Regression and Classification Model Based Predictive Maintenance of Aircrafts Using Neural Network

Abstract

One of the key objectives of today's  businesses and mills is to predict machine problems. Failures  must be avoided, because downtimes represent expensive  expenses and a loss of productivity. This is why the number  of remaining cycles (RULs) until the failure occurs is vital in  machine maintenance. The estimations of the RUL should be  based on earlier observations, whenever possible under the  same conditions. In the research of RUL estimates, the  creation of systems that monitor current equipment  conditions is becoming crucial. I employed Long Short Term  Memory (LSTM) in my project to determine an aircraft's  remaining usable lives. The aircraft's functioning condition is  also forecast. The former is done by a regression method,  using a classification methodology predicted by working  circumstances. In order to estimate operating conditions and  remaining usable life of the aircraft, data utilized for LSTM  models training are derived from 21 aircraft sensor readings  located at different locations with three distinct settings.