Methodology and model for predicting energy consumption in manufacturing at multiple scales

Certain fields of manufacturing, like casting, forming or cutting, may cause high energy load. Especially under the consideration of renewable energy sources it is beneficial to negotiate production schedules and consumption forecasts with the energy supplier. This would enable an optimized management of energy sources and infrastructure components on the supplier side, helping to reduce costs. Optimal and balanced expenses for production would be the consequence. The problem of power consumption prediction in manufacturing was subject of many studies in the past. Most of them either consider the physical modeling of processes at a very detailed level, or they introduce tailored prediction models for specific production processes. Thus, it is hard to apply their results to other uses cases in different scenarios. As a consequence, a generic methodology and model regarding power consumption prediction in manufacturing is required in order to cover the variety of processes, machines and materials. Furthermore, an approach must support flexible levels of granularity for predicting the energy consumption of manufacturing processes. On the one hand, a whole factory may be the object of investigation while, on the other hand, predictions for finer-grained levels, such as certain parts of a machine, are required to allow for specific optimizations. Our contribution is twofold. First, we propose a generic model for the specification of the power-consuming machine. A tree-based compositional approach supports arbitrary levels, depending on the structure of the machine, or external factors, such as company policies. This approach is highly extensible since the models are stored in ontologies. Second, we propose a methodology for static and dynamic modeling of power consumption for every structural level. Based on that model the prediction can be realized. In addition, we provide an example implementation and prediction for a continuous casting machine process. (C) 2018 The Authors. Published by Elsevier B.V.