Author/Authors :
Ab Rahman, Ahmad Faiz Universiti Teknologi Malaysia (UTM) - Faculty of Electrical Engineering, Center for Artificial Intelligence and Robotics (CAIRO), Malaysia , Selamat, Hazlina Universiti Teknologi Malaysia (UTM) - Faculty of Electrical Engineering, Center for Artificial Intelligence and Robotics (CAIRO), Malaysia , Ismail, Fatimah Sham Universiti Teknologi Malaysia (UTM) - Faculty of Electrical Engineering, Center for Artificial Intelligence and Robotics (CAIRO), Malaysia , Khamis, Nurulaqilla Universiti Teknologi Malaysia (UTM) - Malaysia-Japan International Institute of Technology, Malaysia
Abstract :
This paper discusses the development of a building energy optimization algorithm by using multiobjective Particle Swarm Optimization for a building. Particle Swarm Optimization is a well known algorithm that is proven to be effective in many complex optimization problems. Multiobjective PSO is developed by utilizing non-dominated sorting algorithm in tandem with majority-based selection algorithm. The optimizer is written by using MATLAB alongside its GUI interface. Results are then analyzed by using the Binh and Korn benchmark test and natural distance performance metrics. From the results, the optimizer is capable to minimize up to 42 percent of energy consumption and lowering the electricity bills up to 43 percent, while still maintaining comfort at more than 95 percent as well. With this, building owner can save energy with a low-cost and simple solution.
Keywords :
Multiobjective optimization , particle swarm optimization , building energy optimization , paretofront , benchmark test , performance metrics
