Energy monitoring is one major prerequisite for energy efficiency measures. Energy and power data throughout different levels of production allow benchmarking and condition monitoring applications based on insightful energy performance indicators. However, fine-grained measurement concepts for energy and power require high investments with uncertain benefits. This paper presents a low-cost approach to monitor the component-by-component energy consumption with a minimum of sensor technology that can be applied to a variety of production machines. Aggregated energy data combined with components' control signals are the basis for the determination of components' energy consumptions using two system identification algorithms. While one method is realized in an offline-mode after data collection, the second approach utilizes real-time data based on a recursive least squares algorithm. Eventually, the feasibility of the theoretical system identification concepts is shown in a laboratory environment.
Panten, N., Abele, E., & Schweig, S. (2016). A Power Disaggregation Approach for Fine-grained Machine Energy Monitoring by System Identification. In Procedia CIRP (Vol. 48, pp. 325–330). Elsevier B.V. https://doi.org/10.1016/j.procir.2016.03.025