Regression methodology for sensitivity analysis of solar heating walls

Abstract

Passive solar heating of buildings continues to be a great interest of renewable energy applications. Part of this interest focuses on solar heating walls. A solar heating wall (SHW) is a part of building walls that receive, store, and transfer solar thermal energy into the building. SHW sensitivity analysis is often performed to guide optimum designs. For the sensitivity analysis, the methodology used so far is numerical simulations. This work presents a new approach, regression analysis, and develops a general regression model. To show how to develop a specific model for a given SHW from the general one, how to do the regression, and how to validate the model, the lattice solar heating wall (LSHW) is selected as a case study. Detail heat transfer analysis is performed to develop the specific regression model for the LSHW sensitivity analysis. Four side-by-side test cells are constructed to obtain experimental data. The data are then used to determine the regression constants and coefficients and the time series numbers. Validation of the regression analysis shows that the model has very high confidence. The model is also used for LSHW optimization, yielding the same results as those from the simulation with the numerical simulation program, which further demonstrates that the proposed model is reliable.