The effectiveness and accuracy of the methods used in modeling photovoltaic (PV) cells are necessary for predicting the performance of PV systems. This article presents and analyses four methods used to estimate the unknown parameters of photovoltaic devices for one- (four and five parameters) and two-diode (seven parameters) models. The study has three main objectives: (i) extract the unknown parameters by the four methods for the 28 modules of different PV technologies (mono-crystalline, multi-crystalline, HIT, and amorphous) and manufacturers under standard test conditions, (ii) verify the effectiveness and the accuracy of the methods, at each current-voltage curve data point (category 1), and the error relative at each short-circuit current (I-sc), open circuit voltage (V-oc) and peak power (P-p) points (category 2) and (iii) study the accuracy and reliability of the four methods for the energy measurement of a PV plant of 1.75 kW over three days installed in desert environment. The results show that the most accurate method is obtained by the four parameters one-diode model. It was shown that there is a good agreement with all important points of the I-V curves; especially at I-sc, V-oc and P-p points, and with the energy delivered by the PV system in the three days.