Benefits of Using Intelligent Control Systems in Comparison to Classical Methods for Temperature Control in a Greenhouse

Abstract: In the last decades, intelligent control systems have rapidly replaced classical methods to solve problems in many spheres of science and technology. The purpose of this study is to demonstrate the benefits of intelligent control systems in comparison to the classical methods in a low-scale benchmarking problem. The problem chosen in this study is temperature control in a green house. The intelligent control methods of neural network and fuzzy logic have been simulated. These simulations are used to compare with the simulations of classical methodologies such as PID and open loop controllers. Results from the tests on these simulations show that intelligent control methods have better overall performance with better accuracy and a smaller temperature variance. Additionally, neural networks have the benefit of independence from models and does not require extensive knowledge of the problem field. The classical methods are on the other hand easier to implement and do not require a complex selection of training data.