using the bees algorithm showed good performance, a fitness value of 0.95 was achieved with convergence within 15 iterations. (3) Monitoring of the LSP remanufacturing process was designed based on sound and vibration data for optimal remanufacturing performance. (4) The remanufacturing approach in extending the remaining life of cutting tool was validated through FEA analysis and experimental testing, cutting tool life was extended by 29.32% to achieve a sustainable manufacturing process.

Enhancing PV receiver efficiency in laser wireless power transmission through square elliptic hyperboloid concentrator

AbstractLaser wireless energy transmission is a widely utilized method, yet its efficiency is constrained by a variety of factors. In order to improve the conversion efficiency of the receivers of the laser wireless power transmission (LWPT) system, the square elliptic hyperboloid (SEH) concentrating module designed for LWPT system receivers is developed. By analysing the I–V characteristic curves from the results of the experiments and employing non‐linear parameter regression, a corrected battery characteristic curve was derived within a specific laser irradiation range. On this basis, an optical–thermal–electric multi‐field coupling characteristic model was developed. The finite element method is used to simulate the multi‐field coupling characteristics and conversion efficiency of the receiving end under diverse working conditions (including different rotation angles and different divergence angles) of the concentrating photovoltaic module. Research shows: First, the larger the divergence half‐angle β of the laser beam, the more obvious the improvement of the effective optical efficiency of the system by the SEH concentrator. Second, the short‐circuit current and the maximum output power of the PV cell at the receiving end are significantly improved by the SEH concentrator, and the improvement effect is more obvious with the increase of the divergence angle and the rotation angle. Third, SEH concentrators did not significantly affect the fill factor of PV cells.

Forecasting battery degradation trajectory under domain shift with domain generalization

Rechargeable batteries play a pivotal role in the carbon-neutral green environment by electrifying transportation and mitigating the intermittency of renewable energies. Forecasting the degradation of batteries is crucial for the optimal usage of batteries, while predicting battery degradation is not trivial due to diverse working conditions and complex failure mechanisms. To address this challenge, we develop a deep learning model that treats differences in operating conditions as domain shifts and utilizes meta-learning-based and task-driven domain generalization techniques to attack the domain shifts. The model effectiveness is demonstrated on three datasets comprising 203 cells of various operating conditions and chemistries, with improvements in prediction accuracy ranging from 18.1% to 30.0% (23.8% on average). Moreover, the model has gained some generalization capability via learning the correlation between domain gaps in the model and the degradation modes behind various operating conditions. Collectively, our work not only showcases the promise of the high-reliability data-driven model for diverse conditions and chemistries by exploiting domain generalization, but also spotlights the potential interplay between artificial intelligence and domain knowledge.

Towards proactive corrosion management: A predictive modeling approach in pipeline industrial applications

Corrosion is the most significant factor affecting pipeline integrity management. It may lead to facility damage, production interruption, environmental pollution and even explosion. Corrosion management developed in recent years is a critical means to ensure the safety and reliability of pipeline systems. However, developing proactive corrosion management is difficult due to limited availability of data on target pipelines and corrosion factors. The motivation of this study is to develop a proactive corrosion management approach based on a predictive framework. The proposed framework integrates multiple corrosion factors, diverse working conditions, multi-phase degradation, and probabilistic analysis to offer an efficient analysis tool. A numerical analysis of corrosion prediction is conducted using the corrosion data of the training pipeline and the target pipeline, and the corrosion rate, corrosion depth, and remaining useful life (RUL) of the target pipeline are evaluated. Based on the RUL information, the maintenance time of the pipeline is further analyzed to maximize the benefits of maintenance. This study offers a comprehensive approach to pipeline corrosion prediction, facilitating proactive maintenance strategies and improving asset management in the pipeline industry.

Midwifery experiences in rural Southern Aotearoa New Zealand: insights into pre-eclampsia management

aim: This study investigated the experiences of rural midwives in the Southern region of Aotearoa New Zealand, focussing on practices and challenges in caring for pregnant individuals displaying signs of pre-eclampsia (PE). method: Conducted as part of the University of Otago’s Trainee Intern Healthcare Evaluation Project, investigating the efficacy of the soluble FMS-like tyrosine kinase 1 (sFlt-1)/placental growth factor (PlGF) ratio test, this exploratory study employed qualitative research methods. Twenty-three midwives from nine locations across the Southern region were interviewed by trainee intern doctors (TIs) using a semi-structured interview protocol. Thematic analysis was applied to the data. results: The study highlighted the challenging context of rural midwifery, emphasising diverse working conditions, geographic complexities and the impact of the midwifery shortage. Midwives’ decision making about PE depended on location, experience, scientific evidence, holistic model of care and the constant concern about PE. A model illustrating midwifery decision making in PE management was developed. conclusion: Rural midwives in Aotearoa New Zealand’s Southern region managing PE cases face complex challenges. The model derived from this study illustrates the delicate balance that rural midwives navigate, emphasising the need for strategies to support their practice and preserve Aotearoa New Zealand’s distinctive maternity care model.

Ship Autonomous Berthing Strategy Based on Improved Linear-Quadratic Regulator

There has been significant interest in the research field of ship automatic navigation, particularly in the area of autonomous berthing. To address the key challenges of path planning and control during ship berthing, we propose an enhanced Linear−Quadratic Regulator (LQR) control approach, reinforced by the Covariance Matrix Adaptation Evolution Strategy (CMA−ES), along with an adaptive berthing strategy decision model. This integrated framework encompasses ship motion control, path planning, and berthing strategy selection to facilitate adaptive and autonomous ship berthing. Initially, a dynamic mathematical model of ship motion is established, taking into account wind and current interference effects. Subsequently, an adaptive environment−aware berthing strategy model is introduced to enable automatic selection of berthing strategies based on spatial relationships between environmental factors and the berth. By utilizing the refined LQR method, autonomous motion control for ship berthing is achieved. To validate the effectiveness of our controller, comprehensive simulation analyses are conducted under varying operating conditions to encompass crucial factors such as large drift angle characteristics of ships, shallow water effects, and bank effects across seven diverse working conditions. The simulation results underscore the robustness of our proposed method in responding to environmental interference while demonstrating its capability to select appropriate berthing strategies based on varying operational scenarios.

Federated Transfer Fault Diagnosis Method Based on Variational Auto-Encoding with Few-Shot Learning

Achieving accurate equipment fault diagnosis relies heavily on the availability of extensive, high-quality training data, which can be difficult to obtain, particularly for models with new equipment. The challenge is further compounded by the need to protect sensitive data during the training process. This paper introduces a pioneering federated transfer fault diagnosis method that integrates Variational Auto-Encoding (VAE) for robust feature extraction with few-shot learning capabilities. The proposed method adeptly navigates the complexities of data privacy, diverse working conditions, and the cross-equipment transfer of diagnostic models. By harnessing the generative power of VAE, our approach extracts pivotal features from signals, effectively curbing overfitting during training, a common issue when dealing with limited fault samples. We construct a federated learning model comprising an encoder, variational feature generator, decoder, classifier, and discriminator, fortified with an advanced training strategy that refines federated averaging and incorporates regularization when handling non-independent data distributions. This strategy ensures the privacy of data while enhancing the model’s ability to discern subtleties in fault signatures across different equipment and operational settings. Our experiments, conducted across various working conditions and devices, demonstrate that our method significantly outperforms traditional federated learning techniques in terms of fault recognition accuracy. The innovative integration of VAE within a federated learning framework not only bolsters the model’s adaptability and accuracy but also upholds stringent data privacy standards.

Synthesis and tribological characterization of cold-sprayed Ni-based composite coatings containing Ag, MoS2 and h-BN

The tribo components often fail to perform in a desired way due to absence of adequate lubrication in severe conditions and may cause loss of material. Even under room temperature (RT), it is desirable to maintain the proper lubrication for efficient functioning. In recent years, many efforts have been made to develop a coating material which can facilitate low friction and wear properties in diverse working conditions. The proper use of the solid lubricants can deliver utmost lubricity and, therefore, minimize the loss of materials. Solid lubricating coatings have found various industrial applications in harsh environmental conditions, such as in internal combustion engine, gas turbine and aircraft. As far as development of coating is concerned, cold spray (CS) technology has been observed to develop the coatings at relatively low temperature (avoids decomposition) as compared to conventional thermal spray technologies (plasma spray & high velocity oxy fuel). In the current investigation, synergistic response of the participating solid lubricants viz. silver (Ag), molybdenum disulfide (MoS2) and hexagonal boron nitride (h-BN) on the tribological properties of the developed coatings were studied in various working regime of loads. The tribological characteristics of different composite coatings, such as NiAl-MoS2-Ag (NB0), NiAl-MoS2-Ag-5 wt. % hBN (NB5), NiAl-MoS2-Ag-7.5 wt % hBN (NB7.5) and NiAl-MoS2-Ag-10 wt % hBN (NB10) against alumina ball were explored at various loads (6, 11, 16 and 21 N) and at a fixed speed of 0.3 m/s under room temperature (RT). The fixed concentration of Ag and MoS2 with varying weight percentage of h-BN (0, 5, 7.5 and 10 wt %) were introduced as solid lubricants for evaluating the lubricating potential of h-BN. During the tests, coefficient of friction (COF) and wear rate were observed to lessen as the load increased from 6 to 16 N. However, increased COF and wear rate were noticed for all the participating coatings at higher load (21 N). The ideal content of hBN in the deposited coatings was ascertained in order to get the utmost favourable lubricating conditions. It was observed that NB7.5 coating exhibited the enhanced tribological characteristics under the aforesaid operating conditions. Coating NB7.5 shows the best friction and wear characteristics during each testing load and reached a COF (0.19) and wear rate (2.1 × 10−5 mm3/Nm) at 16 N load. The observed wear mechanisms were analysed on the basis of the synergy shown by Ag, MoS2, and hBN as well as the optimal hBN level in the coating (NB7.5) which helped in attaining the improved tribological properties.

Cross-domain bearing fault diagnosis using dual-path convolutional neural networks and multi-parallel graph convolutional networks

Bearing fault diagnosis is significant in ensuring large machinery and equipment's safe and stable operation. However, inconsistent operating environments can lead to data distribution differences between source and target domains. As a result, models trained solely on source-domain data may not perform well when applied to the target domain, especially when the target-domain data is unlabeled. Existing approaches focus on improving domain adaptive methods for effective transfer learning but neglect the importance of extracting comprehensive feature information. To tackle this challenge, we present a bearing fault diagnosis approach using dual-path convolutional neural networks (CNNs) and multi-parallel graph convolutional networks (GCNs), called DPC-MGCN, which can be applied to variable working conditions. To obtain complete feature information, DPC-MGCN leverages dual-path CNNs to extract local and global features from vibration signals in both the source and target domains. The attention mechanism is subsequently applied to identify crucial features, which are converted into adjacency matrices. Multi-parallel GCNs are then employed to further explore the structural information among these features. To minimize the distribution differences between the two domains, we incorporate the multi-kernel maximum mean discrepancy (MK-MMD) domain adaptation method. By applying the DPC-MGCN approach for diagnosing bearing faults under diverse working conditions and comparing it with other methods, we demonstrate its superior performance on various datasets.

Fuzzy PID Control Design of Mining Electric Locomotive Based on Permanent Magnet Synchronous Motor

Achieving precise stopping of electric locomotives is crucial for the realization of intelligent and unmanned auxiliary transportation systems. Presently, human drivers play a central role in ensuring accurate stopping, presenting obstacles to automation and cargo location precision, especially within the coal mining sector. This article centers on achieving the precise stopping of electric locomotives under various conditions through the utilization of permanent magnet synchronous motor-driven locomotives. This approach introduces a novel stopping control method that integrates a fuzzy proportional–integral–derivative (F-PID) controller with a vector control model for permanent magnet synchronous motors (PMSM). Subsequently, we develop the F-PID controller using the PMSM technique, incorporating new fuzzy rules for each subsystem to enhance control accuracy and efficiency. Finally, extensive simulations and real-world experiments are conducted on an electric locomotive stopping test bed to validate the effectiveness of the proposed control method. The results show that the method consistently achieves precise stopping under diverse working conditions, with an error of less than 0.3 m, confirming its robustness and reliability.

A dual‐channel transferable RUL prediction method integrated with Bayesian deep learning and domain adaptation for rolling bearings

AbstractMany deep learning methods typically assume that the marginal probability distribution between the training and testing bearing data is similar or the same. However, the probability distribution of rolling bearings may deviate significantly under diverse working conditions. To address the above limitations, a novel transferable remaining useful life (RUL) prediction method integrated with Bayesian deep learning and unsupervised domain adaptation (DA) is proposed. First, the signal alignment is executed on the data after the first prediction time to maintain the same granularity and scale across both source and target domains. Second, the multi‐domain features are extracted and sent into the dual‐channel Transformer network (DCTN) incorporating the convolutional block attention module (CBAM) to adequately exploit the abundant degradation information. Then, the DA module is incorporated into the model to mitigate the distribution discrepancies of the extracted high‐level merged features between the source and target domains. Finally, by applying the variational inference method, the DCTN‐CBAM is extended to the Bayesian deep neural network, and the RUL prediction and its corresponding confidence intervals can be conveniently derived. In addition, the generalization capability and effectiveness are validated through six bidirectional transfer RUL prediction tasks across two rolling bearing datasets. The experimental results demonstrate that it could provide a more reliable RUL prediction and efficiently account for the prediction uncertainty.

Localizing and tracking of in-pipe inspection robots based on distributed optical fiber sensing

Ensuring the safe transportation of energy relies heavily on the timely safety inspection and pigging of energy pipelines using in-pipe inspection robots known as Smart/Intelligent PIGs. Addressing the potential risks of PIG block incidents and the limitations of excessive speed in the maintenance, real-time localization and tracking of these inspection robots has become imperative. However, traditional localization and tracking methods present a challenge due to their resource-intensive nature, requiring significant manpower and resources. To overcome this limitation and enhance the intelligence of the robot operation monitoring, this paper proposes an innovative artificial intelligence (AI) integration algorithm framework based on distributed optical fiber sensing (DOFS) for real-time localization and tracking of robots. It adopts noise reduction and reconstruction techniques in signal processing, effectively enhancing the quality of optic fiber vibration signals. Notably, the integration of two distinct features that complement each other enables dual verification of tracking detection, ultimately bolstering the system’s effectiveness and trustworthiness. Besides, a logical reasoning-based localization decision strategy further enhances the system’s capabilities, allowing for controlled tracking intervals and step sizes that can be tailored to meet the task requirements under diverse working conditions. The collaboration of three modules makes it feasible to monitor dynamic changes of the detection robot in the pipeline along the direction of operation. The experimental results convincingly demonstrate that the integrated framework possesses several key advantages of remarkable robustness, real-time performance, and minimal error. It underscores the system’s potential to ensure energy pipeline safety efficiently and effectively.

TSoft-Net: A novel transfer soft thresholding network based on self-attention for intelligent fault diagnosis of rotating machinery

In the industrial scenes, the machinery operates under diverse working conditions and generates varying levels of noise, which can hinder the performance of the intelligent fault diagnosis model that is trained in the laboratory. This is because the different working conditions and environmental noise change the distribution of laboratory vibration signal. To tackle this issue, we proposed a novel transfer soft thresholding network (TSoft-Net) based on self-attention for intelligent fault diagnosis of rotating machinery, which has good anti-noise performance and can be transferred to different working conditions with excellent accuracy. This paper constructs a soft residual block for extracting the fault representation and enhancing the residual learning. In this block, we propose a residual factor to learn and enhance domain-invariant fault representation. Furthermore, a representation soft fusion block is built for extracting and fusing the different scale fault representation. In this block, we propose a scale-attention weight to dynamically fuse the different scale fault representation. The experiments show that the TSoft-Net, compared with six existing methods on eight sub-datasets, has better anti-noise ability and achieves an accuracy of up to 100% on the target domain.

EhdNet: Efficient Harmonic Detection Network for All-Phase Processing with Channel Attention Mechanism

The core of harmonic detection is the recognition and extraction of each order harmonic in the signal. The current detection methods are seriously affected by the fence effect and spectrum aliasing, which brings great challenges to the detection of each order harmonic in the signal. This paper proposes an efficient harmonic detection neural network based on all-phase processing. It is based on three crucial designs. First, a harmonic signal-processing module is developed to ensure phase invariance and establish the foundation for subsequent modules. Then, we constructed the backbone network and utilized the feature-extraction module to extract deep abstract harmonic features of the target. Furthermore, a channel attention mechanism is also introduced in the weight-selection module to enhance the energy of the residual convolution stable spectrum feature, which facilitates the accurate and subtle expression of intrinsic characteristics of the target. We evaluate our method based on frequency, phase, and amplitude in two environments with and without noise. Experimental results demonstrate that the proposed EhdNet method can achieve 94% accuracy, which is higher than the compared methods. In comparison experiments with actual data, the RMSE of EhdNet is also lower than that of other recent methods. Moreover, the proposed method outperforms ResNet, BP, and other neural network approaches in data processing across diverse working conditions due to its incorporation of a channel attention mechanism.

Model-based Reinforcement Learning for Ship Path Following with Disturbances

Modelling and control of ships is a challenging task due to their intrinsic nonlinearities and high uncertainty. In the complex and dynamic sea environment, the efficacy of well-designed motion controllers diminishes significantly. To achieve optimal performance, high-performance motion control systems must possess the ability to adapt to diverse working conditions, repel external disturbances, and incorporate learning capabilities. In addressing critical computational challenges encountered in the real-world deployment of autonomous driving agents, Reinforcement Learning (RL) methods have been effectively employed. This paper proposed a reinforcement learning control method for ship path following based on an environmental disturbance model. RL is utilized to learn the dynamic properties of the system to resist environmental disturbances. Experimental results show that the method can follow the path with acceptable accuracy and high robustness.

A Novel Method for State of Charge Estimation in Lithium-Ion Batteries Using Temporal Convolutional Network and Multi-Verse Optimization

This study presents a novel data-driven method for state-of-charge estimation in lithium-ion batteries. It integrates a temporal convolutional network with multi-verse optimization to enhance the accuracy of predicting the state of charge. The temporal convolutional network possesses advantages such as an extended memory window and efficient parallel computation, exhibiting exceptional performance in time-series tasks for state of charge estimation. Its hyperparameters are optimized by adopting multi-verse optimization to obtain better model performance. The driving model utilizes various measurable data as inputs, including battery terminal voltage, current, and surface temperature. To validate the effectiveness of the proposed method, extensive datasets from diverse dynamic working conditions at different ambient temperatures are employed for model training, validation, and testing. The numerical outcomes provide evidence of the proposed method’s superior performance compared to the other two methods, providing a more robust and accurate solution for the state of charge estimation in lithium-ion batteries.

Cross-domain fault diagnosis method for rolling bearings based on contrastive universal domain adaptation

To address the unknown spatial relationship between source and target domain labels, which leads to poor fault diagnosis accuracy, a contrastive universal domain adaptation model and rolling bearing fault diagnosis approach are proposed. The approach introduces bootstrap your own latent network to mine the data-specific structure of the target domain and proposes rejecting unknown class samples using an entropy separation strategy. Simultaneously, a source class weighting mechanism is designed to improve the transferable semantics augmentation method by assigning various class-level weights to source categories, which improves the alignment of the feature distributions in the shared label space to further construct fault diagnosis models. Experimental validation on two rolling bearing datasets confirmed the superior fault diagnosis accuracy of the proposed method under diverse working conditions.