A Fault Location Method for 35kV Distribution Network Based on Natural Oscillation Frequency

Abstract

This invention relates to the field of power system fault detection technology, specifically a fault location method for 35kV distribution networks based on natural oscillation frequencies. The invention involves delaying the arc for a set period after fault isolation to extinguish it; connecting a pre-charged backup capacitor to the fault section via a dual-controllable switch to form an underdamped oscillation circuit; synchronously acquiring the decaying oscillation voltage waveform across the capacitor; using the Prony algorithm to establish an exponentially decaying sine fitting model to extract characteristic angular frequencies; and calculating the fault distance based on the functional relationship between the cable unit length parameter and the fault distance using a reverse mapping function. This invention, by combining an active excitation strategy and the Prony algorithm, improves fault location accuracy and response speed, and enhances anti-interference capabilities.

Description

Technical Field

[0001] This invention relates to the field of power system fault detection technology, specifically a fault location method for 35kV distribution networks based on natural oscillation frequency. Background Technology

[0002] As a critical link in the power system's transmission and transformation, the 35kV AC distribution network is prone to frequent faults, which can easily lead to equipment damage and large-scale power outages. Efficient and accurate fault location is of great significance for quickly restoring power supply and minimizing economic losses.

[0003] Existing fault location technologies have significant limitations: the traditional traveling wave method requires precise capture of the wavefront arrival time, but in 35kV distribution networks, high-frequency traveling waves are severely attenuated by line sag and branch structures, and arc reignition generates multiple reflected waves, causing confusion in wavefront identification. Furthermore, wavefront detection errors are large under noise interference. The impedance method relies on steady-state measurement and has a slow response speed; the high-frequency impedance method is easily affected by noise from power electronic switches. Existing technologies suffer from low location accuracy and slow speed in underground cable environments, complex branch networks, and high-impedance fault scenarios. Moreover, the traveling wave method is severely affected by reflected wave interference in short-distance distribution lines. Traditional methods generally suffer from wavefront capture failure or steady-state measurement delays, failing to meet the requirements for rapid protection and fault recovery. Summary of the Invention

[0004] This invention provides a fault location method for 35kV distribution networks based on natural oscillation frequency, which solves the problems of low positioning accuracy, slow response speed and weak anti-interference ability of traditional methods.

[0005] To achieve the above objectives, the present invention provides the following technical solution:

[0006] A fault location method for 35kV distribution networks based on natural oscillation frequencies includes:

[0007] S100: After detecting a short-circuit fault and isolating the fault section through the circuit breaker, a set delay is made to extinguish the arc at the fault point and attenuate the residual energy.

[0008] S200: After the fault section is isolated, the charging and discharging path is switched through a dual-controllable switch. The backup capacitor, which is pre-charged to a safe voltage, is connected to both ends of the cable in the isolated fault section. The fault circuit is actively excited to form an underdamped oscillating circuit that includes the fault circuit's distributed resistance, distributed inductance, and backup capacitor.

[0009] S300: Synchronously acquire the attenuated oscillating voltage waveform across the two ends of the backup capacitor during the discharge process;

[0010] S400: For the damped oscillation voltage waveform generated by the active excitation, the Prony algorithm is used to establish an exponentially decaying sine fitting model for the damped oscillation voltage waveform, and the characteristic angular frequency representing the oscillation characteristics of the loop is extracted by the least squares method.

[0011] S500: The total resistance of the fault circuit is expressed as the product of the resistance per unit length of the cable and twice the fault distance, and the total inductance of the fault circuit is expressed as the product of the inductance per unit length of the cable and twice the fault distance. The fault distance is calculated through a reverse mapping function based on the characteristic angular frequency, the cable per unit length parameter, and the spare capacitance value.

[0012] As a preferred embodiment of the present invention, the delay setting duration is 0.5 to 2 seconds.

[0013] As a preferred embodiment of the present invention, S200 includes:

[0014] During normal operation, the backup capacitor is connected to the charging power supply via the first controllable switch and its terminal voltage is maintained at a preset safe value.

[0015] After isolating the faulty section, disconnect the first controllable switch and close the second controllable switch to connect the backup capacitor to both ends of the cable in the faulty section.

[0016] As a preferred embodiment of the present invention, the preset safety value is 1% to 10% of the peak value of the rated phase voltage of the system.

[0017] As a preferred embodiment of the present invention, the acquisition requirements of S300 include:

[0018] The sampling frequency is not less than 10 kHz;

[0019] The sampling window duration must cover at least two complete oscillation cycles;

[0020] The analog-to-digital conversion resolution is no less than 16 bits.

[0021] As a preferred technical solution of the present invention, the S400 uses the Prony algorithm to represent the acquired decaying oscillation voltage waveform as the product of an exponential decay term and a sinusoidal oscillation term, and extracts the characteristic angular frequency by the least squares method.

[0022] As a preferred embodiment of the present invention, when the fitting residual of the Prony algorithm is greater than 5%, the sampling window is automatically extended and the parameters are refitted.

[0023] As a preferred embodiment of the present invention, the calculation of fault distance using the reverse mapping function includes:

[0024] The total resistance of the faulted circuit is expressed as the product of the resistance per unit length of the cable and twice the fault distance;

[0025] The total inductance of the faulted circuit is expressed as the product of the inductance per unit length of the cable and twice the fault distance;

[0026] Based on the oscillation characteristic relationship between the characteristic angular frequency and the total circuit resistance, total inductance, and spare capacitor values, a functional expression for the characteristic angular frequency with respect to the fault distance is established.

[0027] The attenuation coefficient correction calculation is introduced, and the fault distance is obtained by solving the function expression.

[0028] The present invention also proposes a computer device, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to implement the above-described method for fault location in a 35kV distribution network based on a natural oscillation frequency.

[0029] The present invention also proposes a readable storage medium storing a computer program, which is executed by a processor to provide the above-described method for fault location in a 35kV distribution network based on natural oscillation frequency.

[0030] The beneficial effects of this invention are:

[0031] 1. This invention employs a two-step decoupling architecture combined with an active excitation strategy, overcoming the technical limitations of traditional passive measurement. After isolating the faulty section, this invention actively injects controllable energy into the faulty circuit via a backup capacitor to excite a damped oscillating response, obtaining transient data suitable for analysis. The two-step decoupling ensures precise positioning is performed only within the isolated linear section, reducing interference from multiple reflections in complex branch networks and improving positioning accuracy.

[0032] 2. This invention applies the Prony algorithm to the backup capacitor excitation-fault loop response scenario, establishing a direct mapping relationship between characteristic angular frequency and fault distance. The extracted characteristic angular frequency directly reflects the fault loop inductance parameters and thus the fault distance, avoiding the accumulation of errors in wavefront acquisition, wave velocity calibration, and other processes.

[0033] 3. The sampling frequency requirement of this invention is much lower than that of the traveling wave method, reducing hardware costs. The pre-charge voltage of the backup capacitor is a low percentage of the system's rated voltage, ensuring operational safety. The method has extremely low sensitivity to transition resistance because it utilizes the inherent characteristics of dynamic oscillation rather than static impedance measurement, a significant advantage over traditional impedance methods. This characteristic makes this method applicable to various distribution network scenarios, including urban underground cables, industrial parks, and mountainous overhead lines, and it is particularly superior to traditional methods in locating high-resistance grounding faults. Attached Figure Description

[0034] The accompanying drawings are provided to further illustrate the invention and form part of the specification. They are used in conjunction with embodiments of the invention to explain the invention and do not constitute a limitation thereof. In the drawings:

[0035] Figure 1 This is a flowchart illustrating a method for fault location in a 35kV distribution network based on natural oscillation frequency, according to the present invention.

[0036] Figure 2 This is a schematic diagram of the structure of the spare capacitor unit of the present invention;

[0037] Figure 3 This is the equivalent circuit model diagram of the fault circuit of the present invention. Detailed Implementation

[0038] The preferred embodiments of the present invention will be described below with reference to the accompanying drawings. It should be understood that the preferred embodiments described herein are for illustration and explanation only and are not intended to limit the present invention.

[0039] Example 1: As Figure 1 As shown, the present invention provides a method for fault location in a 35kV distribution network based on natural oscillation frequency, comprising:

[0040] S100: After detecting a short-circuit fault and isolating the fault section through the circuit breaker, a set delay is made to extinguish the arc at the fault point and attenuate the residual energy.

[0041] Furthermore, the delay setting duration is 0.5 to 2 seconds.

[0042] Specifically, in this embodiment, when a short-circuit fault occurs in the 35kV distribution network, the fault current rises rapidly. The protection device (such as instantaneous overcurrent protection or distance protection) installed in the outgoing switchgear of the substation determines that the fault occurs in the outgoing cable section based on the preset protection settings. Within 20-50 milliseconds after detecting the fault, the protection device issues a trip command to control the circuit breaker to operate and isolate the faulty section from the system.

[0043] After the circuit breaker isolates the faulty section, a continuously burning electric arc may remain at the fault point, and capacitive and inductive energy may remain in the line. To ensure the accuracy and safety of subsequent active excitation operations, a delay waiting period is introduced into the system.

[0044] In this embodiment, the delay duration is set to 1 second. This duration is selected based on the following considerations: For 35kV cable lines, the arc extinguishing time at the fault point is typically within 100-300 milliseconds; the residual energy stored in the line-to-ground capacitance (typically 0.2-0.4 μF / km) is released through ground leakage and surge arresters, with a time constant of approximately 200-500 milliseconds. A 1-second delay is sufficient to ensure that the arc is completely extinguished and the residual energy decays to a level that does not affect the measurement (typically requiring the residual voltage to be less than 5% of the precharge voltage).

[0045] The delay timing is implemented by the control unit of the fault location device. After receiving the circuit breaker trip confirmation signal, the control unit starts an internal timer with an accuracy of ±10 milliseconds. After the timer expires, the control unit outputs a trigger signal to initiate the next step of the backup capacitor switching operation.

[0046] By setting a delay range of 0.5 to 2 seconds, the system can be flexibly adjusted according to specific line parameters, fault types, and field conditions. For situations with short cables or high fault impedance, a shorter delay of 0.5 seconds can be selected to accelerate the location speed; for situations with long lines or complex fault conditions, a longer delay of 2 seconds can be selected to ensure that residual energy is fully released. In this embodiment, field testing has verified that a 1-second setting achieves a good balance between accuracy and response speed.

[0047] S200: After the fault section is isolated, the charging and discharging path is switched through a dual-controllable switch. The backup capacitor, which is pre-charged to a safe voltage, is connected to both ends of the cable in the isolated fault section. The fault circuit is actively excited to form an underdamped oscillating circuit that includes the fault circuit's distributed resistance, distributed inductance, and backup capacitor.

[0048] Further, S200 includes:

[0049] During normal operation, the backup capacitor is connected to the charging power supply via the first controllable switch and its terminal voltage is maintained at a preset safe value.

[0050] After isolating the faulty section, disconnect the first controllable switch and close the second controllable switch to connect the backup capacitor to both ends of the cable in the faulty section.

[0051] Furthermore, the preset safety value is 1% to 10% of the peak value of the system's rated phase voltage.

[0052] Specifically, such as Figure 2 As shown, the backup capacitor unit mainly includes: backup capacitors C1 and C2, charging power supply CB, and a first controllable switch. Second controllable switch And a fault location unit. During normal operation... Closing the circuit keeps the backup capacitor charged during fault location. disconnect, Closing the circuit allows the backup capacitor to discharge into the faulty circuit.

[0053] In this embodiment, the backup capacitor is a metallized film capacitor with a rated capacity of 100μF and a withstand voltage of 3kV. The charging power supply is a DC 110V battery pack, which is approximately 0.38% of the peak rated phase voltage of the 35kV system (28.58kV). This voltage level is within the safe low-voltage range, providing sufficient excitation energy without causing secondary impact on the fault point or equipment.

[0054] During normal operation of the distribution network, the control unit maintains the first controllable switch. Close, second controllable switch Disconnect. The backup capacitor is connected. Connected to a DC 110V charging power supply, the voltage across the backup capacitor is maintained at a constant 110V. A current-limiting resistor is connected in series in the charging circuit to prevent inrush current during the initial charging of the backup capacitor. The voltage across the backup capacitor is monitored in real time by a high-precision voltage sensor. When the voltage falls below a set threshold, the control unit automatically initiates supplementary charging to ensure that the backup capacitor is always fully charged.

[0055] When the delay period of S100 ends, the control unit receives the timer trigger signal and immediately executes the path switching operation: first, it sends a command to disconnect the first controllable switch. The process disconnects the backup capacitor from the charging power supply, taking approximately 5 milliseconds; a subsequent 10 millisecond delay is then made to ensure... Completely disconnect to prevent accidental connection between the charging power supply and the faulty circuit; finally, send a command to close the second controllable switch. Connect the spare capacitor to both ends of the cable in the isolated faulty section. The closing time is approximately 8 milliseconds.

[0056] In this embodiment, the first controllable switch Second controllable switch All use vacuum contactors, whose rated parameters meet the system voltage level and current capacity requirements, and whose operating time is less than 10 milliseconds.

[0057] The spare capacitor passes After connecting to the faulty section, the electrical structure of the fault circuit is as follows: a backup capacitor (100μF, initial voltage 110V) connected in series with the total cable resistance R, the total cable inductance L, and the transition resistance at the fault point. For a single-phase ground fault, the circuit path is: positive terminal of the backup capacitor → faulty phase cable core (length 2d, where d is the fault distance) → fault point → ground → cable metal sheath / armor layer → negative terminal of the backup capacitor. The distributed resistance and distributed inductance of the cable are represented by lumped parameters R and L in this circuit.

[0058] like Figure 3 As shown, the equivalent circuit model after the backup capacitor is connected to the fault circuit is a typical second-order RLC series resonant circuit, where: : Spare capacitor; R: Total resistance of the fault circuit, consisting of the resistance of the cable round trip path; L: Total inductance of the fault circuit, consisting of the inductance of the cable round trip path; R f Fault point transition resistance. This equivalent model clearly shows the electrical structure of the oscillating circuit, where the fault point transition resistance R... f It is included in the total resistance R. The characteristic angular frequency is subsequently extracted using the Prony algorithm. The overall effect of R is implicitly reflected, but due to the oscillation frequency and R... f Irrelevant mathematical properties, R f The impact on ranging accuracy was effectively suppressed.

[0059] Taking the YJV22-3×185 type 35kV cross-linked polyethylene cable as an example, its unit length parameter is: resistance 0.164Ω / km, inductance The value is 0.35 mH / km. Assume the fault distance... If the path length is 1.5km, then the total resistance of the faulty circuit is:

[0060] Ω;

[0061] Total inductance:

[0062] mH;

[0063] According to the theory of second-order RLC series circuits, the condition for underdamped oscillation is: Substitute the parameters from this embodiment into: ,and The resistance is much less than 6.48Ω, therefore the fault circuit meets the underdamped condition.

[0064] when At the instant of closing, the standby capacitor begins to discharge into the faulty circuit, generating a damped oscillating current in the circuit. The voltage across the capacitor exhibits a damped sinusoidal waveform. The characteristic angular frequency of this oscillation process is... It is closely related to the circuit parameters R, L, and C, among which inductance Directly reflects the distance to the fault The information is then used to analyze the oscillating voltage waveform in subsequent steps to extract the characteristic angular frequency. And calculate the fault distance. .

[0065] S300: Synchronously acquire the attenuated oscillating voltage waveform across the two ends of the backup capacitor during the discharge process;

[0066] Furthermore, the data acquisition requirements of the S300 include:

[0067] The sampling frequency is not less than 10 kHz;

[0068] The sampling window duration must cover at least two complete oscillation cycles;

[0069] The analog-to-digital conversion resolution is no less than 16 bits.

[0070] Specifically, the data acquisition system consists of a voltage sensor, a signal conditioning circuit, an analog-to-digital converter (ADC), and a data storage module.

[0071] The voltage sensor is directly connected across the two ends of the spare capacitor, employing a differential measurement method to improve anti-interference capability. The sensor uses a voltage transformer with an accuracy class of 0.2, a measurement range of 0-500V, and an output standard signal of 0-5V. The sensor's frequency response range is DC to 100kHz, and its amplitude-frequency characteristic flatness within 10kHz is better than ±0.5dB, meeting the frequency measurement requirements of this invention.

[0072] The signal conditioning circuit filters and amplifies the sensor output signal. A low-pass filter is used to suppress high-frequency switching noise and power frequency harmonic interference, and the amplifier circuit adjusts the signal to a standard level suitable for the ADC input range.

[0073] The analog-to-digital converter (ADC) uses a 16-bit successive approximation ADC chip with a sampling frequency of 50kHz. This sampling frequency is significantly higher than the minimum requirement of 10kHz, and according to the Nyquist sampling theorem, it can reconstruct signal components with frequencies up to 25kHz without distortion. In practical applications, the characteristic angular frequency of faulty loops is typically in the range of 500Hz to 5kHz, and the 50kHz sampling frequency provides ample margin. The 16-bit resolution corresponds to a quantization accuracy of 1 / 65536, and for a 10V full-scale input, the quantization step size is 0.153mV, with a quantization error of less than 0.0015%.

[0074] The duration of the sampling window is dynamically determined based on the expected characteristic angular frequency range. In this embodiment, considering that the longest possible fault distance in the fault section is 5km, the corresponding minimum characteristic angular frequency is approximately... (Frequency approximately 127Hz). Based on covering at least two complete oscillation cycles. Requirements, sampling window duration In practice, to improve the fitting accuracy and noise resistance of the Prony algorithm, the sampling window duration is set to 50 milliseconds, which can cover approximately 6 to 8 complete oscillation cycles.

[0075] Data acquisition triggering mechanism and second controllable switch The closing action is synchronized. The control unit issues the following: Simultaneously with the closing command, ADC sampling is initiated via a hardware trigger signal. The trigger delay time is less than 1μs, ensuring the complete start-up process of capacitor discharge is captured. During sampling, the ADC continuously acquires data at a frequency of 50kHz, obtaining a total of 2500 data points within a 50ms sampling window.

[0076] The acquired digital signals are temporarily stored in the built-in high-speed cache (SRAM) of the fault location device. The cache capacity is 64KB, which is sufficient to store multiple measurement data. The data format is a 16-bit unsigned integer, with each sampling point occupying 2 bytes. After sampling, the data is transferred to the RAM area of ​​the main processor via DMA for subsequent processing by the Prony algorithm.

[0077] S400: For the damped oscillation voltage waveform generated by the active excitation, the Prony algorithm is used to establish an exponentially decaying sine fitting model for the damped oscillation voltage waveform, and the characteristic angular frequency representing the oscillation characteristics of the loop is extracted by the least squares method.

[0078] Furthermore, in the S400, the Prony algorithm is used to represent the acquired decaying oscillation voltage waveform as a product of an exponential decay term and a sinusoidal oscillation term, and the characteristic angular frequency is extracted by the least squares method.

[0079] Furthermore, when the fitting residual of the Prony algorithm is greater than 5%, the sampling window is automatically extended and the parameters are refitted.

[0080] Specifically, in this embodiment, the processing of the Prony algorithm is completed by the embedded processor of the fault location device. The processor adopts the ARM Cortex-A9 architecture, has a main frequency of 800MHz, and its floating-point operation capability meets the requirements of real-time signal processing.

[0081] According to second-order RLC circuit theory, the voltage response during the discharge process of a standby capacitor exhibits typical damped oscillation characteristics. The core idea of ​​the Prony algorithm is to analyze the acquired discrete voltage waveforms... The fit is expressed as the product of an exponentially decaying term and a trigonometric function term. In this embodiment, the following fitting model is used:

[0082] ;

[0083] in, It is a time variable; This is the amplitude coefficient, reflecting the combined influence of the initial voltage and loop parameters; The attenuation coefficient is determined by the ratio of the loop resistance to the inductance. The characteristic angular frequency contains fault distance information and is a key parameter that needs to be extracted in this invention. The initial phase angle depends on the circuit state at the moment the capacitor is connected.

[0084] The algorithm is implemented in the following steps: First, the 2500 voltage data points obtained by sampling are converted from a discrete time-domain sequence into a format suitable for Prony analysis. Data preprocessing includes removing DC bias (achieved by subtracting the sequence mean) and normalization (divided by the maximum amplitude).

[0085] Secondly, the Prony equations are constructed. For a first-order Prony model, the fitting function is applied at sampling time... (n=0, 1, 2, ..., 2499, Expanding the sample interval (20 μs) yields the discrete form. By introducing a complex exponential form and utilizing Euler's formula, the above equation is transformed into a linear prediction model containing complex poles.

[0086] Then, the parameters are solved using the least squares method. Specifically, the error function is constructed as the sum of the squares of the differences between the actual sampled values ​​and the fitted values. The partial derivatives of the four parameters are calculated and set to zero to obtain the normal equation system. This embodiment uses the QR decomposition method to solve the overdetermined equation system, which has better numerical stability than direct matrix inversion, and iterative calculations are performed until the parameters converge.

[0087] In an actual measurement in this embodiment, the initial amplitude of the acquired voltage waveform was 108V. After processing by the Prony algorithm, the extracted parameters were: Among them, the characteristic angular frequency These are the key parameters required for subsequent distance calculations.

[0088] To ensure the quality of the fit, the algorithm calculates the fitting residuals in real time. The residual is defined as the percentage of the root mean square error between the actual sampled values ​​and the predicted values ​​of the fitted model, relative to the effective value of the signal. The calculation formula is:

[0089] ;

[0090] in, The number of sampling points. For the first The measured voltage values ​​at each sampling point This represents the Prony model's prediction at that time. This represents the effective value of the actual signal.

[0091] When the fitting residual exceeds 5%, it indicates that the current sampling window length is insufficient to fully characterize the signal characteristics, or that it is significantly affected by noise. At this point, the control unit automatically triggers a resampling process: the sampling window duration is doubled, and the number of sampling points is increased accordingly to cover more oscillation periods. After extending the window, steps S300 and S400 are re-executed until the fitting residual is reduced to below 5%. The extracted characteristic angular frequencies... The data is transmitted to the S500 for distance calculation.

[0092] S500: The total resistance of the fault circuit is expressed as the product of the resistance per unit length of the cable and twice the fault distance, and the total inductance of the fault circuit is expressed as the product of the inductance per unit length of the cable and twice the fault distance. The fault distance is calculated through a reverse mapping function based on the characteristic angular frequency, the cable per unit length parameter, and the spare capacitance value.

[0093] Furthermore, the calculation of the fault distance using the inverse mapping function includes:

[0094] The total resistance of the faulted circuit is expressed as the product of the resistance per unit length of the cable and twice the fault distance;

[0095] The total inductance of the faulted circuit is expressed as the product of the inductance per unit length of the cable and twice the fault distance;

[0096] Based on the oscillation characteristic relationship between the characteristic angular frequency and the total circuit resistance, total inductance, and spare capacitor values, a functional expression for the characteristic angular frequency with respect to the fault distance is established.

[0097] The attenuation coefficient correction calculation is introduced, and the fault distance is obtained by solving the function expression.

[0098] Specifically, in this embodiment, the fault distance is calculated based on the physical relationship between the electrical parameters of the fault circuit and the fault distance.

[0099] First, establish the relationship between the fault loop parameters and the fault distance. For a single-phase ground fault, the current starts from the standby capacitor, travels through the faulty phase cable core to the fault point, and then returns through the ground and the cable's metal sheath, forming a closed loop. The total round-trip path length of this loop is twice the fault distance. Therefore, the total resistance of the fault loop... Total Inductance They are represented as follows:

[0100] ;

[0101] ;

[0102] in, The resistance per unit length of the cable. Inductance per unit length of cable This refers to the distance from the fault point to the measurement point. These parameters are derived from the standard parameters of the cable model or factory test data and are pre-stored in the parameter database of the fault location device.

[0103] Secondly, the relationship between the characteristic angular frequency and the loop parameters is established. Based on the theory of second-order underdamped RLC oscillating loops, the characteristic angular frequency... The loop parameters satisfy:

[0104] ;

[0105] in, Let be the capacitance value of the backup capacitor. Substituting the distribution of the fault loop parameters into the above equation, we get:

[0106] ;

[0107] To simplify the expression, constant coefficient terms in the formula that are independent of distance are removed. Defined as attenuation coefficient This attenuation coefficient is determined by the inherent parameters of the cable and is independent of the fault distance. Therefore, the characteristic angular frequency is expressed as:

[0108] ;

[0109] in, The characteristic angular frequency; The inductance per unit length of the cable; This is the capacitance value of the spare capacitor; The distance from the fault point to the measurement point; The resistance per unit length of the cable.

[0110] Then, establish the reverse mapping function. Squaring and rearranging both sides of the above equation, we can solve for the fault distance:

[0111] ;

[0112] The parameters in the formula have the same meaning as above. This formula establishes the relationship between the characteristic angular frequency and the given characteristic angular frequency. The direct mapping relationship to the fault distance d.

[0113] Finally, distance calculation is performed. The fault location device retrieves the cable unit length parameter for the current line from the parameter database. and Combined with the nominal capacitance value of the spare capacitor Calculate the attenuation coefficient Then the characteristic angular frequencies extracted by S400 are... Substitute into the inverse mapping function to obtain the fault distance. .

[0114] The calculation results are displayed through the human-machine interface of the fault location device and simultaneously uploaded to the main station of the power distribution automation system via a communication interface for use by dispatchers and maintenance personnel. The display accuracy is set to 1 meter, and maintenance personnel use the location results to conduct fault inspection and repair work at the corresponding locations along the cable.

[0115] Example 2: A municipal power supply company is responsible for the operation and maintenance of a 35kV distribution network within its jurisdiction. This network includes power lines in the urban core area, industrial development zone, commercial area, and some mountainous areas, totaling over 300 kilometers of 35kV lines. The lines are diverse, including underground cable lines in urban areas, overhead lines in suburban areas, and hybrid cable-overhead lines. Due to the complex line environment and diverse fault types, traditional fault location methods are significantly insufficient. To improve fault handling efficiency, the power supply company adopted a 35kV distribution network fault location method based on natural oscillation frequency, as proposed in this invention, and verified its application in actual fault handling.

[0116] Case 1: A single-phase ground fault occurred on the company's 35kV development zone distribution line. The line uses YJV22-3×185 type cross-linked polyethylene cable, with a total length of approximately 8 kilometers, entirely underground. The fault occurred at 2:15 AM. After isolating the faulty section with the circuit breaker, the method of this invention was used for location. The extracted characteristic angular frequency is... Based on the cable parameters and backup capacitor value, the fault distance was calculated to be 3268 meters. The time from fault isolation to obtaining the location result was 3.2 seconds. Maintenance personnel rushed to the cable well 3.27 kilometers from the substation and found that the cable joint had suffered insulation breakdown due to water ingress. The actual fault location was only 22 meters off from the measured distance. The total power outage time for fault handling was 4 hours and 15 minutes.

[0117] Case 2: A two-phase short-circuit fault occurred on the company's 35kV industrial park distribution line, with a relatively short distance between the fault and the line. This line is a hybrid cable and overhead line, and the fault occurred in the cable section. After isolating the faulty section with the circuit breaker, a location tracking program was initiated. Based on the calculations of this invention, the fault distance was determined to be 758 meters. Maintenance personnel discovered a burst cable termination at a distance of 760 meters from the substation, with an actual error of 2 meters and a location accuracy of 99.7%. The entire repair process, from fault occurrence to completion, took 2 hours and 40 minutes.

[0118] Case 3: A single-phase ground fault occurred on a 35kV power supply line in a mountainous area. The line was 12 kilometers long and had multiple branches. Traditional traveling wave methods are difficult to accurately locate faults in such complex networks due to interference from multiple reflected waves. Using the method of this invention, the fault section was isolated, and precise location was performed within that section. The calculated fault distance was 6685 meters. Maintenance personnel discovered that a tree had fallen and caused the conductor to ground at a distance of 6.72 kilometers from the substation, with an actual error of 35 meters. Although the line had multiple branches, because this invention uses a two-step decoupling architecture, precise location was only performed within the isolated linear section and was not affected by interference from external branches.

[0119] Case 4: A high-resistance grounding fault occurred on the company's 35kV commercial area distribution line, with a transition resistance of approximately 150 ohms. Traditional impedance methods exhibit significant measurement errors under high-resistance faults. Using the method of this invention, the calculated fault distance was 1912 meters. Maintenance personnel discovered leakage current due to cable sheath damage at a distance of 1.89 kilometers from the substation, with an actual error of 22 meters. The high transition resistance has a relatively small impact on the distance measurement accuracy of the method of this invention. This case fully illustrates that although this method includes a transition resistance R in the equivalent circuit, it is still effective. f However, due to the use of Prony's algorithm for dynamic fitting and the essential relationship of the oscillation equation, R... f The impact on ranging accuracy is reduced to less than 1%, and excellent accuracy is maintained even under high-resistance grounding faults.

[0120] The four cases described above cover different scenarios including underground cables, short-distance faults, complex branch networks, and high-resistance grounding. The ranging error range is 2–35 meters, and the positioning accuracy is above 99%. Positioning time is within 3–5 seconds, and the total fault handling time is reduced by an average of 50–70% compared to traditional methods. The method of this invention demonstrates high accuracy, rapid response, and strong anti-interference capability under different fault types and line environments, verifying the method's universality and reliability. Compared to the 200–500 meter positioning error of the traditional traveling wave method in underground cable environments and the kilometer-level error of the impedance method under high-resistance faults, this invention significantly improves the accuracy and efficiency of fault location in 35kV distribution networks.

[0121] Example 3: In the third embodiment of the present invention, based on the same inventive concept, the present invention proposes a computer-readable storage medium storing a computer program. When the computer program is executed by a processor, it implements the steps of the above embodiment of a 35kV distribution network fault location method based on natural oscillation frequency.

[0122] Example 4: The fourth embodiment of the present invention, based on the same inventive concept, proposes a computer device comprising: a processor and a memory; the processor and the memory communicate with each other; the memory is used to store instructions; the processor is used to execute the instructions in the memory to execute a 35kV distribution network fault location method based on natural oscillation frequency as described in the above embodiment.

[0123] It should be understood that various parts of the present invention can be implemented in hardware, software, firmware, or a combination thereof. In the above embodiments, multiple steps or methods can be implemented in software or firmware stored in memory and executed by a suitable instruction execution system. For example, if implemented in hardware, as in another embodiment, it can be implemented using any one or a combination of the following techniques known in the art: discrete logic circuits having logic gates for implementing logical functions on data signals, application-specific integrated circuits (ASICs) having suitable combinational logic gates, programmable gate arrays (PGAs), field-programmable gate arrays (FPGAs), etc.

[0124] Finally, it should be noted that the above descriptions are merely preferred embodiments of the present invention and are not intended to limit the present invention. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art can still modify the technical solutions described in the foregoing embodiments or make equivalent substitutions for some of the technical features. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A method for fault location in a 35 kV distribution network based on natural frequency of oscillation, characterized in that, The method comprises the following steps: S100: after detecting a short-circuit fault and isolating the fault section by a circuit breaker, delaying for a set time length to make the arc at the fault point extinguish and residual energy attenuate; S200: after the fault section is isolated, switching the charging and discharging path by a double-path controllable switch, connecting a backup capacitor pre-charged to a safe voltage to both ends of the cable in the isolated fault section, and actively exciting the fault loop to form an under-damped oscillation loop containing the distributed resistance, distributed inductance of the fault loop and the backup capacitor; S300: synchronously collecting the attenuated oscillation voltage waveform at both ends of the backup capacitor during discharging; S400: for the attenuated oscillation voltage waveform generated by active excitation, establishing an exponential decay sinusoidal fitting model for the attenuated oscillation voltage waveform by using Prony algorithm, and extracting the characteristic angular frequency representing the oscillation characteristics of the loop by the least square method; S500: representing the total resistance of the fault loop as the product of the unit length resistance of the cable and twice the fault distance, representing the total inductance of the fault loop as the product of the unit length inductance of the cable and twice the fault distance, and calculating the fault distance based on the characteristic angular frequency, the unit length parameters of the cable and the backup capacitor value by using a reverse mapping function.

2. The method of claim 1, wherein the method is characterized by, The delay set time length is 0.5-2 seconds.

3. The method of claim 1, wherein the method is characterized by, The S200 comprises the following steps: connecting the backup capacitor to a charging power supply through a first controllable switch and maintaining the terminal voltage of the backup capacitor at a preset safe value during normal operation; after the fault section is isolated, disconnecting the first controllable switch and closing a second controllable switch to connect the backup capacitor to both ends of the cable in the fault section.

4. The method of claim 3, wherein the method is characterized by, The preset safe value is 1%-10% of the peak value of the rated phase voltage of the system.

5. The method of claim 1, wherein the method is based on natural frequency of oscillation of 35 kV distribution network for fault location. The collection requirements of the S300 comprise the following steps: the sampling frequency is not less than 10 kHz; the sampling window time length covers at least two complete oscillation periods; the analog-to-digital conversion resolution is not less than 16 bits.

6. The method of claim 1, wherein the method is based on natural frequency of oscillation of a 35 kV distribution network. In the S400, the collected attenuated oscillation voltage waveform is expressed in the form of the product of an exponential decay term and a sinusoidal oscillation term by using Prony algorithm, and the characteristic angular frequency is extracted by the least square method.

7. The method of claim 6, wherein the method is characterized by, When the fitting residual of Prony algorithm is greater than 5%, the sampling window is automatically lengthened and parameter fitting is re-performed.

8. The method of claim 1, wherein the method is based on natural frequency of oscillation for 35 kV distribution network fault location. The calculation of the fault distance by using the reverse mapping function comprises the following steps: representing the total resistance of the fault loop as the product of the unit length resistance of the cable and twice the fault distance; representing the total inductance of the fault loop as the product of the unit length inductance of the cable and twice the fault distance; establishing a function expression of the characteristic angular frequency with respect to the fault distance according to the oscillation characteristic relationship among the characteristic angular frequency and the total resistance, total inductance of the loop and the backup capacitor value; introducing an attenuation coefficient correction calculation to obtain the fault distance by solving the function expression.

9. A computer device comprising a memory, a processor, and a computer program stored on the memory and executable on the processor, characterized in that, The processor executes the computer program to implement the method for fault distance measurement of a 35kV distribution network based on natural oscillation frequency according to any one of claims 1-8.

10. A readable storage medium, characterized by, The readable storage medium stores the computer program, and the computer program is executed by the processor to implement the method for fault distance measurement of a 35kV distribution network based on natural oscillation frequency according to any one of claims 1-8.

Images