Multiple linear regression based model for solar collectors

Simplified multiple linear regression based model for solar collectors

“…For the Reader's convenience, the physically-based collector model of Buzás et al [2], which will be called physically-based model in short, and the MLR model [3] are recalled in this section. The scheme of the studied solar collector can be seen in Figure 1.…”

“…Buzás et al [2] proposed a simpler model assuming that the collector temperature is homogeneous in space. This model is a linear ordinary differential equation (ODE) validated in [3] and is likely the simplest physically-based model used in the practice (see e.g. [4,5]), which can still describe the transient collector processes with an appropriate accuracy.…”

“…Nevertheless, the model may be rather precise even if it is simple as in the case of [3]. The most frequent blackbox model type is perhaps the artificial neural network (ANN) in the field of collector modelling.…”

“…Because of the above problems, a simple and general but still accurate black-box model, which can be applied easily and fast for a wide range of solar collectors, has been recently worked out in [3]. The model is based on the well-known methods of mathematical statistics, more precisely, the multiple linear regression (MLR).…”

“…Based on the literature, MLR is a rare blackbox modelling technique in the field of collectors despite of its simplicity. Considering the high precision (with an error of 4.6%), simple usability and low computational demand of the mentioned MLR-based model (MLR model in short) in [3], it is definitely worth trying to simplify further the MLR model to gain an even more easy-to-apply model with a still good accuracy. Such a simplification (likely the simplest possible MLR based model) is proposed in the present study.…”