• Energy Research

Optimizing and predicting the energy consumption of industrial manufacturing can increase its cost efficiency. The interaction of different aspects and components is necessary. An overarching framework is currently still missing, and establishing such is the central research approach in this paper. This paper provides an overview of the current demands on the manufacturing industry from the perspective of digitalization and sustainability. On the basis of the developed fundamentals and parameters, a superordinate framework is proposed that allows the modelling and simulation of energy-specific properties on several product and process levels. A detailed description of the individual methods concludes this work and demonstrates their application potential in an industrial context. As a result, this integrated conceptual framework offers the possibility of optimizing the production system, in relation to different energy flexibility criteria.

Le suivi de la durabilité énergétique dans l'industrie manufacturière se heurte à un problème d'échelle. La surveillance d'indicateurs de performance est essentielle, en utilisant cependant le moins de capteurs possible et en limitant leur intrusivité vis-à-vis des systèmes existants. Les capteurs non-intrusifs sont particulièrement adaptés à de telles applications, en cela qu'ils captent de nombreuses sources depuis un lieu distant de celles-ci. La reconstitution des indicateurs-cibles nécessite toutefois davantage de traitement du signal. Les méthodes présentées dans cet article visent avant tout à reconstruire, de manière non-supervisée, le processus de production d'une machine à partir de données issues de capteurs. Une série de mesures est ainsi séquencée temporellement en opérations distinctes. Leur contenu en termes d'actionneurs actifs est ensuite estimé par décomposition. Ces méthodes sont toutes particulièrement adaptées aux signaux apparaissant comme stationnaires par morceaux dans la représentation temps-fréquence. Energy sustainability in the manufacturing industry faces a scalability issue. Monitoring appropriate performance indicators is essential, yet as few sensors as possible should be used, and with limited intrusiveness (software- or hardware-wise). Non-intrusive sensors are well suited to such applications, as multiple sources can be sensed at once. Recovering the desired indicators requires additional signal processing though. This paper focuses on recovering a machine’s process from sensor data in an unsupervised fashion, and unveiling which actuators are active within each operation. The proposed method is particularly well suited to mixed signals which appear as stationary in the time-frequency domainwithin each operation.

With the ever increasing complexity of Industry 4.0 systems, plant energy management systems developed to improve energy sustainability become equally complex. Based on a Model-Based Systems Engineering analysis, this paper aims to provide a general approach to perform holistic development of an autonomous energy management system for manufacturing industries. This Energy Management System (EMS) will be capable of continuously improving its ability to assess, predict, and act, in order to improve by monitoring and controlling the energy sustainability of manufacturing systems. The approach was implemented with the System Modeling Language (SysML).