Parameter identification and uncertainty evaluation in quasi-dynamic test of solar thermal collectors with Monte Carlo method

This work presents a comparison between two methods used for parameter identification and calculation of parameter uncertainty applied to the measured data of solar thermal collectors when tested according to ISO 9806:2017. One method is using a weighted least square (WLS) fit and the partial derivative approach described in the GUM (the Guide to the Expression of Uncertainty in Measurement). The second is using the Monte Carlo (MC) method, also described in GUM. Uncertainty evaluation by Monte Carlo method is based on a probabilistic approach and is an alternative way for identification of parameters and determination of the uncertainties. In this work the results were obtained according to Quasi-Dynamic Test (QDT) method for a flat plate collector and an evacuated tubular collector. The least squares (LS) method and a nonlinear regression method ( MPFit ) are used in the identification of parameters for each iteration of the MC method. For the implemented MC method computation times are also discussed. One disadvantage of the MC method is the computation time which depends on the number of samples in the experimental test quantity files, however with this study we think that the advantages of the MC method outweigh the disadvantages, and it is useful even as a complementary tool in QDT testing of collectors. • Quasi-Dynamic test method data analysis. • Least square methods for parameter identification. • Monte Carlo method in parameter identification. • Nonlinear regression method for parameter identification. • Calculation of parameter uncertainties.