Optimizing Energy Consumption Patterns in Southern California: An AI-Driven Approach to Sustainable Resource Management

Abstract

Southern California is a special case scenario for any energy management study, given its sunny climate, sprawling urban landscapes, and economic strength.  This research project focuses on how artificial intelligence can be applied in energy management by showing its potential toward optimum energy consumption and maximizing sustainability within Southern California. The dataset used for this study was accessed from the Kaggle website. The dataset encompassed various energy consumptions, and the data were collected across different building structures in Southern California between January 2018 and January 2024. These datasets included hourly records of electricity consumption for residential and commercial buildings and industrial buildings. Furthermore, it also provided a record of environmental and operating metrics. This dataset is useful to researchers and practitioners who work on forecasting electricity consumption, energy management, sustainability, and developing AI-based optimization models. To depict an insight into which variable affects strongly within the patterns of energy consumption, machine learning techniques used were logistic regression, Random Forest, and XG-Boost while considering a power outage data set. This study gives evidence that the AI-based models significantly enhance the forecast's accuracy and further allow integration of renewable energy resources, which in turn yield benefits through reduced operational costs and reduction of GHG emissions. The discussion has shown how such development implications could be translated into supportive policy frameworks of advanced studies in the future to electric vehicle charging and even energy storage solutions. Generally, this research underlines the crucial role of AI in changing energy management practices towards sustainable energy use.