26 Fault Diagnosis of Helical Gear Box using Logistic Function and REP Tree 1 Abhinav Aggarwal, 2 V. Sugumaran, 3 M. Amarnath, 4 Hemantha Kumar 1 SMBS, VIT University, Chennai Campus, Vandalur-Kelambakam Road, Chennai 2 SMBS, VIT University, Chennai Campus, Vandalur-Kelambakam Road, Chennai 3 Indian Institute of Information Technology, Design & Manufacturing (IIITDM)-Jabalpur, Jabalpur, Madya Pradesh, India 4 National Institute of Technology-Warangal, Warangal, Andhra Pradesh, India ABSTRACT Most of the mechanical systems use gears as means to transfer power from one shaft to another. Due to continuous operation and heavy load, they tend to develop some faults, which if not treated properly create a permanent damage to the structure. This eventually affects the efficiency of the system, which creates an alarming situation and thus need to be looked into. Various research tasks have already been taken up to study these faulty conditions and their effects. In this study, the vibrations created during the meshing of teeth of gears are used for acquiring the data set. Machine learning algorithms are then applied to this data set to achieve the information which then can be interpreted into a form which people can easily understand. Here, logistic function classifier and REP tree are employed to obtain the output. The whole process was executed in a series of steps namely feature/data extraction, classification of data and then the results obtained are further discussed. Keywords: Fault Diagnosis, Gear Fault Diagnosis, Logistic Function, REP Decision Tree Statistical Analysis. 1. INTRODUCTION Being one of the most important components of the mechanical systems, helical gear box has received considerable attention over past many years. The failures that occur in the gears are caused mainly due to continuous change in gear speed and torque and presence of sizable metal or dust particles. The failure alarm gets active when such types of defects are detected in the gear box in the primary stage [1-3]. Certain parameters like vibration signals [4], sound signal, UV imaging, acoustic emission, wear debris, thermal imaging, oil particle analysis, image analysis, etc. have been considered in detection and diagnosis of faults. Either frequency domain analysis technique which include short time Fourier transform (STFT), wavelet analysis [5], spectral analysis etc. or time domain analysis techniques [6] like time signal averaging and using the statistical analysis of the time domain signal of sound or vibrational signals can be used [7]. Several approaches can be used to process the signals to give an output for a fault diagnosis system. Some of these approaches can be temporal data, model-based reasoning, heuristic reasoning, optimal disturbance de-coupling, machine learning etc. Machine learning is becoming more and more popular as the computational resources needed are relatively easily available and are reliable. In the present study the time domain vibration signal [8] is focused on for the fault diagnosis of a helical gear box. Here, the machine learning approach is used. Various classification algorithms such as logistic function and REP decision tree algorithms are used for the classification. The classification accuracies are presented and compared. International Journal of Research in Mechanical Engineering Volume-2, Issue-1, January-February, 2014, pp. 26-32, © IASTER 2014 www.iaster.com, ISSN Online:2347-5188 Print: 2347-8772