Effective feature set-based high impedance fault detection
Effective feature set-based high impedance fault (HIF) detection is provided. Systems, methods and devices described herein present a systematic design of power feature extraction for HIF detection and classification. For example, power features associated with HIF events are extracted according to when a fault happens, how long it lasts, and the magnitude of the fault. Complementary power expert information is also integrated into feature pools. In another aspect, a ranking procedure is deployed in a feature pool for balancing information gain and complexity in order to avoid over-fitting of features. In aspects described herein, a logic-based HIF detector implements HIF feature extraction. To determine when an HIF occurs, the HIF detector calculates different quantities, such as active power and reactive power, based on a voltage and current time series, and uses the derivative of these quantities to tell when there is a potential change due to HIF.
1. A high impedance fault (HIF) detector, comprising:
a system characteristic averager configured to store instantaneous power characteristic values and provide averaged power characteristic values of the instantaneous power characteristic values;
a decision circuit configured to determine occurrence of an HIF based on the instantaneous power characteristic values and the averaged power characteristic values; and
a Kalman filter configured to provide a first portion of the instantaneous power characteristic values to the system characteristic averager after filtering a received voltage signal and a received current signal, wherein the first portion of the instantaneous power characteristic values comprises at least one of an estimated in-phase component of a third harmonic of the received voltage signal (KF V a cos H V3 , KF V b cos H V3 , or KF V c cos H V3 ) or an estimated in-quadrature component of a third harmonic of the received voltage signal (KF V a sin H V3 , KF V b sin H V3 , or KF V c sin H V3 );
wherein KF is a Kalman Filter coefficient, v a is defined by a voltage of phase A, v b is defined by a voltage of phase B, and v c is defined by a voltage of phase C.
2. The HIF detector of claim 1 , further comprising a comparison circuit configured to compare the instantaneous power characteristic values with the averaged power characteristic values to provide comparison values to the decision circuit.
3. The HIF detector of claim 1 , further comprising a discrete Fourier transform (DFT) circuit configured to provide a second portion of the instantaneous power characteristic values to the system characteristic averager based on a DFT of the received voltage signal and the received current signal.
4. The HIF detector of claim 3 , further comprising a comparison circuit configured to:
receive the instantaneous power characteristic values from the Kalman filter and the DFT circuit;
receive the averaged power characteristic values from the system characteristic averager; and
compare the instantaneous power characteristic values with the averaged power characteristic values to provide comparison values to the decision circuit.
5. The HIF detector of claim 1 , further comprising a discrete Fourier transform (DFT) circuit configured to provide at least a portion of the instantaneous power characteristic values to the system characteristic averager based on a DFT of a received voltage signal and a received current signal.
6. The HIF detector of claim 5 , wherein the at least the portion of the instantaneous power characteristic values provided by the DFT circuit to the system characteristic averager comprises a negative sequence voltage (V 2 ) and a negative sequence current (I 2 ).
7. The HIF detector of claim 6 , wherein the at least the portion of the instantaneous power characteristic values provided by the DFT circuit further comprises an angle difference between the negative sequence voltage and a zero sequence voltage (θ V 2 −θ V 0 ) and an angle difference between the negative sequence current and a zero sequence current (θ I 2 −θ I 0 ).
8. The HIF detector of claim 1 , wherein the instantaneous power characteristic values comprise an effective feature set.
9. The HIF detector of claim 8 , wherein the effective feature set comprises at least one of: a negative sequence voltage (V 2 ), a negative sequence current (I 2 ), an angle difference between the negative sequence voltage and a zero sequence voltage (θ V 2 −θ V 0 ), an angle difference between the negative sequence current and a zero sequence current (θ I 2 −θ I 0 ), or a harmonic of a received voltage signal (KF V cos H V3 or KF V sin H V3 ).
10. A method for detecting a high impedance fault (HIF), comprising:
receiving power measurements from a power distribution system;
extracting an angle difference between a negative sequence voltage and a zero sequence voltage (θ V 2 −θ V 0 ) from the power measurements; and
determining occurrence of an HIF based on the angle difference θ V 2 −θ V 0 .
11. The method of claim 10 , wherein extracting the angle difference θ V 2 −θ V c comprises performing a discrete Fourier transform (DFT) of a received voltage signal.
12. The method of claim 11 , further comprising averaging the angle difference θ V 2 −θ V 0 to produce an averaged angle difference θ V 2 −θ V 0 , wherein the HIF is determined based on comparing the averaged angle difference θ V 2 −θ V 0 with the angle difference θ V 2 −θ V 0 .
13. The method of claim 10 , further comprising:
extracting an effective feature set (EFS) comprising the angle difference θ V 2 −θ V 0 and at least one of the negative sequence voltage (V 2 ), a negative sequence current (I 2 ), an angle difference between the negative sequence current and a zero sequence current (θ I 2 −θ I 0 ), or a harmonic of a received voltage signal (KF V cos H V3 or KF V sin H V3 ); and
determining the occurrence of the HIF based on the EFS.
14. The method of claim 13 , wherein extracting the EFS comprises performing a set of discrete Fourier transforms (DFTs) on a received voltage signal and a received current signal.
15. The method of claim 14 , wherein extracting the EFS further comprises Kalman filtering the received voltage signal and the received current signal.
16. The method of claim 13 , further comprising averaging the EFS to produce an averaged EFS, wherein the occurrence of the HIF is determined based on comparing the averaged EFS with the EFS.
17. A protective relay for a power distribution line, the protective relay comprising:
a power coupler; and
a high impedance fault (HIF) detector connected to the power coupler, comprising:
feature extraction logic configured to extract instantaneous power characteristic values from a signal of the power coupler, the feature extraction logic comprising a Kalman filter that is configured to provide a first portion of the instantaneous power characteristic values after filtering a received voltage signal and a received current signal, the first portion of the instantaneous power characteristic values comprising at least one of an estimated in-phase component of a third harmonic of the received voltage signal (KF V a cos H V3 , KF V b cos H V3 , or KF V c cos H V3 ) or an estimated in-quadrature component of a third harmonic of the received voltage signal (KF V a sin H V3 , KF V b sin H V3 , or KF V c sin H V3 );
wherein KF is a Kalman Filter coefficient, v a is defined by a voltage of phase A, v b is defined by a voltage of phase B, and v c is defined by a voltage of phase C;
a system characteristic averager configured to provide averaged power characteristic values from the instantaneous power characteristic values; and
a decision circuit configured to determine occurrence of an HIF based on a comparison of the instantaneous power characteristic values with the averaged power characteristic values using a trained HIF model.