top of page
Search

High Impedance fault detection classification using Neural Network

High Impedance fault detection classification using Neural Network

Introduction:

We'll explore the detection of high-impedance faults in distribution feeders using MATLAB and Neural Networks. High-impedance faults occur when a conductor comes into contact with a tree or the ground, creating a high-impedance path and potentially leading to fires. This tutorial demonstrates how to create a simulation model of a distribution system, induce high-impedance faults, and employ a neural network to detect and classify these faults.

Creating High Impedance Fault Model:

1. Understanding High Impedance Faults: High impedance faults occur when a conductor touches the ground or another object, creating a high resistance path and potential fire hazards.

2. Simulation Model in MATLAB: Using MATLAB, a simulation model is developed based on a reference paper, introducing high impedance faults in a distribution system. The model includes sending and receiving ends, transformers, feeder lines, and high impedance faults in different phases.

3. Data Collection for Neural Network: Data is collected for different fault scenarios, including phases A, B, C, and normal cases. Input data includes sending end voltage and current, and target data is set to classify faults (one-hot encoding).

Training Neural Network:

1. Data Loading in MATLAB: Input and target data are loaded into MATLAB, and the Neural Network Fitting app is used for training.

2. Neural Network Training: The neural network is trained using the collected data, creating a model capable of detecting and classifying high impedance faults.

Implementing Detection Model:

1. Integration of Neural Network in Final Model: The trained neural network is integrated into the final simulation model. Input parameters include sending end voltage and RMS current.

2. Fault Detection and Classification: The model is simulated with induced faults, and the neural network successfully detects and classifies high impedance faults in different phases (A, B, C).

Conclusion:

This tutorial demonstrated the process of creating a MATLAB simulation model to detect high impedance faults in distribution systems using neural networks. The integration of neural networks enhances the system's capability to identify and classify faults accurately.