Refine
Year of publication
- 2022 (3) (remove)
Document Type
- Article (2)
- Conference Proceeding (1)
Language
- English (3)
Has Fulltext
- yes (3)
Is part of the Bibliography
- no (3)
Keywords
- Gießerei (2)
- Optimale Kontrolle (2)
- optimal scheduling (2)
- Ablaufplanung (1)
- Building automation (1)
- Energiemanagement (1)
- Energy management (1)
- LSTM (1)
- MILP (1)
- Model Predictive Control (1)
Institute
We present a novel long short-term memory (LSTM) approach for time-series prediction of the sand demand which arises from preparing the sand moulds for the iron casting process of a foundry. With our approach, we contribute to qualify LSTM and its combination with feedback-corrected optimal scheduling for industrial processes.
The sand is produced in an energy intensive mixing process which is controlled by optimal scheduling. The optimal scheduling is solved for a fixed prediction horizon. One major influencing factor is the sand demand, which is highly disturbed, for example due to production interruptions. The causes of production interruptions are in general physically unknown. We assume that information about the future behavior of the sand demand is included in current and past process data. Therefore, we choose LSTM networks for predicting the time-series of the sand demand.
The sand demand prediction is performed by our multi model approach. This approach outperforms the currently used naive estimation, even when predicting far into the future. Our LSTM based prediction approach can forecast the sand demand with a conformity up to 38 % and a mean value accuracy of approximately 99%. Simulating the optimal scheduling with sand demand prediction leads to an improvement in energy savings of approximately 1.1% compared to the naive estimation. The application of our novel approach at the real production plant of a foundry proves the simulation results and verifies the capability of our approach.
This paper proposes an extended Petri net formalism as a suitable language for composing optimal scheduling problems of industrial production processes with real and binary decision variables. The proposed approach is modular and scalable, as the overall process dynamics and constraints can be collected by parsing of all atomic elements of the net graph. To conclude, we demonstrate the use of this framework for modeling the moulding sand preparation process of a real foundry plant.
We present a methodology based on mixed-integer nonlinear model predictive control for a real-time building energy management system in application to a single-family house with a combined heat and power (CHP) unit. The developed strategy successfully deals with the switching behavior of the system components as well as minimum admissible operating time constraints by use of a special switch-cost-aware rounding procedure. The quality of the presented solution is evaluated in comparison to the globally optimal dynamic programming method and conventional rule-based control strategy. Based on a real-world scenario, we show that our approach is more than real-time capable while maintaining high correspondence with the globally optimal solution. We achieve an average optimality gap of 2.5% compared to 20% for a conventional control approach, and are faster and more scalable than a dynamic programming approach.