New perturb and observe algorithm based on trapezoidal rule: uniform and partial shading conditions

• Main Authors: Mahmod Mohammad Altwallbah, Neda, Mohd Radzi, Mohd Amran, Azis, Norhafiz, Shafie, Suhaidi, Mohd Zainuri, Muhammad Ammirrul Atiqi
• Format: Article
• Published: Elsevier 2022
• Online Access: http://psasir.upm.edu.my/id/eprint/102310/; https://www.sciencedirect.com/science/article/pii/S0196890422005349
• Institution: Universiti Putra Malaysia

This paper proposes a new version of the perturb and observe (P&O) algorithm which is able to recognize the global maximum power point (GMPP) among the multiple maximum peaks that exist in the P-V curve under the effect of the partial shading condition (PSC). For the first time, Trapezoidal rule concept is employed in P&O approach as a new consideration in the tracking process to calculate the best path for precisely tracking the GMPP. The new algorithm allows the search to continue until the extremum maximum is identified among numerous local maximums, rather than being imprisoned at one of the local maximums as is the case with the standard algorithm. Furthermore, the algorithm is able to achieve the MPP efficiently under the presence of rapid and abrupt changes in irradiation levels. The proposed Trapezoidal rule based P&O algorithm is evaluated under three conditions: (1) under uniform irradiance (2) under shading conditions and (3) under sudden and step changes in irradiation levels as well as the transition between shading patterns. The feasibility and effectiveness of the suggested algorithm, as well as the well-designed boost converter, are validated using two different methodologies: a simulation model in MATLAB/Simulink and hardware implementation by the (TMS320F28335) digital signal processing controller. Both simulation and hardware results confirm that the proposed algorithm provides excellent efficiency of 100% and 99.6% respectively. The GMPP is achieved within less than 100 ms, which is extremely advantageous tracking time, that avoids the power losses. In addition, the minimal steady-state oscillation is shown, with the desired level of simplicity without the need to complicated computations or random particles.