Multi-objective optimization and decision making for greenhouse climate control system considering user preference and data clustering

Optimization of the systems can increase their efficiency with appropriate system performance indices. Nowadays, climate control of industrial greenhouses consists of many control parameters such as light, temperature, CO2\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$_{2}$$\end{document} concentration, humidity, and etc. Most of these systems use the PID control structure due to its simplicity, flexibility, and good performance. The electrical lamps and heaters can be used to provide appropriate light and temperature inside the greenhouse. Optimal tuning of electrical lamps and heaters control system has a significant influence on efficiency and performance improvement of greenhouse cultivation system. For this aim, NSGA-II, a well-known evolutionary algorithm, is applied for the system optimization verified with an exhaustive search approach. Making a final decision to choose the best solution among the optimal solutions is a challenging decision making issue. In this regard, post Pareto-optimal pruning algorithms are employed considering various user preferences and clustering approaches. The final results show and verify the substantial improvement of greenhouse climate control system efficiency and performance.