EARLY ANOMALY PREDICTION OF MACHINE DAMAGE USING C4.5 ALGORITHM BASED ON IOT

Abstract

Machine condition is a problem that is difficult to predict, prediction of machine condition is an important aspect in the application of maintenance because the occurrence of damage can result in a decrease in the productivity of a company. Measurement of vibration, temperature, and machine displacement is a fairly good method to determine the condition of the machine because it is an indicator of mechanical conditions and an early indicator of damage to the machine as a whole, Application of Algorithms in processing vibration, temperature and machine displacement data to improve the prognosis of damage. In this project, predictions of machine condition will be carried out using the C4.5 algorithm. Data taken using sensors at a certain time will be used to predict the decline in the performance of a machine. This data will be used for training and testing data. This project was concluded that the C45 algorithm obtained accuracy with the difference between Training data and Test data, 60.8% for Training data and 76.4% for Testing data. Proving that the C45 algorithm is effective for predicting the initial anomaly of damage to the machine. It is necessary to re-calibrate the sensor limits, and replace the sensors used because this project uses sensors for prototypes.ds yang berhubungan dengan artikel ini. Tuliskan paling tidak tiga buah keywords.