Parameter identification of photovoltaic cells/modules by using an improved artificial ecosystem optimization algorithm and Newton-Raphson method

Abstract

Precise models of photovoltaic (PV) modules are crucial for simulating PV system characteristics. To address the challenges of accurately and promptly acquiring parameters from measured current-voltage (I-V) data of PV modules, an improved artificial ecosystem optimization (IAEO) algorithm was proposed. The IAEO algorithm enhanced the producer diversity within the production operator by introducing a probabilistic selection mechanism and an evolution mechanism. Moreover, a mutation operation based on historical values was introduced into the consumption operator to enhance the search ability. A comparative analysis was conducted among IAEO and the other six algorithms for parameter extraction of PV cells/modules using publicly available datasets. The IAEO algorithm demonstrated high identification accuracy and speed. The identification accuracy was improved by constructing a two-layer optimization structure with Newton-Raphson method and using the double-diode model. The Newton-Raphson method integrated IAEO algorithm achieved root mean squared error (RMSE) values of 7.32392×10-4, 1.67191×10-3, 9.89002×10-3, and 1.48362×10-3 for France cell, STM6-40, STP6-120, and PWP 201, respectively. An I-V tester was designed to measure the I-V data of different PV modules on the experimental platform. Results showed that the IAEO algorithm exhibited superior performance in extracting four specific PV modules, and the relative error of power is below 2.5%.

Keywords

• Artificial ecosystem optimization algorithm
• Newton-Raphson method
• Photovoltaic power generation
• Parameter extraction
• Current-voltage curve