Intelligent Systems for Monitoring and Controlling Electricity Consumption in Industrial Enterprises: Application of Artificial Intelligence and Digital Technologies

Abstract

Relevance: In industrial enterprises with continuous and energy-intensive production, electricity consumption is tightly coupled to the technological regime; therefore, monitoring and management must rely on measurable data and respect process constraints. Digital technologies-smart meters, IoT sensors, SCADA, EMS, MES, ERP and digital twins-provide the technical basis, while artificial intelligence methods derive forecasting and control decisions from accumulated data.

Aim: The paper examines artificial intelligence methods and digital technologies used for consumption monitoring, load forecasting, efficiency assessment and regime control, and identifies the conditions of their applicability, the achievable effect, their limitations and integration problems.

Methods: An analytical review of peer-reviewed publications (mainly 2016–2026) was performed; bibliographic records were verified against Crossref and OpenAlex, standards and legislation against the official ISO, IEC and lex.uz pages. The methods are classified by their function within the monitoring loop and compared by input data, tasks solved, strengths, limitations, real-time operation and integration with enterprise systems.

Results: A four-layer architecture of digital monitoring and a functional map of artificial intelligence methods are proposed, in which method families (neural networks, tree ensembles, clustering, reinforcement learning, fuzzy logic, hybrid and physics-informed models) are matched to data processing, forecasting, efficiency assessment and control tasks. It is shown that the bottleneck of implementation is not so much the models as the quality of primary data, the completeness of measurements and the admissible depth of load impact.