A method and system for software control and management of secondary equipment in an intelligent substation

Abstract

This application relates to the field of data processing, and particularly to a method and system for software control and management of secondary equipment in an intelligent substation. The method includes the steps of: obtaining multiple data points of the preprocessed secondary equipment, each data point representing data in multiple dimensions at the same moment, and calculating the degree of outlier of the data points. The quotient of the standard deviation and covariance between the degree of outlier of any two dimensions and the mean value of the degree of outlier is normalized as the correlation between the dimensions. For one dimension, the average value of the correlations between this dimension and other dimensions is used as the correlation of this dimension, and the ratio of the correlation of this dimension to the cumulative sum of the correlations of all dimensions is used as the weighting coefficient to calculate the weighted Euclidean distance between two data points. The weighted Euclidean distance is used as the local density in the LOF algorithm for anomaly detection of power loads. This application has the effect of comprehensively and accurately detecting anomalies in multi-dimensional data according to the LOF algorithm.

Description

Technical Field

[0001] This application relates to the field of data processing, and in particular, to a method and system for software control and management of secondary equipment in an intelligent substation. Background Art

[0002] An intelligent substation is an important part of the power system. Secondary equipment refers to auxiliary equipment in the power system other than primary equipment directly involved in the production, transmission, and distribution of electric energy. Secondary equipment includes, but is not limited to, measuring instruments, control switches, relay protection devices, automatic devices, signal devices, control cables, and related auxiliary equipment, etc. The main function of secondary equipment is to monitor, control, and protect the working state of primary equipment to ensure the safe and stable operation of the entire power system.

[0003] The LOF algorithm is an anomaly detection algorithm that can be used to detect whether there is an anomaly in the power load. As pointed out in the paper "Research on Anomaly Detection and Repair of Power Load Data", Tang Yuchuan, the LOF algorithm can be used to detect power load anomalies. The prior art uses the nearest neighbor distance (local density) of data points in a single dimension (for example, current or voltage) for anomaly detection.

[0004] The detection data of all secondary equipment are usually multi-dimensional data (for example, current, voltage, power, temperature, etc.). Directly using the Euclidean distance between any two data points as the distance between them for anomaly detection will make the calculation result inaccurate due to different data representations in different dimensions, resulting in misjudgment of the software system. Summary of the Invention

[0005] In order to perform comprehensive and accurate anomaly detection on multi-dimensional data according to the LOF algorithm, this application provides a method and system for software control and management of secondary equipment in an intelligent substation.

[0006] In a first aspect, this application provides a method for software control and management of secondary equipment in an intelligent substation, adopting the following technical solution:

[0007] An intelligent substation secondary equipment software control method includes the steps of: obtaining multiple data points of the secondary equipment after preprocessing, each data point representing data in multiple dimensions at the same moment, and calculating the degree of outlier of the data points, where the dimensions include but are not limited to current, voltage, and / or power; normalizing the quotient of the standard deviation and covariance between the degrees of outlier of any two dimensions and the mean degree of outlier as the correlation between the dimensions; for a dimension, the average of the correlations between this dimension and other dimensions is used as the correlation of this dimension, and the ratio of the correlation of this dimension to the sum of the correlations of all dimensions is used as the weighting coefficient to calculate the weighted Euclidean distance between two data points; using the weighted Euclidean distance as the local density in the LOF algorithm for abnormal detection of power loads.

[0008] The beneficial effects are as follows: In this application, the correlation between dimensions is introduced to calculate the weighting coefficient, and the local density calculation in the LOF algorithm is adjusted according to the weighting coefficient. Compared with the prior art, the correlation between different power parameters is considered, and the similarity between data points can be measured more accurately. Therefore, the comprehensiveness of data dimensions is relatively high, and the situation of misjudgment caused by single dimension in the prior art can be reduced, thereby improving the accuracy of abnormal detection. Further, accurate abnormal detection helps to timely discover and respond to problems in the power system, thereby enhancing the reliability and stability of the entire power system, avoiding unnecessary maintenance and inspection work, and thus reducing the operating cost.

[0009] Optionally, the formula for calculating the degree of outlier is:

[0010] , where represents the degree of outlier of the th data point in the th dimension, represents the value of the th data point in the th dimension, represents the mean value of all data points within the local value range of the th data point in the th dimension, represents the local value range of the data in the th dimension, represents the standard normalization function.

[0011] The beneficial effects are as follows: In the formula for calculating the degree of outlier, the larger the value of the numerator and the smaller the value of the denominator, the higher the degree of outlier. The calculation formula of the numerator reflects the th data point in the The difference between a data point and the data within its local value range. The larger the value, the greater the difference between the data point and the data within its local value range. Conversely, the smaller the difference. The calculation formula of the denominator reflects the data fluctuation degree within the local value range where the data point is located. The larger the value of the fluctuation degree, the smaller the outlier performance within the local value range where the data point is located, that is, the lower the outlier degree. Conversely, the larger the outlier performance within the local value range where the data point is located, that is, the higher the outlier degree.

[0012] Optionally, the calculation formula for the local value range is:

[0013] , where represents the local value range of the data in the th dimension, is a preset reference value, represents the total number of data points in each dimension, represents the th data point in the th dimension, represents the th data point in the represents the mean value of all data points in the th dimension, represents the exponential function with the natural constant

[0014] The beneficial effect is that a fixed local value range cannot well reflect the local characteristics of substation data, such as large data fluctuations or small data fluctuations. Through the above technical solution, obtaining the local value range according to the local fluctuation characteristics of different data can make the calculation result of the outlier degree of the data more accurate and make the calculation result more consistent with the actual data situation.

[0015] Optionally, the weighted Euclidean distance is used as the local density in the LOF algorithm for power load anomaly detection. The power load anomaly detection includes the steps: inputting the detection data of secondary equipment into the LOF algorithm to output the anomaly score of the data point; calculating the warning degree, and when the warning degree is not less than the preset threshold, triggering the alarm system to alarm; where the calculation formula for the warning degree is:

[0016] , represents the warning degree of the current secondary equipment, represents the total number of data points in each dimension, represents the anomaly score of the th data point collected, represents the exponential function with the natural constant as the base.

[0017] Optionally, the quotient of the standard deviation and covariance between the outlier degree of any two dimensions and the average outlier degree is normalized and used as the correlation between dimensions. The normalization function uses the standard normalization function.

[0018] Optionally, among the multiple data points of the secondary equipment after preprocessing, the preprocessing method is: performing signal amplification and filtering on the data points.

[0019] The beneficial effects are: signal amplification of the data points can enhance the signal, and filtering of the data points can remove the noise in the signal.

[0020] Optionally, multiple data points of the secondary equipment are obtained through the self-checking device of the secondary equipment itself.

[0021] The beneficial effects are: compared with additionally setting a detection device, the number of devices arranged is reduced, the design scheme of the entire power system is simplified, the system is made more concise and efficient, and due to the reduction in the number of devices, the corresponding maintenance workload and maintenance cost are also reduced.

[0022] In a second aspect, the present application provides an intelligent substation secondary equipment software control system, adopting the following technical solution:

[0023] An intelligent substation secondary equipment software control system includes a processor and a memory. The memory stores computer program instructions, and when the computer program instructions are executed by the processor, it realizes an intelligent substation secondary equipment software control method according to the above.

[0024] The present application has the following effects:

[0025] 1. The present application considers the factor of the correlation between different power parameters, and helps to analyze whether there is a certain correlation in the data change trends between different dimensions. Taking the ratio of the correlation of each dimension to the sum of the correlations of all dimensions as the weighting coefficient, calculating the weighted Euclidean distance between data points, and using the weighted Euclidean distance as the distance between data points in the LOF algorithm for outlier detection. Compared with the method in the prior art that directly uses the Euclidean distance between data points as the distance between data points for outlier detection, the situation of misjudgment caused by single data dimension is reduced, the comprehensiveness of the data is relatively high, and the accuracy and comprehensiveness of the LOF algorithm for outlier detection are improved.

[0026] 2. When calculating the outlier degree, the local value range is obtained according to the local fluctuation characteristics of different data. Compared with using a fixed local value range, it can better reflect the local characteristics of the data, and make the calculation result of the outlier degree more consistent with the actual data situation. Description of the Drawings

[0027] 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 application will become readily understandable. In the drawings, several embodiments of the present application are shown in an exemplary rather than restrictive manner, and the same or corresponding reference numerals represent the same or corresponding parts.

[0028] Figure 1 It is a flowchart of a method for software control and management of secondary equipment in an intelligent substation according to an embodiment of the present application. Specific embodiments

[0029] Next, the technical solutions in the embodiments of the present application will be clearly and completely described in conjunction with the accompanying drawings in the embodiments of the present application. Obviously, the described embodiments are some, but not all, of the embodiments of the present application. All other embodiments obtained by those skilled in the art based on the embodiments of the present application without creative efforts fall within the protection scope of the present application.

[0030] It should be understood that when terms such as "first" and "second" are used in the claims, the specification, and the drawings of the present application, they are only used to distinguish different objects and not to describe a specific order. The terms "including" and "comprising" used in the specification and claims of the present application indicate the presence of the described features, wholes, steps, operations, elements, and / or components, but do not exclude the presence or addition of one or more other features, wholes, steps, operations, elements, components, and / or their combinations.

[0031] An embodiment of the present application discloses a method for software control and management of secondary equipment in an intelligent substation. In the software system of the preset secondary equipment, the LOF (Local Outlier Factor) algorithm is used for anomaly monitoring. The present application analyzes the change of the outlier degree of data in each dimension, further obtains the correlation between dimensions, obtains the weighting coefficient of each dimension according to the correlation between dimensions, and performs weighted calculation on the Euclidean distance between any two data points according to the weight of each dimension. The weighted Euclidean distance is used for anomaly monitoring by the LOF algorithm, so as to make the anomaly detection result more accurate and reduce the possibility of misjudgment of the software system. The technical solution of this embodiment is as follows:

[0032] Refer to Figure 1 , a method for software control and management of secondary equipment in an intelligent substation, including steps S1 - S4, as follows:

[0033] S1: Obtain multiple data points of the secondary equipment after preprocessing. Each data point represents data of multiple dimensions at the same moment, and calculate the outlier degree of the data point.

[0034] The dimensions include but are not limited to current, voltage, power, temperature, and frequency. Taking the three dimensions of current, voltage, and power as an example, the signals formed after preprocessing the current, voltage, and power signals are called secondary signals.

[0035] Different from conventional substations, the secondary signal input mode of the relay protection device in an intelligent substation is the digital sampled value message transmitted by fiber optic Ethernet, and the protection action is not the tripping and reclosing contacts, but the GOOSE digital quantity information transmitted by fiber optic Ethernet.

[0036] Therefore, the status monitoring of secondary equipment is different from that of primary equipment. For primary equipment status monitoring, additional monitoring equipment generally needs to be installed to monitor the main equipment; while for secondary equipment, such as relay protection and safety automatic devices, they generally have online self-checking functions and communication functions, belonging to embedded status monitoring. The online monitoring is realized by using the self-checking device of the secondary equipment itself. Among them, the monitoring period is set to 2S, that is, it is monitored once every 2S.

[0037] In one embodiment, the preprocessing method is as follows: first, amplify the signal of the data point to enhance the signal, and then perform filtering processing to remove the noise in the signal.

[0038] For the data of any one dimension, there will be a certain correlation between the data of different dimensions. For example, under ideal conditions, when the resistance is constant, the current data and voltage data of the secondary equipment show a proportional change. However, in actual situations, the collected data is easily affected by external environmental factors, resulting in the difference between the current and voltage data not meeting the strict proportional change. The same is true for other dimensions. Therefore, it is necessary to calculate the outlier degree of the data of a certain dimension according to the local value of the data of any one dimension.

[0039] In one embodiment, the existing method for calculating the outlier degree can be applied to this application, that is, calculating the LOF value of the data point. Since it is prior art, it will not be elaborated here.

[0040] In one embodiment, another way can be used to calculate the outlier degree. Specifically, the calculation formula for the outlier degree is:

[0041] , where represents the outlier degree of the th data point in the th dimension, represents the value of the th data point in the th dimension, represents the mean value of all data points within the local value range of the th data point in the th dimension, Indicates the local value range of data in the th dimension. Indicates the standard normalization function.

[0042] Indicates the th data point in the th dimension and the data difference within its local value range. The larger the value, the greater the data difference between this data point and the data within its local value range. Conversely, the smaller the data difference between this data point and the data within its local value range.

[0043] Indicates the degree of data fluctuation within the local value range where this data point is located. The larger the degree of fluctuation, the smaller the outlier performance within the local value range where this data point is located, that is, the lower the degree of outlier. Conversely, the larger the outlier performance within the local value range where this data point is located, that is, the higher the degree of outlier.

[0044] Among them, the calculation formula for the local value range is:

[0045] , where Indicates the local value range of data in the th dimension. Is a preset reference value. For example, . Indicates the total number of data points in each dimension. Indicates the th data point in the th dimension. Indicates the th mean value of all data points in the Indicates the exponential function with the natural constant as the base. Indicates the ceiling function.

[0046] Indicates the degree of data fluctuation in the th dimension. If the data fluctuation in the th dimension is smaller, it indicates that the overall distribution change of the data in this dimension is more stable, that is, a larger range can better reflect the outlier degree of the data in this dimension.

[0047] If the data fluctuation in the th dimension is larger, it indicates that the data change in this dimension is unstable. That is, at this time, a smaller range needs to be used to reflect the outlier degree of local data points. Therefore, if the degree of data point fluctuation in this dimension is smaller, the local value range corresponding to the data in this dimension is larger. Conversely, if the degree of data point fluctuation in this dimension is larger, the local value range corresponding to the data in this dimension is smaller.

[0048] In one embodiment, the local value range can also adopt a fixed local value range, and the definition of the range is limited by the user according to the actual application scenario, which will not be elaborated here. However, the fixed local value range cannot well reflect the local characteristics of the substation data. For example, the data fluctuates greatly or the data fluctuates little. Compared with the fixed local value range, the local value range calculated by the above calculation formula is more in line with the real data situation.

[0049] S2: Normalize the quotient of the standard deviation and covariance between the outlier degrees of any two dimensions and the mean value of the outlier degrees as the correlation between the dimensions.

[0050] For different dimensions, by calculating the correlation between the dimensions, the correlation can help analyze whether there is a certain correlation in the data change trends between different dimensions. If the changes in the outlier degrees between certain dimensions have a high correlation, then their data may be interdependent or have similar change patterns, which may be due to real correlations or due to errors or biases in the data collection or processing process. Therefore, correlation analysis can help identify data quality problems and discover potential data anomalies or errors.

[0051] The calculation formula for the correlation is:

[0052] , where represents the correlation between the th dimension and the th dimension, represents the total number of data points in each dimension, represents the outlier degree of the th data point in the th dimension, represents the mean value of the outlier degrees of all data points in the th dimension, represents the outlier degree of the th data point in the th dimension represents the mean value of the outlier degrees of all data points in the th dimension, represents the local value range of the data in the th dimension, represents the local value range of the data in the th dimension, represents the standard normalization function. The th dimension and the th dimension are both any dimension, and the th dimension and the The two dimensions are different and will not be elaborated hereinafter.

[0053] denotes the th dimension and the th dimension's covariance. denotes the th dimension and the th dimension's standard deviation. The local value range of each data point in different dimensions may be different. Therefore, it is necessary to divide the standard deviation of the dimension by the size of its corresponding local range to exclude the influence caused by different local ranges.

[0054] S3: For one dimension, the average value of the correlations between this dimension and other dimensions is used as the correlation of this dimension. Taking the ratio of the correlation of this dimension to the sum of the correlations of all dimensions as the weighting coefficient, calculate the weighted Euclidean distance between two data points.

[0055] The calculation formula of the weighted Euclidean distance is:

[0056] , denotes the th data point and the th data point in the th dimension's weighted Euclidean distance. denotes the th dimension in the collected data. denotes the th dimension's data correlation. denotes the th data point's value in the th dimension. denotes the th data point's value in the th dimension. The th data point and the th data point represent any two data points at different times in the same dimension.

[0057] denotes the th dimension's data correlation. denotes the total number of dimensions of the collected data. denotes the th dimension in the collected data. denotes any dimension except the th dimension. denotes except for the th dimension's total number of dimensions. denotes the th dimension and the The correlation of each dimension.

[0058] Indicates the data correlation weight of the nth dimension, that is, the weighting coefficient. The larger the value, the higher the importance of the data in this dimension, that is, the larger the Euclidean distance of the data in this dimension after weighting. Conversely, the lower the importance of the data in this dimension, that is, the smaller the Euclidean distance of the data in this dimension after weighting.

[0059] According to the Euclidean distance after weighting of two data points, as the local density of the data points in the LOF anomaly detection algorithm, use the LOF algorithm in the prior art to perform anomaly monitoring, so as to obtain the anomaly score of each data. The prior art will not be elaborated here. The anomaly score is a value distributed around 1, and the value range of the anomaly score can be , for example, the anomaly scores are 0.6, 0.8, 1, 1.2, 1.6, etc.

[0060] S4: Use the weighted Euclidean distance as the local density in the LOF algorithm to detect anomalies in power load.

[0061] Input the detection data of secondary equipment into the LOF algorithm to output the anomaly score of the data point; calculate the warning level. When the warning level is not less than the preset threshold, trigger the alarm system to alarm.

[0062] Among them, the calculation formula of the warning level is: , represents the warning level of the current secondary equipment, represents the total number of data points in each dimension, represents the anomaly score of the nth data point collected, represents the exponential function with the natural constant e as the base.

[0063] If the anomaly scores of all data points are higher, the warning level of the current secondary equipment is higher. The threshold is set to , for example, , which can be determined according to the specific implementation situation.

[0064] If , it means that the data collected currently is abnormal. The software system of the secondary equipment triggers the alarm system to generate a sound signal. The sound signal is to emit continuous or intermittent beeps to attract the attention of the operation and maintenance personnel; the alarm system can also generate a text signal. The text signal is to send a short message containing alarm information to the mobile device of the operation and maintenance personnel, and the operation and maintenance personnel perform fault repair, so as to realize remote software control.

[0065] If , the software system of the secondary equipment operates normally.

[0066] An embodiment of the present application also discloses a software control and management system for secondary equipment of an intelligent substation, including a processor and a memory. The memory stores computer program instructions, and when the computer program instructions are executed by the processor, the software control and management method for secondary equipment of the intelligent substation according to the present application is implemented.

[0067] The above system also includes other components well-known to those skilled in the art, such as a communication bus and a communication interface. Their settings and functions are known in the art, so they will not be elaborated here.

[0068] In the present application, the aforementioned memory can be any tangible medium that contains or stores a program, and this program can be used by or in combination with an instruction execution system, device, or component. For example, a computer-readable storage medium can be any suitable magnetic storage medium or magneto-optical storage medium, such as a resistive random access memory (RRAM), a dynamic random access memory (DRAM), a static random access memory (SRAM), an enhanced dynamic random access memory (EDRAM), a high bandwidth memory (HBM), a hybrid memory cube (HMC), etc., or any other medium that can be used to store the required information and can be accessed by an application program, module, or both. Any such computer storage medium can be part of the device or accessible or connectable to the device.

[0069] Although this specification has shown and described multiple embodiments of the present application, it is obvious to those skilled in the art that such embodiments are provided only by way of example. Those skilled in the art will think of many changes, alterations, and alternative ways without departing from the idea and spirit of the present application. It should be understood that various alternative solutions to the embodiments of the present application described herein can be adopted in the process of practicing the present application.

[0070] The above are all preferred embodiments of the present application, and the protection scope of the present application is not limited thereby. Therefore, all equivalent changes made according to the structure, shape, and principle of the present application should be covered within the protection scope of the present application.

Claims

1. A software control method for secondary equipment of a smart substation, characterized in that: Includes steps: Acquire multiple data points of the preprocessed secondary device, each data point represents data of multiple dimensions at the same time, and calculate the degree of outliers of the data points, wherein the dimensions include but are not limited to current, voltage and / or power; The quotient of the standard deviation and covariance between the outlier degree and the mean of the outlier degree of any two dimensions is normalized as the correlation between the dimensions; For a dimension, the average value of the correlations between the dimension and other dimensions is taken as the correlation of the dimension, and the ratio of the correlation of the dimension to the cumulative sum of the correlations of all dimensions is taken as the weighting coefficient to calculate the weighted Euclidean distance between two data points; The weighted Euclidean distance is used as the local density in the LOF algorithm to detect abnormal power loads. The abnormal power load detection includes the following steps: The detection data of the secondary equipment is input into the LOF algorithm, which outputs the abnormality score of the data point; Calculate the warning degree. When the warning degree is not less than the preset threshold, the alarm system is triggered to alarm. The calculation formula of the warning degree is: , Indicates the current warning level of the secondary equipment. represents the total number of data points in each dimension, Indicates the collected The anomaly score of the data point, Indicated by natural constant The exponential function of base .

2. A method for controlling software of secondary equipment in a smart substation according to claim 1, characterized in that: The calculation formula of outlier degree is: ,in, Indicates The first dimension The degree of outlier of a data point, Indicates The dimension The value of the data point, Indicates The first dimension The mean of all data points within the local value range of a data point, Indicates The local value range of the data in each dimension, Represents the standard normalization function.

3. A method for controlling software of secondary equipment in a smart substation according to claim 2, characterized in that: The calculation formula for the local value range is: ,in, Indicates The local value range of the data in the dimension, is the preset reference value, represents the total number of data points in each dimension, Indicates The dimension The value of the data point, Indicates The mean of all data points in the dimension, Indicated by natural constant is the exponential function of the base, Represents the ceiling function.

4. A method for controlling software of secondary equipment in a smart substation according to claim 1, characterized in that: The quotient of the standard deviation and covariance between the outlier degree and the mean of the outlier degree of any two dimensions is normalized and used as the correlation between the dimensions. The normalization function adopts the standard normalization function.

5. A method for controlling software of secondary equipment in a smart substation according to claim 1, characterized in that: In obtaining a plurality of data points of the secondary device after preprocessing, the preprocessing method is: performing signal amplification and filtering processing on the data points.

6. A method for controlling software of secondary equipment in a smart substation according to claim 1, characterized in that: Multiple data points of the secondary device are obtained through the self-test device of the secondary device itself.

7. A software management and control system for secondary equipment of smart substation, characterized in that: include: A processor and a memory, wherein the memory stores computer program instructions, and when the computer program instructions are executed by the processor, a method for software management and control of secondary equipment of a smart substation according to any one of claims 1 to 6 is implemented.

Images