top of page
Search

# Fault Classification Location and Detection in Power System using ANFIS

Fault Classification Location and Detection in Power System using ANFIS

IntroductionÂ

We delve into the advanced topic of fault detection, classification, and location in power systems using the Adaptive Neuro-Fuzzy Inference System (ANFIS). ANFIS combines the neural network and fuzzy logic system to enhance fault detection capabilities in power systems.

System OverviewÂ The power system under consideration comprises three buses:

• Bus 1: Source feeder, 11 kV, 30 MVA

• Bus 2: Source feeder, 11 kV, 30 MVA

• Bus 3: Connected to Bus 1 via a 10 km distribution line

Fault ScenariosÂ Various fault scenarios are simulated in the system:

• Underground fault

• Double line-to-ground fault

• Line-to-line fault

• Bus faults

Data Collection for ANFISÂ To train the ANFIS model effectively, the following data is collected:

• RMS voltage of Bus 1

• RMS current of Bus 1

• Zero sequence voltage of Bus 1

• Zero sequence current of Bus 1

Program for Data CollectionÂ A program is used to collect and preprocess the data for both normal and fault scenarios. This data includes specifying fault locations at different distances along the distribution line.

ANFIS Model DesignÂ The ANFIS model is designed to perform two main tasks:

1. Fault Classification: Determines the type of fault (e.g., single line-to-ground, double line-to-ground) based on collected data.

2. Fault Location: Determines the precise location of the fault along the distribution line.

Classification ModelÂ The ANFIS model for fault classification takes inputs of RMS voltage, RMS current, zero sequence voltage, and zero sequence current of Bus 1. It outputs a binary classification identifying the type of fault detected.

Location ModelÂ The ANFIS model for fault location takes the RMS voltage of Bus 1 as input and outputs the exact distance (in kilometers) from Bus 1 where the fault is detected.

Simulation and ResultsÂ The ANFIS models are simulated using different fault scenarios:

• Results show accurate detection and classification of faults such as double line-to-ground faults, line-to-line faults, etc.

• Fault location accuracy is demonstrated by pinpointing the exact kilometer distance where the fault occurred.

ConclusionÂ In conclusion, the ANFIS-based approach proves effective for fault detection, classification, and location in power systems. It combines the strengths of neural networks and fuzzy logic to enhance the reliability and efficiency of fault management in electrical grids.