ISSN Print : 2621-3745 ISSN Online : 2621-3753 (Page.7-19) Jurnal Journal of Science and Applied Engineering (JSAE) Vol. 5, No. 1, April 2022 DOI : 10.31328/jsae.v5i1.3263 Hasan et al., Optimizing Maximum Power Point Tracking ... 7 OPTIMIZING MAXIMUM POWER POINT TRACKING ON PHOTOVOLTAIC ARRAYS USING ANT COLONY OPTIMIZATION AND PARTICLE SWARM OPTIMIZATION ALGORITHMS F. Hasan 1, *, H. Suyono 2,* , A. Lomi³ 1,2,3 Department of Electrical Engineering, Faculty of Engineering Universitas Brawijaya *Email: hadis@ub.ac.id Submitted : 23 November 2021; Revision : 14 January 2022; Accepted : 2 March 2022 ABSTRACT Solar power plants, in general, cannot produce maximum power by themselves; the characteristics of the PV voltage generally follow the battery voltage or the load that is connected directly to the PV. The intensity of light received by the PV modules does not all get uniform irradiation, so the power produced is not optimal and causes multi-peak. A Maximum Power Point Tracking (MPPT) system is needed to optimize power from PV. However, the often used methods are still trapped in local peaks and long convergence times. In this study, we compare the performance of each algorithm to find the maximum power point (MPP) and tracking time using two methods, namely Ant Colony Optimization (ACO) and Particle Swarm Optimization (PSO). This study uses 6 selected cases that can occur in 6 solar panel modules arranged in series. Characteristic curves in 6 cases were generated using MATLAB SIMULINK for tracking to find the maximum power point using the ACO and PSO algorithms. The ACO has an efficiency of 99.4910% and tracking failure 7 times in 6 cases in 10 trials of each case, while the PSO algorithm has an efficiency of 99.1043% and tracking failure seven times in 6 cases in 10 trials each case. The efficiency comparison of the ACO algorithm is 0.39% better than the PSO algorithm, while the PSO method is faster in tracking. Keywords Ant Colony Optimization; Particle Swarm Optimization; Maximum Power Point Tracking; Convergent Time. Paper type Research paper INTRODUCTION In general, PV cannot produce optimally because the PV voltage will usually follow the battery's voltage connected to the PV. Maximum Power Point Tracking (MPPT) technology can help control the PV module to work at the maximum power point or Maximum Power Point (MPP) to produce optimal PV power. For PV to produce a higher maximum current and voltage, it is necessary to use several PV modules connected in series or parallel to get a higher current and voltage. The modules connected in parallel or series are called PV arrays. Suppose shadows of trees partially cover the PV array modules, clouds, buildings and so on. In that case, not all of the modules get the same irradiation, where each PV module has different or unbalanced results, so the total output power of the PV array is very high decrease. In addition, the hot-spot effect caused by partial shading tends to damage the PV cells and affect the safety of the PV system [1]. Therefore, more comprehensive tracking is needed when experiencing partial shadows. Conventional methods widely used for MPPT include Perturb and Observe (P&O), fractional open-circuit voltage, and incremental conductance. This conventional method has a slow response and is unsatisfactory in solving the problem of rapid environmental change and overcoming the non-linearity of PV. The rapid change in irradiation due to weather factors, the P&O method failed to track MPP [2]. The latest research related to the manufacture of the MPPT algorithm in Partial Shading conditions is using Artificial Intelligence (AI) based control methods such as Artificial Neural Network [3] Adaptive-Neuro-Fuzzy Inference System [4] and Fuzzy Logic Control [5]. However, the very large data is needed for fuzzification process in fuzzy logic control will burden the computational process. Likewise, the large amount of data makes the training process slow with the neural network method.