A Multi-Strategy Optimized Framework for Health Status Assessment of Air Compressors DOI Creative Commons
Dali Hou,

Xiaoran Wang

Machines, Год журнала: 2025, Номер 13(3), С. 248 - 248

Опубликована: Март 20, 2025

Air compressors play a crucial role in industrial production, and accurately assessing their health status is vital for ensuring stable operation. The field of assessment has made significant progress; however, challenges such as dataset class imbalance, feature selection, accuracy improvement remain require further refinement. To address these issues, this paper proposes novel algorithm based on multi-strategy optimization, using air the research subject. During data preprocessing, Synthetic Minority Over-sampling Technique (SMOTE) introduced to effectively balance distribution. By integrating Squeeze-and-Excitation (SE) mechanism with Convolutional Neural Networks (CNNs), key features within are extracted emphasized, reducing impact irrelevant model efficiency. Finally, Bidirectional Long Short-Term Memory (BiLSTM) networks employed classification compressor. Ivy (IVYA) optimize BiLSTM’s hyperparameters improve avoid local optima. Through comparative ablation experiments, effectiveness proposed SMOTE-IVY-SE-CNN-BiLSTM validated, demonstrating its ability significantly enhance compressor assessment.

Язык: Английский

A Machine Learning Implementation to Predictive Maintenance and Monitoring of Industrial Compressors DOI Creative Commons
Ahmad Aminzadeh, Sasan Sattarpanah Karganroudi, Soheil Majidi

и другие.

Sensors, Год журнала: 2025, Номер 25(4), С. 1006 - 1006

Опубликована: Фев. 8, 2025

Integrating machine learning algorithms leveraged by advanced data acquisition systems is emerging as a pivotal approach in predictive maintenance. This paper presents the deployment of such an integration on industrial air compressor unit. research combines updated concepts from Internet Things, learning, multi-sensor collection, structured mining, and cloud-based analysis. To this end, temperature, pressure, flow rate were acquired sensors contact with compressor. The observed sent to Structured Query Language database. Then, Linear Regression model was fitted training data, optimized stored for real-time inference. Afterward, passed through model, if exceeded determined threshold, warning email operator. Adopting Things enhances surveillance specialists, decreasing failure damage probabilities. achieved 98% accuracy Mean Squared Error metric our regression model. By analyzing gathered implemented system demonstrates capabilities predict potential equipment failures promising accuracy, facilitating shift reactive proactive maintenance strategies. findings reveal substantial improvements efficiency, uptime, cost savings.

Язык: Английский

Процитировано

1
A Multi-Strategy Optimized Framework for Health Status Assessment of Air Compressors DOI Creative Commons
Dali Hou,

Xiaoran Wang

Machines, Год журнала: 2025, Номер 13(3), С. 248 - 248

Опубликована: Март 20, 2025

Air compressors play a crucial role in industrial production, and accurately assessing their health status is vital for ensuring stable operation. The field of assessment has made significant progress; however, challenges such as dataset class imbalance, feature selection, accuracy improvement remain require further refinement. To address these issues, this paper proposes novel algorithm based on multi-strategy optimization, using air the research subject. During data preprocessing, Synthetic Minority Over-sampling Technique (SMOTE) introduced to effectively balance distribution. By integrating Squeeze-and-Excitation (SE) mechanism with Convolutional Neural Networks (CNNs), key features within are extracted emphasized, reducing impact irrelevant model efficiency. Finally, Bidirectional Long Short-Term Memory (BiLSTM) networks employed classification compressor. Ivy (IVYA) optimize BiLSTM’s hyperparameters improve avoid local optima. Through comparative ablation experiments, effectiveness proposed SMOTE-IVY-SE-CNN-BiLSTM validated, demonstrating its ability significantly enhance compressor assessment.

Язык: Английский

Процитировано

0