Fault Diagnosis of a Multistage Centrifugal Pump Using Explanatory Ratio Linear Discriminant Analysis

Sensors, Journal Year: 2024, Volume and Issue: 24(6), P. 1830 - 1830

Published: March 13, 2024

This study introduces an innovative approach for fault diagnosis of a multistage centrifugal pump (MCP) using explanatory ratio (ER) linear discriminant analysis (LDA). Initially, the method addresses challenge background noise and interference in vibration signals by identifying fault-sensitive frequency band (FSFB). From FSFB, raw hybrid statistical features are extracted time, frequency, time–frequency domains, forming comprehensive feature pool. Recognizing that not all adequately represent MCP conditions can reduce classification accuracy, we propose novel ER-LDA method. evaluates importance calculating between interclass distance intraclass scatteredness, facilitating selection discriminative through LDA. fusion ER-based assessment LDA yields technique. The resulting selective set is then passed into k-nearest neighbor (K-NN) algorithm condition classification, distinguishing normal, mechanical seal hole, scratch, impeller defect states MCP. proposed technique surpasses current cutting-edge techniques classification.

Language: Английский

---

Research on roller bearing fault diagnosis based on robust smooth constrained matrix machine under imbalanced data

Advanced Engineering Informatics, Journal Year: 2024, Volume and Issue: 62, P. 102667 - 102667

Published: June 25, 2024

Language: Английский

---

Experimental study on the unsteady evolution mechanism of centrifugal pump impeller wake under solid–liquid two-phase conditions: Impact of particle concentration

Physics of Fluids, Journal Year: 2024, Volume and Issue: 36(11)

Published: Nov. 1, 2024

To study the spatiotemporal evolution process of particle wakes behind impeller in centrifugal pump, this paper utilized high-speed photography to capture motion characteristics under different solid-phase concentrations (1%, 1.5%, and 2%). First, studies changes hydraulic performance pump solid–liquid two-phase flow conditions. It then introduces wake, comparing differences wake varying concentrations. Finally, impact concentration on wear volute's partitions is investigated. This found that as increases, gradually declines. Under design conditions, when increases by 0.5%, efficiency decreases 0.56% 0.35%. There mutual transport particles between adjacent wakes, movement within volute passage not equidistant over time. As cutting occurs at partitions, there a significant separation wakes. The spatial significantly influenced concentration. Wear intensifies with increasing also affected research results provide basis for further exploration dynamics pumps.

Language: Английский

---

Artificial intelligence-driven ensemble deep learning models for smart monitoring of indoor activities in IoT environment for people with disabilities

Scientific Reports, Journal Year: 2025, Volume and Issue: 15(1)

Published: Feb. 5, 2025

Disabled persons demanding healthcare is a developing global occurrence. The support in longer-term care includes nursing, intricate medical, recovery, and social help services. price large, but advanced technologies can aid decreasing expenditure by certifying effective health services enhancing the superiority of life. transformative latent Internet Things (IoT) prolongs existence nearly one billion worldwide with disabilities. By incorporating smart devices technologies, IoT provides solutions to tackle numerous tasks challenged individuals disabilities promote equality. Human activity detection methods are technical area which studies classification actions or movements an individual achieves over recognition signals directed smartphones wearable sensors images video frames. They efficient functions actions, observing crucial functions, tracking. Conventional machine learning deep approaches effectively detect human activity. This study develops designs metaheuristic optimization-driven ensemble model for monitoring indoor activities disabled (MOEM-SMIADP) model. proposed MOEM-SMIADP concentrates on detecting classifying using applications physically people. First, data preprocessing performed min-max normalization convert input into useful format. Furthermore, marine predator algorithm employed feature selection. For activities, utilizes three classifiers, namely graph convolutional network model, long short-term memory sequence-to-sequence (LSTM-seq2seq) method, autoencoder. Eventually, hyperparameter tuning accomplished improved coati optimization enhance outcomes models. A wide range experiments was accompanied endorse performance technique. validation technique portrayed superior accracy value 99.07% existing methods.

Language: Английский

---

Hybrid Deep Learning Model for Fault Diagnosis in Centrifugal Pumps: A Comparative Study of VGG16, ResNet50, and Wavelet Coherence Analysis

Machines, Journal Year: 2024, Volume and Issue: 12(12), P. 905 - 905

Published: Dec. 10, 2024

Significant in various industrial applications, centrifugal pumps (CPs) play an important role ensuring operational efficiency, yet they are susceptible to faults that can disrupt production and increase maintenance costs. This study proposes a robust hybrid model for accurate fault detection classification CPs, integrating Wavelet Coherence Analysis (WCA) with deep learning architectures VGG16 ResNet50. WCA is initially applied vibration signals, creating time–frequency representations capture both temporal frequency information, essential identifying subtle characteristics. These enhanced signals processed by ResNet50, each contributing unique complementary features enhance feature representation. The approach fuses the extracted features, resulting more discriminative set optimizes class separation. proposed achieved test accuracy of 96.39%, demonstrating minimal overlap t-SNE plots precise confusion matrix. When compared ResNet50-based VGG16-based models from previous studies, which reached 91.57% 92.77% accuracy, respectively, displayed better performance, particularly distinguishing closely related classes. High F1-scores across all categories further validate its effectiveness. work underscores value combining multiple CNN advanced signal processing reliable diagnosis, improving real-world CP applications.

Language: Английский

---

Auscultation-Based Pulmonary Disease Detection through Parallel Transformation and Deep Learning

Bioengineering, Journal Year: 2024, Volume and Issue: 11(6), P. 586 - 586

Published: June 8, 2024

Respiratory diseases are among the leading causes of death, with many individuals in a population frequently affected by various types pulmonary disorders. Early diagnosis and patient monitoring (traditionally involving lung auscultation) essential for effective management respiratory diseases. However, interpretation sounds is subjective labor-intensive process that demands considerable medical expertise, there good chance misclassification. To address this problem, we propose hybrid deep learning technique incorporates signal processing techniques. Parallel transformation applied to adventitious sounds, transforming sound signals into two distinct time-frequency scalograms: continuous wavelet transform mel spectrogram. Furthermore, parallel convolutional autoencoders employed extract features from scalograms, resulting latent space fused feature pool. Finally, leveraging long short-term memory model, used as input classifying Our work evaluated using ICBHI-2017 dataset. The experimental findings indicate our proposed method achieves promising predictive performance, average values accuracy, sensitivity, specificity, F1-score 94.16%, 89.56%, 99.10%, respectively, eight-class diseases; 79.61%, 78.55%, 92.49%, 78.67%, four-class 85.61%, 83.44%, 84.21%, binary-class (normal vs. abnormal) sounds.

Language: Английский

---

Deep learning for fault diagnosis of monoblock centrifugal pumps: a Hilbert–Huang transform approach

International Journal of Systems Assurance Engineering and Management, Journal Year: 2024, Volume and Issue: unknown

Published: Sept. 4, 2024

Language: Английский

---

Advanced Fault Detection in Power Systems Using Wavelet Transform: SIMULINK-Based Implementation and Analysis

Journal of Energy Engineering and Thermodynamics, Journal Year: 2024, Volume and Issue: 43, P. 12 - 25

Published: April 23, 2024

Traditional methods struggle to find faults in power transmission lines. This paper presents an approach for short lines, leveraging the of wavelet transforms. analyze time-domain signals, limiting their ability differentiate fault transients. Wavelet transforms, offering a combined time-frequency analysis, provide deeper understanding these A detailed line model is built SIMULINK. Diverse scenarios are meticulously simulated, and current signals undergo transform analysis. Key features extracted from coefficients act as fingerprints potential faults. These then utilized develop robust detection algorithm specifically designed The proposed method promises enhanced capabilities compared existing techniques this domain. results, presented subsequent sections, will shed light on effectiveness transforms empowering smarter more reliable operations.

Language: Английский

---