Prognosis of bearing damage performance to industrial system using nonlinear autoregressive with exogenous (NARX)

Monitoring the engine´s performance carried out with the aim of the business process can run smoothly without any process stalled because if the business process stops only a few seconds, the industry will experience a loss. In this research, monitoring of industrial systems using an intelligent system that will do the prognosis of damage to the components contained in the industrial machine is the bearing performance. The method used is the nonlinear autoregressive with exogenous (NARX), based on the linear ARX model, which is commonly used in time-series modeling. This model uses a recurrent dynamic network, with feedback connections and enclosing several layers of tissues. The results of this research remaining useful life (RUL) in machine tools industry. It means the fatal damage in a machine can be avoidable. It is also part of maintenance on industrial machinery, in order to become more efficient in costing and replacement of equipment to be effective.