Study on ice phase change prediction model for transmission lines based on multiphysics coupling

Area of Science:

• Meteorology and atmospheric science
• Electrical engineering
• Materials science

Background:

• Predicting ice phase changes on transmission lines is crucial for preventing failures.
• Existing models lack accuracy due to ignoring interactions between meteorological factors and physical fields.

Purpose of the Study:

• To develop a multi-physical field coupling model for accurate ice phase change prediction on transmission lines.
• To enhance the reliability of prediction algorithms through kernel function selection for Support Vector Machines (SVM).

Main Methods:

• Developed a multi-physical field coupling model with specific boundary conditions.
• Simulated diverse meteorological conditions to create a 3D geometric model of ice formation.
• Validated the Support Vector Machine (SVM) algorithm using various kernel functions.
• Conducted prediction experiments and compared results with field data and SVM predictions.

Main Results:

• The multi-physical field coupling model achieved an average prediction accuracy of 98.12%.
• The model demonstrated an average precision of 98.54% in ice phase change prediction.
• Results significantly outperformed traditional prediction methods.

Conclusions:

• The developed model offers a substantial improvement in predicting ice phase changes on transmission lines.
• This high-precision model provides valuable decision support for early warning and protection against ice disasters.
• The findings are highly relevant for practical engineering applications in power systems.