Performance measurement of various AI techniques for energy estimation and its optimisation using sensitivity analysis
Yashish Swami, Navjot Singh, Umang Soni · International Journal of Intelligent Enterprise · 2022
The objective of this research is to predict energy performance of a building (EPB) in terms of heating and cooling load by using various artificial intelligence (AI) techniques then measuring the corresponding strength of each input and its effect on the output in order to identify the most significant input from the lot by using sensitivity analysis. EPB can help in efficient construction of buildings as well as put a leash on dwindling natural resources and global warming. The various intelligent techniques used in this project are artificial neural network (ANN), adaptive neuro-fuzzy inference system (ANFIS), ANFIS-GA (genetic algorithm) and ANFIS-PSO (particle swarm optimisation). In order to identify the most significant input, we are using a technique based on sensitivity analysis, which is called the connection weight algorithm. In the end, performance of the AI techniques is compared to select the best performing model.