Machine Learning algorithms for prediction of energy consumption and IoT modeling in complex networks

In this paper, the internal energy consumption on the Internet of Things (IoT) has been studied. The purpose of this paper is to predict the factors affecting energy consumption in buildings by considering machine learning algorithms such as k-nearest neighbors (KNN), Ada boost, random forest and ne...

Full description

Saved in:
Bibliographic Details
Published in:Microprocessors and microsystems Vol. 89; p. 104423
Main Authors: Fard, Rehane Hafezi, Hosseini, Soodeh
Format: Journal Article
Language:English
Published: Kidlington Elsevier B.V 01.03.2022
Elsevier BV
Subjects:
Ada boost algorithm
Algorithms
Building energy consumption
Buildings
Computer simulation
Cooling loads
Energy
Energy consumption
Heating load
Internal energy
Internet of Things
K-nearest neighbor algorithm (KNN) algorithm
Machine learning
Neural network algorithm
Neural networks
Random forest algorithm
Regression
Internet of things
Energy
Building energy consumption
Random forest algorithm
Regression
K-nearest neighbor algorithm (KNN) algorithm
Neural network algorithm
Ada boost algorithm
ISSN:0141-9331, 1872-9436
Online Access:Get full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:In this paper, the internal energy consumption on the Internet of Things (IoT) has been studied. The purpose of this paper is to predict the factors affecting energy consumption in buildings by considering machine learning algorithms such as k-nearest neighbors (KNN), Ada boost, random forest and neural network. These algorithms are implemented in the Orange tool. Also, the univariate regression algorithm is used to select the best feature. This algorithm determines the most important factors affecting energy consumption and their impact. Then, with the help of Gephi tool, these data are simulated in the IoT environment as a complex network. The simulated network in the Internet of things is also a small world network. This network shows the relationships between the features. The results of this paper show that the overall height, roof area, surface area and relative compaction have the greatest impact on the energy consumption of buildings. It can be seen that the predicted error percentages for these data for cooling loads and heating loads are 0.911 and 0.292, respectively. It should be noted that the best algorithm for cooling load and heating load is Ada boost algorithm.
Bibliography:ObjectType-Article-1
SourceType-Scholarly Journals-1
ObjectType-Feature-2
content type line 14
ISSN:0141-9331
1872-9436
DOI:10.1016/j.micpro.2021.104423