Prediction of Robot Welding Damages with Markov Chain

Abstract

Mathematical modeling is widely used by researchers in interpreting various events into mathematical language. There are several methods included in mathematical modeling including neural networks, naive bayes, monte carlo and markov. From some of these methods used in predicting future events. For example, Markov chain is used to predict hospital users, rain and population movements. So in this research, the authors try to use Markov Chain method to predict the damage that robots welding experiences in the coming year. Since there is no description of the damage that will occur, a prediction is needed to get a picture of the damage. In this study, a calibration process is also presented to get an error value between the predicted results and their actual results. The error value of the calibration process is a reference in determining the error value of predicted results for the coming year.