Refine
Document Type
- Article (2)
Language
- English (2)
Has Fulltext
- yes (2)
Is part of the Bibliography
- no (2)
Keywords
- Energieversorgung (1)
- Gebäude (1)
- Steuerung (1)
- batch control (1)
- feedforward control (1)
- industrial production system (1)
- intelligent control (1)
- neural control (1)
- neural network model (1)
- nonlinear model predictive control (1)
- prediction methods (1)
- target control (1)
Institute
Mixed-integer NMPC for real-time supervisory energy management control in residential buildings
(2023)
In recent years, building energy supply and distribution systems have become more complex, with an increasing number of energy generators, stores, flows, and possible combinations of operating modes. This poses challenges for supervisory control, especially when balancing the conflicting goals of maximizing comfort while minimizing costs and emissions to contribute to global climate protection objectives. Mixed-integer nonlinear model predictive control is a promising approach for intelligent real-time control that is able to properly address the specific characteristics and restrictions of building energy systems. We present a strategy that utilizes a decomposition approach, combining partial outer convexification with the Switch-Cost Aware Rounding procedure to handle switching behavior and operating time constraints of building components in real-time. The efficacy is demonstrated through practical applications in a single-family home with a combined heat and power unit and in a multi-family apartment complex with 18 residential units. Simulation studies show high correspondence to globally optimal solutions with significant cost savings potential of around 19%.
This paper presents a databased approach for improving the precision of the moulding sand compressibility in the moulding sand mixer of a foundry. In this approach, the deviation between the measured and the target compressibility is reduced by controlling the water addition. The complex dynamic behaviour of the process variables and their influence on the water addition is modelled with a long short-term memory (LSTM) network. Another LSTM network as control path simulates the impact of the water addition on the compressibility. Simulation and experimental results with the applied model for water prediction in a feedforward control yield relevant improvements of the moulding sand compressibility.