International Journal of Research in Advanced Electronics Engineering
2026, Vol. 7, Issue 1, Part A
Comparative performance analysis of PID and fuzzy logic controllers for single-axis solar tracker systems
Author(s): Michael J Thompson
Abstract: Solar tracking systems enhance photovoltaic energy capture by maintaining optimal panel orientation relative to the sun throughout the day, yet the choice of control algorithm significantly influences tracking precision, energy yield, and system reliability under varying environmental conditions. This research presents a comprehensive comparative analysis of Proportional-Integral-Derivative and Fuzzy Logic control strategies implemented on identical single-axis solar tracker hardware, evaluating performance across multiple metrics including tracking accuracy, energy harvest, response characteristics, and robustness to environmental disturbances. Both controllers were implemented on an Arduino Mega 2560 platform driving a 100W polycrystalline solar panel through a DC motor and gear reduction system. The PID controller employed Ziegler-Nichols tuning methodology yielding gains of Kp=2.5, Ki=0.8, and Kd=0.3, while the Fuzzy Logic controller utilized a 5×5 rule base with triangular membership functions processing error and change-in-error inputs through Mamdani inference with centroid defuzzification. Both systems received identical sun position inputs from a dual light-dependent resistor sensor array providing differential feedback proportional to tracking error. Field trials conducted at Toronto Institute of Applied Sciences Solar Research Facility from September to November 2024 compared controller performance under diverse Canadian autumn conditions including clear skies, intermittent cloud cover, and varying wind speeds. The Fuzzy Logic controller achieved mean tracking error of 1.05 degrees compared to 1.78 degrees for PID, representing 41% improvement in positioning accuracy. This enhanced precision translated to 8.1% higher daily energy yield, with the Fuzzy system harvesting an average of 990.2 Wh compared to 915.8 Wh for the PID-controlled tracker over matched daylight periods. The performance differential amplified under challenging conditions. During high wind events exceeding 20 km/h, the PID controller exhibited tracking errors up to 3.2 degrees with visible oscillation, while the Fuzzy controller maintained errors below 2.0 degrees through its inherently damped response characteristics. Similarly, rapid irradiance fluctuations during partly cloudy conditions revealed PID overshooting tendencies that the Fuzzy controller's graduated rule-based response avoided, demonstrating superior disturbance rejection without requiring parameter retuning. Computational analysis confirmed that both algorithms executed within the Arduino's processing constraints, with PID requiring approximately 0.8 milliseconds per control cycle compared to 2.4 milliseconds for Fuzzy inference. While the Fuzzy controller's higher computational load reduced maximum achievable sampling rate, the 100 Hz update frequency employed proved more than adequate for solar tracking dynamics where sun position changes gradually over minutes rather than milliseconds. The research demonstrates that Fuzzy Logic control offers tangible advantages for solar tracking applications where environmental variability challenges conventional linear control approaches, providing a validated implementation framework for practitioners seeking to enhance tracking system performance without sophisticated mathematical modeling requirements. The 8.1% energy gain achieved justifies the modest additional implementation complexity, particularly for installations in locations experiencing variable weather conditions similar to southern Ontario's climate.
DOI: 10.22271/27084558.2026.v7.i1a.73
Pages: 26-32 | Views: 44 | Downloads: 17
Download Full Article: Click Here
How to cite this article:
Michael J Thompson. Comparative performance analysis of PID and fuzzy logic controllers for single-axis solar tracker systems. Int J Res Adv Electron Eng 2026;7(1):26-32. DOI: 10.22271/27084558.2026.v7.i1a.73
