The Usage of Fuzzy Logic for Detecting Mechanical Faults in Gearboxes of Robotic System

Using machine learning or artificial intelligence enhances the quality of control and the degree of diagnostics of mechanical system states. However, in cases of insufficient data, these methods cannot fully realize their potential. In such instances, fuzzy logic algorithms can serve as a complete alternative to these methods. Fuzzy logic does not require the calculation of a precise mathematical model. In case of unexpected faults, some data can be lost, or additional noise appears, which can lead to destabilization of the control system. Working with fuzzy sets and linguistic variables allows the construction of a high-level control and diagnostic system. This article explores the possibility of using a fuzzy logic algorithm to create a diagnostic system for mechanical damage based on a Cartesian robot and underscores the importance of this method in addressing such issues. The article provides a description of the method and its working principles, along with conclusions and results of damage detection in the robot.

I. Introduction

The development of artificial intelligence [1] and machine learning leads [2], [3] to the emergence of new technologies, control systems, and diagnostic methods [4]. The development of these technologies has significantly impacted the diagnosis of mechanical faults across various industrial sectors [5].