Study on Performance Prediction Method for Axial Flow Fans Using Machine Learning

To improve cooling performance, higher efficiency of axial flow fans is required for electric motors. While the design method based on Euler's law can accommodate various fan shapes, predicting losses in axial flow fans remains challenging. This study explores a method to predict the performance characteristics of axial flow fans at the design stage by correlating past experimental and analytical data with design parameters using machine learning. As a result, favorable outcomes were obtained, which are reported in this paper.