Fog-Computing-Enabled Intelligent Control and Fault Diagnosis of Power Converters for Distributed Energy Systems

Abstract

This paper proposes an intelligent control and faultdiagnosis framework based on fog computing for power converters in Distributed Energy System (DES) applications, which combines distributed energy resources, a grid interface, an Energy Storage System (ESS), a sensing unit, a fogcomputing node, an intelligent controller, and a protection unit that enables local decision-making while reducing dependency on communications. This paper proposes an intelligent control and fault-diagnosis framework based on fog computing for power converters in DES applications. It combines distributed energy resources, a grid interface, an energy storage system (ESS), a sensing unit, a fog-computing node, an intelligent controller and a protection unit that is able to allow local decision-making while reducing dependency on communications. At the fog layer, high-bandwidth converter variables, as well as harmonic components, are preprocessed for filtering and feature extraction to perform adaptive control and fault classification. The mathematical model contains statespace converter dynamics, DC-link voltage regulation, power balance, efficiency estimation and Total Harmonic Distortion (THD) measures; ESS charge dynamics, including duty-ratio control and power-loss know-how; and residual-based diagnosis of faults. Performance analysis shows good DC-link voltage stabilization, with a minimum settling time of 0.078 s, lower harmonic distortion, a minimum fault-detection delay of 0.015 s and fog-node command-update indices within the range from 0.91 to 0.97 through different operating modes. The proposed framework is developed to support low-latency monitoring, adaptive control, robust fault diagnosis and enhanced power quality in converter-fed DES environments.