Abstract
This paper proposes an approach to develop building dynamic thermal models that are of paramount importance for controller application. In this context, controller requires a low-order, computationally efficient, and accurate models to achieve higher performance. An efficient building model is developed by having proper structural knowledge of low-order model and identifying its parameter values. Simplified low-order systems can be developed using thermal network models using thermal resistances and capacitances. In order to determine the low-order model parameter values, a specific approach is proposed using a stochastic particle swarm optimization. This method provides a significant approximation of the parameters when compared to the reference model whilst allowing low-order model to achieve 40% to 50% computational efficiency than the reference one. Additionally, extensive simulations are carried to evaluate the proposed simplified model with solar radiation and identified model parameters. The developed simplified model is afterward validated with real data from a case study building where the achieved results clearly show a high degree of accuracy compared to the actual data.
Highlights
In the current situation, buildings in EU countries consume more primary energy than other sectors such as: transportation and industry [1]
Considering all the above hypothesis, a thermal model is developed with 6 sets of 3R2C networks and the equations obtained by energy balance at each node from the schematic shown in Figure 9: Cz
The results show that the simplified thermal network model with parameter identification is well suited approach to model the building thermal dynamics, and the models can be used for the controller purpose
Summary
Buildings in EU countries consume more primary energy than other sectors such as: transportation and industry [1]. Fraisse et al [19], proposed a technique to evaluate resistors and capacitors values analytically, analysed the performance of various RC combinations (2R1C, 1R2C, 3R2C, and 3R4C) These parameter values can be used to develop a reduced order model to realize thermal behavior of buildings while considering radiative and convective loads from the internal and external sources. The calibrated model with the optimized parameter values has actual dynamics of the building envelope, and 40% to 50% computationally faster than the reference model Using this approach a simplified model developed for a classroom in case study building (container building) and predicted indoor temperature of the model is validated against the real data obtained from the container building
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have