Submarine observation network power supply system protection method based on traveling wave extreme value time
By dividing the power supply system of the submarine observation network into protection zones and using the extreme time of traveling waves for fault location and isolation, the reliability and accuracy problems of protection methods under communication failure in the existing technology are solved, and fast and reliable fault isolation is achieved.
Patent Information
- Application Number
- CN202511110032.1
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2025-08-08
- Publication Date
- 2025-11-11
AI Technical Summary
The existing power supply system for submarine observation networks cannot operate reliably in scenarios where communication fails. Traditional protection methods become less accurate when faced with remote faults, high-impedance faults, and secondary wave reflection interference, and cannot meet the requirements for robustness and selectivity of protection for complex topologies.
The protection method based on traveling wave extreme time divides the protection area, collects voltage signals in real time and performs first-order differential processing, calculates the traveling wave extreme time by combining Gaussian filtering and differential gradient algorithm, and uses the dual criteria of time difference and extreme time to determine the fault direction and location, and generates protection action commands.
It enables accurate fault location and rapid isolation of complex submarine power supply networks under communication-free conditions, improving high reliability and rapid response capabilities, and solving the problems of communication dependence and difficulty in identifying high-impedance faults in traditional methods.
Smart Images

Figure CN120933873A_ABST
Abstract
Description
Technical Field
[0001] This invention relates to the field of submarine power grid protection technology, specifically to a protection method for a submarine observation network power supply system based on the extreme time of traveling waves. Background Technology
[0002] Existing power supply systems for submarine observation networks typically employ a unipolar DC power supply structure. These networks are massive in scale, have numerous nodes, and power cables can stretch for thousands of kilometers. In such systems, a ground fault often damages the fiber optic communication lines within the cables, rendering communication-based unit protection methods unreliable.
[0003] Currently, dual-terminal differential protection is mainly used as the primary protection, with overcurrent protection from the shore base station serving as backup. However, due to its strong dependence on communication, selective tripping will be impossible once communication is interrupted. Furthermore, existing non-communication protection methods (such as those based on current wave differentiation, integration, and waveform amplitude changes) show decreased accuracy when facing remote faults, high-resistance faults, and secondary wave reflection interference. Their extracted features are easily affected by transition resistance, measurement noise, and converter boundary impedance, failing to meet the robustness and selectivity requirements for protection in the complex environment of the submarine observation network.
[0004] Therefore, there is currently a lack of a protection method that can still maintain high reliability and high-speed operation in communication failure scenarios and is applicable to submarine power supply systems with complex topologies. Summary of the Invention
[0005] In view of this, the present invention provides a protection method for the power supply system of the seabed observation network based on the extreme time of traveling waves, which has the advantages of achieving high-reliability fault location and rapid isolation even in communication failure scenarios, and effectively solving the problem of remote fault and high-resistance fault identification.
[0006] This invention provides a protection method for a power supply system of a seabed observation network based on the extreme time of traveling waves, comprising:
[0007] S1. Based on the network structure, power supply path and node importance of the power supply system of the submarine observation network, determine the location of the protection device and divide the power supply system of the submarine observation network into multiple protection zones. Each protection zone is defined by two adjacent protection devices.
[0008] S2. Real-time acquisition of voltage signals at the deployment location, and first-order differentiation processing of the voltage signals to obtain the corresponding voltage change rate; and comparison of the voltage change rate with a preset voltage threshold to determine whether a fault has occurred.
[0009] S3. When a fault is determined to occur, record the current traveling wave data within a preset time period before and after the fault occurs, and use a Gaussian filter to perform noise reduction processing on the current traveling wave data.
[0010] S4. Based on the denoised current traveling wave data, calculate the extreme value time t of the traveling wave using the differential gradient algorithm. 极值 ;
[0011] S5. Compare the extreme time t of the traveling wave in the branch units on both sides of the protected area. Mi and t Mo According to the time difference Δt = t Mi -t Mo With direction criterion threshold [t] low , t up The relationship between the fault and the direction of the fault is determined.
[0012] S6, the extreme time t of the traveling wave 极值 Compared with the preset remote fault threshold t of the section re Compare the data to determine whether the fault is within the protected area;
[0013] S7. When the fault simultaneously satisfies the forward direction criterion Δt≤t low and the criteria within the area t 极值 ≤t re If the fault occurs, it is determined to be a fault within the zone, and a protection action command is generated to isolate the protected zone where the fault is located.
[0014] In one optional implementation, the arrangement of the protection device in S1 must meet the following requirements:
[0015] Redundancy principle: Install protection devices in branch units that connect multiple trunk lines;
[0016] Critical load protection principle: Protective devices must be configured for branch units connected to high-priority observation equipment;
[0017] Waveform observability principle: The number of branch units between adjacent protection devices is less than or equal to a preset threshold number N. BU ;
[0018] The preset quantity threshold N BU The following constraints must be met:
[0019]
[0020] in, and These represent the current traveling wave passing through N under high-impedance fault conditions. BU Amplitude and extreme time after each branch unit; I set.low The smallest current amplitude that the measuring device can resolve; L lineThis represents the maximum length of the line between two adjacent protection devices, where v is the traveling wave propagation speed.
[0021] Link uniqueness principle: the traveling wave propagation path between adjacent protection devices is unique; when there is a branch line between adjacent protection devices, a protection device is added at the branch point.
[0022] In one optional implementation, the criterion for determining the occurrence of a fault in step S2 is:
[0023]
[0024] Δu(k)=u fault (t k )-u rated
[0025] Where k is the sampling point number, and k = 1, 2, ..., P; P is the number of sampling points in a data window; t k The time corresponding to the kth sampling point; u fault (t k ) represents the fault state at t k The voltage value measured at any time; u rated The rated voltage is Δu(k); Δu(k) represents the voltage change; Δu set This is a preset voltage threshold.
[0026] The voltage threshold is set to ensure that the protection device can still reliably start the protection function when a remote fault occurs under high grounding resistance conditions.
[0027] In one optional implementation, the Gaussian filter expression in S3 is:
[0028]
[0029] Where σ is the standard deviation; G(i) is the time-domain response function of the Gaussian filter; and i is the time variable, with the fault occurrence time t=0 as the starting point.
[0030] In one optional implementation, the calculation of the traveling wave extremum time in S4 employs a differential gradient algorithm, the expression of which is:
[0031]
[0032] Where i(t) represents the current traveling wave data; t0 represents the extreme point of the current traveling wave data after noise reduction; and Δt represents the time window.
[0033] When the gradient value changes from positive to negative, it is marked as an extreme point. The time corresponding to the first occurrence of the extreme point is the extreme time t of the traveling wave. 极值 .
[0034] In one optional implementation, the criterion for determining the fault direction in S5 is:
[0035] If Δt≥t up The fault was determined to be a back-side fault, and the protection action was not triggered.
[0036] If Δt≤t low The fault is identified as a forward fault, and the process proceeds to the area determination procedure.
[0037] In one optional implementation, the criterion for determining whether the fault is within the protection zone in step S6 is:
[0038] When the extreme time of the traveling wave is t 极值 >t re When the fault is located outside the protected area, it is determined that the fault is outside the protected area.
[0039] When the extreme time of the traveling wave is t 极值 ≤t re When the fault is located outside the protected area, it is determined that the fault is outside the protected area.
[0040] Among them, t re The preset fault threshold at the far end of the section is set according to the maximum traveling wave extreme time when a fault occurs at the end of the protected area.
[0041] In an optional implementation, S7 further includes:
[0042] S71. Number the lines in the network structure of the power supply system for the submarine observation network.
[0043] S72. Preset the extreme time intervals for each line;
[0044] S73. Obtain the traveling wave extreme time detected by the protection devices at both ends of the protection area where the fault is located, and match the traveling wave extreme time detected by the protection devices at both ends to the target extreme time intervals respectively, and output the target line number corresponding to the target extreme time interval.
[0045] As can be seen from the above, the seabed observation network power supply system protection method based on traveling wave extreme time provided in this application achieves accurate fault location and rapid isolation of complex seabed power supply networks under conditions without communication by dividing the protection area, acquiring voltage signals in real time, calculating traveling wave extreme time and combining multi-dimensional criteria. It effectively solves the problems of strong dependence on communication and difficulty in identifying high-impedance faults in existing technologies, and has high reliability and rapid action capability. Attached Figure Description
[0046] To more clearly illustrate the specific embodiments of the present invention or the technical solutions in the prior art, the drawings used in the description of the specific embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of the present invention. For those skilled in the art, other drawings can be obtained from these drawings without creative effort.
[0047] Figure 1 This is a flowchart illustrating a method for protecting a submarine observation network power supply system based on the extreme time of traveling waves, according to an embodiment of the present invention.
[0048] Figure 2 This is a network topology diagram of the power supply system for the submarine observation network according to an embodiment of the present invention;
[0049] Figure 3 This is a diagram illustrating the layout of the protection device according to an embodiment of the present invention;
[0050] Figure 4 These are current traveling wave diagrams measured by the same protection device under different faults according to embodiments of the present invention;
[0051] Figure 5 It is a current traveling wave diagram measured by the protection devices at various locations under the same fault according to an embodiment of the present invention. Detailed Implementation
[0052] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.
[0053] In existing technologies, the power supply system of submarine observation networks typically adopts a unipolar DC power supply structure. These networks are large in scale, have numerous nodes, and power cables can reach lengths of thousands of kilometers. When a ground fault occurs, fiber optic communication lines often fail simultaneously, causing communication-based unit protection methods to fail. Existing technologies mainly rely on dual-end differential protection and overcurrent protection at shore base stations, but these methods have extremely high requirements for communication reliability and cannot achieve selective tripping when communication is interrupted. Traditional non-communication protection methods rely on the differentiation, integration, and amplitude changes of current waves for judgment, but their accuracy significantly decreases in scenarios involving remote faults, high-resistance faults, and secondary wave reflection interference. Feature extraction is easily affected by transition resistance, measurement noise, and converter boundary impedance, making it difficult to meet the robustness and selectivity requirements of protection in complex submarine environments.
[0054] To address the aforementioned issues, a single-ended protection method independent of two-way communication needs to be developed. Considering the time-difference propagation characteristics of traveling wave signals in cables, an attempt was made to locate faults by analyzing the extreme time of the traveling wave. However, three main challenges are faced: selective protection under complex network topologies, feature extraction under high-impedance faults, and signal processing under noise interference. Through studying network structural characteristics, it was found that rationally dividing the protection zone can limit the fault judgment range; using the voltage differential initiation criterion can quickly capture fault characteristics; and combining Gaussian filtering and differential gradient algorithms can effectively extract the extreme time of the traveling wave. Ultimately, a dual-criteria mechanism based on protection zone division and traveling wave extreme time analysis is formed, achieving communication-independent fault protection.
[0055] Therefore, as Figure 1 As shown, this application proposes a protection method for the power supply system of a seabed observation network based on the extreme time of traveling waves, including:
[0056] Step S1: Based on the network structure, power supply path and node importance of the power supply system of the seabed observation network, determine the location of the protection device and divide the power supply system of the seabed observation network into multiple protection zones. Each protection zone is defined by two adjacent protection devices.
[0057] Among these, the division of protection zones refers to dividing the power supply system of the seabed observation network into multiple logical segments defined by adjacent protection devices. This can be specifically achieved by analyzing the network topology (e.g., Figure 2 As shown in the diagram, the importance of the equipment (i.e., node importance) is used to define the physical boundaries of fault diagnosis. The protection device is a non-communication traveling wave protection device, which acts as a "smart circuit breaker" for the submarine power grid. It is configured in some branch units (BUs) and can realize the integrated function of "detection-location-isolation" to prevent local faults from causing the entire network to be paralyzed.
[0058] Specifically, such as Figure 2 As shown, the power supply system of the entire submarine observation network includes shore base stations, lower-level junction boxes, branch unit (BU) units, and protection devices (e.g., Figure 2 The equipment at the bottom of the black dashed box and the equipment within the white dashed box). There are multiple branch units, but not all branch units are equipped with independent protection devices. In this case, it is necessary to select the branch units that need protection based on the network structure, power supply path, and node importance of the submarine observation network power supply system, and configure them with independent protection devices, thus forming distributed protection nodes, such as... Figure 3 As shown.
[0059] Step S2: Real-time acquisition of voltage signals at the deployment location, and first-order differentiation processing of the voltage signals to obtain the corresponding voltage change rate; and comparison of the voltage change rate with a preset voltage threshold to determine whether a fault has occurred.
[0060] Voltage differential processing refers to performing first-order derivative operations on the acquired voltage signal, which can be implemented using a digital differential algorithm to quickly capture voltage change characteristics.
[0061] Specifically, under normal system operation, voltage signals are continuously collected at the deployment locations (i.e., protection nodes), and the collected voltage signals are processed by first-order differentiation to obtain the voltage change rate. A preset voltage threshold for determining the occurrence of a fault is set. When the voltage change rate at any sampling point exceeds the preset voltage threshold, a fault is determined to have occurred, triggering the protection device to enter the subsequent fault identification process.
[0062] It should be noted that the preset voltage threshold can be set according to the noise level and normal fluctuation range of the on-site voltage signal acquisition device to ensure the sensitivity and anti-interference capability of the protection device.
[0063] Step S3: When a fault is determined to occur, record the current traveling wave data within a preset time period before and after the fault occurs, and use a Gaussian filter to perform noise reduction processing on the current traveling wave data.
[0064] Among them, the Gaussian filter refers to a time-domain filter constructed using a Gaussian function. Specifically, the filtering strength can be adjusted by setting the standard deviation parameter to eliminate high-frequency noise interference.
[0065] Specifically, when the protection device at a certain protection node determines that a fault has occurred, it records the fault traveling wave data (i.e., current traveling wave data) at that protection node from the time the fault is determined to occur to 5ms after the time the fault is determined to occur.
[0066] Step S4: Based on the denoised current traveling wave data, calculate the extreme value time t of the traveling wave using the differential gradient algorithm. 极值 .
[0067] Among them, the differential gradient algorithm refers to determining the extreme point by calculating the rate of change of current between adjacent sampling points. Specifically, it can be implemented using the three-point difference method to accurately extract the arrival time of traveling waves.
[0068] Step S5: Compare the traveling wave extreme time t of the branch units on both sides of the protected area. Mi and t Mo According to the time difference Δt = t Mi -t Mo With direction criterion threshold [t] low , t up The relationship between the two can be used to determine the direction of the fault.
[0069] Among them, the direction criterion threshold [t] low , t up[ ] refers to a pre-set time difference judgment interval, which can be calculated and determined based on the traveling wave propagation speed and the length of the protected area, used to distinguish between forward and back-side faults. The fault direction refers to the relative direction between the fault and the protection device, denoted as t. Mi t represents the extreme time of the traveling wave measured at the forward protection node. Mo The extreme time of the traveling wave is measured at the back-side node protection point.
[0070] Step S6: Set the extreme time t of the traveling wave. 极值 Compared with the preset remote fault threshold t of the section re The comparison determines whether the fault is within the protection zone.
[0071] Among them, the remote fault threshold t of the section re This refers to the maximum traveling wave propagation time corresponding to a fault at the end of the protected area. It can be specifically calculated and set by the line length and traveling wave speed, and is used to define faults inside and outside the protected area.
[0072] Step S7: When the fault simultaneously satisfies the forward direction criterion Δt≤t low and the criteria within the area t 极值 ≤t re If the fault occurs, it is determined to be a fault within the zone, and a protection action command is generated to isolate the protected zone where the fault is located.
[0073] The branch unit is also equipped with a protection signal detection device, which is used to receive protection action commands and isolate the protection area where the fault is located according to the protection action commands.
[0074] Specifically, this method achieves communication-free protection through a collaborative working mechanism. First, protection zones are divided according to the network topology, establishing clear physical boundaries for each zone. When a fault occurs, voltage differentiation processing quickly captures voltage surges, avoiding the shortcomings of traditional amplitude criteria which are affected by high-impedance faults. After Gaussian filtering eliminates noise interference, the differential gradient algorithm accurately extracts the extreme time of the traveling wave, a time feature less affected by transition resistance. The fault direction is determined by comparing the time difference between the two sides of the protection zone, and the fault area is determined by comparing the extreme time with a preset threshold. This dual-criteria mechanism ensures that action is triggered only when the fault direction points to the protection zone and the extreme time matches the zone's characteristics, effectively avoiding false triggering and failure to trigger. The protection zone division strategy solves the selectivity problem of complex networks, the time difference criterion replaces the traditional communication-dependent differential protection, and the extreme time analysis overcomes the insufficient sensitivity of the amplitude criterion.
[0075] Compared to existing technologies, traditional dual-ended differential protection requires continuous bidirectional communication, while this method only requires single-ended data to complete the judgment. Existing non-communication protection methods mostly rely on amplitude comparison, which leads to decreased sensitivity during high-resistance faults. This invention utilizes time characteristics, resulting in better resistance to transition resistance. Traditional methods face difficulties in selective coordination in complex networks; this invention transforms the global problem into a local judgment by dividing the protection area. Existing traveling wave protection is susceptible to reflected wave interference; this invention improves anti-interference capability through extreme value time extraction and a dual-criteria mechanism. Compared to protection based on waveform integration, this invention employs differential processing and gradient algorithms, resulting in a faster response speed.
[0076] Through the above technical solutions, this invention can achieve selective protection even in the event of a complete communication outage, solving the communication dependency problem of traditional differential protection. The use of dual criteria—time difference and extreme time—effectively improves the detection accuracy of high-resistance and remote faults. The combination of protection zone division and single-end feature analysis adapts to the protection needs of complex topologies in seabed observation networks. The combined application of Gaussian filtering and gradient algorithms significantly improves the reliability of traveling wave feature extraction in noisy environments. The entire method requires no modification to existing equipment structures; improved protection performance can be achieved solely through algorithm improvements.
[0077] In one optional implementation, the arrangement of the protection device in step S1 must meet the following requirements:
[0078] Redundancy principle: Install protection devices in branch units that connect multiple trunk lines;
[0079] Critical load protection principle: Protective devices must be configured in branch units that connect to high-priority observation equipment;
[0080] Waveform observability principle: The number of branch units between adjacent protection devices is less than or equal to a preset threshold number N. BU ;
[0081] The preset quantity threshold N BU The following constraints must be met:
[0082]
[0083] in, and These represent the current traveling wave passing through N under high-impedance fault conditions. BU Amplitude and extreme time after each branch unit; I set.low The smallest current amplitude that the measuring device can resolve; L lineThe first formula represents the maximum length of the line between two adjacent protection devices, and v is the traveling wave propagation speed. The first formula is used to ensure that the traveling wave can still be effectively extracted when a high-resistance fault occurs at the end of the protected section. At the same time, the second formula is used to ensure that only the secondary reflected wave affects the extreme time of the waveform measured at the protection device.
[0084] Link uniqueness principle: the traveling wave propagation path between adjacent protection devices is unique; when there is a branch line between adjacent protection devices, a protection device is added at the branch point.
[0085] Specifically, the deployment of protection devices must meet the principles of redundancy, critical load protection, waveform observability, and link uniqueness. Redundancy means deploying protection devices within branch units connecting multiple trunk lines. This can be achieved by setting multiple protection devices at topology intersection nodes, preventing the trunk line from losing protection due to the failure of a single protection device, enhancing power reconfiguration capabilities after fault clearing, and ensuring overall system reliability. Critical load protection means mandating protection devices for branch units connecting high-priority observation equipment. This is achieved by using a preset equipment priority list and associated protection device configuration strategies, ensuring the continuity of power supply to important observation equipment and preventing power outages of important loads after non-unit protection actions. Shore base stations must be designated as critical units and configured with protection. Waveform observability means that the number of branch units between adjacent protection devices is less than or equal to a preset threshold N. BU Considering that the traveling wave will experience significant attenuation after passing through multiple branch units during transmission, affecting the accuracy of extreme time extraction, the number of branch units between any two adjacent protection devices is limited to a preset threshold N. BU Among them, the preset quantity threshold N BU It is determined by establishing an attenuation model of the current traveling wave amplitude and extreme time under high resistance faults, combined with the minimum resolution capability of the measuring device, to ensure that the traveling wave characteristics can be effectively detected; the link uniqueness principle means that the traveling wave propagation path between adjacent protection devices is unique, which is specifically achieved by adding protection devices at the bifurcation point to eliminate multi-path interference and avoid ambiguity in traveling wave propagation.
[0086] Specifically, an optimal deployment scheme for protection devices is constructed through the synergistic constraints of four principles. The redundancy principle sets up multiple protections for key intersection nodes in the network topology; the critical load guarantee principle enforces configuration strategies to ensure the power supply continuity of important observation equipment; the waveform observability principle determines the threshold for the number of branch units through mathematical modeling, ensuring that the traveling wave signal can still be accurately identified by the detection equipment after passing through multiple branch units, while avoiding resource waste caused by over-deployment; and the link uniqueness principle eliminates path ambiguity during traveling wave propagation through path uniqueness constraints, establishing a clear signal propagation link by adding protection devices when branch lines exist. These four principles together form a dual guarantee mechanism of spatial layout and signal propagation characteristics, ensuring reliable capture of traveling wave characteristics even under high-impedance faults and complex network topologies.
[0087] Compared with existing technologies, current protection device deployment only considers a single line length or simple topology, without establishing a multi-dimensional constraint model. This leads to the traveling wave signal attenuation exceeding detection capability during high-impedance faults, or branching lines causing non-unique traveling wave propagation paths. This invention effectively solves the reliability problem of protection device deployment in complex network environments by establishing a multi-physical quantity constraint model that includes amplitude attenuation, time delay, and path uniqueness, combined with a redundancy protection strategy for key nodes.
[0088] Through the above technical solutions, the present invention realizes the reasonable deployment of protection devices in high-impedance fault and complex network topology scenarios, ensuring that the amplitude and time characteristics of traveling wave signals can be accurately detected, avoiding fault misjudgment caused by signal attenuation or path ambiguity, while improving the reliability of trunk line protection through redundant configuration of key nodes, and optimizing resource allocation efficiency through threshold constraints.
[0089] In one optional implementation, the criterion for determining the occurrence of a fault in step S2 is:
[0090]
[0091] Δu(k)=u fault (t k )-u rated
[0092] Where k is the sampling point number, and k = 1, 2, ..., P; P is the number of sampling points in a data window; t k The time corresponding to the kth sampling point; u fault (t k ) represents the fault state at t k The voltage value measured at any time; u rated The rated voltage is Δu(k); Δu(k) represents the voltage change; Δu set This is a preset voltage threshold.
[0093] The voltage threshold is set to ensure that the protection device can still reliably start the protection function when a remote fault occurs under high grounding resistance conditions.
[0094] Specifically, voltage change refers to the absolute deviation between the instantaneous fault voltage and the rated voltage. This can be calculated using the absolute value of the difference between the real-time sampled voltage and the rated voltage, used to capture voltage surge characteristics. Cumulative change within a time window refers to the sum of voltage changes at multiple consecutive sampling points. This can be calculated by accumulating the voltage changes at each sampling point within a sliding window, used to enhance the stability of the criterion. The preset voltage threshold is a critical value calibrated based on high-resistance fault conditions. It can be determined through simulation or experimentation to obtain the minimum detectable voltage change of a remote fault under high-resistance fault conditions, used to adapt to sensitivity requirements under different operating conditions.
[0095] Specifically, by calculating the absolute deviation between the instantaneous voltage value and the rated value at each sampling point in real time, the sudden voltage change characteristics can be quickly captured. Furthermore, a criterion for the cumulative change of multiple consecutive sampling points within a time window is introduced to avoid false triggering at a single moment. Simultaneously, combined with a dynamic voltage threshold calibrated based on high-impedance operating conditions, it ensures effective fault identification even under weak voltage changes. For example, when a high-impedance fault causes a small voltage change, the absolute deviation calculation can eliminate the influence of voltage polarity fluctuations, the cumulative criterion enhances signal characteristics through the superposition of multi-point data, and the dynamic threshold adjusts the sensitivity according to the actual operating conditions. The synergistic effect of these three elements solves the problem of insufficient sensitivity of traditional criteria in high-impedance scenarios.
[0096] Through the above technical solution, this invention can accurately identify weak voltage change characteristics in scenarios with high grounding resistance and remote faults, avoiding the problem of protection device startup failure due to unclear signal characteristics. At the same time, it suppresses noise interference through multi-dimensional criteria, improving the robustness of the criteria. For example, even when the voltage change caused by a high-resistance fault is only 5% of the rated value, the protection can still be reliably triggered to start through the synergistic effect of cumulative criteria and dynamic threshold.
[0097] In one optional implementation, the Gaussian filter expression in step S3 is:
[0098]
[0099] Where σ is the standard deviation; G(i) is the time-domain response function of the Gaussian filter; and i is the time variable, with the fault initiation time t=0 as the origin.
[0100] Specifically, the standard deviation refers to the width parameter of the Gaussian function's distribution in the time domain. It can be determined by matching numerical simulations with measured data. The magnitude of the standard deviation is related to the noise spectrum characteristics, and its value affects the passband cutoff frequency of the filter. The time-domain response function refers to the impulse response form of the Gaussian filter in the time domain. It is specifically implemented through discretization to achieve digital signal filtering. The symmetry of the time-domain response function can avoid phase distortion.
[0101] During the filtering process, the exponential decay characteristic of the Gaussian function effectively suppresses high-frequency noise components while preserving the key frequency bands of the traveling wave's abrupt change characteristics. The standard deviation parameter is determined by analyzing the noise spectrum distribution under typical fault scenarios. For example, for low-frequency oscillation interference caused by the converter boundary impedance, the standard deviation can be appropriately increased to extend the low-pass characteristics of the filter. The symmetrical time-domain response of the Gaussian filter avoids the waveform distortion introduced by the asymmetric filter when suppressing secondary wave reflection interference, thus accurately extracting the true traveling wave extrema.
[0102] In one optional implementation, the calculation of the traveling wave extremum time in step S4 employs the differential gradient algorithm, the expression of which is:
[0103]
[0104] Where i(t) represents the current traveling wave data; t0 represents the extreme point of the current traveling wave data after noise reduction; and Δt represents the time window.
[0105] When the gradient value changes from positive to negative, it is marked as an extreme point. The time corresponding to the first occurrence of the extreme point is the extreme time t of the traveling wave. 极值 .
[0106] Among them, the traveling current data refers to the transient current signal propagating in the line after a fault occurs. Specifically, it can be collected using a high-frequency sampling device at a sampling frequency of not less than 1MHz to reflect the abrupt changes in the propagation process of the traveling wave. The extreme point refers to the first amplitude abrupt change point generated by the traveling current signal during propagation, which can be identified by the change in gradient sign, and is used to characterize the time when the traveling wave front arrives at the measurement point. The time window Δt refers to the interval used to calculate the gradient change between consecutive sampling points. Specifically, it can be the time difference between two adjacent sampling points, used to balance calculation sensitivity and noise resistance.
[0107] Specifically, after a fault is detected, the denoised current traveling wave data is input into a differential gradient algorithm. This algorithm dynamically tracks the abrupt changes in the current traveling wave signal by calculating the gradient changes between adjacent sampling points. When the gradient value changes from positive to negative, it indicates that the current traveling wave has reached a local maximum, and this point is marked as an extremum. By prioritizing the time corresponding to the first occurrence of an extremum as the traveling wave extremum time, false extremum interference caused by reflected waves or noise can be effectively avoided. This process uses mathematical differential operations to capture the precise arrival time of the traveling wave front, providing a reliable time reference for fault direction and area determination.
[0108] In one optional implementation, the criterion for determining the fault direction in step S5 is:
[0109] If Δt≥t up The fault was determined to be a back-side fault, and the protection action was not triggered.
[0110] If Δt≤t low The fault is identified as a forward fault, and the process proceeds to the area determination procedure.
[0111] In one optional implementation, the criterion for determining whether the fault is within the protection zone in step S6 is:
[0112] When the extreme time of the traveling wave is t 极值 >t re When the fault is located outside the protected area, it is determined that the fault is outside the protected area.
[0113] When the extreme time of the traveling wave is t 极值 ≤t re When the fault is located outside the protected area, it is determined that the fault is outside the protected area.
[0114] Among them, t re The preset fault threshold at the far end of the section is set according to the maximum traveling wave extreme time when a fault occurs at the end of the protected area.
[0115] The traveling wave extreme time refers to the time when the current traveling wave first reaches its extreme point. Specifically, it can be determined by using a differential gradient algorithm to detect extreme points in the denoised current traveling wave data, and by calculating the inflection point where the gradient value changes from positive to negative. The segment far-end fault threshold refers to the maximum traveling wave propagation time corresponding to a fault occurring at the end of the protected area. Specifically, it can be determined by obtaining the maximum traveling wave propagation time at the end of the protected area under different fault conditions through simulation or actual measurement, and then setting a safety margin. This threshold setting method considers the relationship between line length and traveling wave propagation speed, ensuring that the criterion can adapt to the protection area division of different topologies.
[0116] Specifically, in the fault area determination process, the comparison between the extreme travel wave time and the fault threshold at the far end of the section forms a dual verification mechanism. When the extreme travel wave time exceeds the threshold, it indicates that the fault point is located outside the end of the protection area, thus ruling out the possibility of a fault within the area. When the extreme travel wave time does not exceed the threshold, the fault is confirmed to be located within the protection area by combining the forward direction criteria. The threshold setting is based on the longest travel wave propagation time of the fault at the end of the protection area, covering the most unfavorable operating conditions such as high-resistance faults at the end of the line, avoiding the problem of traditional amplitude criteria being affected by transition resistance. At the same time, this criterion does not rely on the communication channel to transmit information from the other end; the area boundary identification can be completed through single-end detection alone.
[0117] In an optional implementation, step S7 further includes:
[0118] S71. Number the lines in the network structure of the power supply system for the submarine observation network.
[0119] S72. Preset the extreme time intervals for each line;
[0120] S73. Obtain the traveling wave extreme time detected by the protection devices at both ends of the protection area where the fault is located, and match the traveling wave extreme time detected by the protection devices at both ends to the target extreme time intervals respectively, and output the target line number corresponding to the target extreme time interval.
[0121] In this context, "line number" refers to assigning a unique identifier to each transmission line in the power supply system network. This can be implemented using a tree-like coding structure; for example, the main line can be defined as a first-level number, and branch lines can be further subdivided into second-level numbers based on the main line number. "Extreme value time interval" refers to the time range calculated based on the line length and the traveling wave propagation speed. This can be achieved by establishing a mapping table between line length and traveling wave propagation speed. For example, for a line of length L, its extreme value time interval can be set as [L / (v+Δv), L / (v-Δv)], where Δv is the speed fluctuation tolerance value. "Target line number matching" refers to bidirectionally comparing the measured extreme value time of the traveling wave with a preset interval. This can be implemented using an interval coverage algorithm. For example, when the detection time falls simultaneously into the intersection of the forward propagation time interval and the reverse propagation time interval of a line, it is determined that the line has a fault.
[0122] Specifically, after a power supply failure in the seabed observation network, pre-stored network topology data is first invoked to generate globally unique line identifiers according to the line hierarchy. Based on the actual line length and traveling wave propagation speed parameters, the theoretical extreme time range is calculated for each line, and a time-line mapping database is established. When the branch units at both ends of the protected area detect the traveling wave extreme time t... Mi With t MoThe two time values are matched against the extreme time intervals of each line stored in the database. By verifying whether the time data at both ends simultaneously meet the forward propagation time constraints and backward propagation time constraints of the same line, interference characteristics of adjacent lines are eliminated, and the line number that meets the bidirectional time matching condition is finally output as the fault location result.
[0123] Compared to existing technologies, traditional fault location methods rely on two-way data transmission via communication channels or on single-end traveling wave waveform feature identification. In complex branch networks, these methods are susceptible to signal attenuation and waveform distortion, leading to misjudgments. This invention, through a pre-set line feature database and a localized time matching mechanism, utilizes the deterministic characteristics of traveling wave propagation time to achieve accurate fault line identification without real-time communication, thus solving the problem of location ambiguity caused by reflected wave interference from branch lines.
[0124] Through the above technical solution, this application effectively overcomes the technical obstacle of difficulty in identifying faulty lines in the multi-branch topology of the submarine observation network. By establishing a line feature database and a two-way time matching mechanism, it can still accurately locate the specific line where the fault occurred in the case of communication interruption, providing a reliable local decision-making basis for the selective protection of complex power supply networks.
[0125] In addition, the present invention also provides a protection system for the power supply system of a submarine observation network based on the extreme time of traveling waves, including a shore base station, a lower-level junction box, and a branch unit;
[0126] The branch unit includes a protection signal detection device and a protection device.
[0127] The protection device includes a signal acquisition unit, a protection start unit, a traveling wave extreme value time calculation unit, a fault direction judgment unit, a fault area judgment unit, and a protection action command issuance unit.
[0128] The signal acquisition unit is used to acquire the voltage signal on the branch unit where the protection is located, and is connected to the protection start unit and the traveling wave extreme value time calculation unit. The protection start unit is used to calculate the voltage derivative based on the voltage signal recorded by the signal acquisition unit to determine whether a fault has occurred. The traveling wave extreme value time unit is connected to the signal acquisition unit, the protection start unit, the fault direction judgment unit, and the fault area judgment unit. It is used to calculate the traveling wave extreme value time after the fault is determined to have occurred, and send the calculation result to the lower-level unit. The fault direction judgment unit is used to determine the fault direction based on the magnitude of the traveling wave extreme value time on both sides of the protection. The fault area judgment unit is used to determine whether the fault has occurred within the fault area based on the traveling wave extreme value time. The protection action command issuing unit is connected to the traveling wave extreme value time unit and the fault area judgment unit. It is used to generate protection action commands based on the judgment results of the two units.
[0129] The system comprises several modules: a signal acquisition unit (hardware module for real-time acquisition of branch unit voltage signals, typically using Hall effect sensors and high-speed sampling circuits), a protection activation unit (logic module for detecting fault characteristics through voltage differentiation, typically using differential circuits and threshold comparators), a traveling wave extreme value time calculation unit for extracting the time of the first extreme value of the current traveling wave, typically using a sliding time window differential gradient algorithm, and a fault location unit for identifying faults based on millisecond-level voltage change rates, enabling rapid activation of the protection process), a fault direction determination unit for comparing the extreme value time differences of adjacent protection devices, typically using time difference calculation and direction threshold comparison logic, and a fault location unit for identifying the fault side based on the traveling wave arrival time difference, eliminating reflected wave interference), and a fault area determination unit for matching extreme value time with preset thresholds, typically using a time interval mapping algorithm, to distinguish between faults within and outside the designated area, ensuring selective tripping. The protection action command issuing unit refers to the output module that generates control commands based on the comprehensive judgment results. It can be implemented using a programmable logic controller. Its function is to perform circuit breaker tripping operations and achieve fault area isolation.
[0130] Specifically, each branch unit is configured with an independent protection device, forming a distributed protection node. The signal acquisition unit continuously monitors the voltage signal of its node. When an abnormal fluctuation is detected, the protection activation unit calculates the voltage differential value and compares it with a preset threshold to determine that a fault has occurred. After confirming the fault, the traveling wave extreme value time calculation unit performs noise reduction processing on the current traveling wave and uses a differential gradient algorithm to extract the time corresponding to the first extreme point. The fault direction determination unit acquires the time data of adjacent nodes, calculates the time difference and compares it with the direction criterion threshold to determine whether the fault is located on the forward or backward side of the protection area. The fault area determination unit compares the extreme value time of its node with the preset segment far-end fault threshold to determine whether the fault is within its jurisdiction. When both the direction criterion and the area criterion meet the action conditions simultaneously, the protection action command issuing unit triggers the circuit breaker to trip, completing the fault isolation. The entire process requires no coordination from the master station; each node independently completes data acquisition, feature extraction, and logical judgment.
[0131] Compared to existing technologies, traditional solutions rely on two-way communication for differential protection, losing selective tripping capability when communication is interrupted, and are susceptible to high-impedance faults due to amplitude comparison. This solution eliminates communication dependence by independently acquiring local signals through distributed nodes; it uses the extreme time of traveling waves as a characteristic quantity, whose propagation characteristics are unaffected by transition resistance, ensuring reliable high-impedance fault detection; it effectively distinguishes between faults inside and outside the protection zone through a dual criterion of time difference and threshold, avoiding misjudgments caused by reflected wave interference; and each protection node makes autonomous decisions, achieving rapid fault isolation without master station coordination.
[0132] Through the above technical solutions, this invention solves the technical problem of unreliable protection in the power supply system of the submarine observation network when communication fails, and achieves selective fault isolation under complex network topologies. By using localized traveling wave extreme time analysis, it overcomes the insufficient sensitivity of the traditional amplitude comparison method in high-resistivity fault scenarios, improving the operational reliability of the protection system. The distributed architecture design ensures that the failure of a single node does not affect the overall protection function, enhancing system robustness. The dual criterion mechanism of time difference and threshold effectively suppresses interference caused by traveling wave reflection, improving the judgment accuracy in complex fault scenarios.
[0133] To verify the technical effect of the present invention, simulation was performed using the simulation model parameters of the power supply system of the submarine observation network shown in Table 1 under the specific environmental settings of the PSCAD / EMTDC simulation platform.
[0134] Table 1
[0135]
[0136] For example Figure 3 The NEPTUNE seabed observation network example shown is a topology example, with metallic short-circuit faults set in the middle of lines I, II, III, IV, and V, respectively. All faults are set to occur at t = 2s. To verify the operational characteristics of the proposed method, a BU was selected. 12 BU 15 BU 18 and BU 22 The protective devices on it are used for observation. Figure 4 The BU provides the following response when different lines experience failures: 15 The current traveling wave results measured at the location. For BU 15 The protection device on the device has set upper and lower limits for the extreme time of the traveling wave as t. up =0.05ms, t low = -0.05ms, the extreme time t corresponding to a remote fault in the section. re = 0.8ms. For faults occurring on lines II, III, IV, and V, the extreme time of the traveling wave satisfies M. 15_1 -M 15_2 ≤tlow Therefore, it is determined to be a forward fault; while for a fault on line I, the extreme time satisfies M 15_1 -M 15_2 ≥t up Based on this, the protection device judges it as a backward fault. Figure 5 This demonstrates what happens when a fault occurs at the midpoint of Line II, and what happens to BU. 12 BU 15 BU 18 and BU 22 The current traveling wave waveform measured at point BU. As can be seen from the figure, BU 15 With BU 18 The extreme times are all lower than t re And due to BU 12 and BU 22 Due to the distance from the fault point, the measured traveling wave waveform could not effectively extract clear extreme value times. Based on the fault direction criterion, the faults on lines II, III, and IV can be determined to be caused by BU. 15 With BU 18 For faults within the designated protection zone, the two protection devices should isolate the faulty zone, and the protection action results are shown in Table 2.
[0137] Table 2
[0138]
[0139] This invention acquires the current traveling wave waveform after a fault occurs, extracts extreme points using traveling wave filtering and differential algorithms, and calculates the extreme time of the traveling wave at each protection device. Furthermore, by comparing the extreme time differences between adjacent protection nodes, a fault identification criterion for internal and external protection zones is constructed, and combined with a preset extreme time threshold, the fault direction and the protected section where the fault is located are determined. Because this invention reveals a propagation law where the extreme time of the traveling wave is linearly related to the number of branch units traversed and is independent of transition resistance, stable and accurate section location can be achieved even under high-resistance or remote fault conditions. Experimental results show that, compared to existing protection technologies that rely on communication or steady-state quantities, this invention has stronger robustness and protection coverage in scenarios without communication conditions and with damaged measuring devices.
[0140] Although embodiments of the invention have been described in conjunction with the accompanying drawings, those skilled in the art can make various modifications and variations without departing from the spirit and scope of the invention, and such modifications and variations all fall within the scope defined by the appended claims.
Claims
1. A protection method for a submarine observation network power supply system based on traveling wave extreme time, characterized in that, include: S1. Based on the network structure, power supply path and node importance of the power supply system of the submarine observation network, determine the location of the protection device and divide the power supply system of the submarine observation network into multiple protection zones. Each protection zone is defined by two adjacent protection devices. S2. Real-time acquisition of voltage signals at the deployment location, and first-order differentiation processing of the voltage signals to obtain the corresponding voltage change rate; In addition, the voltage change rate is compared with a preset voltage threshold to determine whether a fault has occurred; S3. When a fault is determined to occur, record the current traveling wave data within a preset time period before and after the fault occurs, and use a Gaussian filter to perform noise reduction processing on the current traveling wave data. S4. Based on the denoised current traveling wave data, calculate the extreme value time t of the traveling wave using the differential gradient algorithm. 极值 ; S5. Compare the extreme time t of the traveling wave in the branch units on both sides of the protected area. Mi and t Mo According to the time difference Δt = t Mi -t Mo With direction criterion threshold [t] low , t up The relationship between the fault and the direction of the fault is determined. S6, the extreme time t of the traveling wave 极值 Compared with the preset remote fault threshold t of the section re Compare the data to determine whether the fault is within the protected area; S7. When the fault simultaneously satisfies the forward direction criterion Δt≤t low and the criteria within the area t 极值 ≤t re If the fault occurs, it is determined to be a fault within the zone, and a protection action command is generated to isolate the protected zone where the fault is located.
2. The method according to claim 1, characterized in that, The layout of the protection device in S1 must meet the following requirements: Redundancy principle: Install protection devices in branch units that connect multiple trunk lines; Critical load protection principle: Protective devices must be configured for branch units connected to high-priority observation equipment; Waveform observability principle: The number of branch units between adjacent protection devices is less than or equal to a preset threshold number N. BU ; The preset quantity threshold N BU The following constraints must be met: in, and These represent the current traveling wave passing through N under high-impedance fault conditions. BU Amplitude and extreme time after each branch unit; I set.low The smallest current amplitude that the measuring device can resolve; L line This represents the maximum length of the line between two adjacent protection devices, where v is the traveling wave propagation speed. Link uniqueness principle: the traveling wave propagation path between adjacent protection devices is unique; when there is a branch line between adjacent protection devices, a protection device is added at the branch point.
3. The method according to claim 1, characterized in that, The criteria for determining the occurrence of a fault in S2 are as follows: Δu(k)=u fault (t k )-u rated Where k is the sampling point number, and k = 1, 2, ..., P; P is the number of sampling points in a data window; t k The time corresponding to the kth sampling point; u fault (t k ) represents the fault state at t k The voltage value measured at any time; u rated The rated voltage is Δu(k); Δu(k) represents the voltage change; Δu set This is a preset voltage threshold. The voltage threshold is set to ensure that the protection device can still reliably start the protection function when a remote fault occurs under high grounding resistance conditions.
4. The method according to claim 1, characterized in that, The expression for the Gaussian filter in S3 is: Where σ is the standard deviation; G(i) is the time-domain response function of the Gaussian filter; and i is the time variable, with the fault occurrence time t=0 as the starting point.
5. The method according to claim 1, characterized in that, The calculation of the traveling wave extremum time in S4 uses the differential gradient algorithm, and its expression is: Where i(t) represents the current traveling wave data; t0 represents the extreme point of the current traveling wave data after noise reduction; and Δt represents the time window. When the gradient value changes from positive to negative, it is marked as an extreme point. The time corresponding to the first occurrence of the extreme point is the extreme time t of the traveling wave. 极值 .
6. The method according to claim 1, characterized in that, The criterion for determining the fault direction in S5 is as follows: If Δt≥t up The fault was determined to be a back-side fault, and the protection action was not triggered. If Δt≤t low The fault is identified as a forward fault, and the process proceeds to the area determination procedure.
7. The method according to claim 1, characterized in that, The criterion for determining whether a fault is within the protection zone in S6 is: When the extreme value of the traveling wave is t 极值 >t re When the fault is located outside the protected area, it is determined that the fault is outside the protected area. When the extreme value of the traveling wave is t 极值 ≤t re When the fault is located outside the protected area, it is determined that the fault is outside the protected area. Among them, t re The preset fault threshold at the far end of the section is set according to the maximum traveling wave extreme time when a fault occurs at the end of the protected area.
8. The method according to claim 7, characterized in that, The S7 also includes: S71. Number the lines in the network structure of the power supply system for the submarine observation network. S72. Preset the extreme time intervals for each line; S73. Obtain the traveling wave extreme time detected by the protection devices at both ends of the protection area where the fault is located, and match the traveling wave extreme time detected by the protection devices at both ends to the target extreme time intervals respectively, and output the target line number corresponding to the target extreme time interval.
Citation Information
Cited By
Multi-branch area protection method for direct current power supply system in extreme scene
CN121584504A
A multi-branch area protection method for an extreme scenario direct current power supply system
CN121584504B