Optimization of Design Parameters in LSTM Model for Predictive Maintenance

Predictive maintenance conducts actions according to the prognostic state of machinery, which can be demonstrated by a model. Due this characteristic, choosing proper model for describing machinery is important. Among various model-based approaches, we address an artificial intelligence (AI) approach uses AI models obtained from collected data. Specifically, optimize design parameters predictive based on long short-term memory (LSTM). To define effective and efficient health indicator, suggest method feature reduction correlation analysis stepwise comparison features. Then, hyperparameters determining structure LSTM are optimized using genetic algorithm. Through numerical experiments, performance suggested validated.