Remaining useful life estimation of mechanical systems by mixed method of mathematical method and evolutionary computational framework

An accurate prediction of the remaining useful life of the equipment is necessary for use, repairs and maintenance. Useful life prediction has been widely used, while the data obtained from it is not functional in different conditions. Many data-driven algorithms have been proposed and good results have been obtained in the field of predictive troubleshooting. Therefore, in this article, the relevant parameters are optimized using the meta-heuristic algorithm, so that the moving time window is used along with the mathematical model. Setting parameters related to data in the optimization framework allows the use of simple models such as neural networks with a small number of hidden layers and a small number of neurons in each layer, which can be used in environments with limited resources such as embedded systems. To evaluate the effectiveness of the proposed method, the root mean square error scoring index and useful life health score have been used. For this purpose, a random data set has been considered and the results show the acceptability of the method.