An Intelligent Approach to Develop, Assess and Optimize Energy Consumption Models for Air-Cooled Chillers using Machine Learning Algorithms

Abstract

The building sector accounts for more than 70% of the total electricity use. Chillers consume more than 50% of electrical energy during seasonal periods of building use. With the growth of the building sector and climate change, it's essential to develop energy-efficient HVAC systems that optimize the ever-increasing energy demand. This study aims to develop an energy consumption prediction model for air-cooled chillers using machine learning algorithms. This is done by developing different static and dynamic data-driven regressive and neural network models and comparing the accuracy of their prediction to identify the most accurate modeling algorithm using 3 main inputs chilled water return temperature, outside dry bulb temperature, and cooling load. The proposed model structure was then optimized in terms of the number of neurons, epochs, time delays as well as the number of input variables using a genetic algorithm. Training and testing were done using real data obtained from a fully instrumented 4-tonair-cooled chiller. Results of the study show that the optimized artificial neural network model can predict energy consumption with a high level of accuracy compared to conventional modeling techniques. The development of highly accurate self-tuning models can be a powerful tool to use for other applications such as fault detection and diagnosis, assessment, and system optimization. Further studies are necessary to evaluate the effectiveness of using deep learning algorithms with more hidden layers and cross-validation techniques.