Power network information security detection method and system based on artificial intelligence
By calculating the segmented distortion rate and coherence of voltage data of power equipment, and combining it with the local outlier factor algorithm, the problem of misjudgment in low-density areas of traditional algorithms is solved, the accuracy of power grid information security detection and equipment fault prediction capability are improved, and maintenance and interruption costs are reduced.
Patent Information
- Application Number
- CN202411154166.9
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2024-08-22
- Publication Date
- 2024-11-19
- Estimated Expiration
- 2044-08-22
AI Technical Summary
Traditional local outlier factor algorithms neglect the correlation between data points and outliers caused by environmental factors in power equipment data detection, leading to misjudgment of data in low-density areas and affecting the accuracy of power grid information security detection.
By collecting voltage data from power equipment, calculating the voltage distortion rate and coherence of each segment, and combining the local outlier factor algorithm to calculate the anomaly score of the sampling points, the correlation between data points is considered to improve detection accuracy.
It reduces misjudgments of data in low-density areas, improves the accuracy and reliability of power grid information security detection, promptly detects potential equipment failures, and reduces maintenance and production interruption costs caused by equipment failures.
Smart Images

Figure CN118673446B_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the field of data processing technology. More specifically, the present invention relates to an artificial intelligence-based power network information security detection method and system. Background Art
[0002] During the operation of power equipment, if certain parameter data changes abnormally, this may indicate that the equipment status is unstable or there are potential faults. In this case, it is necessary to conduct real-time security detection of power grid information to avoid equipment failures that lead to unsustainable power supply.
[0003] For the security detection of power grid information, the Chinese patent application document with publication number CN110288383B discloses a method for detecting abnormal power consumption in distribution networks based on group behavior of user attribute labels. This method obtains abnormal data through the local outlier factor algorithm and compares it with historical abnormal data to complete the security detection of power grid information. The main idea of the local outlier factor algorithm is to assign an outlier factor LOF that depends on the neighborhood density to each data point. If the LOF value of a data point is much greater than 1, it means that the data point is an abnormal point.
[0004] However, the traditional local outlier factor algorithm mainly focuses on the local density of data points, but pays less attention to the correlation between data points. In low-density areas, even if a data point has some correlation with other data points (for example, they may belong to the same potential subcluster), it may be misjudged as an outlier by the local outlier factor algorithm due to its relatively low local density. Moreover, since the data obtained from power equipment may contain outliers due to environmental factors and other factors in the process of acquiring data, the outliers caused by these environmental factors will interfere with the accurate calculation of the correlation and local density between data points, thereby reducing the outlier detection ability of the local outlier factor algorithm. Summary of the invention
[0005] In order to solve the problem that the local outlier factor algorithm ignores the correlation between data points and environmental factors, resulting in outliers and thus misjudgment of data in low-density areas, the present invention provides solutions in the following aspects.
[0006] In a first aspect, the present invention provides an artificial intelligence-based power network information security detection method, comprising: collecting voltage data of power equipment, the voltage data including multiple segments; determining the voltage distortion of each segment; calculating the coordination degree of each sampling point in the voltage data, the calculation expression of the coordination degree is as follows:
[0007] ;
[0008] In the formula, Indicates The coordination degree of the i-th sampling point in a segment is Indicates The difference between the voltage distortion rate of a segment and the voltage distortion rate of other segments, Indicates The difference between the data value of the i-th sampling point and the data value of the k-th sampling point in the segment, n represents the The total number of sampling points other than the i-th sampling point in the segment; the anomaly score of each sampling point is calculated according to the local outlier factor algorithm, the final anomaly score of each sampling point is obtained by multiplying the coordination degree and the anomaly score, and detection is performed based on the final anomaly score.
[0009] Beneficial effect: Since the correlation between data is taken into account, the possibility of misjudgment of the local outlier factor algorithm when detecting data in low-density areas is reduced, the accuracy of the local outlier factor algorithm in detection is improved, thereby improving the accuracy of power grid information security detection.
[0010] In one embodiment, the voltage distortion rate is calculated as follows:
[0011] ;
[0012] In the formula, Indicates The voltage distortion rate of each segment is Represents the effective value of the hth harmonic voltage, where h=2,3,4..., Indicates the effective value of the fundamental voltage.
[0013] Beneficial effect: When the voltage distortion rate of the equipment is too high, it indicates that there may be potential equipment failure or operation problems. By calculating the voltage distortion rate of the power equipment from time to time, prevention or repair can be carried out in time, thereby reducing the cost of equipment production interruption.
[0014] In one embodiment, ,in, Indicates The voltage distortion rate of each segment is Indicates The voltage distortion rate of each segment.
[0015] In one embodiment, the multiple segments are divided into time intervals of 5 minutes.
[0016] Beneficial effect: By segmenting the voltage data, the flexibility in conducting power grid information security detection is improved.
[0017] In one embodiment, the method further includes: performing detection according to a total final anomaly score within a segment.
[0018] In one embodiment, the method further includes: performing data cleaning on the collected voltage data.
[0019] Beneficial effects: By cleaning the voltage data, outliers are removed, ensuring the accuracy and reliability of the data, thereby improving the accuracy of power grid information security detection.
[0020] In one embodiment, the electrical device is a circuit breaker.
[0021] Beneficial effects: By conducting information security testing on circuit breakers, safety issues with circuit breakers can be discovered in advance and intervention can be made before they become serious, reducing maintenance costs and production interruption costs caused by equipment failures.
[0022] In a second aspect, the present invention provides an artificial intelligence-based power network information security detection system, comprising a processor and a memory, wherein the memory stores computer program instructions, and when the computer program instructions are executed by the processor, the artificial intelligence-based power network information security detection method described in the first aspect is implemented.
[0023] The beneficial effect of the present invention is that the power network information security detection method of the present invention determines the correlation between the sampling point and other sampling points by calculating the coordination degree of each sampling point, avoiding the misjudgment that may occur when the local outlier factor algorithm detects sampling points in low-density areas, thereby improving the accuracy and reliability of power grid information security detection. BRIEF DESCRIPTION OF THE DRAWINGS
[0024] By reading the following detailed description with reference to the accompanying drawings, the above and other objects, features and advantages of the exemplary embodiments of the present invention will become readily understood. In the accompanying drawings, several embodiments of the present invention are shown in an exemplary and non-limiting manner, and the same or corresponding reference numerals represent the same or corresponding parts, wherein:
[0025] Figure 1 is a flow chart schematically illustrating an artificial intelligence-based power network information security detection method according to an embodiment of the present invention;
[0026] Figure 2 It is a block diagram schematically showing the structure of an artificial intelligence-based power network information security detection system according to an embodiment of the present invention. DETAILED DESCRIPTION
[0027] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative work are within the scope of protection of the present invention.
[0028] The specific embodiments of the present invention are described in detail below with reference to the accompanying drawings.
[0029] Figure 1 It is a flow chart schematically showing an artificial intelligence-based power network information security detection method according to an embodiment of the present invention.
[0030] like Figure 1 As shown, the power network information security detection method based on artificial intelligence of the present invention includes:
[0031] S101. Collect voltage data of power equipment, where the voltage data includes multiple segments.
[0032] In this embodiment, the power equipment is a circuit breaker. Further, in order to collect voltage data of the power equipment in real time, a voltage sensor or other monitoring device may be installed on the line where the power equipment is located.
[0033] Furthermore, when installing the voltage sensor, the sampling frequency can be set to 500Hz (i.e., 500 samples per second) to sample the voltage of the power equipment to avoid signal distortion caused by insufficient sampling. Furthermore, the collected voltage data is transmitted to the database, and then the collected voltage data is cleaned. By cleaning the collected voltage data, outliers in the voltage data that may be caused by collection errors, noise interference, etc. are removed, thereby avoiding the outliers from interfering with the calculation results of the local outlier factor algorithm, thereby improving the accuracy of the local outlier factor algorithm.
[0034] In other optional embodiments, the abnormal value can be directly deleted or data padded. For example, the abnormal value can be replaced by the average of two normal values adjacent to the abnormal value. Technical personnel in this field can make a choice based on actual needs. In addition, the sampling frequency of collecting voltage data can also be set according to actual needs.
[0035] In other optional embodiments, the collected voltage data may also be filtered, denoised, and other operations to ensure the accuracy and reliability of the voltage data, thereby improving the accuracy and reliability of subsequent voltage distortion rate calculations and avoiding noise interference with the local outlier factor algorithm during detection.
[0036] Furthermore, the collected voltage data is divided into a plurality of segments at a time interval of 5 minutes. In other optional embodiments, those skilled in the art can select a suitable time interval according to actual needs, for example, setting the time interval to 3 minutes.
[0037] S102: Determine the voltage distortion rate of each segment.
[0038] The voltage distortion rate of each segment can be obtained by Fourier transform. Specifically, Fast Fourier Transform (FFT) is applied to the collected voltage signal to convert the voltage signal from the time domain to the frequency domain, thereby decomposing the voltage signal into sinusoidal components of different frequencies, and then analyzing the harmonic components in the voltage signal, and then calculating the amplitude of each harmonic component to determine the effective value of each harmonic voltage. In other optional embodiments, the voltage data can also be converted from the time domain to the frequency domain by discrete Fourier Transform (DFT).
[0039] Furthermore, the voltage distortion rate of each segment is determined according to the acquired effective value of each harmonic voltage and the effective value of the fundamental voltage.
[0040] Specifically, the calculation expression of voltage distortion rate is as follows:
[0041] ;
[0042] In the formula, Indicates The voltage distortion rate of each segment is Represents the effective value of the hth harmonic voltage, where h=2,3,4..., Indicates the effective value of the fundamental voltage.
[0043] In one embodiment, obtaining The method can be achieved by calculating the mean value of the fundamental voltage of all sampling points in a segment, and the mean value is the effective value of the fundamental voltage.
[0044] In one embodiment, the voltage distortion rate of a segment may be obtained by calculating up to the 10th harmonic voltage, or the voltage distortion rate of a segment may be calculated by calculating up to other harmonic voltages.
[0045] In other optional embodiments, other methods may be selected to calculate the voltage distortion rate, such as designing a suitable filter to filter out the harmonic components in the voltage signal, and then calculating the ratio of the amplitude of the fundamental voltage to the effective value to obtain the voltage distortion rate.
[0046] By calculating the voltage distortion rate of a segment, the degree of waveform deviation of the voltage in the segment can be determined. Specifically, the higher the voltage distortion rate of a segment, the greater the degree to which the voltage waveform of the segment deviates from the ideal sine waveform, and the greater the difference between the sampling points, that is, the smaller the correlation.
[0047] In addition, the higher the voltage distortion rate, the more harmonic components the equipment has, which means there may be potential equipment failure or operation problems. Therefore, by calculating the voltage distortion rate of the equipment from time to time, problems can be discovered in a timely manner and appropriate measures can be taken to prevent or repair them, thereby reducing the maintenance costs and production interruption costs caused by equipment failure.
[0048] S103: Calculate the coordination degree of each sampling point in the voltage data.
[0049] In this embodiment, the coordination degree is used to characterize the correlation between a sampling point and other sampling points in the segment where the sampling point is located.
[0050] Specifically, the calculation expression of the coordination degree is as follows:
[0051] ;
[0052] In the formula, Indicates The coordination degree of the i-th sampling point in a segment; Indicates The difference between the voltage distortion rate of a segment and the voltage distortion rate of other segments; Indicates The difference between the data value of the i-th sampling point and the data value of the k-th sampling point in the segment, n represents the The total number of sampling points in a segment except the i-th sampling point.
[0053] Further, Indicates The difference between the voltage distortion rate of a segment and the voltage distortion rate of other segments, specifically, It can be expressed as: ,in, Indicates The voltage distortion rate of each segment is Indicates The voltage distortion rate of each segment. It should be noted that The segment is one of the other segments mentioned above.
[0054] From the above calculation expression of coordination degree, we can know that The larger the value of The smaller the value of is, the The larger the value, the The i-th sampling point in the segment and the The greater the difference between the other sampling points in the segment, that is, the greater the difference between the i-th sampling point and the The smaller the correlation between other sampling points in the segment, the smaller the coordination degree of the i-th sampling point. In addition, The larger the value of The smaller the value of is, the The larger the value, the The greater the difference between the voltage distortion rate of the first segment and the voltage distortion rate of other segments, the greater the difference between the first segment and the voltage distortion rate of other segments. The differences between the sampling points in the first segment are greater than those in other segments, indicating that The correlation between the sampling points in a segment is smaller than that between the sampling points in other segments.
[0055] In summary, a sampling point The value of and / or The larger the value of , the smaller the coordination degree of the sampling point.
[0056] S104, calculating the anomaly score of each sampling point according to the local outlier factor algorithm, multiplying the coordination degree and the anomaly score to obtain the final anomaly score of each sampling point, and performing detection according to the final anomaly score.
[0057] The above anomaly score is the LOF value obtained according to the local outlier factor algorithm.
[0058] Furthermore, when performing detection based on the final abnormality score of each sampling point, a safety threshold can be preset for each sampling point in the same segment. When detecting whether a sampling point is abnormal, the final abnormality score of the sampling point can be compared with the safety threshold corresponding to the sampling point. If the final abnormality score of the sampling point is higher than its safety threshold, the sampling point is an abnormal point, otherwise it is a normal point. Then, all the abnormal points in this segment are counted. If all the abnormal points in a segment exceed half of the number of all the sampling points, it indicates that there may be unsafe phenomena in the power equipment in this segment, and the nearby personnel are notified to carry out maintenance.
[0059] Since the correlation between sampling points is taken into account when calculating the final anomaly score, the local outlier factor algorithm avoids possible misjudgment of data in low-density areas, thereby improving the accuracy of information security detection.
[0060] In other optional embodiments, those skilled in the art may select the same or different preset safety thresholds for each sampling point according to actual needs. In addition, an unsafe phenomenon may be present when the number of abnormal points accounts for a certain percentage, for example, an unsafe phenomenon occurs when the number of abnormal points in a segment accounts for 60%.
[0061] In one embodiment, the final abnormality score of each sampling point in the same segment is summed up, and then a safety threshold is preset. If the total final abnormality score of each sampling point in this segment is higher than the preset safety threshold, it is determined that the power equipment in this segment is unsafe, and nearby personnel are notified for maintenance.
[0062] In one embodiment, a power grid information security warning score can also be calculated based on the final abnormality score. Specifically, the power grid information security warning score can be obtained by calculating the average of the total final abnormality scores of all circuit breakers, transformers, distribution cabinets and other equipment in a power grid topology and the corresponding weights of each power equipment, and then performing statistics.
[0063] For example, assuming that there are only circuit breakers and transformers in a power grid topology, the calculation expression for the power grid information security early warning score is as follows:
[0064] + ;
[0065] In the formula, Indicates The power grid information security early warning score in each segment, M is the number of circuit breakers, Indicates The sum of the final abnormal scores of each sampling point of the mth circuit breaker in the segment, D is the number of transformers, Indicates The sum of the final abnormal scores of each sampling point of the d-th transformer in the segment, is the weight of the circuit breaker in the power grid information security early warning score, is the weight of the transformer in the power grid information security early warning score.
[0066] In this embodiment, and It can be obtained by calculating the ratio between the circuit breaker and the transformer.
[0067] Furthermore, a warning threshold can be preset. When the power grid information security warning score exceeds the warning threshold, it indicates that there is an unsafe phenomenon in the power equipment. In this regard, it is necessary to detect the safety status of each power equipment separately and take warning or repair measures for the abnormal equipment to ensure continuous power supply.
[0068] Figure 2 It is a block diagram schematically showing the structure of the power network information security detection system based on artificial intelligence in this embodiment.
[0069] The present invention also provides an artificial intelligence-based power network information security detection system. Figure 2As shown, the power network information security detection system includes a processor and a memory, and the memory stores computer program instructions. When the computer program instructions are executed by the processor, the power network information security detection method based on artificial intelligence according to the first aspect of the present invention is implemented.
[0070] The artificial intelligence-based power network information security detection system also includes other components familiar to technicians in this field, such as communication interfaces. Their settings and functions are known in the art, so they will not be repeated here.
[0071] In the present invention, the aforementioned memory may be any tangible medium containing or storing a program that can be used by or in combination with an instruction execution system, apparatus or device. For example, a computer-readable storage medium may be any appropriate magnetic storage medium or magneto-optical storage medium, such as a resistive random access memory RRAM (Resistive Random Access Memory), a dynamic random access memory DRAM (Dynamic Random Access Memory), a static random access memory SRAM (Static Random-Access Memory), an enhanced dynamic random access memory EDRAM (Enhanced Dynamic Random Access Memory), a high-bandwidth memory HBM (High-Bandwidth Memory), a hybrid memory cube HMC (Hybrid Memory Cube), etc., or any other medium that can be used to store the required information and can be accessed by an application, a module, or both. Any such computer storage medium may be part of a device or accessible or connectable to a device. Any application or module described in the present invention may be implemented using computer-readable / executable instructions that may be stored or otherwise maintained by such a computer-readable medium.
[0072] In the description of this specification, "plurality" means at least two, such as two, three or more, etc., unless otherwise clearly and specifically defined.
[0073] Although this specification has shown and described a number of embodiments of the present invention, it will be apparent to those skilled in the art that such embodiments are provided by way of example only. Those skilled in the art will conceive of many modifications, changes and alternatives without departing from the ideas and spirit of the present invention. It should be understood that in the practice of the present invention, various alternatives to the embodiments of the present invention described herein may be employed.
Claims
1. A power network information security detection method based on artificial intelligence, characterized in that: include: Collecting voltage data of the electric power equipment, wherein the voltage data includes a plurality of segments; Determine the voltage distortion rate of each segment; The coordination degree of each sampling point in the voltage data is calculated, and the calculation expression of the coordination degree is as follows: ; In the formula, Indicates The coordination degree of the i-th sampling point in a segment is Indicates The difference between the voltage distortion rate of a segment and the voltage distortion rate of any other segment, Indicates The difference between the data value of the i-th sampling point and the data value of the k-th sampling point in the segment, n represents the The total number of sampling points in a segment except the i-th sampling point; The anomaly score of each sampling point is calculated according to the local outlier factor algorithm, the coordination degree and the anomaly score are multiplied to obtain the final anomaly score of each sampling point, and detection is performed based on the final anomaly score.
2. The power network information security detection method based on artificial intelligence according to claim 1 is characterized in that: The calculation expression of the voltage distortion rate is as follows: ; In the formula, Indicates The voltage distortion rate of each segment is Represents the effective value of the hth harmonic voltage, where h=2,3,4..., Indicates the effective value of the fundamental voltage.
3. The power network information security detection method based on artificial intelligence according to claim 1 is characterized in that: ,in, Indicates The voltage distortion rate of each segment is Indicates The voltage distortion rate of each segment.
4. The power network information security detection method based on artificial intelligence according to claim 1 is characterized in that: The multiple segments are divided at time intervals of 5 minutes.
5. The power network information security detection method based on artificial intelligence according to claim 1 is characterized in that: Also includes: Detection is performed based on the total final anomaly score within a segment.
6. The power network information security detection method based on artificial intelligence according to claim 1 is characterized in that: Also includes: Perform data cleaning on the collected voltage data.
7. The power network information security detection method based on artificial intelligence according to claim 1 is characterized in that: The electric power equipment is a circuit breaker.
8. An artificial intelligence-based power network information security detection system, characterized in that: It includes a processor and a memory, wherein the memory stores computer program instructions, and when the computer program instructions are executed by the processor, the power network information security detection method based on artificial intelligence as described in any one of claims 1 to 7 is implemented.
