Data-Mining-Based Intelligent Differential Relaying for Transmission Lines Including UPFC and Wind Farms

Area of Science:

• Electrical Engineering
• Power Systems Protection

Background:

• Traditional differential relaying schemes face challenges with complex power systems.
• Integration of flexible AC transmission system (FACTS) devices and renewable energy sources complicates fault detection.

Purpose of the Study:

• To develop a data-mining-based intelligent differential relaying scheme for enhanced transmission line protection.
• To ensure reliable fault detection in power systems with unified power flow controllers (UPFC) and wind farms.

Main Methods:

• Phasor estimation using extended Kalman filter phasor measurement unit.
• Extraction of 21 features and differential features at line ends.
• Application of a decision tree (DT) data-mining model for relaying decisions.

Main Results:

• The proposed scheme demonstrated high reliability and accuracy in simulations and real-time digital simulator tests.
• The decision tree-based scheme achieved a rapid response time of 2.25 cycles post-fault inception.
• Successful testing on single and double-circuit lines with UPFC and wind farm integration under varied operating conditions.

Conclusions:

• The data-mining-based intelligent differential relaying scheme offers a robust solution for modern transmission line protection.
• The decision tree approach effectively handles the complexities introduced by FACTS devices and wind farms.
• The scheme provides a reliable and fast protective solution for power grids.