Intelligent Image Recognition Based Automatic Modeling System for Power Equipment Equivalent Circuits
ID：16
Submission ID：17 View Protection：ATTENDEE
Updated Time：2025-10-11 22:12:27 Hits：52
Poster Presentation
Start Time：2025-11-09 09:01 （Asia/Shanghai）
Duration：1min
Session：[P] Poster presentation » [P6] 6.AI-driven technology
No files
Abstract
Equivalent circuit modeling of power equipment is a crucial foundation for power system analysis, fault diagnosis, and monitoring equipment health. Traditional circuit modeling methods rely on manual identification and calculations. These approaches are inefficient and prone to errors when handling complex circuits. This paper proposes an automatic modeling and simulation system for equivalent circuits of power equipment based on intelligent image recognition. The system uses deep learning technology for automatic component recognition in circuit diagrams and applies image processing algorithms to detect node connection relationships. It automatically establishes Kirchhoff’s Current Law (KCL) and Kirchhoff’s Voltage Law (KVL) equation systems—standard electrical circuit principles to describe current and voltage flows—and then solves circuit frequency-domain characteristics using state-space equations. The system features a triple verification framework. It cross-validates the results from intelligent recognition and automatic calculations with simulation results from manually constructed Multisim models and automatically generated Simulink simulation models via a JSON (JavaScript Object Notation) interface. Using the equivalent circuit of a transformer winding as an example, experimental results show that the system can accurately identify repetitive unit structures, extract circuit features, and establish mathematical models. The high-frequency impedance spectra calculated by the intelligent method agree with results from manual simulation in Multisim and automatic modeling in Simulink. This validates the intelligent modeling method's effectiveness and accuracy. The system automates circuit modeling, providing an efficient and intelligent solution for a wide range of power equipment.
Keywords
Circuit image recognition, Equivalent circuit modeling, State space analysis
Speaker
Comment submit