LAUSR

Smart grid (SG) is an emerging subject in power engineering that requires interdisciplinary skills. SG implementation has challenges in accurate prediction of electricity consumption, power stability, power quality, and critical infrastructure security. This study uses predictive modeling and different tools to introduce problem‐based learning techniques through household electricity consumption (HEC) analysis. The problem statement is to understand the effect of the weather conditioning data set on the accuracy of HEC prediction and select the most fitting predictive model for further decision‐making. Five predictive models, namely random forest, linear regression, support vector machine, neural network, and adaptive boosting, are utilized, and their prediction results are compared based on a performance matrix. The paper also briefs the usage of different tools, including Jeffrey's Amazing Statistics Program tool for finding the correlations within variables, the Power Business Intelligence tool for data visualization, and the Orange tool for building predictive modeling. This study focuses on three aspects. Firstly, to demonstrate energy predictive modeling analysis without knowing the coding knowledge. Secondly, to develop the creative thinking of engineering students to solve complex, interdisciplinary engineering problems using tools. Thirdly, to motivate students to be actively involved in their problem‐based research projects.