Monitoring data processing method of electric power instrument

Abstract

The invention relates to the technical field of data cleaning, in particular to a monitoring data processing method of an electric power instrument, which comprises the following steps: acquiring monitoring data of the electric power instrument, and determining parameter change curves of different dimensions according to the monitoring data; determining a power consumption trend characteristic of each time sequence point by using parameter change curves of different dimensions, and further determining a parameter change characteristic value of each time sequence point; determining parameter detection disturbance degrees under different dimensions by using the parameter change characteristic values; the disturbance degree is detected through the parameters, the dimension distance of at least two to-be-detected data points under different dimensions is obtained, a data monitoring result is determined according to the dimension distance, and then abnormal data is cleaned according to the monitoring result; according to the method, through analysis of the power utilization data change characteristics, interference of power data change on anomaly detection is reduced, the detection accuracy is improved, and the accuracy of power system analysis is further improved.

Description

Technical Field

[0001] The present invention relates to the technical field of data cleaning, and in particular to a monitoring data processing method of an electric power instrument. Background Art

[0002] Power instruments are special equipment used to measure, monitor and analyze various parameters of power systems. They play an important role in the field of power engineering and are used to ensure the stable operation of power systems, evaluate power quality, monitor equipment performance, etc. The current power system is relatively complex, and the entire system contains a large number of control links. In order to ensure the reasonable and reliable operation of the system, it is necessary to monitor every link of the system through power instruments, and evaluate the operation status of the power system in combination with the monitoring data of the power system.

[0003] In the process of power instruments actually monitoring the power system, the power parameter information of each link of the power system is collected and converted into data for evaluation and analysis. However, as the instrument is used for a long time, the collection process is easily affected by electrical noise, resulting in abnormal monitoring data. The parameter information of the power system changes with the changes in the system operation process, and the existing abnormal data detection algorithm often ignores this change process, resulting in the detection results being difficult to meet the cleaning needs of the power system monitoring data. Summary of the invention

[0004] In order to solve the technical problem of abnormal monitoring data, the purpose of the present invention is to provide a monitoring data processing method for an electric power instrument, which can effectively clean the abnormal data of the monitoring data.

[0005] The technical solution adopted is specifically as follows: a method for processing monitoring data of an electric power instrument is provided, the method comprising: obtaining monitoring data of the electric power instrument, and determining parameter change curves of different dimensions based on the monitoring data; using the parameter change curves of different dimensions to determine the power consumption trend characteristics of each time point, and then determining the parameter change characteristic value of each time point; using the parameter change characteristic value to determine the parameter detection disturbance degree in different dimensions; using the parameter detection disturbance degree to obtain the dimensional distance of at least two data points to be tested in different dimensions, determining the data monitoring results based on the dimensional distance, and then cleaning the abnormal data based on the monitoring results.

[0006] In one embodiment of the present application, the acquisition of monitoring data of an electric power instrument and determination of parameter change curves of different dimensions based on the monitoring data include: detecting a target node using an electric power instrument to acquire the monitoring data, wherein the monitoring data includes monitoring data of different dimensions; and performing curve fitting using the monitoring data of different dimensions respectively to acquire parameter change curves of different dimensions.

[0007] In one embodiment of the present application, the parameter change curves of different dimensions are used to determine the power consumption trend characteristics of each time point, and then determine the parameter change characteristic value of each time point, including: using the parameter change curves of different dimensions to determine the power consumption fitting curve; obtaining the derivative of each time point on the power consumption fitting curve, and using the derivative to determine the derivative mean of each time point on the power consumption fitting curve; using a preset activation function, the derivative and the derivative mean to determine the power consumption trend characteristics of each time point; obtaining the relative time position of each time point in a preset interval in the parameter change curves of different dimensions, and then using the relative time position and the power consumption trend characteristics to determine the parameter change characteristic value of each time point.

[0008] In one embodiment of the present application, the parameter change curve includes a current change curve and a voltage change curve; the relative timing position of each timing point in a preset interval in the parameter change curve of different dimensions is obtained, and then the parameter change characteristic value of each timing point is determined by using the relative timing position and the power consumption trend characteristics, including: obtaining the first relative timing position of each timing point in the preset interval in the current change curve, and obtaining the second relative timing position of each timing point in the preset interval in the voltage change curve; obtaining the first derivative corresponding to each timing point in the current change curve, and obtaining the second derivative corresponding to each timing point in the voltage change curve, wherein the first derivative and the second derivative are obtained through standardization; using the first relative timing position, the second relative timing position, the first derivative, the second derivative and the power consumption trend characteristics, the parameter change characteristic value of each timing point is determined.

[0009] In one embodiment of the present application, the use of the first relative timing position, the second relative timing position, the first derivative, the second derivative and the power consumption trend characteristics to determine the parameter change characteristic value at each timing point includes: using the first relative timing position, the second relative timing position and the power consumption trend characteristics to determine the overall change characteristics of the power parameters; using the first relative timing position, the second relative timing position, the first derivative and the second derivative to determine the data change characteristics; determining the parameter change characteristic value at each timing point based on the overall change characteristics of the power parameters and the data change characteristics.

[0010] In one embodiment of the present application, the parameter change curve includes a current change curve and a voltage change curve; the use of the parameter change characteristic value to determine the parameter detection disturbance degree in different dimensions includes: obtaining a first derivative time series sequence of the current change curve, and obtaining a second derivative time series sequence of the voltage change curve; obtaining the amount of monitoring data in each dimension, obtaining the difference mean of the monitoring data between the target time point and the adjacent time point in each dimension, and obtaining the difference mean of the parameter change characteristic values ​​between the target time point and the adjacent time point; using the first derivative time series sequence, the second derivative time series sequence, the monitoring data amount, the difference mean and the difference mean, determine the parameter detection disturbance degree in each dimension.

[0011] In one embodiment of the present application, the first derivative timing sequence, the second derivative timing sequence, the amount of monitored data, the difference mean and the difference mean are used to determine the parameter detection disturbance degree of each dimension, including: using the difference mean and the difference mean to determine the data change frequency parameter of the target timing point, and then using the data change frequency parameter and the amount of monitored data to determine the first calculation parameter of each dimension; using the first derivative timing sequence and the second derivative timing sequence to determine the derivative timing sequence coefficients of the voltage change curve and the current change curve; using a preset exponential function, the first calculation parameter and the derivative timing sequence coefficient to determine the parameter detection disturbance degree of each dimension.

[0012] In one embodiment of the present application, the parameter detection disturbance degree is used to obtain the dimensional distance of at least two data points to be tested in different dimensions, the data monitoring result is determined by the dimensional distance, and then the abnormal data is cleaned according to the monitoring result, including: obtaining data values ​​of at least two data points to be tested in different dimensions, and obtaining parameter change characteristic values ​​of at least two data points to be tested; using the data values, the parameter change characteristic values ​​and the parameter detection disturbance degree to determine the dimensional distance of at least two data points to be tested in different dimensions; using the dimensional distances in different dimensions to determine the abnormal data, and then determining the data monitoring result.

[0013] In one embodiment of the present application, a first data mean of the data points included in each monotonic interval in the current change curve is obtained, and a second data mean of the data points included in each monotonic interval in the voltage change curve is obtained; and the first data mean and the second data mean are used to optimize the data monitoring results.

[0014] In an embodiment of the present application, the cleaning of abnormal data according to the monitoring results includes: deleting the abnormal data from the monitoring results; and then filling in the abnormal data in chronological order using an interpolation method.

[0015] The present invention has the following beneficial effects: by establishing a fitting curve for monitoring data, analyzing the overall trend of data changes, and providing basis parameters for the abnormal detection judgment process; then combining the power consumption change trend and the current and voltage curve characteristics to analyze the detection disturbance degree of the two-dimensional data, further improving the accuracy of the data change feature analysis, while evaluating the detection interference characteristics, providing basis parameters for the subsequent adjustment of the abnormal detection process; finally, combining the disturbance degree of each dimensional data and the data change characteristics to optimize the distance measurement method, and optimizing the abnormal detection algorithm according to the change characteristics of the fitting curve. Compared with the prior art, the detection process adds the analysis of the change characteristics of the power consumption data, reduces the interference of the power data change on the abnormal detection, improves the detection accuracy, and further increases the accuracy of the power system analysis. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] In order to more clearly illustrate the technical solutions and advantages in the embodiments of the present invention or the prior art, the drawings required for use in the embodiments or the prior art descriptions are briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0017] Figure 1 A schematic flow chart of a method for processing monitoring data of an electric power instrument provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0018] In order to further explain the technical means and effects adopted by the present invention to achieve the predetermined invention purpose, the following is a detailed description of the monitoring data processing method of an electric power instrument proposed by the present invention, its specific implementation method, structure, characteristics and effects, in combination with the accompanying drawings and preferred embodiments. In the following description, different "one embodiment" or "another embodiment" does not necessarily refer to the same embodiment. In addition, specific features, structures or characteristics in one or more embodiments may be combined in any suitable form.

[0019] Unless defined otherwise, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention belongs.

[0020] A specific scheme of a monitoring data processing method for an electric power instrument provided by the present invention is described in detail below with reference to the accompanying drawings.

[0021] See also Figure 1 , which shows a flow chart of a monitoring data processing method for an electric power instrument provided by an embodiment of the present invention.

[0022] like Figure 1 As shown, a monitoring data processing method for an electric power instrument comprises the following steps:

[0023] S10. Acquire monitoring data of the power instrument, and determine parameter change curves of different dimensions based on the monitoring data.

[0024] Among them, power instruments are special equipment used to measure, monitor and analyze various parameters of the power system, and are used to ensure the stable operation of the power system, evaluate the power quality, monitor equipment performance, etc.; monitoring data refers to the data obtained by measuring the various power consumption parameters of the power system through power instruments; different dimensions can include current dimension, voltage dimension, power consumption dimension, etc.

[0025] Specifically, different nodes of the power system are tested by power instruments to obtain various power parameters corresponding to different nodes, such as current, voltage, power consumption, etc., as data of different dimensions; and then, a current variation curve is determined according to the obtained change of current over time, a voltage variation curve is determined according to the obtained change of voltage over time, and a power consumption variation curve is determined according to the obtained change of power consumption over time.

[0026] S20. Using parameter change curves of different dimensions, determine the power consumption trend characteristics of each time point, and then determine the parameter change characteristic value of each time point.

[0027] Among them, the parameter change curve of each dimension can include many time points according to the division of the timeline. Therefore, the electricity consumption trend characteristics of each time point can be determined based on the parameter change curves corresponding to each dimension; the parameter change characteristic value refers to the comprehensive change characteristics of the power parameters of each dimension corresponding to the time point.

[0028] Specifically, after obtaining the parameter change curves of different dimensions, the parameters corresponding to each time point can be determined from the parameter change curve according to the division of each time point on the timeline, which can be used as the power consumption trend characteristics of each time point, and then the parameter changes of each time point in each dimension can be analyzed to determine the parameter change characteristic value of each time point.

[0029] S30. Determine the parameter detection disturbance degree in different dimensions by using the parameter change characteristic value.

[0030] Among them, parameter detection disturbance refers to the data changes reflected by power parameter fluctuations.

[0031] Specifically, after obtaining the parameter change characteristic value of each time point, the parameter change characteristic values ​​corresponding to the current time point and the adjacent time points are used to determine the frequency of changes corresponding to each dimensional data, and then obtain the parameter detection disturbance degree under different dimensions.

[0032] S40, using parameters to detect disturbance degree, obtaining the dimensional distance of at least two data points to be tested in different dimensions, determining the data monitoring result by the dimensional distance, and then cleaning the abnormal data according to the monitoring result.

[0033] Among them, dimensional data refers to the distances corresponding to different time points in the data space, which is used to reflect the local density in the data space.

[0034] Specifically, after obtaining the parameter detection disturbance degree in different dimensions, the dimensional data between different data points to be tested are obtained through the data values ​​of different data points to be tested in different dimensions and the corresponding power parameter change characteristic values ​​and the corresponding parameter detection disturbance degree, so as to highlight the distance between the abnormal data and the normal data, filter out the abnormal data, obtain the data monitoring results, and then clean the abnormal data.

[0035] In this embodiment, by adding analysis of the characteristics of power consumption data changes during the detection process, the interference of power data changes on abnormal detection is reduced, the detection accuracy is improved, and the accuracy of power system analysis is further increased.

[0036] In some embodiments, step S10 acquires monitoring data of the power instrument and determines parameter change curves of different dimensions according to the monitoring data, and may also include the following operations:

[0037] First, the target node is detected using an electric power instrument to obtain monitoring data, wherein the monitoring data includes monitoring data of different dimensions.

[0038] The target node refers to the node that can be monitored in the power system.

[0039] Specifically, the power instrument is connected to the target node of the power system to obtain the various power parameters corresponding to the target node, such as current, voltage, power consumption, etc.; and obtain the historical monitoring data within the detection time, as well as the various power consumption parameters corresponding to each time point.

[0040] Next, curve fitting is performed using monitoring data of different dimensions to obtain parameter change curves of different dimensions.

[0041] Among them, because during the operation of the power system, the power data will change with the changes of electrical equipment in the power system, therefore, in order to analyze the trend characteristics of the data changes, it is necessary to fit the various power data obtained by the power instrument into a curve, wherein the horizontal axis of the curve can be the detection time, and the vertical axis of the curve can be the current dimension data value.

[0042] Specifically, the curve corresponding to the current can be a current change curve, with the horizontal axis being the current detection time and the vertical axis being the current detection value; the curve corresponding to the voltage can be a voltage change curve, with the horizontal axis being the voltage detection time and the vertical axis being the voltage detection value; the curve corresponding to the power consumption can be a power consumption change curve.

[0043] It is understandable that the data detected by the current change curve and the voltage change curve are current values. Therefore, the power consumption change curve can be determined by the current change curve and the voltage change curve. The data detected by the power consumption change curve is usually a cumulative value.

[0044] In some embodiments, step S20 uses parameter change curves of different dimensions to determine the power consumption trend characteristics of each time point, and then determines the parameter change characteristic value of each time point, which may include the following operations:

[0045] First, the power consumption change curve is determined using parameter change curves of different dimensions; wherein the parameter change curve includes a current change curve and a voltage change curve.

[0046] As mentioned above, the data detected by the current change curve and the voltage change curve are current values. Therefore, the power consumption change curve can be determined through the current change curve and the voltage change curve. The data detected by the power consumption change curve is usually a cumulative value.

[0047] Next, the derivative of each time point on the power consumption change curve is obtained, and the derivative mean of each time point on the power consumption change curve is determined using the derivative.

[0048] Specifically, assume that there are multiple time points on the power consumption change curve, including the t-th time point, and obtain the derivative p of the t-th time point on the power consumption change curve t ′ , and then obtain the derivative mean corresponding to multiple time points

[0049] Next, the power consumption trend characteristics of each time point are determined using the preset activation function, derivative and derivative mean;

[0050] Among them, the monitoring data in the power system changes with the changes in the power equipment in the power system, and the changes in the power equipment are mainly reflected by the overall power consumption of the power system. Therefore, the current power consumption trend characteristics can be analyzed by combining the changes in the power consumption curve. The preset activation function can use a commonly used activation function, such as sigmoid(.).

[0051] Specifically, the difference between the derivative and the mean of the derivative at each time point is calculated using a preset activation function to determine the power consumption trend characteristics corresponding to each time point; it is difficult to reflect the power consumption trend characteristics between different time points directly through the derivative: to distinguish the power consumption trend characteristics between different time points, it is necessary to calculate the difference between the power consumption characteristics of each time point and the overall power consumption characteristics. The overall characteristics of power consumption are represented by the mean of the derivative of the power consumption curve, and then the power consumption of a certain point is subtracted from the overall power consumption characteristics. This reflects the electricity consumption trend of this time point compared with the overall electricity consumption data. Finally, the difference result is placed between (0, 1) through the sigmoid function, and the β t Represents the power consumption trend characteristics at the tth time point. The larger the value of the power consumption trend characteristics, the greater the power consumption at the current time point compared with the overall power consumption situation.

[0052] The electricity consumption trend characteristics are calculated as follows:

[0053]

[0054] Among them, β t is the electricity consumption trend feature, sigmoid(.) is the preset activation function, and p t ′ is the derivative of the t-th time point on the power consumption change curve, It is the mean value of the derivative of each time point on the electricity consumption change curve.

[0055] It can be understood that after taking the derivative of the power consumption fitting curve, its derivative p t ′ It can represent the power consumption at the tth time point.

[0056] Then, the relative time position of each time point in the preset interval in the parameter change curve of different dimensions is obtained, and then the parameter change characteristic value of each time point is determined by using the relative time position and the power consumption trend characteristics.

[0057] Among them, after obtaining the power consumption trend characteristics of the power system, the actual monitored power parameters can be analyzed: in the process of rapid changes in power consumption trends, electrical noise that affects the detection accuracy of power instruments is more likely to appear in the power system; in addition, for the operation of the power system, in order to ensure the stability of the power system, power parameters are usually not adjusted frequently. In order to increase the accuracy of the power data analysis process, after obtaining the power consumption trend characteristics of the power system, it is necessary to further analyze its detection disturbance degree in combination with the change characteristics of the power monitoring data value.

[0058] Specifically, the voltage and current data are standardized using the range normalization method; the monotonic intervals of the current change curve and the voltage change curve are obtained using derivative values; among them, to analyze the monitoring disturbance degree of data of different dimensions, it is first necessary to analyze the characteristic values ​​of the power parameter changes of each dimensional data in the monitoring data points.

[0059] A first relative timing position of each timing point in a preset interval in the current variation curve is obtained, and a second relative timing position of each timing point in a preset interval in the voltage variation curve is obtained.

[0060] Among them, the preset interval in the current change curve can be an increase or decrease interval in the current change curve, that is, a monotonic interval; the preset interval in the voltage change curve can be an increase or decrease interval in the voltage change curve, that is, a monotonic interval.

[0061] Specifically, taking the tth time point as an example, the first relative time position of the tth time point in the preset interval in the current change curve is obtained. and obtaining a second relative timing position of the tth timing point in a preset interval in the voltage variation curve

[0062] Next, a first derivative corresponding to each time point in the current variation curve is obtained, and a second derivative corresponding to each time point in the voltage variation curve is obtained, wherein the first derivative and the second derivative are obtained through standardization.

[0063] The first derivative may be a derivative value obtained by normalizing the range at each time point in the current change curve; and the second derivative may be a derivative value obtained by normalizing the range at each time point in the voltage change curve.

[0064] Specifically, taking the tth time point as an example, the first derivative ‖I obtained by the range normalization process in the current change curve of the tth time point is obtained: t ′ ‖, and obtain the second derivative ‖U obtained by range normalization in the voltage change curve at the tth time point t ′ ‖.

[0065] Next, the parameter change characteristic value at each timing point is determined using the first relative timing position, the second relative timing position, the first derivative, the second derivative and the power consumption trend characteristic.

[0066] Specifically, taking the tth time point as an example, using the power consumption trend feature β corresponding to the tth time point t , the first relative timing position The second relative timing position First derivative ‖I t′ ‖, the second derivative ‖U t ′ ‖, determine the parameter change characteristic value at the tth time point.

[0067] Among them, the first relative timing position can be used The second relative timing position and electricity consumption trend characteristics β t Determine the overall change characteristics of the power parameters; use the first relative timing position The second relative timing position First derivative ‖I t ′ ‖ and the second derivative ‖U t ′ ‖, determine the data change characteristics; then determine the parameter change characteristic value at each time point based on the overall change characteristics of the power parameters and the data change characteristics.

[0068] The calculation of parameter variation eigenvalue is as follows:

[0069]

[0070] Among them, d t represents the parameter change characteristic value at the t-th time point, β t represents the electricity consumption trend characteristics at the tth time point, represents the first relative timing position of the tth timing point in the preset interval in the current variation curve, represents the second relative timing position of the tth timing point in the preset interval in the voltage variation curve; ‖I t ′ ‖ represents the first derivative of the current change curve obtained by range normalization at the tth time point, ‖U t ′ ‖ represents the second derivative of the voltage change curve at the tth time point obtained by range normalization.

[0071] It is understandable that the first relative timing position and the second relative timing position The calculation method is Among them, t M ,t m They are respectively the maximum and minimum time series values ​​of the increase and decrease interval to which the tth time series point belongs.

[0072] It is understandable that for the t-th time point, in order to show its change characteristics, in addition to the power consumption trend characteristics reflected in the power consumption change curve, it is also necessary to analyze it in combination with the change characteristics of the monitoring data. The power consumption trend characteristics are the result data obtained by further analysis of the current and voltage. The change of this data depends on the change of current and voltage, reflecting the overall characteristics of power consumption; however, for the abnormal monitoring process, it is difficult to resist the influence of abnormal data by only analyzing the data change characteristics of a certain time point through the result data. The power change process also needs to be reflected by its change range: the data at the end of the monotonic interval of the data has a weaker ability to reflect the change characteristics of power parameters. In addition to the overall power consumption trend changes reflected by the result data (i.e., power consumption), the degree of change needs to be reflected through two actual monitoring data.

[0073] In order to further improve the accuracy of the change characteristics of the t-th time point and improve its noise resistance, in the monotonic interval of the fitting curve corresponding to a certain monitoring data, The larger the value is, the more obvious the trend of the parameter change characteristic value of the power parameter in the time series direction of the monitoring data of the tth time series point is about to change, and it is necessary to Amplify its rate of change during monitoring Indicates the characteristics of its data change; when The smaller the value, the weaker the change trend of the parameter change characteristic value of the power parameter at the time point t in the time direction, that is, the closer the change characteristic of the tth time point is to the overall characteristic of power consumption (power consumption), so through Amplify β t The overall change characteristics of the power parameters represented by are taken as the parameter change characteristic value at the t-th time point. Finally, the parameter representing the monotonic interval After the adjusted data change features are added (similar to the weighted averaging process), the t Indicates the parameter change characteristic value at the tth time point.

[0074] Furthermore, after obtaining the parameter change characteristic value of the power parameter at each point, combined with the above analysis, the monitoring disturbance degree of data in different dimensions is analyzed.

[0075] In some embodiments, step S30 uses parameter change characteristic values ​​to determine parameter detection disturbance degrees in different dimensions, which may include the following contents:

[0076] First, a first derivative time series sequence of the current variation curve is obtained, and a second derivative time series sequence of the voltage variation curve is obtained.

[0077] The first derivative time series is a derivative time series corresponding to the current change curve, and the second derivative time series is a derivative time series corresponding to the voltage change curve.

[0078] Next, the amount of monitoring data in each dimension is obtained, the mean difference of the monitoring data between the target time point and the adjacent time points in each dimension is obtained, and the mean difference of the parameter change characteristic values ​​between the target time point and the adjacent time points is obtained.

[0079] Among them, because it contains multi-dimensional data, it is necessary to obtain the amount of monitoring data in each dimension, the corresponding difference mean and the corresponding difference mean.

[0080] Specifically, take the tth time point as the target time point, obtain the amount of monitoring data in different dimensions, and obtain the difference mean Δx between the monitoring data of the tth time point and the adjacent time points in different dimensions. t , and obtain the difference mean of the parameter change characteristic values ​​between the t-th time point and the adjacent time points under different dimensions, where the difference mean and the difference mean can be obtained through standardization.

[0081] Next, the parameter detection disturbance degree of each dimension is determined using the first derivative time series sequence, the second derivative time series sequence, the amount of monitoring data, the mean of the difference and the mean of the difference.

[0082] Specifically, the data frequent change parameters of the target timing points are determined using the difference mean and the difference mean, and then the first calculation parameter of each dimension is determined using the data frequent change parameter and the amount of monitored data; the derivative timing sequence coefficients of the voltage change curve and the current change curve are determined using the first derivative timing sequence and the second derivative timing sequence; the parameter detection disturbance degree in each dimension is determined using the preset exponential function, the first calculation parameter and the derivative timing sequence coefficient.

[0083] The calculation of parameter detection disturbance is as follows:

[0084]

[0085] in, is the parameter detection disturbance degree, tanh is the hyperbolic tangent function, I ′ is the first derivative time series corresponding to the current change curve, U ′ is the second derivative time series of the voltage change curve, r is the correlation calculation function, n is the amount of monitoring data in the current dimension, Δx t Indicates the standardized mean difference between the monitoring data of the tth time point and the adjacent time points in the current dimension data, Δd t It represents the mean difference between the parameter change characteristic values ​​of the tth time point and the adjacent time points.

[0086] According to the above analysis, the electrical equipment in the power system will not frequently adjust its power parameters during use. To analyze the detection disturbance degree of a certain dimension of data, it is necessary to analyze the frequency of changes in the dimension of data: Δx t The difference between the tth time point in the current dimension and the adjacent time points is calculated, Δx t The larger the value, the higher the degree of fluctuation of the power parameter at the current time sequence point; in addition, in addition to the numerical fluctuation, the greater the difference in the parameter change characteristic value corresponding to the change characteristics of the current time sequence point and the adjacent time sequence point, the more frequent the data changes reflected by the power parameter fluctuation, that is, the greater the current parameter detection disturbance. t ×Δd t As the frequent data change parameter of the t-th time point, after averaging it, use As one of the parameters for calculating the disturbance degree of a certain dimension of power data.

[0087] In this embodiment, the power consumption data is the embodiment of the result parameters of the power system, and the detection data mainly refers to the current and voltage data. For the power system, the change process of current and voltage data should be proportional; however, when the monitoring process is disturbed, the proportional relationship will also be affected. ′ ,U ′ ) calculated the correlation coefficient of the derivative time series of the voltage change curve and the current change curve. According to the above analysis, under normal circumstances, r(I ′ ,U ′ ) should be larger; and when the current data detection is more disturbed, r(I ′ ,U ′ ) is smaller. Therefore, by After adjusting the logical relationship and setting the value range to (0, 1), it is used as another parameter for detecting disturbance degree calculation.

[0088] Finally, after combining the two parameters, the value range is set between (0, 1) through tanh, and the Indicates the parameter detection disturbance degree of power data of a certain dimension. The larger the value is, the stronger the interference is to the data detection process of the current dimension.

[0089] Furthermore, after obtaining the power data detection disturbance degree of each dimension of data, an anomaly detection algorithm for the monitoring data of power instruments is designed by adjusting the LOF anomaly monitoring detection algorithm. The LOF algorithm analyzes abnormal data by calculating the local density of data points in the data space, and the distance between data points is an important parameter that directly reflects the local density of data points. The traditional LOF algorithm mainly uses the Euclidean distance as the distance measurement method between data points, but for power monitoring data, this method ignores the data change characteristics of power data in time series, so the distance measurement method needs to be further optimized.

[0090] In step S40, the disturbance degree is detected by using parameters to obtain the dimensional distance of at least two data points to be tested in different dimensions, and the data monitoring result is determined by the dimensional distance, and then the abnormal data is cleaned according to the monitoring result, which may include the following steps:

[0091] The data values ​​of at least two data points to be tested in different dimensions are obtained, and the parameter change characteristic values ​​of at least two data points to be tested are obtained.

[0092] The data value refers to the data value of different data points to be tested in the current dimension, and the data points to be tested can also be time series points.

[0093] Specifically, the data values ​​of two different data points to be tested in the current dimension are obtained, and the parameter change characteristic values ​​corresponding to the two different data points to be tested are respectively obtained.

[0094] The dimensional distances of at least two data points to be tested in different dimensions are determined by using data values, parameter change characteristic values ​​and parameter detection disturbance degrees.

[0095] Among them, the parameter change characteristic value and the parameter detection disturbance degree can be obtained from the above calculation process.

[0096] Specifically, taking data point a and data point b as an example, the first difference between the data value of the j-th dimension of data point a and the data value of the j-th dimension of data point b is obtained, and the second difference between the parameter change characteristic value of data point a and the parameter change characteristic value of data point b is obtained, and the dimensional distance between data point a and data point b in the data space is determined using the dimensional data value, parameter detection disturbance degree, the first difference and the second difference.

[0097] The dimension distance is calculated as follows:

[0098]

[0099] Among them, s(a,b) is the dimensional distance between multidimensional data points a and b in the data space, and is also the abnormal distance of power data; is the data value of the jth dimension of data point a, is the data value of the jth dimension of data point b, d a is the parameter change characteristic value of data point a, d b is the parameter change characteristic value of data point b, m is the number of dimensions, Detect the perturbation degree for the parameter of the jth dimension.

[0100] in, The distance between the data points a and b in each dimension is calculated, and adjusted by the detection disturbance degree of the dimension data. When the detection disturbance degree of the dimension data is higher, the weight of the distance calculation between the data points a and b in this dimension is greater, so as to highlight the distance between the abnormal data and the normal data. In addition, by (d a -d b ) 2 The power parameter change characteristic distance of data points a and b is combined and used as another parameter for distance calculation. That is, in addition to the difference between the actual detection values, the power parameter change characteristic is also taken into account in the data point distance.

[0101] In some embodiments, after obtaining the distance measurement method, the LOF algorithm also needs to specify a neighborhood range k. In order to further reduce the impact of different data change characteristics on the anomaly detection process, the mean number of data points contained in each monotonic interval in the voltage and current curves is used as the k parameter to improve the accuracy of anomaly detection, that is, to optimize the data monitoring results.

[0102] That is, a first data mean of the data points included in each monotonic interval in the current change curve is obtained, and a second data mean of the data points included in each monotonic interval in the voltage change curve is obtained; the first data mean and the second data mean are used to optimize the data monitoring result; after obtaining the detection parameters, the current power instrument monitoring data is detected by the LOF algorithm to obtain abnormal data.

[0103] In some embodiments, abnormal data is determined using monitoring results, deleted, and filled in according to the time series order using interpolation. That is, after the detected abnormal data is deleted, the commonly used interpolation method is used to fill in the corresponding time series order to achieve data cleaning.

[0104] In some embodiments, the cleaned data is acquired and the cleaned data of the electric power instrument is output.

[0105] In this application, by establishing a fitting curve for monitoring data and analyzing the overall trend of data changes, a basis parameter is provided for the abnormal detection judgment process; then, the detection disturbance degree of the two-dimensional data is analyzed in combination with the power consumption change trend and the current and voltage curve characteristics, and the accuracy of the data change feature analysis is further improved while evaluating the detection interference characteristics, providing a basis parameter for the subsequent adjustment of the abnormal detection process; finally, the distance measurement method is optimized in combination with the disturbance degree of each dimensional data and the data change characteristics, and the abnormal detection algorithm is optimized according to the change characteristics of the fitting curve. Compared with the prior art, the detection process adds the analysis of the change characteristics of the power consumption data, reduces the interference of the power data change on the abnormal detection, improves the detection accuracy, and further increases the accuracy of the power system analysis.

[0106] It should be noted that the sequence of the above embodiments of the present invention is only for description and does not represent the advantages and disadvantages of the embodiments. The processes depicted in the accompanying drawings do not necessarily require the specific order or continuous order shown to achieve the desired results. In some embodiments, multitasking and parallel processing are also possible or may be advantageous.

[0107] The various embodiments in this specification are described in a progressive manner, and the same or similar parts between the various embodiments can be referenced to each other, and each embodiment focuses on the differences from other embodiments.

Claims

1. A method for processing monitoring data of an electric power instrument, characterized in that: The method comprises: Acquire monitoring data of power instruments, and determine parameter change curves of different dimensions based on the monitoring data; Using the parameter change curves of different dimensions, determine the power consumption trend characteristics of each time point, and then determine the parameter change characteristic value of each time point; Determine the parameter detection disturbance degree in different dimensions by using the parameter change characteristic value; The disturbance degree is detected by using the parameters to obtain the dimensional distance of at least two data points to be tested in different dimensions, the data monitoring result is determined by the dimensional distance, and then the abnormal data is cleaned according to the monitoring result.

2. The monitoring data processing method of the electric power instrument according to claim 1, characterized in that: The acquiring of monitoring data of the electric power instrument and determining parameter change curves of different dimensions according to the monitoring data include: Using an electric power instrument to detect the target node to obtain the monitoring data, wherein the monitoring data includes monitoring data of different dimensions; Curve fitting is performed using the monitoring data of different dimensions to obtain parameter change curves of different dimensions.

3. The monitoring data processing method of the electric power instrument according to claim 1 is characterized in that: The method of using the parameter change curves of different dimensions to determine the power consumption trend characteristics of each time point, and then determining the parameter change characteristic value of each time point, includes: Determine the power consumption change curve using parameter change curves of different dimensions; wherein the parameter change curve includes a current change curve and a voltage change curve; Obtaining the derivative of each time point on the power consumption change curve, and using the derivative to determine the mean value of the derivative of each time point on the power consumption change curve; Determine the power consumption trend characteristics of each time point by using a preset activation function, the derivative and the derivative mean; The relative time sequence position of each time sequence point in the preset interval in the parameter change curve of different dimensions is obtained, and then the parameter change characteristic value of each time sequence point is determined by using the relative time sequence position and the power consumption trend characteristic.

4. The monitoring data processing method of the electric power instrument according to claim 3 is characterized in that: The obtaining of the relative time sequence position of each time sequence point in a preset interval in the parameter change curve of different dimensions, and then determining the parameter change characteristic value of each time sequence point by using the relative time sequence position and the power consumption trend characteristic, includes: Obtaining a first relative timing position of each timing point in a preset interval in the current change curve, and obtaining a second relative timing position of each timing point in a preset interval in the voltage change curve; Obtaining a first derivative corresponding to each time point in the current change curve, and obtaining a second derivative corresponding to each time point in the voltage change curve, wherein the first derivative and the second derivative are obtained through standardization; The parameter change characteristic value at each timing point is determined by using the first relative timing position, the second relative timing position, the first derivative, the second derivative and the power consumption trend characteristic.

5. The monitoring data processing method of the electric power instrument according to claim 4 is characterized in that: The method of determining the parameter change characteristic value at each time point by using the first relative time sequence position, the second relative time sequence position, the first derivative, the second derivative and the power consumption trend characteristic comprises: Determine the overall change characteristics of the power parameters using the first relative time sequence position, the second relative time sequence position and the power consumption trend characteristics; Determining a data change characteristic using the first relative temporal position, the second relative temporal position, the first derivative, and the second derivative; The parameter change characteristic value at each time point is determined based on the overall change characteristic of the power parameter and the data change characteristic.

6. The monitoring data processing method of the electric power instrument according to claim 1, characterized in that: The parameter change curve includes a current change curve and a voltage change curve; and using the parameter change characteristic value to determine the parameter detection disturbance degree in different dimensions includes: Acquire a first derivative time series sequence of the current variation curve, and acquire a second derivative time series sequence of the voltage variation curve; Obtain the amount of monitoring data in each dimension, obtain the mean difference of monitoring data between the target time point and the adjacent time points in each dimension, and obtain the mean difference of parameter change characteristic values ​​between the target time point and the adjacent time points; The parameter detection disturbance degree of each dimension is determined by using the first derivative time series sequence, the second derivative time series sequence, the monitoring data volume, the difference mean and the difference mean.

7. The monitoring data processing method of the electric power instrument according to claim 6, characterized in that: The determining the parameter detection disturbance degree of each dimension by using the first derivative time series sequence, the second derivative time series sequence, the monitoring data volume, the difference mean and the difference mean includes: Determine the data change frequency parameter of the target time sequence point by using the difference mean and the difference mean, and then determine the first calculation parameter of each dimension by using the data change frequency parameter and the monitoring data volume; Determining derivative time series coefficients of the voltage change curve and the current change curve by using the first derivative time series sequence and the second derivative time series sequence; The parameter detection disturbance degree of each dimension is determined by using a preset exponential function, the first calculation parameter and the derivative time series coefficient.

8. The monitoring data processing method of the electric power instrument according to claim 1, characterized in that: The method of detecting the disturbance degree by using the parameter, obtaining the dimensional distance of at least two data points to be measured in different dimensions, and determining the data monitoring result by using the dimensional distance includes: Obtaining data values ​​of at least two data points to be tested in different dimensions, and obtaining parameter change characteristic values ​​of at least two data points to be tested; Determine the dimensional distance of at least two data points to be tested in different dimensions by using the data value, the parameter change characteristic value and the parameter detection disturbance degree; Dimension distances in different dimensions are used to identify abnormal data and then determine the data monitoring results.

9. The monitoring data processing method of the electric power instrument according to claim 8, characterized in that: Also includes: Obtaining a first data mean value of data points included in each monotonic interval in the current variation curve, and obtaining a second data mean value of data points included in each monotonic interval in the voltage variation curve; The data monitoring result is optimized by using the first data mean and the second data mean.

10. The monitoring data processing method of the electric power instrument according to claim 1, characterized in that: The cleaning of abnormal data according to the monitoring results includes: Deleting the abnormal data from the monitoring results; Then, interpolation is used to fill in the data in chronological order.

Images