Stéphane Thil

Black-box modeling of buildings thermal behavior using system identification

Abstract In this paper the modeling of buildings thermal behavior is studied. The main goal is to develop a modeling procedure that can be used at different scales (a thermal zone, a floor or a whole building) and on different buildings. The scalability of the chosen black-box model structure is first assessed; simulation experiments are then conducted in order to test if the modeling procedure is reusable. As these tests are hardly feasible in practice, a real university building is first modeled using an energy simulation software. This model is then used to validate the proposed approach.

Modelling of a multi-zone building and assessment of its thermal behaviour using an energy simulation software

This paper deals with the modelling of a multizone university building in the town of Perpignan, in southern France. An energy simulation software (EnergyPlus) is used to create a model of the buildings thermal behaviour. The building has been instrumented and the comparison between simulated and measured temperatures validates the obtained model.

Intra-Day DNI Forecasting Under Clear Sky Conditions Using ANFIS

Abstract The present paper deals with the developement of a new intra-day Direct Normal Irradiance (DNI) forecasting methodology, under clear sky conditions. Indeed, one challenge of the CSPIMP (Concentrated Solar Power plant efficiency IMProvement) research project is to forecast the suns resource accurately and design efficient plant control approaches. First, a quick review of the different formulations available for the atmospheric turbidity coefficient (from which DNI can be calculated) is performed. The data selection and filtering process is then described. Finally, the new forecasting approaches are compared to persistent and autoregressive models. The most efficient model presented here is based on side-by-side Adaptive Network-based Fuzzy Inference Systems (ANFIS). The selected configuration achieves very good results and validates the proposed forecasting methodology.

Towards the intrahour forecasting of direct normal irradiance using sky-imaging data

Increasing power plant efficiency through improved operation is key in the development of Concentrating Solar Power (CSP) technologies. To this end, one of the most challenging topics remains accurately forecasting the solar resource at a short-term horizon. Indeed, in CSP plants, production is directly impacted by both the availability and variability of the solar resource and, more specifically, by Direct Normal Irradiance (DNI). The present paper deals with a new approach to the intrahour forecasting (the forecast horizon Δtf is up to 30min ahead) of DNI, taking advantage of the fact that this quantity can be split into two terms, i.e. clear-sky DNI and the clear sky index. Clear-sky DNI is forecasted from DNI measurements, using an empirical model (Ineichen and Perez, 2002) combined with a persistence of atmospheric turbidity. Moreover, in the framework of the CSPIMP (Concentrating Solar Power plant efficiency IMProvement) research project, PROMES-CNRS has developed a sky imager able to provide High Dynamic Range (HDR) images. So, regarding the clear-sky index, it is forecasted from sky-imaging data, using an Adaptive Network-based Fuzzy Inference System (ANFIS). A hybrid algorithm that takes inspiration from the classification algorithm proposed by Ghonima et al. (2012) when clear-sky anisotropy is known and from the hybrid thresholding algorithm proposed by Li et al. (2011) in the opposite case has been developed to the detection of clouds. Performance is evaluated via a comparative study in which persistence models – either a persistence of DNI or a persistence of the clear-sky index – are included. Preliminary results highlight that the proposed approach has the potential to outperform these models (both persistence models achieve similar performance) in terms of forecasting accuracy: over the test data used, RMSE (the Root Mean Square Error) is reduced of about 20Wm−2, with Δtf=15min, and 40Wm−2, with Δtf=30min.

A Bayesian approach to closed-loop system identification

The challenging issue of identifying a closed-loop system from short and/or non-informative data records is addressed. A bayesian approach is developed within this framework. It is shown that accurate estimates and realistic confidence intervals can be obtained by taking into account prior knowledge on the system. The performances of the proposed method are illustrated with a simulation example.

Towards a generic procedure for modeling buildings and their thermal zones

In this paper, a procedure to model buildings and their thermal zones is studied. Since, ideally, the developed methodology should be applicable to different buildings, at different scales and in different locations, a feasibility study is necessary. The experiments used to test the flexibility of the model structure and the adaptability of the modeling procedure itself are presented. The obtained results show that the chosen model class (linear state-space models) is complex enough to model thermal zones at different scales. Finally, the adaptability of the modeling procedure is been tested, by applying it to different buildings with various Heating, Ventilation and Air-Conditioning (HVAC) systems and in different meteorological conditions. Again, the obtained results are satisfying.