SISTEM MAXIMUM POWER POINT TRACKING BERBASIS HYBRID PARTICLE SWARM OPTIMIZATION - GENETIC ALGORITHM (PSO-GA) PADA PANEL SURYA

Abstract

Solar panels used as a device to convert sun light energy into electrical energy have non-linear characteristics that depend on the variables of sunlight intensity and temperature. Varying environmental conditions such as temperature and sunlight intensity, indicate inconsistent PV Maximum Power Point (MPP) characteristic curves, thus creating challenges when tracking MPP, namely in the condition of solar panels exposed to partial shading with the condition that one part or all PV modules receive unequal irradiation. This unequal irradiation can occur due to partial shading. Partial shading problems can reduce the level of efficiency and affect the output power produced by solar panels. Therefore, partial shading problems can be handled with the Maximum Power Point Tracking (MPPT) system. In previous studies, by combining the PSO and GA methods to avoid the MPP point falling into the local optimum. This method can track global peaks quickly compared to conventional P&O MPPT methods. The advantages of the PSO method include fast tracking speed, more flexibility, and the ability to find global point solutions. Therefore, in this study the PSO-GA method was used for better performance improvement. From the research results, MPPT testing using the PSO-GA (Particle Swarm Optimization-Genetic Algorithm) algorithm is able to track maximum and stable power with conditions of irradiation variation, partial shading, temperature variation with a time of 0.027 seconds - 0.035 seconds with an average accuracy of 98.01% - 98.62%. The performance of PSO-GA MPPT compared to PSO and GA using 4 different cases, namely case1 with (1000 W / m2, 1000 W / m2), case2 (500 W / m2, 500 W / m2), case3 (1000 W / m2, 700 W / m2), and case4 (1000 W / m2, 700 W / m2). The results of the PSO-GA tracking time reached 0.029 seconds faster than PSO (0.071 seconds) and GA (0.142 seconds).