Electric power system fault analysis method and system based on artificial intelligence

Abstract

The invention relates to the technical field of power faults, and provides a power system fault analysis method and system based on artificial intelligence, and the method comprises the steps: obtaining a relay protection action message of power equipment under a preset protection mechanism through an interactive relay protection device; wherein the relay protection action message is provided with a first timestamp; fault recording information is obtained through an interactive fault recorder; wherein the fault recording information carries a second timestamp; taking the second timestamp as a convergence target, and correcting the first timestamp based on an adaptive filtering algorithm to obtain a corrected first timestamp; the relay protection action message is updated according to the corrected first timestamp; and comparing and analyzing the fault recording information and the updated relay protection action message, and generating a power system fault report according to a comparison and analysis result. According to the method and the system, the difficulty and the complexity of artificial intelligence analysis of power system faults are remarkably reduced, and meanwhile, the accuracy of fault analysis is improved.

Description

Technical Field

[0001] The present invention relates to the technical field of power fault analysis, and in particular to an artificial intelligence-based power system fault analysis method and system. Background Art

[0002] The reliability and safety of the power system are important factors in ensuring the stable operation of the power grid. After a power system fault occurs, the relay protection device performs a protection action to protect the safety of equipment and personnel and prevent the expansion of local problems. The relay protection device generates an action message when performing a protection action. Combined with the fault waveform and data recorded by the fault recorder when the fault occurs, it can help professionals quickly diagnose the cause of the fault, evaluate the system status and formulate a recovery strategy.

[0003] Current power system fault analysis mainly relies on relay protection professionals to manually compare the action messages of relay protection devices with the fault waveforms and data recorded by fault recorders, which is not only time-consuming but also easily affected by human factors.

[0004] With the development of artificial intelligence technology, many studies have begun to try to use intelligent algorithms to replace manual methods to achieve automated fault analysis and report generation. Although this direction has great application potential, it still faces severe challenges in practical applications. Among them, when the devices are physically and strictly synchronized, the data returned by the fault recorder and the relay protection device still have timestamp deviations. For example, the action message of the relay protection device has 643ms "longitudinal differential protection action" and 644ms "distance acceleration action"; while the fault waveform and data show 645ms "distance acceleration action". The same action protection has timestamp deviations on the fault recorder and relay protection device action messages, which brings difficulties to artificial intelligence analysis. When the timestamp deviation is large or the same timestamp has different categories of information, it may even lead to incorrect fault analysis results, thereby affecting the timeliness and accuracy of fault handling. Summary of the invention

[0005] To this end, the present invention aims to overcome the problem in the prior art that the deviation in the timestamps of the data returned by the fault recorder and the relay protection device causes inaccurate artificial intelligence analysis, and to provide an artificial intelligence-based power system fault analysis method and system, which significantly reduces the difficulty and complexity of artificial intelligence analysis of power system faults, while improving the accuracy of fault analysis.

[0006] In the first aspect, in order to solve the above technical problems, the present invention provides a power system fault analysis method based on artificial intelligence, comprising:

[0007] Obtaining a relay protection action message of the power equipment under a predetermined protection mechanism through an interactive relay protection device; wherein the relay protection action message carries a first timestamp;

[0008] Acquiring fault recording information through an interactive fault recorder; wherein the fault recording information carries a second timestamp;

[0009] Taking the second timestamp as a convergence target and correcting the first timestamp based on an adaptive filtering algorithm to obtain a corrected first timestamp; updating the relay protection action message according to the corrected first timestamp;

[0010] Comparing and analyzing the fault recording information and the updated relay protection action message, and generating a power system fault report according to the comparison and analysis results;

[0011] Among them, comparing and analyzing the fault recording information and the updated relay protection action message, and generating a power system fault report according to the comparison and analysis results, including:

[0012] Step 1: determine whether the sampling data of the relay protection action message is correct, if it is correct, proceed to step 2; otherwise, generate the first fault information;

[0013] Step 2: Determine whether the fault type of the relay protection action message is correct, if it is correct, proceed to step 3; otherwise, generate the second fault information;

[0014] Step three: determine whether the fault current and fault voltage of the relay protection action message are within the range of the set value list. If so, generate the power system fault report; otherwise, generate the third fault information.

[0015] In one embodiment of the present invention, taking the second timestamp as a convergence target and correcting the first timestamp based on an adaptive filtering algorithm to obtain a corrected first timestamp includes:

[0016] Get the first timestamp Z k and the second timestamp D k , and calculate the difference ΔT for each pair of timestamps k =D k -Z k ;

[0017] Construct a minimum mean square error filter and initialize its weight vector W(0);

[0018] The weight vector W(k) is iteratively updated according to the following operator:

[0019] W(k)=W(k-1)-μ(W T (k-1)u(k))*u(k);

[0020] Use the updated weight vector W(k) to update the first timestamp Z k Make corrections and obtain the first timestamp after correction

[0021] Among them, Z k represents the first timestamp at the kth time step; D k Indicates the second timestamp at the kth time step; ΔT k represents the difference at the kth time step; u(k) is the input signal of the minimum mean square error filter, u(k) = ΔT k ; μ is the step size parameter of the minimum mean square error filter, which is set between 0.01 and 0.1; W(k) is the weight vector of the kth iteration; W T (k) is the transpose of W(k); W(k-1) is the weight vector for the k-1th iteration; W T (k-1) is the transpose of W(k-1); is the first corrected timestamp at the kth time step.

[0022] In one embodiment of the present invention, the method further includes adjusting the number of iterations and / or the step size parameter so that the difference between each pair of timestamps converges to zero.

[0023] In one embodiment of the present invention, the step 1 includes configuring a circuit analysis module, the circuit analysis module includes at least a first analysis channel, the first analysis channel is used to compare the maximum amplitude of the fault current, the maximum amplitude of the fault voltage and the fault recording information;

[0024] Among them, the fault recording information includes the recorded current amplitude and the recorded voltage amplitude; when the difference between the maximum amplitude of the fault current and the recorded current amplitude meets the current difference threshold, and the difference between the maximum amplitude of the fault voltage and the recorded voltage amplitude meets the voltage difference threshold, it is judged that the sampling of the relay protection action message is correct.

[0025] In one embodiment of the present invention, the circuit analysis module further includes a second analysis channel, through which it is determined whether the phase variable of the relay protection device satisfies a predetermined phase variable constraint; the phase variable constraint includes a phase current constraint, a phase voltage constraint and a phase constraint;

[0026] wherein, judging whether the phase current of the relay protection device satisfies the phase current constraint, and obtaining a first constraint judgment result;

[0027] Determine whether the phase voltage of the relay protection device satisfies the phase voltage constraint, and obtain a second constraint judgment result;

[0028] Determining whether the phase of the relay protection device satisfies the phase constraint, and obtaining a third constraint judgment result;

[0029] When the first constraint judgment result, the second constraint judgment result and the third constraint judgment result are all satisfied, it is determined that the phase variable of the relay protection device satisfies the phase variable constraint;

[0030] When it is determined that the phase variable of the relay protection device satisfies the phase variable constraint, it is determined that the fault type generated by the relay protection device is correct.

[0031] In one embodiment of the present invention, a third analysis channel is further included, through which it is determined whether the fault electrical parameter of the relay protection device meets the range of the setting list;

[0032] wherein, determining whether the fault current of the relay protection device is within the range of the setting list to obtain a first determination result;

[0033] Determine whether the fault voltage of the relay protection device meets the range of the setting list, and obtain a second determination result;

[0034] Determine whether the measured impedance of the relay protection device meets the range of the setting list, and obtain a third determination result;

[0035] When the first judgment result, the second judgment result and the third judgment result are all in compliance, it is determined that the fault electrical parameter of the relay protection device is in compliance with the range of the setting list;

[0036] When it is determined that the fault electrical parameters of the relay protection device are within the range of the setting list, the power system fault report is generated.

[0037] In one embodiment of the present invention, the setting list includes protection setting, and the protection setting is configured by the following method:

[0038] Configure the monitoring period and the initial protection setting of the relay protection device;

[0039] Execute the initial protection setting to operate the relay protection device, and determine the protection action frequency of the relay protection device according to the relay protection action message and the fault recording information;

[0040] Determining the protection action frequency of the relay protection device according to the monitoring period and the protection action frequency of the relay protection device;

[0041] Determine a protection setting adjustment strategy according to the protection action frequency of the relay protection device, adjust the initial protection setting according to the protection setting adjustment strategy, and obtain an adjusted protection setting;

[0042] Execute step three according to the adjusted protection setting.

[0043] In one embodiment of the present invention, the protection setting value includes an overload protection setting value; the protection setting value adjustment strategy includes:

[0044] Compare the protection action frequency of the relay protection device with the first threshold value. When the protection action frequency of the relay protection device is greater than or equal to the first threshold value, the protection setting adjustment strategy is to increase the protection setting strategy. When the protection action frequency of the relay protection device is less than the first threshold value, determine the target excessive protection setting as the initial protection setting.

[0045] In one embodiment of the present invention, the protection setting includes an under-voltage protection setting; the protection setting adjustment strategy includes:

[0046] Compare the protection action frequency of the relay protection device and the second threshold value. When the protection action frequency of the relay protection device is greater than or equal to the second threshold value, the protection setting adjustment strategy is a protection setting reduction strategy; when the protection action frequency of the relay protection device is less than the second threshold value, determine the target under-quantity protection setting as the initial protection setting.

[0047] In a second aspect, in order to solve the above-mentioned technical problems, the present invention also provides an artificial intelligence-based power system fault analysis system for executing the above-mentioned artificial intelligence-based power system fault analysis method.

[0048] The above technical solution of the present invention has the following beneficial effects compared with the prior art:

[0049] The artificial intelligence-based power system fault analysis method and system described in the present invention take the second timestamp as the convergence target and correct the first timestamp based on an adaptive filtering algorithm to obtain a corrected first timestamp; update the relay protection action message according to the corrected first timestamp; thereby aligning the timestamps of the information carried by the fault recorder and the action message of the relay protection device before comparing and analyzing the fault recording data of the fault recorder and the action message of the relay protection device with the help of artificial intelligence, thereby solving the problem of inaccurate artificial intelligence analysis caused by the timestamp deviation of the data returned by the fault recorder and the relay protection device, significantly reducing the difficulty and complexity of artificial intelligence analysis of power system faults, and improving the accuracy of fault analysis. BRIEF DESCRIPTION OF THE DRAWINGS

[0050] In order to make the content of the present invention more clearly understood, the present invention is further described in detail below according to specific embodiments of the present invention in conjunction with the accompanying drawings, wherein:

[0051] Figure 1It is a flow chart of a power system fault analysis method based on artificial intelligence in a preferred embodiment of the present invention;

[0052] Figure 2 Flow chart of correcting the first timestamp in a preferred embodiment of the present invention

[0053] Figure 3 The flowchart of generating a power system fault report in a preferred embodiment of the present invention. DETAILED DESCRIPTION

[0054] The present invention is further described below in conjunction with the accompanying drawings and specific embodiments so that those skilled in the art can better understand the present invention and implement it, but the embodiments are not intended to limit the present invention.

[0055] In one embodiment, referring to Figure 1 As shown, a power system fault analysis method based on artificial intelligence is disclosed, the method comprising:

[0056] Step S100: obtaining a relay protection action message of the power equipment under a predetermined protection mechanism through an interactive relay protection device; wherein the relay protection action message carries a first timestamp;

[0057] In a specific application scenario, the relay protection device operates under a predetermined protection mechanism and generates an action message, extracting key data of the relay protection device action message, including fault type; fault phase; fault occurrence time; fault duration; action behavior of the relay protection device; fault distance measurement; fault voltage and current amplitude; fault phase impedance; each phase current value, each phase voltage value and each phase phase. Each data of the relay protection action message carries a first timestamp, for example,

[0058] 0000ms protection starts

[0059] 0008ms B power frequency variation impedance action

[0060] 0016ms B longitudinal differential protection action

[0061] 0016ms B distance I stage action

[0062] 0553ms reclosing action.

[0063] 0000ms, 0008ms, 0016ms, and 0553ms are the first timestamps.

[0064] Step S200: acquiring fault recording information through an interactive fault recorder; wherein the fault recording information carries a second timestamp;

[0065] In a specific application scenario, the fault recorder starts recording after a fault occurs and generates a fault recording waveform and recording data, and extracts characteristic values ​​of the waveform recorded by the fault recorder, including the duration of the fault waveform; the amplitude of the fault voltage and current; the fault phase difference; the fault distance measurement, etc. Each piece of data recorded by the fault recorder carries a second timestamp.

[0066] Step S300: taking the second timestamp as the convergence target and correcting the first timestamp based on the adaptive filtering algorithm to obtain a corrected first timestamp; updating the relay protection action message according to the corrected first timestamp; comparing and analyzing the fault recording information and the updated relay protection action message, and generating a power system fault report according to the comparison and analysis results;

[0067] In specific application scenarios, when the devices (relay protection device and fault recorder) are physically and strictly time synchronized, the data returned by the fault recorder and the relay protection device still have timestamp deviations. For example, the action message of the relay protection device shows 0643ms "longitudinal differential protection action" and 0644ms "distance acceleration action"; while the fault waveform and data show 0645ms "distance acceleration action". The same action protection has timestamp deviations in the fault recorder and relay protection device action messages. In order to overcome the problem of timestamp deviations in the data returned by the relay protection device and the fault recorder, the first timestamp is corrected after obtaining the action message of the relay protection device and before comparing and processing the action message and the recorded fault information. The goal is to make the first timestamp consistent with the second timestamp, and the first timestamp received by correction replaces the first timestamp before correction to update the relay protection action message.

[0068] Specifically, refer to Figure 2 As shown, the process of correcting the first timestamp is as follows:

[0069] Get the first timestamp Z k and the second timestamp D k , and calculate the difference ΔT for each pair of timestamps k =D k -Z k ;

[0070] Construct a minimum mean square error filter and initialize its weight vector W(0);

[0071] The weight vector W(k) is iteratively updated according to the following operator:

[0072] W(k)=W(k-1)-μ(W T (k-1)u(k))*u(k);

[0073] Use the updated weight vector W(k) to update the first timestamp Zk Make corrections and obtain the first timestamp after correction

[0074] The difference between each pair of time stamps is made to converge to zero by adjusting the number of iterations of the minimum mean square error filter and / or the step size parameter.

[0075] Among them, Z k represents the first timestamp at the kth time step; D k Indicates the second timestamp at the kth time step; ΔT k represents the difference at the kth time step; u(k) is the input signal of the minimum mean square error filter, u(k) = ΔT k ; μ is the step size parameter of the minimum mean square error filter, which is set between 0.01 and 0.1; W(k) is the weight vector of the kth iteration; W T (k) is the transpose of W(k); W(k-1) is the weight vector for the k-1th iteration; W T (k-1) is the transpose of W(k-1); is the first corrected timestamp at the kth time step.

[0076] The first timestamp and the second timestamp usually have the following characteristics: 1. There is a deviation between the two timestamps of the same event, and the deviation between the two timestamps is at the millisecond level; 2. There may be other action events between the two timestamps of the same event. In view of the above characteristics of the first timestamp and the second timestamp, it is preferred to use a minimum mean square error filter to correct the first timestamp, which has the following technical advantages:

[0077] Since the deviation between the two timestamps is in the millisecond level and the resolution is high, the minimum mean square error filter can respond quickly and make adjustments to ensure that the error of the timestamp can be corrected within a reasonable range. At the same time, for multi-event environments, the minimum mean square error filter is adjusted based on minimization of the square of the error, which can effectively filter out the influence of noise and non-target events; in the case of other action events between timestamps, this robustness can ensure that the timestamps of the main events are optimized first. Through continuous iterative optimization, the error of the timestamp can be significantly reduced, reducing the analysis errors caused by time differences, thereby improving the accuracy and reliability of subsequent fault analysis. At the same time, in an environment with high real-time requirements for power systems, the minimum mean square error filter can achieve real-time updates with low computational overhead and adapt to dynamically changing situations.

[0078] The artificial intelligence-based power system fault analysis method described in the present invention takes the second timestamp as the convergence target and corrects the first timestamp based on an adaptive filtering algorithm to obtain a corrected first timestamp; updates the relay protection action message according to the corrected first timestamp; thereby, before comparing and analyzing the fault recording data of the fault recorder and the action message of the relay protection device with the help of artificial intelligence, the information carried by both is timestamp aligned, thereby solving the problem of inaccurate artificial intelligence analysis caused by timestamp deviation in the data returned by the fault recorder and the relay protection device, significantly reducing the difficulty and complexity of artificial intelligence analysis of power system faults, and improving the accuracy of fault analysis.

[0079] Further, refer to Figure 3 As shown, the fault recording information and the updated relay protection action message are compared and analyzed, and a power system fault report is generated according to the comparison and analysis results, including

[0080] Step 1: determine whether the sampling data of the relay protection action message is correct, if it is correct, proceed to step 2; otherwise, generate first fault information; the first fault information is used to describe the sampling data error of the relay protection action message;

[0081] Step 2: Determine whether the fault type of the relay protection action message is correct, if correct, proceed to step 3; otherwise, generate second fault information; the second fault information is used to describe that the fault type of the relay protection action message is wrong;

[0082] Step three: Determine whether the fault current and fault voltage of the relay protection action message are within the range of the set value list. If so, generate the power system fault report; otherwise, generate the third fault information; the third fault information is used to describe that the fault current and / or fault voltage of the relay protection action message are not within the requirements of the set value list.

[0083] In the specific application scenario, step one, step two and step three have a time sequence relationship. Step two is executed after step one, and step three is executed after step two, until the comparison and analysis is completed to generate a power system fault report; wherein, the power system fault report can be generated in combination with the results of each step in conjunction with the template. Only after judging that the relay protection device has correctly sampled, further judge whether the fault type of the relay protection action message is correct to ensure the authenticity, validity and reliability of the relay protection action message; after judging that the fault type is correct, continue to identify and judge whether the fault current and fault voltage are within the requirements of the set value sheet to ensure that the protection action of the relay protection device is correct.

[0084] Specifically, the step 1 includes configuring a circuit analysis module, the circuit analysis module includes a first analysis channel, and the first analysis channel is used to compare the maximum amplitude of the fault current, the maximum amplitude of the fault voltage and the fault recording information;

[0085] Among them, the fault recording information includes the recorded current amplitude and the recorded voltage amplitude; when the difference between the maximum amplitude of the fault current and the recorded current amplitude meets the current difference threshold, and the difference between the maximum amplitude of the fault voltage and the recorded voltage amplitude meets the voltage difference threshold, it is judged that the sampling of the relay protection action message is correct.

[0086] The circuit analysis module also includes a second analysis channel, through which it is determined whether the phase variable of the relay protection device meets a predetermined phase variable constraint; the phase variable constraint includes a phase current constraint, a phase voltage constraint and a phase constraint;

[0087] Among them, judging whether the phase current of the relay protection device satisfies the phase current constraint, and obtaining a first constraint judgment result; judging whether the phase voltage of the relay protection device satisfies the phase voltage constraint, and obtaining a second constraint judgment result; judging whether the phase of the relay protection device satisfies the phase constraint, and obtaining a third constraint judgment result; when the first constraint judgment result, the second constraint judgment result and the third constraint judgment result are all satisfied, it is determined that the phase variable of the relay protection device satisfies the phase variable constraint; when it is determined that the phase variable of the relay protection device satisfies the phase variable constraint, it is determined that the fault type generated by the relay protection device is correct.

[0088] In a specific application scenario of one of the embodiments, the fault type recorded in the relay protection action message is B phase distance I section action; the corresponding fault recording data extracts 1) the B phase current increases and the voltage decreases; zero-sequence voltage and zero-sequence current appear at the same time; 2) the B phase voltage leads the fault phase current by about 80 degrees, and the zero-sequence current leads the zero-sequence voltage by about 100 degrees; 3) the zero-sequence current phase is the same as the B phase current phase, and the zero-sequence voltage phase is the same as the B phase voltage phase). By extracting the above three sets of information from the fault recording data, it can be determined that the fault type recorded in the relay protection action message is correct.

[0089] As described above, accurate judgment of whether sampling is correct and whether fault type recording is correct is achieved through the first analysis channel and the second analysis channel of the circuit analysis module, with an accuracy rate of more than 99%.

[0090] Furthermore, it also includes a third analysis channel, through which it is determined whether the fault electrical parameter of the relay protection device meets the range of the setting list;

[0091] Among them, judging whether the fault current of the relay protection device is within the range of the set value list, and obtaining a first judgment result; judging whether the fault voltage of the relay protection device is within the range of the set value list, and obtaining a second judgment result; judging whether the measured impedance of the relay protection device is within the range of the set value list, and obtaining a third judgment result; when the first judgment result, the second judgment result and the third judgment result are all in compliance, it is determined that the fault electrical parameters of the relay protection device are within the range of the set value list; when it is determined that the fault electrical parameters of the relay protection device are within the range of the set value list, the power system fault report is generated.

[0092] Only when the fault current, fault voltage and measured impedance all meet the requirements of their respective set value sheets, can the fault electrical parameters be deemed to be within the range of the set value sheets, thereby generating an accurate power system fault report and realizing artificial intelligence generation of power system fault reports.

[0093] The scope of the above-mentioned set value list is pre-set according to the circuit topology of the power system, as well as the working principle and working process of the relay protection device. The conventional set value list requires setting, that is, fixing. However, as the operating time of the relay protection device increases, its performance indicators will change dynamically, for example, line aging, equipment operating fatigue, etc. The fixity of the set value list will affect the accuracy of fault analysis to a certain extent.

[0094] In order to solve the above problems, in the embodiment of the present invention, the set value single requirement is dynamically adjusted according to the set value adjustment strategy according to the actual operation needs. The solution is as follows:

[0095] The setting list includes protection setting values, which are configured by the following method:

[0096] Configure a monitoring period and an initial protection setting of the relay protection device; execute the initial protection setting to run the relay protection device, and determine the protection action frequency of the relay protection device according to the relay protection action message and the fault recording information; determine the protection action frequency of the relay protection device according to the monitoring period and the protection action frequency of the relay protection device; determine a protection setting adjustment strategy according to the protection action frequency of the relay protection device, adjust the initial protection setting according to the protection setting adjustment strategy, and obtain the adjusted protection setting; execute step three according to the adjusted protection setting.

[0097] Among them, the protection settings include over-protection settings and under-protection settings. For over-protection, the measured value needs to be greater than the set value (i.e., over-protection setting); for under-protection, the measured value needs to be less than the set value (i.e., under-protection setting) to determine whether the relay protection device is operating correctly.

[0098] When the line is aged or the equipment is fatigued, there is a deviation between the measured value and the true value of the relay protection device. For example, the measured value of the differential current of the relay protection device is 36.90A, while the true value of the differential current is 38.81A. When the initial differential current setting is set to 37.63, the relay protection device malfunctions. In another case, the measured impedance of the relay protection device is 0.75 ohms, while the true value of the impedance is 1.45 ohms. When the initial impedance setting is set to 1.26 ohms, the relay protection device malfunctions.

[0099] When the protection setting is an excessive protection setting, the protection setting adjustment strategy includes comparing the protection action frequency of the relay protection device and a first threshold value; when the protection action frequency of the relay protection device is greater than or equal to the first threshold value, the protection setting adjustment strategy is an increase protection setting strategy; when the protection action frequency of the relay protection device is less than the first threshold value, the target excessive protection setting is determined to be the initial protection setting.

[0100] When the protection setting includes an under-quantity protection setting, the protection setting adjustment strategy includes comparing the protection action frequency of the relay protection device and a second threshold value; when the protection action frequency of the relay protection device is greater than or equal to the second threshold value, the protection setting adjustment strategy is a protection setting reduction strategy; when the protection action frequency of the relay protection device is less than the second threshold value, the target under-quantity protection setting is determined to be the initial protection setting.

[0101] The first threshold and the second threshold are preset based on experience or experimental results.

[0102] Based on the protection action frequency of the relay protection device, a fixed value adjustment strategy is formulated to dynamically adjust the fixed values ​​in the fixed value list, so as to adapt to the operation of the relay protection device, improve the action accuracy of the relay protection device, and thus improve the accuracy and effectiveness of fault analysis.

[0103] In another embodiment, the present invention also discloses an artificial intelligence-based power system fault analysis system, which is used to execute the above-mentioned artificial intelligence-based power system fault analysis method, and can effectively realize artificial intelligence analysis of white power system faults. The technical effects that can be achieved are as described in the above-mentioned embodiments and will not be repeated here.

[0104] In summary, the artificial intelligence-based power system fault analysis method and system described in the present invention takes the second timestamp as the convergence target and corrects the first timestamp based on an adaptive filtering algorithm to obtain a corrected first timestamp; updates the relay protection action message according to the corrected first timestamp; thereby aligning the timestamps of the information carried by the fault recorder and the action message of the relay protection device before comparing and analyzing the fault recording data of the fault recorder and the action message of the relay protection device with the help of artificial intelligence, thereby solving the problem of inaccurate artificial intelligence analysis of the data returned by the fault recorder and the relay protection device due to the timestamp deviation, significantly reducing the difficulty and complexity of artificial intelligence analysis of power system faults, and improving the accuracy of fault analysis.

[0105] Those skilled in the art will appreciate that the embodiments of the present application may be provided as methods, systems, or computer program products. Therefore, the present application may adopt the form of a complete hardware embodiment, a complete software embodiment, or an embodiment in combination with software and hardware. Moreover, the present application may adopt the form of a computer program product implemented in one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) that include computer-usable program code.

[0106] The present application is described with reference to the flowcharts and / or block diagrams of the methods, devices (systems), and computer program products according to the embodiments of the present application. It should be understood that each process and / or box in the flowchart and / or block diagram, as well as the combination of the processes and / or boxes in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to generate a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowchart and / or block diagram. Figure 1 A process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0107] These computer program instructions may also be stored in a computer-readable memory capable of directing a computer or other programmable data processing device to operate in a specific manner, so that the instructions stored in the computer-readable memory produce an article of manufacture comprising an instruction device, which implements the process Figure 1 A process or multiple processes and / or boxes Figure 1 A function specified in one or more boxes.

[0108] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operating steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing instructions for implementing the process. Figure 1 A process or multiple processes and / or boxes Figure 1 The steps for the functions specified in one or more boxes.

[0109] Obviously, the above embodiments are merely examples for clear explanation and are not intended to limit the implementation methods. For those skilled in the art, other different forms of changes or modifications can be made based on the above description. It is not necessary and impossible to list all the implementation methods here. The obvious changes or modifications derived from these are still within the protection scope of the invention.

Claims

1. A power system fault analysis method based on artificial intelligence, characterized in that: include, Obtaining a relay protection action message of the power equipment under a predetermined protection mechanism through an interactive relay protection device; wherein the relay protection action message carries a first timestamp; Acquiring fault recording information through an interactive fault recorder; wherein the fault recording information carries a second timestamp; Taking the second timestamp as a convergence target and correcting the first timestamp based on an adaptive filtering algorithm to obtain a corrected first timestamp; updating the relay protection action message according to the corrected first timestamp; Comparing and analyzing the fault recording information and the updated relay protection action message, and generating a power system fault report according to the comparison and analysis results; Among them, comparing and analyzing the fault recording information and the updated relay protection action message, and generating a power system fault report according to the comparison and analysis results, including: Step 1: determine whether the sampling data of the relay protection action message is correct, if it is correct, proceed to step 2; otherwise, generate the first fault information; Step 2: Determine whether the fault type of the relay protection action message is correct, if it is correct, proceed to step 3; otherwise, generate the second fault information; Step three: determine whether the fault current and fault voltage of the relay protection action message are within the range of the set value list. If so, generate the power system fault report; otherwise, generate the third fault information.

2. The power system fault analysis method based on artificial intelligence according to claim 1 is characterized in that: Taking the second timestamp as a convergence target and correcting the first timestamp based on an adaptive filtering algorithm to obtain a corrected first timestamp, comprising: Get the first timestamp Z k and the second timestamp D k , and calculate the difference ΔT for each pair of timestamps k =D k -Z k ; Construct a minimum mean square error filter and initialize its weight vector W(0); The weight vector W(k) is iteratively updated according to the following operator: W(k)=W(k-1)-μ(W T (k-1)u(k))*u(k); Use the updated weight vector W(k) to update the first timestamp Z k Make corrections and obtain the first timestamp after correction Among them, Z k represents the first timestamp at the kth time step; D k Indicates the second timestamp at the kth time step; ΔT k represents the difference at the kth time step; u(k) is the input signal of the minimum mean square error filter, u(k) = ΔT k ; μ is the step size parameter of the minimum mean square error filter, which is set between 0.01 and 0.1; W(k) is the weight vector of the kth iteration; W T (k) is the transpose of W(k); W(k-1) is the weight vector for the k-1th iteration; W T (k-1) is the transpose of W(k-1); is the first corrected timestamp at the kth time step.

3. The power system fault analysis method based on artificial intelligence according to claim 2 is characterized in that: The method also includes adjusting the number of iterations and / or the step size parameter so that the difference between each pair of time stamps converges to zero.

4. The power system fault analysis method based on artificial intelligence according to claim 1 or 2, characterized in that: The step 1 includes configuring a circuit analysis module, wherein the circuit analysis module includes at least a first analysis channel, wherein the first analysis channel is used to compare the maximum amplitude of the fault current, the maximum amplitude of the fault voltage, and the fault recording information; Among them, the fault recording information includes the recorded current amplitude and the recorded voltage amplitude; when the difference between the maximum amplitude of the fault current and the recorded current amplitude meets the current difference threshold, and the difference between the maximum amplitude of the fault voltage and the recorded voltage amplitude meets the voltage difference threshold, it is judged that the sampling of the relay protection action message is correct.

5. The power system fault analysis method based on artificial intelligence according to claim 4 is characterized in that: The circuit analysis module also includes a second analysis channel, through which it is determined whether the phase variable of the relay protection device meets a predetermined phase variable constraint; the phase variable constraint includes a phase current constraint, a phase voltage constraint and a phase constraint; wherein, judging whether the phase current of the relay protection device satisfies the phase current constraint, and obtaining a first constraint judgment result; Determining whether the phase voltage of the relay protection device satisfies the phase voltage constraint, and obtaining a second constraint judgment result; Determining whether the phase of the relay protection device satisfies the phase constraint, and obtaining a third constraint judgment result; When the first constraint judgment result, the second constraint judgment result and the third constraint judgment result are all satisfied, it is determined that the phase variable of the relay protection device satisfies the phase variable constraint; When it is determined that the phase variable of the relay protection device satisfies the phase variable constraint, it is determined that the fault type generated by the relay protection device is correct.

6. The power system fault analysis method based on artificial intelligence according to claim 1 is characterized in that: It also includes a third analysis channel, through which it is determined whether the fault electrical parameter of the relay protection device meets the range of the setting list; wherein, determining whether the fault current of the relay protection device is within the range of the setting list to obtain a first determination result; Determine whether the fault voltage of the relay protection device meets the range of the setting list, and obtain a second determination result; Determine whether the measured impedance of the relay protection device meets the range of the setting list, and obtain a third determination result; When the first judgment result, the second judgment result and the third judgment result are all in compliance, it is determined that the fault electrical parameter of the relay protection device is in compliance with the range of the setting list; When it is determined that the fault electrical parameters of the relay protection device are within the range of the setting list, the power system fault report is generated.

7. The power system fault analysis method based on artificial intelligence according to claim 1 or 6, characterized in that: The setting list includes protection setting values, which are configured by the following method: Configure the monitoring period and the initial protection setting of the relay protection device; Execute the initial protection setting to operate the relay protection device, and determine the protection action frequency of the relay protection device according to the relay protection action message and the fault recording information; Determining the protection action frequency of the relay protection device according to the monitoring period and the protection action frequency of the relay protection device; Determine a protection setting adjustment strategy according to the protection action frequency of the relay protection device, adjust the initial protection setting according to the protection setting adjustment strategy, and obtain an adjusted protection setting; Execute step three according to the adjusted protection setting.

8. The power system fault analysis method based on artificial intelligence according to claim 7 is characterized in that: The protection setting value includes an overload protection setting value; the protection setting value adjustment strategy includes: Comparing the protection action frequency of the relay protection device with a first threshold, when the protection action frequency of the relay protection device is greater than or equal to the first threshold, the protection setting value adjustment strategy is to increase the protection setting value strategy; When the protection action frequency of the relay protection device is less than a first threshold, the target over-protection setting is determined to be the initial protection setting.

9. The power system fault analysis method based on artificial intelligence according to claim 7 is characterized in that: The protection setting includes an under-voltage protection setting; the protection setting adjustment strategy includes: Compare the protection action frequency of the relay protection device and the second threshold value. When the protection action frequency of the relay protection device is greater than or equal to the second threshold value, the protection setting adjustment strategy is a protection setting reduction strategy; when the protection action frequency of the relay protection device is less than the second threshold value, determine the target under-quantity protection setting as the initial protection setting.

10. An artificial intelligence-based power system fault analysis system, characterized in that: It is used to execute the artificial intelligence-based power system fault analysis method as described in any one of claims 1-9.

Images