Comparative Study of Optimization Techniques for Tuning of PI Gains for Greenhouse Climate Control

Greenhouses are highly nonlinear, non-isolated systems which are very difficult to model and control. For desired plant growth, such greenhouses demand a very specific form of climatic conditions which can vary due to the impact of highly stochastic disturbances in the form of solar radiation, temperature and humidity of outside environment. Greenhouse climate control designs are often put into use to increase the quality of growth of the plants cultured inside the greenhouse while keeping a check on the production costs. Such designs need to be efficient at setpoint tracking while providing good disturbance rejection. In this paper a greenhouse's climate is controlled with conventional PI controller. The controller is tuned with different global optimization algorithms, namely Genetic Algorithm (GA), Cuckoo Search Algorithm (CSA) and Grey Wolf Optimization (GWO) Algorithm. The tuning has been performed through minimization of a weighted sum of integral of the absolute error. Simulation study clearly reveals that the GWO tuned PI controller offers much better results.