Fault diagnosis of self-aligning troughing rollers in belt conveyor system using k-star algorithm

Abstract A belt conveyor system is a mechanical bulk material handling equipment which conveys millions of tons of material from one end to another, in core mechanical industries. Self-Aligning Troughing Roller (SATR) is one of the critical components in a belt conveyor system, It is pretty important in running the belt conveyor in good condition. Self-aligning troughing roller has machine elements like ball bearing, central shaft and the outer shell. In belt conveyor system, certain faults such as ball bearing fault (BF), central shaft fault (CSF), combined shaft and ball bearing fault (SBF) occurs frequently. For further research in this area, a prototype experimental set up has been made with the above mentioned faults, and the vibration signals were acquired from the experimental set up. From the raw vibration signals acquired, a set of significant parameters was identified, called as features. All the identified features may not be used for classification of faults. Hence, the significant features used for classification were found out using decision tree method. These descriptive statistical features such as mean, median, sum etc were calculated and used for classification. In the successive stage k-star algorithm was used for fault diagnosis. The K-star algorithm achieved 91.7% fault classification accuracy, which acknowledged that the algorithm has lots of potential in the field of fault diagnosis applications. The results were compared with the benchmarked algorithm such as Artificial Neural Network (ANN).