Energy Consumption Estimation Using Machine Learning with Data from Smart Meters in a Residential Complex Building in Iraq

Demand for energy production, energy conservation techniques, and now a significant international problem due to the exponential expansion in population and their total dependency on the use of electrical and electronic equipment. The construction industry is a significant user of energy, hence. This research paper aims to determine the energy usage of a residential complex building in Iraq by using machine learning strategies to the data collected by smart meters in that building. The critical focus is producing accurate projections of future loads in the residential complex over two or three years to surpass any potential artificial intelligence barriers. Data gathering, pre-processing, algorithm selection, model training, model assessment, and energy consumption estimation are the steps involved in the technique. The dataset has been corrected, and the three algorithms used are linear regression, decision trees, random forests, gradient boosting, bagging regressor, extra trees regressor,SVR, lasso, ridge, elastic net, K-nearest neighbours and neural network. The protection of users’ privacy and the safety of their data are among the ethical issues. An accurate calculation of the energy consumed is required to manage energy and reduce costs effectively. The findings of this study can be used to address the problem of inefficient management of energy resources in residential structures in Iraq and other countries with comparable conditions.