Akademik Veri Yönetim Sistemi
Journal of the Faculty of Engineering and Architecture of Gazi University, cilt.40, sa.2, ss.937-950, 2025 (SCI-Expanded)
In the developing world where energy consumption is dominant, energy estimation, efficiency and cost effectiveness have gained importance with artificial intelligence solutions. The energy consumption in businesses is a significant part of this consumption. The importance of this study was revealed by estimating heat transfer values for an automobile factory with the Internet of Things (IoT) and machine learning. The study was applied to a real industrial system with a combination of IoT and machine learning models. The integrated use of two different Industry 4.0 topics working on this system is the original aspect of the study. Machine learning-based regression models were developed with an expanded data set by generating new variables from the existing temperature data for the prediction of the selected pilot plant temperature in the factory. Temperature regulation was provided with the developed model, and costs were reduced by preventing negative factors such as heat losses of the plant, changes in external environmental conditions, overheating or cooling of the environment, and losses of heat during transfer. The heat to be sent to the pilot plant was predicted using Linear, Random Forest, Polynomial, Decision Tree, Support Vector, Extra Tree, Adaboost, Gradient Boosting, Voting Regression models and Artificial Neural Network algorithms. Of these algorithms, the Linear Regression model has the highest prediction value. Finally, the study was integrated into the business's live monitoring system, SCADA (Supervisory Control and Data Acquisition), and tested in real time. In the results of the study, it was observed that with the proposed approach, overheating costs caused by the heating system decreased by 90% yearly in the pilot plant and employee satisfaction increased.