Axis Orbit Recognition of the Hydropower Unit Based on Feature Combination and Feature Selection

Area of Science:

• Mechanical Engineering
• Signal Processing
• Fault Diagnosis

Background:

• Axis-orbit recognition is crucial for identifying faults in hydropower units.
• Existing methods often suffer from low accuracy, poor robustness, and inefficiency.
• There is a need for advanced techniques to improve axis-orbit recognition.

Purpose of the Study:

• To develop a novel axis-orbit recognition method for hydropower units.
• To address the limitations of existing methods in terms of accuracy, robustness, and efficiency.
• To enhance the fault diagnosis capabilities of hydropower units.

Main Methods:

• Extraction and combination of contour, moment, and geometric features from axis orbit data.
• Application of Random Forest (RF)-Fisher for feature dimensionality reduction.
• Optimization of Support Vector Machine (SVM) using the Gravitational Search Algorithm (GSA) for classification.

Main Results:

• The proposed method demonstrates high recognition accuracy for axis-orbit data.
• The approach exhibits excellent robustness in identifying axis-orbit patterns.
• Significant improvements in recognition efficiency were achieved compared to existing methods.

Conclusions:

• The developed axis-orbit recognition method offers a robust and efficient solution for hydropower unit fault diagnosis.
• The combination of feature engineering, selection, and optimized SVM classification is effective.
• This method provides a reliable tool for ensuring the operational integrity of hydropower systems.