Design and Modeling of an AI-Powered Industrial Maximum Demand Controller With Web Service Interface

Abstract

Industrial facilities face challenges in managing energy consumption and minimizing peak demand charges. Managing maximum demand is a critical aspect of energy management in industrial facilities. The proposed MDC is based on a predictive analysis method using time-series models such as long short-term memory (LSTM) and XGBoost. However, this chapter also discusses the pros and cons of TCN and k-NN model oriented models. Further, this chapter discusses development and implementation of the proposed MDC as a web service. Finally, development of a user-friendly web interface for dataset uploads, system configuration, and alerts are explored.