High-Voltage Cable Buffer Layer Ablation Fault Identification Based on Artificial Intelligence and Frequency Domain...

Area of Science:

• Electrical Engineering
• Materials Science
• Artificial Intelligence

Background:

• High-voltage cable buffer layer ablation faults are increasingly frequent and dangerous.
• Existing fault detection methods have limitations.
• Prompt detection is crucial to prevent cable breakdowns and power system disruptions.

Purpose of the Study:

• To develop an advanced method for identifying buffer layer ablation faults in high-voltage cables.
• To overcome limitations of current fault detection techniques.
• To improve the safety and reliability of power cable operations.

Main Methods:

• Derivation of a mathematical model for cable input impedance with buffer layer ablation faults using distributed parameter models and frequency domain impedance spectroscopy.
• Simulation of input impedance spectroscopy for normal, ablation, aging, and inductive faults to differentiate inductive and capacitive faults.
• Utilizing frequency domain amplitude spectroscopy data for buffer layer ablation and local aging faults to train and validate a neural network model (MLP).

Main Results:

• The study successfully simulated input impedance spectroscopy to distinguish between different fault types.
• A neural network model was trained and validated using frequency domain amplitude spectroscopy data.
• The MLP neural network demonstrated superiority in identifying cable faults compared to other models.

Conclusions:

• The proposed method effectively identifies buffer layer ablation and local aging faults in high-voltage cables.
• The MLP neural network proves effective for cable fault identification.
• The study confirms the practical effectiveness of the combined frequency domain impedance spectroscopy and AI approach for enhancing power cable safety.