Machine learning based energy evaluation using air temperature and air velocity

Climate change has been a hot topic in the past decade. Impacts of such climate change includes the rise in temperature. Poor thermal conditions can result in discomfort and lead to potential health hazards which have an impact on a person productivity. Air condition systems are used to resolve the poor thermal conditions. Building’s energy consumption has a significant carbon footprint. Extensive research has been done to reduce the impact on climate change. Hence, this has give rise to technology such as green and energy efficient building. These improvements have also led to energy savings for companies. Heating, Ventilation, Air-Conditioning (HVAC) systems of a building are the focus of smart building’s research to find a more sustainable way of utilizing them. Hence, smart building technology is the direction forward to achieve energy efficiency and energy savings. The focus of this paper is on machine learning techniques to analyze the relationship and develop a model to optimize the energy consumption and the thermal comfort of buildings. Predicted Mean Vote (PMV) will be used to analysis the thermal comfort level of the occupants. Artificial Neural Network (ANN) will be created using air velocity and air temperature data to predict the energy consumption and thermal comfort. Through the use of the energy consumption and PMV model function, a optimization a model is determined. Optimization methods such as Genetic Algorithm (GA) and Particle Swarm Optimization (PSO) was then applied and compared to identify a suitable optimization model. MATLAB was then used to simulate both the GA and PSO to identify the optimal solution for the HVAC system. Using the identified optimal model, the building owner could operate the HVAC system based on their requirement and adjust their optimal operating frequency to reduce energy consumption and maintain thermal comfort level.Bachelor of Engineerin