Classification of offshore wind grid-connected power quality disturbances based on fast S-transform and CPO-optimized...

Area of Science:

• Electrical Engineering
• Renewable Energy Systems
• Signal Processing

Background:

• Large-scale integration of offshore wind power presents significant power quality challenges.
• Accurate detection and classification of power quality disturbances are crucial for grid stability.

Purpose of the Study:

• To develop and validate a novel algorithm for detecting and classifying power quality disturbances caused by offshore wind power integration.
• To enhance the accuracy and efficiency of power quality disturbance analysis in power grids.

Main Methods:

• Analysis of offshore wind power grid-connected disturbance mechanisms and waveform characteristics.
• Feature extraction and time-frequency diagram generation using Fast S-transform.
• Optimization of Convolutional Neural Network (CNN) hyperparameters using the Crested Porcupine Optimizer (CPO).
• Classification of power quality disturbances using the CPO-optimized CNN model.

Main Results:

• The proposed CPO-CNN model achieved improved classification accuracy compared to standard CNN methods.
• Simulation results demonstrated the effectiveness of the algorithm in identifying various power quality disturbance signals.
• The method successfully performed feature extraction, selection, and classification of time-frequency diagrams.

Conclusions:

• The developed Fast S-transform and CPO-optimized CNN algorithm is effective for power quality disturbance detection and classification.
• The approach offers a significant improvement in classification accuracy, aiding in power quality assessment and control.
• This method provides a robust solution for managing power quality issues arising from offshore wind energy integration.