US-12620797-B2 - Computer-implemented method of power line protection, intelligent electronic device and electric power system

US12620797B2US 12620797 B2US12620797 B2US 12620797B2US-12620797-B2

Abstract

Techniques for distance protection of a transmission line include determining a fault inception time from a voltage and/or current, determining rate of change sample values indicative of a rate of change of the voltage and/or of a rate of change of the current for at least one sample time that is dependent on the fault inception time, and using the rate of change sample values to generate a phase classifier for fault classification of a zone classifier for faulted zone identification.

Inventors

OD NAIDU
Dinesh Patil
Neethu GEORGE
Vedanta Pradhan
Suresh Maturu

Assignees

HITACHI ENERGY LTD

Dates

Publication Date: 20260505
Application Date: 20220422
Priority Date: 20210611

Claims (18)

1 . A computer-implemented method comprising: computing features for use in generating a decision logic operative to generate a signal operative for use with a power system, wherein computing the features comprises determining a fault inception time from at least one electric characteristics of data sampled for a given type of power system, and determining rate of change values indicative of a rate of change of the at least one electric characteristics for at least one sample time, the at least one sample time being dependent on the fault inception time; using the features to generate the decision logic, wherein using the features to generate the decision logic comprises performing a machine learning (ML) model training using the computed features to generate one or both of a phase classifier and a zone classifier, wherein the phase classifier receives rates of change of voltages for three phases and rates of change of currents for the three phases as inputs and outputs a fault type, and/or wherein the zone classifier receives rates of change of voltages for three phases and rates of change of currents for the three phases as inputs and outputs a zone; and deploying the decision logic to an intelligent electronic device (IED) to generate the signal.
2 . The computer-implemented method of claim 1 , wherein the ML model has an input layer that receives the rate of change values.
3 . The computer-implemented method of claim 1 , wherein the decision logic is or comprises at least one of: a protection function; the phase classifier for fault classification; or the zone classifier for faulted zone identification.
4 . The computer-implemented method of claim 1 , wherein the signal is at least one of: a circuit breaker control signal; a switch control signal; an alarm; a warning; status information; or output for outputting via a human machine interface (HMI).
5 . The computer-implemented method of claim 1 , wherein determining the rate of change values comprises, for each phase, determining a filtered voltage, determining a filtered current, and calculating the rate of change values from the filtered voltage and from the filtered current.
6 . The computer-implemented method of claim 5 , wherein the filtered voltage is determined by averaging a number N>1 of sample values of the voltage and the filtered current is determined by averaging a number N>1 of sample values of the current.
7 . The computer-implemented method of claim 6 , wherein calculating the rate of change values comprises determining a voltage difference between a sample value of the filtered voltage at a time k and a sample value of the filtered voltage at a time k−N, and a current difference between a sample value of the filtered current at a time k and a sample value of the filtered current at a time k−N, and wherein the rate of change values for the voltage are determined as a change of the voltage difference between a sample time and a previous sample time, and the rate of change values for the current are determined as a change of the current difference between the sample time and the previous sample time.
8 . The computer-implemented method of claim 1 , wherein deploying the decision logic comprises deploying the phase classifier and/or the zone classifier for execution by the IED for distance protection.
9 . The computer-implemented method of claim 1 , wherein generating the decision logic comprises using the computed features to generate one or both of the phase classifier and the zone classifier.
10 . The computer-implemented method of claim 9 , wherein using the computed features to generate the decision logic comprises generating the phase classifier and generating the zone classifier using an ensemble machine learning (ML) method, and/or wherein generating the phase classifier and generating the zone classifier comprises a random forest training using the computed features to generate a random forest, and/or wherein using the rate of change values to generate one or both of the phase classifier and the zone classifier comprises training a first machine learning model, using the computed features, to generate the phase classifier, and training a second machine learning model, using the computed features, to generate the zone classifier.
11 . The computer-implemented method of claim 1 , wherein the phase classifier and/or zone classifier are generated in a self-setting manner.
12 . The computer-implemented method of claim 1 , wherein the phase classifier and/or zone classifier are continually updated during field operation.
13 . The computer-implemented method of claim 1 , wherein the decision logic is a decision logic for transmission line protection, and/or wherein generating the phase classifier and/or zone classifier is performed using a dataset including data for several distinct source to line impedances.
14 . An intelligent electronic device (IED) comprising: an interface to receive a voltage and/or current for at least one phase of a power transmission line; wherein the IED is operative to determine a rate of change of at least one electric characteristics and to input the determined rate of change of the at least one electric characteristics as input into a decision logic that has been generated and deployed to the IED according to the computer-implemented method of claim 1 .
15 . An electric power system, comprising: a power transmission line; and the IED of claim 14 operative to perform a distance protection function for the power transmission line.
16 . A computer-implemented method comprising: computing features for use in generating a decision logic operative to generate a signal operative for use with a power system, wherein computing the features comprises determining a fault inception time from at least one electric characteristics of data sampled for a given type of power system, wherein the fault inception time is identified as earlier one of a time at which a modulus of a deviation of a voltage from a sliding window moving average of the voltage reaches or exceeds a phase-specific voltage threshold and a time at which a modulus of a deviation of a current from a sliding window moving average of the current reaches or exceeds a phase-specific current threshold, and determining rate of change values indicative of a rate of change of the at least one electric characteristics for at least one sample time, the at least one sample time being dependent on the fault inception time; using the features to generate the decision logic; and deploying the decision logic to an intelligent electronic device (IED) to generate the signal.
17 . A computer-implemented method comprising: computing features for use in generating a decision logic operative to generate a signal operative for use with a power system, wherein computing the features comprises determining a fault inception time from at least one electric characteristics of data sampled for a given type of power system, wherein determining the fault inception time comprises determining a sliding window moving average and comparing a modulus of a deviation of the at least one electric characteristic from the sliding window moving average to at least one threshold, and determining rate of change values indicative of a rate of change of the at least one electric characteristics for at least one sample time, the at least one sample time being dependent on the fault inception time; using the features to generate the decision logic; and deploying the decision logic to an intelligent electronic device (IED) to generate the signal.
18 . The computer-implemented method of claim 17 , further including: determining a sliding window standard deviation of the at least one electric characteristic, wherein the at least one threshold depends on the sliding window standard deviation, and/or wherein the fault inception time is identified as earlier one of a time at which a modulus of a deviation of a voltage from a sliding window moving average of the voltage reaches or exceeds a phase-specific voltage threshold and a time at which a modulus of a deviation of a current from a sliding window moving average of the current reaches or exceeds a phase-specific current threshold.

Description

CROSS-REFERENCE TO RELATED APPLICATIONS The present application is a national stage entry of International Application No. PCT/EP2022/060654, filed on Apr. 22, 2022, which claims priority to Indian Application No. 202141018766, filed on Apr. 23, 2021, and European Application No. 21179002.7, filed on Jun. 11, 2021, which are all hereby incorporated herein by reference as if set forth in full. FIELD OF THE INVENTION The invention relates to power transmission systems and devices and methods for protecting power transmission lines. The invention relates in particular to methods and devices that are operative to perform at least one of fault classification, zone classification, and/or distance protection. BACKGROUND OF THE INVENTION Electric power grids are undergoing a significant change in generation mix, from synchronous AC rotating machines to Inverter-Based Resource (IBR) technologies. A consequence of this trend is more deployment of renewables both at transmission and distribution networks. In addition to environmental benefits, introduction of renewable energy sources changes the operation of power systems: e.g. reduced inertia, less stability margins, and increased unpredictability. Since most of these renewable technologies are inverter interfaced, their behavior under fault conditions is different than conventional rotating machines. This creates new challenges in power system protection. One issue with higher penetration of inverter-based resources such as wind and solar photovoltaic (PV) generation is a reduction in fault current levels and short circuit strength of the power grid. Moreover, the power system stability margin depends on the inertia of the system and it determines the desired speed of the protection scheme. IBR systems often have low inertia and hence lesser stability margins. A reduced stability margin implies lower critical clearing time. Faults must be cleared faster than the critical clearing time or otherwise system may lose transient stability and it leads to power system blackout. A distance relay is a protective device that is designed to provide primary and backup protection of transmission lines. Conventional distance relays are not self-adjusting as and when the system changes. Routine manual adjustments and automated adaptivity may be afforded to some extent in protective relaying. WO 2005/076428 A9 discloses a system and method for detecting high impedance faults (HIF) in electrical power lines using Artificial Neural Network (ANN). The method is wavelet transform based and hence can be computationally intensive for practical applications. A detection rate of 70.83% with a 22.06% false alarm rate may not be a satisfactory performance for a protection application. The use of a spectrum of a 3-cycle window of data may not be desirable for protection application. U.S. Pat. No. 7,720,619 B2 discloses a method for detecting high impedance faults. Artificial intelligence (AI) methods of classification and pattern recognition, such as neural networks, expert systems or decision trees may be used to differentiate a HIF condition from other system conditions, such as switching operations and noisy loads. U.S. Pat. No. 6,405,184 B1 discloses a method for generating fault classification signals which identify faulty loops in the event of a fault. A neural network is used which is trained using input variables simulating faulty loops. Measured values from currents and voltages of loops of the energy supply system are derived and fault classification signals are generated by the trained model in the event of a fault. WO 1995/009465 A1 discloses a method for generating a direction signal which indicates the direction of a short-circuit current. The direction signal indicates whether a short circuit occurred in forward direction from the measuring point. A neural network is used for generating the direction signal which is formed using normalized sampling values of the currents. An artificial neural network (ANN) based approach for three-zone distance protection of transmission lines is presented in A. Feilat and K. Al-Tallaq, “A new approach for distance protection using artificial neural network,” 39th International Universities Power Engineering Conference, 2004. UPEC 2004, Bristol, U K, 2004, pp. 473-477 Vol. 1. The technique handles only fault detection and classification aspects of distance protection. The input features of the neural network are the fundamental frequency voltage and current magnitudes extracted by Discrete Fourier Transform. Hence, the solution is not expected to improve the speed of the existing phasor-based distance protection algorithms. J. R. de Carvalho et al., “Development of detection and classification stages for a new distance protection approach based on cumulants and neural networks,” 2011 IEEE Power and Energy Society General Meeting, Detroit, MI, USA, 2011, pp. 1-7 suggests the use of Artificial Intelligence (AI) for distance relaying. Fault detect