Machine learning based analysis of factory energy load curves with focus on transition times for anomaly detection

摘要: Abstract An accurate understanding of energy load curves is the key for effective management factory systems and basis several applications (e.g. forecasts, anomaly detection). While curve analysis has been a research topic with practical significance in many areas, there lack methods particularly to evaluate different temporal transitions between states. Consequently, related saving potentials on level remain undetected. Against this background, paper presents methodology combining unsupervised univariate clustering multivariate prediction based methods. Within an automotive use case detection performance management, those are getting applied validated real data.