Abstract
Applications of hydraulic systems are found today in a wide variety of devices, mostly in industrial and mobile machines. When extending the life and ensuring the correct operation of these machines is critical, analytical tools that provide more accurate information about the functioning and operation of these systems must be integrated to make proactive decisions. In industrial and mobile applications, there are many sensors and methods for measuring and determining the state of process variables (e.g., flow, pressure, force). However, little has been done to implement a system that can provide users with equipment status information related to on-machine hydraulics status. Implementing artificial intelligence (AI) technology and machine learning (ML) models in hydraulic system components is presented here as a solution to the challenges many industries face today, optimizing processes and making them safer and more efficient. This research paper presents the implementation of a solution for characterizing and estimating anomalies in hydraulic cylinders, one of the most versatile and widely used components in fluid-powered systems. This work describes AI and ML models implemented to determine the operating state of cylinders and whether they function normally, in specific failure modes, or in abnormal conditions that can be predicted before a catastrophic failure occurs. The models applied demonstrated over 95% level of accuracy in predicting a malfunction of the component studied and presented in this work.
