Indian Journal of Science and Technology
Year: 2023, Volume: 16, Issue: 35, Pages: 2879-2888
Rinku Chavda1, Sohil Pandya2*, Chetan Kotwal3
1Research Scholar, Gujarat Technological University, Gujarat, India
2Assistant Professor, Department of Computer Applications, CMPICA, Charotar University of Science and Technology (Charusat), Charusat Campus, Changa, India
3Professor, Department of Electrical Engineering, Sardar Vallabhbhai Patel Institute of Technology, Vasad, Gujarat, India
Email: [email protected]
Received Date:24 March 2023, Accepted Date:13 August 2023, Published Date:21 September 2023
Objectives: Total electricity consumed by Residential Households contributes remarkably in the domain of Electricity Consumption Patterns and Profiles. Discovering the correlation between electricity consumption with housing and demographic characteristics of households will be helpful to identify the influence of various features to generate consumer electricity consumption patterns and identification of load profiles. Methods: Using the SOMKMeans Clustering Algorithm, Pearson’s, Spearman’s Rank, and Kendall’s tau Correlation techniques, a feature correlation and dependency between input and output features have been analyzed. Additionally, a statistical study has been conducted to determine the impact of housing and demographic features on electricity consumption using correlation coefficients, correlation matrix, and internal evaluation metrics. Findings: As per the SOMKMeans Clustering Analysis, the evaluation metrics’ scores with and without regard to household and demographic characteristics have been compared, and the scores of Davies-Bouldin, Calinski-Harabasz, and Silhouette are compared to one another. It has been found that the impact of demographic characteristics, family habits, and physical characteristics of houses have an indirect presence in the recorded daily electricity consumption of consumers in the experiment dataset of 4942 households for the year 2013 whereas, it has been shown that each variable, such as family structure and age group, is present in each cluster, indicating that there was no discernible influence of housing and demographic features during the training of the classification model. As per the used ANOVA Test, Chi2 Test, and Mutual Information techniques, the impact of Electricity Consumption to predict load profile is 98%, 95%, and 89%, respectively. Novelty: Clustering and Statistical analysis of Consumption give insight into which relevant features are useful for deciding the consumption class. An important factor in estimating the electricity consumption profile of consumers is daily electricity consumption to reduce model complexity and achieve higher accuracy to predict the load profile of consumers.
Keywords: Electricity Consumption; Housing Features; Demographic Features; Correlation; Feature Dependency
© 2023 Chavda et al. This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited. Published By Indian Society for Education and Environment (iSee)
Subscribe now for latest articles and news.