Temperature and irradiance levels are two examples of environmental variables that affect the power value produced by photovoltaic panels. Therefore, in order to transfer the maximum power value from the PV panel to the load under varying climatic conditions, maximum power point tracking (MPPT) algorithms and DC-DC converter topologies are used. In this study, the performances of boost converter and CUK converter circuit topologies are investigated under variable irradiance and variable load conditions by using a neural network-based MPPT algorithm learning particle swarm optimization (PSO). As the first scenario, it is analyzed assuming that the temperature and irradiance values coming to the panel are constant. As the second scenario, the performance evaluation of the converter topologies according to the current, voltage and power parameters is made for the variable load situation. As the last scenario, the difference in the irradiance value coming to the panel depending on the sun's condition during the day has been examined. Canadian Solar CS6P-250P PV panel is used in the study. 50 kHz is selected as the switching frequency. According to the results obtained, it has been observed that the CUK converter circuit topology reaches the maximum power point faster than the boost converter circuit topology both in dynamic environmental conditions and load change, and the oscillation at this point is less. It is aimed to increase the performance of this method, which uses boost converter topology and MPPT in the literature, by applying CUK converter topology.