AS_03_2020

tecnica 94 Aprile 2020 Automazione e Strumentazione CONTROLLO reduced by 17%, the overheating average temperature reduced by 0.1 °C and the overall thermal energy reduced of 3304 kWh over three months. 4.3 Air recirculation feed-forward The last control strategy presented envisages the manipulation of the air recirculation command through the people occupancy measurement by predicting the consequent CO2 level incre- ase. An additive command is sent to the room air recirculation system to compensate such an increase beforehand as presented in the following equation: In such an equation, ur PID (t) is the CO2 PID output, #PPL is the measured people occupancy and PPL MAX is a tuning parameter. This control strategy has been tested over the multi-room model to validate its performances with respect to the baseline PID control strategy, for which a comparison over a time window of one week is presented in υ figure 4 . The rooms average CO2 has been reduced of 2.1 PPMx100, the high CO2 frequency has been reduced of 8% and the high CO2 average has been reduced of 0.7 PPMx100. However, the overall consumption over three months was increased of 688 kWh. This effect is due to the higher exploitation of the air handling units. The study regar- ding this control strategy can be further deepened so to ensure better air quality performance without compromising the ove- rall energy consumption. 5. Conclusions The presented case study envisages the development of dyna- mic models of the building behaviour tuned and validated over two sets of real data. These models have been developed in a gray-box fashion, so to investigate the impact of any real buil- ding parameter on the overall performances, both in terms of user comfort and energy consumption. Three easy to implement advanced control strategies have been developed and tested by exploiting a new measurement allowed by the installed IoT multi-sensor network: the people occupancy profile. The developed advanced control strategies have proven an appreciable performance increase in both the user comfort requirements and the overall energy consumption . A future work idea envisages the exploitation of the developed grey-box models to study the impact of an investment on the building structure and thermal devices. Moreover, the proposed control techniques will be integrated with the thermal genera- tion side (heat pumps and air handling units). 6. References [1] L. G. Swan, V. I. Ugursal, “Modeling of end-use energy consumption in the residential sector: a review of modeling techniques”, Renewable and sustainable energy reviews , vol. 13, n. 8 (2009), pp. 1819-1835. [2] S. Rastegarpour, L. Ferrarini, “A distributed predictive con- trol of energy resources in radiant floor buildings”, Journal of Dynamic Systems, Measurement and Control , vol. 62, n. 4 (2019). [3] G. Mantovani, L. Ferrarini, “Temperature Control of a Commercial Building with Model Predictive Control techni- ques”, IEEE Transactions on Industrial Electronics , vol. 62, n. 4, (2014), pp. 2651-2660. [4] eSmartCity European research project https://esmartcity. interreg-med.eu/ [5] A. P. Kalogeras, H. Rivano, L. Ferrarini, C. Alexakos, O. Iova, S. Rastegarpour, A. A. Mbackez, “Cyber Physi- cal Systems and Internet of Things: Emerging Paradigms on Smart Cities”, 1st International Conference on Societal Automation - SA2019 , September 4-6, 2019, Krakow, Poland. [6] A. Afram, F. Janabi- Sharifi, “Black-boxmodel- ling of residential HVAC system and comparison of gray-box and black- box modelling methods”, Energy Build. , 94 (2015), pp. 121-149. [7] C. Koulamas, A.P. Kalogeras, R. Pacheco- Torres, J. Casillas, L. Ferrarini, “Suitability analysis of modeling and assessment approaches in energy efficiency in buildings”, Energy and Buildings , 158 (2018), pp. 1662-1682. Figure 4 - Advanced control strategy 3 performance