The article considers the method of evaluating the parameters of equivalent schemes of solar panels using data provided by their manufacturers. The technique involves the use of a digitized volt-ampere characteristic of the solar panel and standard test conditions parameters (STC) to estimate the numerical derivatives at the points of short and open circuit. Digitizing the volt-ampere characteristic introduces some error, which significantly affects the accuracy of determining numerical derivatives. To solve this problem, it is advisable to average the values of the derivatives. It was found that it is sufficient to take 20-25% of the first points and 8-15% of the last points of the digitized curve. In this case, to estimate the value of the derivative at the point of short circuit, it is advisable to use linear fit, and at the point of open circuit – quadratic fit. The peculiarities of using the obtained formulas and the block diagram of the algorithm that implements this technique were also given.
A fixed point algorithm is used to solve the transcendental equation of the external characteristic of the solar panel. It was found that in some cases this numerical method may not convergences near the open circuit point. To solve this problem, in the proposed to use the relaxation method. This increases the required number of iterations, but guarantees the convergence of numerical method. The block diagram of the simple iteration algorithm using the relaxation method is given.
Modeling the developed technique, one- and two-diode solar panel replacement schemes were used on the example of KC200GT and ST40 solar panels. A comparative analysis of these schemes showed that a single-diode circuit is more relevant, because its use simplifies calculations and eliminates the need for a number of assumptions. The accuracy of the approximation provided by the single-diode circuit is proportional to the accuracy of the two-diode circuit. A comparative analysis of the developed methodology with other works was performed. The results indicate that the proposed algorithm provides the best approximation accuracy among the considered works. For monocrystalline solar panels, the average standard error is 7 times less and the modulus of relative error - 4 times. For polycrystalline panels, the average standard error is 1.5 times less, and the modulus of relative error is 1.75 times. For polycrystalline panels, the approximation error increases due to a non-unit value of the ideality factor for this manufacturing technology of the solar panels.