Predictive Maintenance using LSTM Network Model Optimized Time-Lag with Decision Tree

Predictive Maintenance integrates time series forecasting and data analysis techniques to predict equipment failures through continuous monitoring and data collection. This study investigates the enhancement of fault prediction using the Long Short-Term Memory (LSTM) deep learning technique combined with a decision tree algorithm for optimizing time-lag. Experimental data analysis compares predictive maintenance models with and without additional features, revealing a significant ۹۷.۰۵% improvement in accuracy when incorporating these additional features. The decision tree algorithm efficiently identifies a near-optimal lag in a short timeframe, highlighting the enhanced performance of LSTM model in optimizing fault prediction models.

کلیدواژه ها:

Predictive maintenance ، Time series forecasting ، Fault prediction ، Decision tree ، Optimal lag ، LSTM