ADVANCED MONITORING IN WATER COOLING SYSTEMS USING LSTM NETWORKS AND EXTERNAL FACTOR INTEGRATION

This paper presents an advanced approach to monitoring and predicting the performance of water cooling systems in industrial settings. Traditional methods often relied on fixed setpoints and overlooked the impact of external factors like wind speed, treating these variations as errors. By leveraging Long Short-Term Memory (LSTM) networks, this study integrates these external influences into the predictive model, transforming what was previously seen as noise into actionable data. The result is a more accurate prediction of potential system failures and an improved understanding of how various factors affect the overall reliability of water cooling systems. Keywords: water cooling systems, LSTM networks, predictive maintenance, external factors integration, industrial monitoring.