Wyniki wyszukiwania - BazTech

1

Temperature Prediction and Performance Comparison of Permanent Magnet Synchronous Motors Using Different Machine Learning Techniques for Early Failure Detection

Aslan Emrah

Eksploatacja i Niezawodność

|

2025

|

Vol. 27, no. 1

art. no. 192164

EN

Electric motors are increasingly used in various products, including turbines and electric vehicles. Precise temperature measurement is essential for the safe operation of a Permanent Magnet Synchronous Motor. Direct temperature detection of the permanent magnet and stator involves significant costs and hardware requirements. To overcome these challenges, Machine Learning models can eliminate the need for specialized sensors. This study used four diverse regression algorithms: Linear, K-Nearest Neighbor, XGBoost, and AdaBoost. The objective of this study is to model a Permanent Magnet Synchronous Motor used in electric vehicles and predict the temperatures of some of its parameters. The K-Nearest Neighbor Regressor outperformed the other algorithms, achieving a training accuracy of 99.65%, test accuracy of 98.72%, root-mean-square error of 2.16, R2 score of 98.72, and Cross-Validation R2 of 97.77%. These results enable low-cost, real-time temperature monitoring of electrical machinery, enhancing power density, safety, and efficiency.

2

Comparison and optimization of machine learning methods for fault detection in district heating and cooling systems

Çinar Mehmet, Aslan Emrah, Özüpak Yıldırım

Bulletin of the Polish Academy of Sciences. Technical Sciences

|

2025

|

Vol. 73, nr 3

art. no. e154063

EN

In this study, the methods used for the detection of sub-station pollution failures in district heating and cooling (DHC) systems are analyzed. In the study, high, medium, and low-level pollution situations are considered and machine learning methods are applied for the detection of these failures. Random forest, decision tree, logistic regression, and CatBoost regression algorithms are compared within the scope of the analysis. The models are trained to perform fault detection at different pollution levels. To improve the model performance, hyperparameter optimization was performed with random search optimization, and the most appropriate values were selected. The results show that the CatBoost regression algorithm provides the highest accuracy and overall performance compared to other methods. The CatBoost model stood out with an accuracy of 0.9832 and a superior performance. These findings reveal that CatBoost-based approaches provide an effective solution in situations requiring high accuracy, such as contamination detection in DHC systems. The study makes an important contribution as a reliable fault detection solution in industrial applications.

3

Prediction and comparative analysis of emissions from gas turbines using random search optimization and different machine learning-based algorithms

Aslan Emrah

Bulletin of the Polish Academy of Sciences. Technical Sciences

|

2024

|

Vol. 72, nr 6

art. no. e151956

EN

Gas turbines are widely used for power generation globally, and their greenhouse gas emissions have increasingly drawn public attention. Compliance with environmental regulations necessitates sophisticated emission measurement techniques and tools. Traditional sensors used for monitoring emission gases can provide inaccurate data due to malfunction or miscalibration. Accurate estimation of gas turbine emissions, such as particulate matter, carbon monoxide, and nitrogen oxides, is crucial for assessing the environmental impact of industrial activities and power generation. This study used five different machine learning models to predict emissions from gas turbines, including AdaBoost, XGBoost, k-nearest neighbour, and linear and random forest models. Random search optimization was used to set the regression parameters. The findings indicate that the AdaBoost regressor model provides superior prediction accuracy for emissions compared to other models, with an accuracy of 99.97% and a mean squared error of 2.17 on training data. This research offers a practical modelling approach for forecasting gas turbine emissions, contributing to the reduction of air pollution in industrial applications.