Abstract:Multiple features fusion through artificial intelligence holds the potential to enhance the reliability of transformer protection. However, the existing methods encounter challenges due to the data scarcity of inrush current and internal fault. To address this, the paper proposes a blocking method of non-fault scenarios for differential protection, employing particle swarm optimization-support vector data description (PSO-SVDD). The approach utilizes several geometric features extracted from the differential current-excitation voltage curve (DEC) of normal operation to fully characterize similarities among non-fault scenarios, such as inrush current, and CT saturation caused by external fault. By exclusively utilizing normal operation data to represent all non-fault scenarios, the challenge posed by the scarcity of non-fault data is overcome. PSO-SVDD, selected as a one-class classification algorithm, is trained using geometric features from normal operation data, thereby avoiding the limited availability of fault data. The region inside the SVDD hypersphere serves as the blocking domain of differential protection. Additionally, a fine-tuning method of the feature boundary is introduced to enhance the robustness of the blocking domain by the small sample of scarce scenarios. Before the model is applied to unseen transformers, offline detection using its normal operation data is conducted to evaluate the performance of the blocking domain in. In case of misidentification, the misidentified sample is utilized as the support vector for fine-tuning the feature boundary. In practical application, if iron core saturation or fault scenarios are misidentified, the feature boundary is fine-tuned offline using the misidentified sample with the similar method above. PSCAD simulations and dynamic simulation experiments validate the superior performance of the proposed protection method through the comparison with several existing methods.

Novel Power Transformer Differential Protection Scheme Based on Improved Short-Window Algorithm

The Principle of Transformer Differential Protection Based on Wave Comparison

Anti-saturation algorithm in differential protection based on the phaselet

A Novel Transformer Protection Scheme Based on Equivalent Excitation Impedance Characteristics

A Novel Algorithm Applied in the Transient Saturation of Current Transformer

A Novel Method for Current Transformer Saturation Countermeasure Based on Comparing Two Currents' Polarities

A Novel Differential Protection Scheme for the Generator

A NEW METHOD TO IDENTIFY INRUSH CURRENT BY PHASOR COMPARISON

A Data-Aided Power Transformer Differential Protection Without Inrush Blocking Module

A Novel Scheme of Generator Differential Protection

A Novel Adaptive Single-Phase Reclosure Scheme Using Dual-Window Transient Energy Ratio And Mathematical Morphology

Steady-state Analysis of Internal Short Circuits of Multi-winding Power Transformer Part Ⅱ Experiment Verification and Sensitivity Analysis of Differential Relay Protection

A Novel Transient Saturation Detection Method for P-Class Current Transformer Using Improved Hausdorff Distance Algorithm

An Adaptive Variable Window Algorithm for Transformer Differential Protection Relay

THE INFLULENCE AND COUNTERMEASURE TO TRANSFORMER DIFFERENTIAL PROTECTION OF CT PARTIAL TRANSIENT SATURATION CAUSED BY EXTERNAL FAULT REMOVAL

Blocking Method with PSO-SVDD for Differential Protection of Power Transformer

Research on differential protection algorithm about special phase-shifting angle rectifier transformer

A method of adaptive transformer turn-turn fault protection based on power loss

Identification of Cross-Country Fault of Power Transformer for Fast Unblocking of Differential Protection

A New Protection Scheme for DC Substation System Based on Single-Ended Measurement

A Novel Single-Phase Adaptive Reclosure Scheme For Transmission Lines With Shunt Reactors