Ensemble Learning Algorithm-Based Air Leakage Detection Technology in Pneumatic Machines

The issue of air leakage in machines has occurred for a long time due to imbalances in compressed air supply and instability of pneumatic machines. As a solution to this problem, real-time monitoring and control methods have been proposed to address the air leakage issue. However, such solutions result in monetary loss due to the need to stop equipment operation. In addition, conventional solutions focus on post-processing, which has the limitation of not being able to maintain the operational efficiency of the system. To solve the air leakage issue without decreasing machine efficiency, this paper proposes a method to detect air leakage several minutes in advance using various IoT sensor data collected from the machine. Using the proposed method, it is confirmed that it is possible to detect air leakage several minutes in advance with at least 90% accuracy.