Sensor placement and parameter identi?ability in grey-box models of building thermal behaviour

Ole Magnus Brastein
Department of Electrical Engineering, Information Technology and Cybernetics, University of South-Eastern Norway, Porsgrunn, Norway

Roshan Sharma
Department of Electrical Engineering, Information Technology and Cybernetics, University of South-Eastern Norway, Porsgrunn, Norway

Nils-Olav Skeie
Department of Electrical Engineering, Information Technology and Cybernetics, University of South-Eastern Norway, Porsgrunn, Norway

Ladda ner artikelhttps://doi.org/10.3384/ecp2017051

Ingår i: Proceedings of The 60th SIMS Conference on Simulation and Modelling SIMS 2019, August 12-16, Västerås, Sweden

Linköping Electronic Conference Proceedings 170:8, s. 51-58

Visa mer +

Publicerad: 2020-01-24

ISBN: 978-91-7929-897-5

ISSN: 1650-3686 (tryckt), 1650-3740 (online)


Building energy management systems can reduce energy consumption for space heating in existing buildings, by utilising Model Predictive Control. In such applications, good models of building thermal behaviour is important. A popular method for creating such models is creating Thermal networks, based cognitively on naive physical information about the building thermal behaviour. Such models have lumped parameters which must be calibrated from measured temperatures and weather conditions. Since the parameters are calibrated, it is important to study the identi?ability of the parameters, prior to analysing them as physical constants derived from the building structure. By utilising a statistically founded parameter estimation method based on maximizing the likelihood function, identi?ability analysis can be performed using the Pro?le Likelihood method. In this paper, the effect of different sensor locations with respect to the buildings physical properties is studied by utilising likelihood pro?les for identi?ability analysis. The extended 2D pro?le likelihood method is used to compute two-dimensional pro?les which allows diagnosing parameter inter-dependence, in addition to analyzing the identi?ability. The 2D pro?les are compared with con?dence regions computed based on the Hessian.


building energy management systems, thermal behavior, parameter estimation, parameter identifiability, pro?le likelihood


Peder Bacher and Henrik Madsen. Identifying suitable models for the heat dynamics of buildings. Energy and Buildings, 43(7):1511 – 1522, 2011. ISSN 0378-7788. doi:https://doi.org/10.1016/j.enbuild.2011.02.005.

Thomas Berthou, Pascal Stabat, Raphael Salvazet, and Dominique Marchio. Development and validation of a gray box model to predict thermal behavior of occupied of?ce buildings. Energy and Buildings, 74:91–100, 2014.

Ole Magnus Brastein, Bernt Lie, Roshan Sharma, and Nils-Olav Skeie. Parameter estimation for externally simulated thermal network models. Energy and Buildings, 2019. ISSN 0378-7788. doi:10.1016/j.enbuild.2019.03.018.

An-Heleen Deconinck and Staf Roels. Is stochastic grey-box modelling suited for physical properties estimation of building components from on-site measurements? Journal of Building Physics, 40(5):444–471, 2017.

Cristina Sarmiento Ferrero, Qian Chai, Marta Dueñas Díez, Sverre H Amrani, and Bernt Lie. Systematic analysis of parameter identi?ability for improved ?tting of a biological wastewater model to experimental data. Modeling, Identi?cation and Control, 27(4):219, 2006.

Samuel F Fux, Araz Ashouri, Michael J Benz, and Lino Guzzella. EKF based self-adaptive thermal model for a passive house. Energy and Buildings, 68:811–817, 2014.

R.A. Johnson and D.W. Wichern. Applied Multivariate Statistical Analysis. Applied Multivariate Statistical Analysis. Pearson Prentice Hall, 2007. ISBN 9780131877153.

M Killian and M Kozek. Ten questions concerning model predictive control for energy ef?cient buildings. Building and Environment, 105:403–412, 2016.

Niels Rode Kristensen, Henrik Madsen, and Sten Bay Jørgensen. Parameter estimation in stochastic grey-box models. Automatica, 40(2):225–237, 2004.

Henrik Madsen. Time series analysis. Chapman and Hall/CRC, 2007.

Henrik Madsen and Jan Holst. Estimation of continuous-time models for the heat dynamics of a building. Energy and buildings, 22(1):67–79, 1995.

D.W.U. Perera, Carlos F. Pfeiffer, and Nils-Olav Skeie. Modelling the heat dynamics of a residential building unit: Application to Norwegian buildings. Modeling, Identi?cation and Control, 35(1):43–57, 2014. doi:10.4173/mic.2014.1.4.

Michael JD Powell. A direct search optimization method that models the objective and constraint functions by linear interpolation. In Advances in optimization and numerical analysis, pages 51–67. Springer, 1994.

Michael JD Powell. The BOBYQA algorithm for bound constrained optimization without derivatives. Cambridge NA Report NA2009/06, University of Cambridge, Cambridge, pages 26–46, 2009.

William H Press, Saul A Teukolsky, William T Vetterling, and Brian P Flannery. Numerical recipes in C++, volume 2. Cambridge University Press, 1992.

Andreas Raue, Clemens Kreutz, Thomas Maiwald, Julie Bachmann, Marcel Schilling, Ursula Klingmüller, and Jens Timmer. Structural and practical identi?ability analysis of partially observed dynamical models by exploiting the pro?le likelihood. Bioinformatics, 25(15):1923–1929, 2009.

Glenn Reynders, Jan Diriken, and Dirk Saelens. Quality of greybox models and identi?ed parameters as function of the accuracy of input and observation signals. Energy and Buildings, 82:263–274, 2014.

C. Runge. Ueber die numerische Au?ösung von Differentialgleichungen. Mathematische Annalen, 46(2):167–178, Jun 1895. ISSN 1432-1807. doi:10.1007/BF01446807.

D. J. Venzon and S. H. Moolgavkar. A method for computing pro?le-likelihood-based con?dence intervals. Journal of the Royal Statistical Society: Series C (Applied Statistics), 37(1):87–94, 1988. doi:10.2307/2347496.

Citeringar i Crossref