Vol. 1, No. 3 (2025)
Performance Evaluation of Modified Internal Model Control-tuned Proportional Integral Derivative Controller for Controlling Greenhouse Temperature
Johnson O. Abiola +, David O. Aborisade +, Muniru O. Okelola +
Johnson O. Abiola
Mechatronics Engineering Department, Bells University of Technology, Ota, Nigeria
David O. Aborisade
Electronic and Electrical Engineering Department, Ladoke Akintola University of Technology, Ogbomoso, Nigeria
Muniru O. Okelola
Electronic and Electrical Engineering Department, Ladoke Akintola University of Technology, Ogbomoso, Nigeria
Abstract
Greenhouses are important in providing favorable conditions for the growth of plants through regulation of internal climatic factors. During the rainy season, the continuous fall of rain causes low temperatures that retard the growth of plants and make them susceptible to attacks by pests. Traditional temperature control methods using Proportional-Integral-Derivative (PID) controllers, with tuning techniques such as Ziegler-Nichols (ZN) and Fuzzy Logic, tend to have poor accuracy and slow response. To overcome these limitations, this paper proposes a Modified Internal Model Control (MIMC)-based PID tuning technique specifically tailored for greenhouse temperature regulation. The thermal dynamics of the greenhouse were modeled analytically using an energy balance approach based on Gustav Kirchhoff's radiation law and the first law of thermodynamics. The model accounts for internal heat sources, convective losses, solar radiation, and external weather influences. The suggested MIMC method adds an adaptation factor to the conventional IMC framework in order to maximize the controller robustness and responsiveness in the face of changing ambient conditions. A MIMC-PID control scheme has been developed by identifying the nominal plant model, selecting a suitable filter time constant from desired closed-loop behavior, and tweaking the control parameters to find a compromise between disturbance rejection and noise attenuation. Simulations were carried out in MATLAB R2023a by comparing the MIMC-PID controller with that of ZN-PID and Fuzzy-PID controllers under similar conditions of disturbances. The results are that MIMC-PID outperforms others as its results show a rise time of 0.180 s, a settling time of 0.490 s, and maximum overshoot of 8.152%. This validates that the MIMC-based PID method is promising to make the control effectiveness of greenhouse temperature more enhanced when adverse conditions dominate during various seasons, promising an agricultural environment.
Click here to download PDFKeywords
- Temperature control
 - Proportional Integral Derivative
 - Greenhouse microclimate
 - Temperature regulation
 - Internal model control
 - Vegetables thermal zone.
 
How to Cite
Johnson O. Abiola, David O. Aborisade & Muniru O. Okelola (2025), Performance Evaluation of Modified Internal Model Control-tuned Proportional Integral Derivative Controller for Controlling Greenhouse Temperature, Nigerian Journal of Applied Science and Innovative Technology, 1(3), 312–327, Retrieved from https://nijasit.vercel.app/article/21