CN116613895B - Smart grid power data anomaly detection method and system - Google Patents
Smart grid power data anomaly detection method and system Download PDFInfo
- Publication number
- CN116613895B CN116613895B CN202310896244.1A CN202310896244A CN116613895B CN 116613895 B CN116613895 B CN 116613895B CN 202310896244 A CN202310896244 A CN 202310896244A CN 116613895 B CN116613895 B CN 116613895B
- Authority
- CN
- China
- Prior art keywords
- data
- power data
- standard
- representing
- feature
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
- 238000001514 detection method Methods 0.000 title claims abstract description 202
- 230000005856 abnormality Effects 0.000 claims abstract description 118
- 230000002159 abnormal effect Effects 0.000 claims abstract description 88
- 238000000034 method Methods 0.000 claims abstract description 29
- 238000013507 mapping Methods 0.000 claims description 71
- 238000012937 correction Methods 0.000 claims description 19
- 230000008859 change Effects 0.000 claims description 10
- 230000000977 initiatory effect Effects 0.000 claims description 3
- 102100037651 AP-2 complex subunit sigma Human genes 0.000 claims 1
- 101000806914 Homo sapiens AP-2 complex subunit sigma Proteins 0.000 claims 1
- 238000004519 manufacturing process Methods 0.000 abstract description 4
- 239000004973 liquid crystal related substance Substances 0.000 description 16
- 238000005070 sampling Methods 0.000 description 12
- 238000012545 processing Methods 0.000 description 11
- 230000009286 beneficial effect Effects 0.000 description 6
- 230000008569 process Effects 0.000 description 6
- 238000003860 storage Methods 0.000 description 5
- 230000000694 effects Effects 0.000 description 3
- 241000854291 Dianthus carthusianorum Species 0.000 description 2
- 238000004590 computer program Methods 0.000 description 2
- 238000010276 construction Methods 0.000 description 2
- 238000000354 decomposition reaction Methods 0.000 description 2
- 230000007547 defect Effects 0.000 description 2
- 238000005516 engineering process Methods 0.000 description 2
- 238000004364 calculation method Methods 0.000 description 1
- 238000004891 communication Methods 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 238000010586 diagram Methods 0.000 description 1
- 238000005265 energy consumption Methods 0.000 description 1
- 230000003993 interaction Effects 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000012544 monitoring process Methods 0.000 description 1
- 230000000877 morphologic effect Effects 0.000 description 1
- 230000003287 optical effect Effects 0.000 description 1
- 238000002360 preparation method Methods 0.000 description 1
- 238000006467 substitution reaction Methods 0.000 description 1
- 230000001360 synchronised effect Effects 0.000 description 1
- 238000012546 transfer Methods 0.000 description 1
Classifications
-
- H—ELECTRICITY
- H02—GENERATION; CONVERSION OR DISTRIBUTION OF ELECTRIC POWER
- H02J—CIRCUIT ARRANGEMENTS OR SYSTEMS FOR SUPPLYING OR DISTRIBUTING ELECTRIC POWER; SYSTEMS FOR STORING ELECTRIC ENERGY
- H02J13/00—Circuit arrangements for providing remote indication of network conditions, e.g. an instantaneous record of the open or closed condition of each circuitbreaker in the network; Circuit arrangements for providing remote control of switching means in a power distribution network, e.g. switching in and out of current consumers by using a pulse code signal carried by the network
- H02J13/00001—Circuit arrangements for providing remote indication of network conditions, e.g. an instantaneous record of the open or closed condition of each circuitbreaker in the network; Circuit arrangements for providing remote control of switching means in a power distribution network, e.g. switching in and out of current consumers by using a pulse code signal carried by the network characterised by the display of information or by user interaction, e.g. supervisory control and data acquisition systems [SCADA] or graphical user interfaces [GUI]
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01D—MEASURING NOT SPECIALLY ADAPTED FOR A SPECIFIC VARIABLE; ARRANGEMENTS FOR MEASURING TWO OR MORE VARIABLES NOT COVERED IN A SINGLE OTHER SUBCLASS; TARIFF METERING APPARATUS; MEASURING OR TESTING NOT OTHERWISE PROVIDED FOR
- G01D21/00—Measuring or testing not otherwise provided for
- G01D21/02—Measuring two or more variables by means not covered by a single other subclass
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y04—INFORMATION OR COMMUNICATION TECHNOLOGIES HAVING AN IMPACT ON OTHER TECHNOLOGY AREAS
- Y04S—SYSTEMS INTEGRATING TECHNOLOGIES RELATED TO POWER NETWORK OPERATION, COMMUNICATION OR INFORMATION TECHNOLOGIES FOR IMPROVING THE ELECTRICAL POWER GENERATION, TRANSMISSION, DISTRIBUTION, MANAGEMENT OR USAGE, i.e. SMART GRIDS
- Y04S10/00—Systems supporting electrical power generation, transmission or distribution
- Y04S10/50—Systems or methods supporting the power network operation or management, involving a certain degree of interaction with the load-side end user applications
Landscapes
- Engineering & Computer Science (AREA)
- Human Computer Interaction (AREA)
- Power Engineering (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Testing And Monitoring For Control Systems (AREA)
Abstract
The invention provides a smart grid power data anomaly detection method and a smart grid power data anomaly detection system, wherein the smart grid power data anomaly detection method comprises the following steps: s1, acquiring power data acquired by a sensor node, wherein the power data comprises multidimensional characteristic data; s2, carrying out local abnormality detection according to the acquired power data to obtain a local abnormality detection result; s3, if the local abnormality detection result is abnormal, outputting the power data detection result as local abnormality; if the local abnormality detection result is normal, further carrying out global abnormality detection on the power data to obtain a global abnormality detection result; s4, if the global abnormality detection result is normal, outputting the power data detection result to be normal; if the global abnormality detection result is abnormal, the output power data detection result is global abnormality. The method and the device help to improve the reliability and the robustness of the power data anomaly detection in a scene based on large-scale power production by means of local anomaly detection and global anomaly detection.
Description
Technical Field
The invention relates to the technical field of intelligent interaction, in particular to a method and a system for detecting power data anomalies of a smart grid.
Background
Along with the construction and development of the smart grid, a large amount of power data needs to be collected by power enterprises under different application scenes so as to support the control and scheduling processing of the smart grid. At present, a data acquisition system based on a wireless sensor network is built in a power production environment to acquire and process power data in a distributed mode based on the characteristics of low energy consumption and flexibility of the wireless sensor network system, so that the system is favored by power enterprises.
At present, abnormal data exists in power data collected by a wireless sensor, which is inevitably influenced by factors such as environment or node performance. In the prior art, there is also a technology for performing localized abnormality detection and correction with respect to collected power data. However, in the existing technology for detecting localized anomalies of power data, single anomaly detection is usually performed for specific types of data, but single-dimensional anomaly data detection has the defect of insufficient adaptability in practical application, and the robustness of power data acquisition and anomaly detection is affected.
Disclosure of Invention
In view of the above problems, the present invention aims to provide a smart grid power data anomaly detection method and system.
The aim of the invention is realized by adopting the following technical scheme:
in a first aspect, the present invention provides a smart grid power data anomaly detection method, including:
s1, acquiring power data acquired by a sensor node, wherein the power data comprises multidimensional characteristic data;
s2, carrying out local abnormality detection according to the acquired power data to obtain a local abnormality detection result;
s3, if the local abnormality detection result is abnormal, outputting the power data detection result as local abnormality;
if the local abnormality detection result is normal, further carrying out global abnormality detection on the power data to obtain a global abnormality detection result;
s4, if the global abnormality detection result is normal, outputting the power data detection result to be normal;
if the global abnormality detection result is abnormal, the output power data detection result is global abnormality.
Preferably, the method is applied to upper nodes in a wireless sensor network of a smart grid, wherein the upper nodes comprise gateway nodes, base station nodes, edge computing nodes, local cluster head nodes, sink nodes and other upper nodes for converging power data acquired by a plurality of sensor nodes.
Preferably, in step S2, local anomaly detection is performed according to the acquired power data, including: based on the acquired power data over a period of timeRespectively extracting feature data sequences corresponding to different features；/>Representing the power acquired at time t over a period of timeData,/->Representing the length of the time period, +.>Representing the power data acquired at time t, wherein +.>,/>Characteristic data representing the d-th characteristic dimension at time t, < >>Representing the total number of feature dimensions; />A feature data sequence representing a d-th feature dimension;
from the obtained characteristic data sequenceAnd carrying out local abnormality detection, wherein the adopted local abnormality detection function is as follows: />；
Wherein, the liquid crystal display device comprises a liquid crystal display device,a local abnormal condition detection function representing time t, wherein +.>When (when)When at least one condition of (1) is satisfied, marking the characteristic data +.>And corresponding power data->Is local abnormal data; />Representing a characteristic data sequence->Average value of>Representing a set deviation threshold corresponding to the d-th feature dimension; />And->Respectively indicate->And->Characteristic data of the time of day>Representing a set variation threshold value corresponding to the d-th feature dimension,/or->Representing a characteristic data sequence->Standard deviation of>Representing a set standard deviation threshold corresponding to the d-th feature dimension; respectively carrying out local abnormality detection results according to the characteristic sequence data to obtain local abnormality detection results of the power data;
if the acquired power data does not contain the local abnormal data, the local detection result of the power data is normal.
Preferably, in step S3, global anomaly detection is further performed on the power data, including:
extracting power data at each timeWherein->Representing power data acquired at time t, where，/>Characteristic data representing the d-th characteristic dimension at time t, < >>Representing the total number of feature dimensions;
from the extracted power dataPerforming feature mapping to obtain power data->Mapping to hypercubeThe method comprises the steps of carrying out a first treatment on the surface of the Wherein the side length of the hypercube is +.>Diagonal length is +.>；
The mapping function adopted is as follows:；
wherein, the liquid crystal display device comprises a liquid crystal display device,index representing the mapping position of the feature data in the hypercube,/->Feature data representing a feature dimension a, wherein +.>，/>Representing the mapping constant of the setting +.>Representing a downward rounding function;
acquiring corresponding characteristic areas according to mapping points of power data in hypercube spaceThe feature region acquisition function adopted is as follows:；
wherein, the liquid crystal display device comprises a liquid crystal display device,representing all mapping points satisfying the condition in the hypercube space +.>A composed region; wherein the variables->；/>Index representing the mapping position of feature data in hypercube space,/for>Representing a set characteristic value, wherein ∈>；
Based on the obtained characteristic regionDetecting the number of standard mapping points contained in the feature region, wherein the standard mapping points are obtained by mapping the standard mapping points to the hypercube space according to preset standard data; if the number of standard mapping points contained in the feature area is greater than the set standard threshold +.>Then mark the power data +.>If the number of standard mapping points contained in the feature region is less than or equal to the set standard threshold +.>Then mark the power data +.>The global anomaly detection result of (2) is anomaly. Preferably, the preset standard data is stored in a standard data set; and mapping the standard data into the hypercube space to obtain standard mapping points.
Preferably, the method further comprises: the Sa1 updates the data in the standard data set, and maps the updated standard data to the hyperspectral space according to the updated standard data to obtain an updated standard mapping point.
Preferably, updating the data in the standard dataset includes:
every set time period, when the detection result of the detected power data is normal, initiating a replacement request aiming at the normal power data to acquire a replacement characteristic parameterThe method comprises the steps of carrying out a first treatment on the surface of the When the replacement characteristic parameter is larger than the set threshold value, marking the normal power data as standard data and inputting the standard data into a standard data set;
further detecting the data capacity of the standard data set, and randomly selecting one standard data to be removed from the standard data set if the number of the standard data contained in the standard data set is larger than a set threshold value;
the updating of the standard data set is completed.
Preferably, the method further comprises: and Sa2, mapping the newly added standard data in the standard data set into the hypercube space to obtain a standard mapping point.
Preferably, the method further comprises:
s5, correcting the abnormal power data to obtain corrected power data. In a second aspect, the present invention shows a smart grid power data anomaly detection system, including an acquisition unit, a local anomaly detection unit, an output unit, and a global anomaly detection unit;
the acquisition unit is used for acquiring the power data acquired by the sensor nodes, wherein the power data comprises multidimensional characteristic data;
the local abnormality detection unit is used for carrying out local abnormality detection according to the acquired power data to obtain a local abnormality detection result;
if the local abnormality detection result is abnormal, the output unit outputs the power data detection result as local abnormality;
if the local abnormality detection result is normal, the global abnormality detection unit further carries out global abnormality detection on the power data to obtain a global abnormality detection result;
if the global abnormality detection result is normal, the output unit outputs the power data detection result as normal;
if the global abnormality detection result is abnormal, the output unit outputs the power data detection result as the global abnormality.
Preferably, the system further comprises a standard management unit; the standard management unit is used for storing the standard data set and updating the data in the standard data set.
Preferably, the system further comprises a correction unit; the correction unit is used for correcting the abnormal power data to obtain corrected power data.
The beneficial effects of the invention are as follows: the invention provides a method and a system for detecting abnormality of data acquired by a sensor node in a smart grid, wherein the method comprises the steps of firstly, detecting local abnormality based on local characteristics of characteristic data according to acquired multidimensional circuit data; the power data is further subjected to global abnormality detection based on multidimensional features through local abnormality detection. The method can adapt to the abnormal detection of the multidimensional characteristic data in the power production scene by means of local abnormal detection and global abnormal detection, and is beneficial to improving the reliability and the robustness of the abnormal detection of the power data based on the sensor nodes.
Drawings
The invention will be further described with reference to the accompanying drawings, in which embodiments do not constitute any limitation of the invention, and other drawings can be obtained by one of ordinary skill in the art without inventive effort from the following drawings.
Fig. 1 is a schematic flow chart of a smart grid power data anomaly detection method according to an embodiment of the present invention;
fig. 2 is a frame structure diagram of a smart grid power data anomaly detection system according to an embodiment of the present invention.
Detailed Description
The invention is further described in connection with the following application scenario.
As shown in fig. 1, the invention discloses a smart grid power data anomaly detection method, which comprises the following steps:
s1, acquiring power data acquired by a sensor node, wherein the power data comprises multidimensional characteristic data;
preferably, the method is applied to upper nodes in a wireless sensor network of a smart grid, wherein the upper nodes comprise gateway nodes, base station nodes, edge computing nodes, local cluster head nodes, sink nodes and other upper nodes for converging power data acquired by a plurality of sensor nodes.
After the localized upper node acquires the power data acquired by the sensor node, firstly, abnormality detection is carried out on the power data, so that distributed and localized abnormality detection processing of the power data is finished, the data processing pressure of the upper computer is reduced, and the requirement of mass power data abnormality detection in a large-scale smart grid scene is met.
The power data comprise monitoring data acquired in a smart grid scene such as temperature, humidity, voltage, current and decibel.
S2, carrying out local abnormality detection according to the acquired power data to obtain a local abnormality detection result;
the local abnormality of the acquired power data can be detected quickly, and the abnormal power data can be identified quickly, by detecting the local abnormality of the acquired power data, for example, detecting the abnormality of specific type of data in the power data, or detecting the abnormal value according to the conventional detection modes of the power data, such as lack of value, abnormal value and the like.
Preferably, in step S2, local anomaly detection is performed according to the acquired power data, including: based on the acquired power data over a period of timeRespectively extracting feature data sequences corresponding to different features；/>Representing the power data during a time period acquired at time t,/for a time period>Representing the length of the time period, +.>Representing the power data acquired at time t, wherein +.>,/>Characteristic data representing the d-th characteristic dimension at time t, < >>Representing the total number of feature dimensions; />A feature data sequence representing a d-th feature dimension;
from the obtained characteristic data sequenceAnd carrying out local abnormality detection, wherein the adopted local abnormality detection function is as follows: />；
Wherein, the liquid crystal display device comprises a liquid crystal display device,a local abnormal condition detection function representing time t, wherein +.>When (when)When at least one condition of (1) is satisfied, marking the characteristic data +.>And corresponding power data->Is local abnormal data; />Representing a characteristic data sequence->Average value of>Representing a set deviation threshold corresponding to the d-th feature dimension; />And->Respectively indicate->And->Characteristic data of the time of day>Representing a set variation threshold value corresponding to the d-th feature dimension,/or->Representing a characteristic data sequence->Standard deviation of>Representing a set standard deviation threshold corresponding to the d-th feature dimension; respectively carrying out local abnormality detection results according to the characteristic sequence data to obtain local abnormality detection results of the power data;
if the acquired power data does not contain the local abnormal data, the local detection result of the power data is normal. In the above embodiment, a technical solution for performing local anomaly detection on power data is also provided, in this embodiment, by acquiring characteristic data in power data in a period of time to perform anomaly data detection based on time sequence variation, rapid anomaly data point detection can be performed based on morphological characteristics of the data, and efficiency of anomaly data detection is effectively improved.
S3, if the local abnormality detection result is abnormal, outputting the power data detection result as local abnormality;
if the local abnormality detection result is normal, further carrying out global abnormality detection on the power data to obtain a global abnormality detection result; aiming at the problem that the abnormality detection of single-dimensional characteristic data is usually carried out in the local abnormality detection process, the adaptability is easy to be insufficient (for example, the single-dimensional abnormality detection can only carry out variation trend abnormality detection through power data acquired by a single node, or specific point detection is carried out, but in the actual abnormality detection process, the judgment standard is certainly not invariable, the standard of partial abnormality detection is also changed along with the change of environment or the change of node performance, and meanwhile, the abnormality detection can not be carried out on data of an unconventional type), therefore, the invention further provides a global abnormality detection method based on multiple characteristics after the local abnormality detection is completed, so that the accuracy and the adaptability of the abnormality detection of the power data are further improved.
Preferably, in step S3, global anomaly detection is further performed on the power data, including: extracting power data at each timeWherein->Representing the power data acquired at time t, wherein +.>，/>Characteristic data representing the d-th characteristic dimension at time t, < >>Representing the total number of feature dimensions;
from the extracted power dataPerforming feature mapping to obtain power data->Mapping to hypercubeThe method comprises the steps of carrying out a first treatment on the surface of the Wherein the side length of the hypercube is +.>Diagonal length is +.>；
The mapping function adopted is as follows:；
wherein, the liquid crystal display device comprises a liquid crystal display device,index representing the mapping position of the feature data in the hypercube,/->Feature data representing a feature dimension a, wherein +.>，/>Representing the mapping constant of the setting +.>Representing a downward rounding function;
acquiring corresponding characteristic areas according to mapping points of power data in hypercube spaceThe feature region acquisition function adopted is as follows:；
wherein, the liquid crystal display device comprises a liquid crystal display device,representing all mapping points satisfying the condition in the hypercube space +.>A composed region; wherein the variables->；/>Index representing the mapping position of feature data in hypercube space,/for>Representing a set characteristic value, wherein ∈>；
Based on the obtained characteristic regionDetecting the number of standard mapping points contained in the feature region, wherein the standard mapping points are obtained by mapping the standard mapping points to the hypercube space according to preset standard data; if the number of standard mapping points contained in the feature area is greater than the set standard threshold +.>Then mark the power data +.>If the number of standard mapping points contained in the feature region is less than or equal to the set standard threshold +.>Then mark the power data +.>The global anomaly detection result of (2) is anomaly. The multi-dimensional power data is mapped based on the hypercube space, the characteristic region detection is carried out according to the mapping position, the relation between the characteristic region and the standard mapping point is judged to carry out global anomaly detection on the power data, and the method can be suitable for comprehensive anomaly detection of the multi-dimensional power data. The method is beneficial to improving the accuracy and adaptability of the power data anomaly detection.
Preferably, the preset standard data is stored in a standard data set; and mapping the standard data into the hypercube space to obtain standard mapping points.
In practical application, only a standard data set is required to be set at an upper node for global anomaly detection, and the anomaly detection process of the power data can be completed in a self-adaptive manner based on the standard data set.
In the preparation stage, firstly, standard data (for example, standard data obtained by standard anomaly detection of power data acquired by the same group of sensor nodes in the wireless sensor network construction stage) are set for normal power data under different conditions, the standard data are mapped to a unified hypercube according to the standard data, standard mapping points of the standard data in the hypercube are obtained, and the standard mapping points are used as a basis for carrying out global anomaly detection on the power data according to the obtained standard mapping points. Meanwhile, in the subsequent global anomaly detection process, in order to adapt to the change of the data in the actual acquisition environment, the unchanged standard data cannot adapt to the change of the environment, so that in the global anomaly detection process, the standard data set is further updated aiming at the detected normal power data, and the adaptability of the standard data set is improved.
Preferably, the method further comprises: the Sa1 updates the data in the standard data set, and maps the updated standard data to the hyperspectral space according to the updated standard data to obtain an updated standard mapping point.
Preferably, updating the data in the standard dataset includes:
every set time period, when the detection result of the detected power data is normal, initiating a replacement request aiming at the normal power data to acquire a replacement characteristic parameterThe method comprises the steps of carrying out a first treatment on the surface of the When the replacement characteristic parameter is larger than the set threshold value, marking the normal power data as standard data and inputting the standard data into a standard data set;
further detecting the data capacity of the standard data set, and randomly selecting one standard data to be removed from the standard data set if the number of the standard data contained in the standard data set is larger than a set threshold value;
the updating of the standard data set is completed.
After the standard data is removed from the standard data set, the corresponding standard mapping points are also eliminated correspondingly.
Preferably, the method further comprises: and Sa2, mapping the newly added standard data in the standard data set into the hypercube space to obtain a standard mapping point.
The foregoing embodiment also particularly proposes a scheme for adaptively updating the standard data set, which can update the standard data set in real time based on the normal power data obtained in real time, so as to adaptively adjust the standard of global anomaly detection according to the standard data set, and thereby, the method is beneficial to improving the adaptability and reliability of global anomaly detection. Meanwhile, the standard data updating mode based on the local data can adapt to the scenes of distributed data acquisition and anomaly detection.
S4, if the global abnormality detection result is normal, outputting the power data detection result to be normal;
if the global abnormality detection result is abnormal, the output power data detection result is global abnormality.
And outputting the power data through the local abnormality detection and the global abnormality detection, for example, further transmitting the data to an upper computer for further data processing, and further processing and utilizing data resources on the collected power data.
The abnormal power data is further marked according to the detected abnormal power data, or corresponding early warning prompt information is sent out.
S5, correcting the abnormal power data to obtain corrected power data.
And aiming at the detected abnormal power data, carrying out correction processing on the marked power abnormal data, and being beneficial to improving the quality of power data acquisition.
Preferably, step S5 performs correction processing on the abnormal power data, and includes:
intercepting data of continuous moments around the abnormal power data as a center according to the power data marked as abnormal to form an abnormal data segment;
EMD decomposition is carried out according to the obtained abnormal data segment, so that a plurality of IMF components and margins RES with different frequencies of the abnormal data segment are obtained;
according to the obtained IMF components, respectively calculating abnormal characteristic factors of abnormal data points in each IMF component, wherein the adopted abnormal characteristic factor calculation function is as follows:
；
an abnormality feature factor representing an nth-dimension IMF component,>representing a counting function, i.e. counting the number of eligible elements,/->A variable representing a counting function, wherein ∈>，/>Representing the sampling time corresponding to the abnormal data point, +.>Representing a symbolic function +_>Representing the magnitude, ++f of the kth sample time in the nth dimension IMF component>Representing the magnitude of the kth-1 sample instant in the nth dimension IMF component, +.>Representing the average absolute amplitude of each sampling moment in the IMF component of the nth dimension; according to the abnormal characteristic factors of the IMF components, recording the IMF component dimension with the largest abnormal characteristic factor change:
；
wherein N represents the total number of IMF components;
will be the firstReconstructing IMF component of dimension to obtain fluctuation data component +.>Will->Reconstructing the IMF component of the dimension to obtain a basic data component +.>；
Based on the resulting base data componentCorrecting abnormal data point segments in sequence, and correcting the abnormal data points as follows:
；
wherein, the liquid crystal display device comprises a liquid crystal display device,representing +.>Amplitude of sampling instant, wherein，/>Representing the sampling time corresponding to the abnormal data point, +.>Representing +.>Amplitude of sampling time; />Representing +.>Amplitude of sampling time; />Representing +.>Amplitude of sampling time; />Representing +.>Amplitude of sampling time;
from the resulting fluctuating data componentCorrecting the abnormal data point, and correcting the abnormal data point as follows:
；
wherein, the liquid crystal display device comprises a liquid crystal display device,representing +.>Amplitude of sampling time，/>Representing the sampling time corresponding to the abnormal data point, +.>Representing the +.>Amplitude of sampling moment, +.>Representing a symbolic function +_>Representing the average absolute amplitude of each sampling instant in the fluctuating data component;
the corrected basic data componentModified wave data component +.>And reconstructing the residual RES to obtain corrected power data.
The abnormal power data obtained by the local abnormality detection and the global abnormality detection is further subjected to correction processing, and the conventional abnormal data point processing is usually performed by means of mean interpolation, linear interpolation or the like, but the data points which are easy to correct by means of simple interpolation have the defects of distortion and low coincidence degree in correction effect. Therefore, the above embodiment also proposes a technical solution for performing the abnormal data correction according to the abnormal power data local context information, so as to improve the effect of the abnormal data correction. Firstly, according to the detected abnormal data point, extracting an abnormal data segment formed in a period of time before and after the data point to carry out EMD decomposition, and extracting the change characteristic of the abnormal data point based on the obtained IMF component; the local abnormal characteristics of the abnormal data points are characterized by the abnormal characteristic factors, so that the abnormal interference condition of the abnormal data points can be further reflected, the obtained IMF component is further divided into a fluctuation data component with the dominant abnormal characteristics and a basic data component with the dominant data context characteristics according to the abnormal characteristic factors, wherein certain interference is caused to the overall context change characteristics of the basic data component by considering the occurrence of the abnormal data points, and therefore, when the abnormal data points are corrected according to the basic data components, the synchronous correction is carried out particularly for the neighbor points before and after the abnormal data points, so that the influence of the context feature change caused by the abnormal data is reduced; and meanwhile, aiming at the obtained fluctuation data component, the abnormal characteristics leading by the abnormal data are directly removed, so that the original data is restored. And carrying out final component reconstruction after carrying out basic data component correction and fluctuation data component correction on the abnormal data points to obtain corrected power data. The corrected power data can adapt to the context characteristics of the normal power data, and meanwhile, abrupt abnormal conditions in abnormal data points can be effectively eliminated; compared with the traditional mean value interpolation or linear interpolation and other modes, the data correction and restoration degree is higher, and meanwhile, the method can adapt to the correction of data with different change frequency characteristics, and improves the adaptability and effect of abnormal data correction.
According to the smart grid power data anomaly detection method provided by the invention, local anomaly detection based on local characteristics of characteristic data is performed according to the acquired multidimensional circuit data; the power data is further subjected to global abnormality detection based on multidimensional features through local abnormality detection. The method can adapt to the abnormal detection of the multidimensional characteristic data in the power production scene by means of local abnormal detection and global abnormal detection, and is beneficial to improving the reliability and the robustness of the abnormal detection of the power data based on the sensor nodes.
As shown in fig. 2, the invention discloses a smart grid power data anomaly detection system, which comprises an acquisition unit, a local anomaly detection unit, an output unit and a global anomaly detection unit;
the acquisition unit is used for acquiring the power data acquired by the sensor nodes, wherein the power data comprises multidimensional characteristic data;
the local abnormality detection unit is used for carrying out local abnormality detection according to the acquired power data to obtain a local abnormality detection result;
if the local abnormality detection result is abnormal, the output unit outputs the power data detection result as local abnormality;
if the local abnormality detection result is normal, the global abnormality detection unit further carries out global abnormality detection on the power data to obtain a global abnormality detection result;
if the global abnormality detection result is normal, the output unit outputs the power data detection result as normal;
if the global abnormality detection result is abnormal, the output unit outputs the power data detection result as the global abnormality.
Preferably, the system further comprises a standard management unit; the standard management unit is used for storing the standard data set and updating the data in the standard data set.
Preferably, the system further comprises a correction unit; the correction unit is used for correcting the abnormal power data to obtain corrected power data.
It should be noted that, in the smart grid power data anomaly detection system shown in fig. 2 of the present invention, each functional unit is further configured to implement the steps and the specific embodiments corresponding to the smart grid power data anomaly detection method shown in fig. 1, and the disclosure is not repeated here.
It should be noted that, in each embodiment of the present invention, each functional unit/module may be integrated in one processing unit/module, or each unit/module may exist alone physically, or two or more units/modules may be integrated in one unit/module. The integrated units/modules described above may be implemented either in hardware or in software functional units/modules.
From the description of the embodiments above, it will be apparent to those skilled in the art that the embodiments described herein may be implemented in hardware, software, firmware, middleware, code, or any suitable combination thereof. For a hardware implementation, the processor may be implemented in one or more of the following units: an Application Specific Integrated Circuit (ASIC), a Digital Signal Processor (DSP), a Digital Signal Processing Device (DSPD), a Programmable Logic Device (PLD), a Field Programmable Gate Array (FPGA), a processor, a controller, a microcontroller, a microprocessor, other electronic units designed to perform the functions described herein, or a combination thereof. For a software implementation, some or all of the flow of an embodiment may be accomplished by a computer program to instruct the associated hardware. When implemented, the above-described programs may be stored in or transmitted as one or more instructions or code on a computer-readable medium. Computer-readable media includes both computer storage media and communication media including any medium that facilitates transfer of a computer program from one place to another. A storage media may be any available media that can be accessed by a computer. The computer readable media can include, but is not limited to, RAM, ROM, EEPROM, CD-ROM or other optical disk storage, magnetic disk storage media or other magnetic storage devices, or any other medium that can be used to carry or store desired program code in the form of instructions or data structures and that can be accessed by a computer.
Finally, it should be noted that the above embodiments are only for illustrating the technical solution of the present invention, and not for limiting the scope of the present invention, and although the present invention has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can be made to the technical solution of the present invention without departing from the spirit and scope of the technical solution of the present invention.
Claims (9)
1. The utility model provides a smart power grid power data anomaly detection method which is characterized by comprising the following steps:
s1, acquiring power data acquired by a sensor node, wherein the power data comprises multidimensional characteristic data;
s2, carrying out local abnormality detection according to the acquired power data to obtain a local abnormality detection result;
s3, if the local abnormality detection result is abnormal, outputting the power data detection result as local abnormality;
if the local abnormality detection result is normal, further carrying out global abnormality detection on the power data to obtain a global abnormality detection result;
s4, if the global abnormality detection result is normal, outputting the power data detection result to be normal;
if the global abnormality detection result is abnormal, outputting the power data detection result as global abnormality;
in step S3, global anomaly detection is further performed on the power data, including:
extracting power data x at each time t Wherein x is t Representing power data acquired at time t, where x t ＝[y 1t ，y 2t ，...y dt ]，y dt Feature data representing a d-th feature dimension at time t, d representing the total number of feature dimensions;
from the extracted power data x t Performing feature mapping to obtain power data x t Mapping to hypercube HC (HC 1t ，...，hc dt ) The method comprises the steps of carrying out a first treatment on the surface of the Wherein the hypercube has a side length dc and a diagonal length dc
The mapping function adopted is as follows:
wherein hc at Representing the mapped position index of feature data in the hypercube, y at Feature data representing an a-th feature dimension, where a=1, 2,..d, con represents a set mapping constant, floor represents a downward rounding function;
acquiring corresponding characteristic region CR (x) according to mapping points of power data in hypercube space t ) The feature region acquisition function adopted is as follows:
wherein CR (x) t ) Representing all mapped points cr satisfying the condition in the hypercube space at A composed region; wherein variable a = 1,2,..d; hc at Representing the mapped position index of the feature data in the hypercube space,representing a set characteristic value, wherein ∈>
Based on the obtained feature region CR (x t ) Detecting the number of standard mapping points contained in the feature region, wherein the standard mapping points are obtained by mapping the standard mapping points to the hypercube space according to preset standard data;
if the number of standard mapping points included in the feature region is greater than a set standard threshold Thnum, the power data x is marked t If the number of standard mapping points included in the feature region is less than or equal to the set standard threshold Thnum, marking the power data x t The global anomaly detection result of (2) is anomaly.
2. The smart grid power data anomaly detection method according to claim 1, wherein in step S2, local anomaly detection is performed according to the acquired power data, including:
according to the acquired power data X in a time period t ＝[x t-K+1 ，...，x t ]Respectively extracting feature data sequences Y corresponding to different features dt ＝[y dt-k+1 ，...，y dt ]；X t Represents the power data in a time period acquired at the time t, K represents the length of the time period, x t Representing the power data collected at time t, whichX in the middle t ＝[y 1t ，y 2t ，...y dt ]，y dt Feature data representing a d-th feature dimension at time t, d representing the total number of feature dimensions; y is Y dt A feature data sequence representing a d-th feature dimension;
from the obtained characteristic data sequence Y dt And carrying out local abnormality detection, wherein the adopted local abnormality detection function is as follows:
wherein ip (y) dt ) A local abnormal condition detection function representing a time t, where t=t-k+1,..t, when ip (y dt ) When at least one condition of (2) is satisfied, the feature data y is marked dt And corresponding power data x t Is local abnormal data; mean (Y) dt ) Representing a characteristic data sequence Y dt Average value of Thm d Representing a set deviation threshold corresponding to the d-th feature dimension; y is dt-1 And y dt+1 Characteristic data, th sigma 1, representing time t-1 and time t+1, respectively d Represents a change threshold value, σ (Y dt ) Representing a characteristic data sequence Y dt Standard deviation of Th sigma 2 d Representing a set standard deviation threshold corresponding to the d-th feature dimension;
respectively carrying out local abnormality detection results according to the characteristic sequence data to obtain local abnormality detection results of the power data;
if the acquired power data does not contain the local abnormal data, the local detection result of the power data is normal.
3. The smart grid power data anomaly detection method of claim 1, wherein the preset standard data is stored in a standard data set; and mapping the standard data into the hypercube space to obtain standard mapping points.
4. A smart grid power data anomaly detection method as claimed in claim 3, further comprising: the Sa1 updates the data in the standard data set, and maps the updated standard data to the hyperspectral space according to the updated standard data to obtain an updated standard mapping point, wherein the updating the data in the standard data set comprises the following steps:
each set time period, when the detection result of the power data is detected to be normal, initiating a replacement request aiming at the normal power data to acquire a replacement characteristic parameter rp=random (0, 1); when the replacement characteristic parameter is larger than the set threshold value, marking the normal power data as standard data and inputting the standard data into a standard data set;
further detecting the data capacity of the standard data set, and randomly selecting one standard data to be removed from the standard data set if the number of the standard data contained in the standard data set is larger than a set threshold value;
the updating of the standard data set is completed.
5. The smart grid power data anomaly detection method of claim 4, further comprising: and Sa2, mapping the newly added standard data in the standard data set into the hypercube space to obtain a standard mapping point.
6. The smart grid power data anomaly detection method of claim 1, further comprising:
s5, correcting the abnormal power data to obtain corrected power data.
7. The intelligent power grid power data anomaly detection system is characterized by comprising an acquisition unit, a local anomaly detection unit, an output unit and a global anomaly detection unit;
the acquisition unit is used for acquiring the power data acquired by the sensor nodes, wherein the power data comprises multidimensional characteristic data;
the local abnormality detection unit is used for carrying out local abnormality detection according to the acquired power data to obtain a local abnormality detection result;
if the local abnormality detection result is abnormal, the output unit outputs the power data detection result as local abnormality;
if the local abnormality detection result is normal, the global abnormality detection unit further carries out global abnormality detection on the power data to obtain a global abnormality detection result;
if the global abnormality detection result is normal, the output unit outputs the power data detection result as normal;
if the global abnormality detection result is abnormal, the output unit outputs the power data detection result as global abnormality;
the global abnormality detection unit further performs global abnormality detection on the power data, including:
extracting power data x at each time t Wherein x is t Representing power data acquired at time t, where x t ＝[y 1t ，y 2t ，...y dt ]，y dt Feature data representing a d-th feature dimension at time t, d representing the total number of feature dimensions;
from the extracted power data x t Performing feature mapping to obtain power data x t Mapping to hypercube HC (HC 1t ，...，hc dt ) The method comprises the steps of carrying out a first treatment on the surface of the Wherein the hypercube has a side length dc and a diagonal length dc
The mapping function adopted is as follows:
wherein hc at Representing the mapped position index of feature data in the hypercube, y at Feature data representing an a-th feature dimension, where a=1, 2,..d, con represents a set mapping constant, floor represents a downward rounding function;
acquiring corresponding characteristic region CR (x) according to mapping points of power data in hypercube space t ) The feature region acquisition function adopted is as follows:
wherein CR (x) t ) Representing all mapped points cr satisfying the condition in the hypercube space at A composed region; wherein variable a = 1,2,..d; hc at Representing the mapped position index of the feature data in the hypercube space,representing a set characteristic value, wherein ∈>
Based on the obtained feature region CR (x t ) Detecting the number of standard mapping points contained in the feature region, wherein the standard mapping points are obtained by mapping the standard mapping points to the hypercube space according to preset standard data;
if the number of standard mapping points included in the feature region is greater than a set standard threshold Thnum, the power data x is marked t If the number of standard mapping points included in the feature region is less than or equal to the set standard threshold Thnum, marking the power data x t The global anomaly detection result of (2) is anomaly.
8. The smart grid power data anomaly detection system of claim 7, further comprising a standard management unit; the standard management unit is used for storing the standard data set and updating the data in the standard data set.
9. The smart grid power data anomaly detection system of claim 7, further comprising a correction unit; the correction unit is used for correcting the abnormal power data to obtain corrected power data.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202310896244.1A CN116613895B (en) | 2023-07-21 | 2023-07-21 | Smart grid power data anomaly detection method and system |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202310896244.1A CN116613895B (en) | 2023-07-21 | 2023-07-21 | Smart grid power data anomaly detection method and system |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN116613895A CN116613895A (en) | 2023-08-18 |
| CN116613895B true CN116613895B (en) | 2023-10-31 |
Family
ID=87680482
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202310896244.1A Active CN116613895B (en) | 2023-07-21 | 2023-07-21 | Smart grid power data anomaly detection method and system |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN116613895B (en) |
Families Citing this family (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN117522211B (en) * | 2023-11-17 | 2024-04-16 | 国网河北省电力有限公司物资分公司 | Carbon footprint information acquisition method and device |
| CN117575176B (en) * | 2024-01-15 | 2024-04-12 | 北京浩然五洲软件技术有限公司 | Processing method and system for abnormal value in power data |
Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102045746A (en) * | 2011-01-24 | 2011-05-04 | 中国人民大学 | Event detection method based on layered grid in wireless sensor network |
| CN109560984A (en) * | 2018-11-13 | 2019-04-02 | 苏宁易购集团股份有限公司 | A kind of network service response time method for detecting abnormality and device |
| CN110362608A (en) * | 2019-06-11 | 2019-10-22 | 广东工业大学 | Energy consumption method for detecting abnormality based on rain flow method and local outlier factor |
| CN111695620A (en) * | 2020-06-08 | 2020-09-22 | 中国电力科学研究院有限公司 | Method and system for detecting and correcting abnormal data of time sequence of power system |
| CN115202314A (en) * | 2021-04-12 | 2022-10-18 | 通用电气公司 | System, method and computer readable medium for protecting network physical system |
| CN116028930A (en) * | 2023-03-28 | 2023-04-28 | 睿至科技集团有限公司 | Defense detection method and system for energy data in Internet of things |
Family Cites Families (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US8660368B2 (en) * | 2011-03-16 | 2014-02-25 | International Business Machines Corporation | Anomalous pattern discovery |
-
2023
- 2023-07-21 CN CN202310896244.1A patent/CN116613895B/en active Active
Patent Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102045746A (en) * | 2011-01-24 | 2011-05-04 | 中国人民大学 | Event detection method based on layered grid in wireless sensor network |
| CN109560984A (en) * | 2018-11-13 | 2019-04-02 | 苏宁易购集团股份有限公司 | A kind of network service response time method for detecting abnormality and device |
| CN110362608A (en) * | 2019-06-11 | 2019-10-22 | 广东工业大学 | Energy consumption method for detecting abnormality based on rain flow method and local outlier factor |
| CN111695620A (en) * | 2020-06-08 | 2020-09-22 | 中国电力科学研究院有限公司 | Method and system for detecting and correcting abnormal data of time sequence of power system |
| CN115202314A (en) * | 2021-04-12 | 2022-10-18 | 通用电气公司 | System, method and computer readable medium for protecting network physical system |
| CN116028930A (en) * | 2023-03-28 | 2023-04-28 | 睿至科技集团有限公司 | Defense detection method and system for energy data in Internet of things |
Non-Patent Citations (1)
| Title |
|---|
| 基于特征符号表示的网络异常流量检测算法;展鹏 等;浙江大学学报（工学版）;第54卷(第7期);第1281-1288页 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN116613895A (en) | 2023-08-18 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN116613895B (en) | Smart grid power data anomaly detection method and system | |
| WO2021129086A1 (en) | Traffic prediction method, device, and storage medium | |
| CN110602105B (en) | Large-scale parallelization network intrusion detection method based on k-means | |
| CN113052868B (en) | Method and device for training matting model and image matting | |
| CN116703183B (en) | Novel power distribution network system power load assessment method | |
| CN106570790B (en) | Wind power plant output data restoration method considering wind speed data segmentation characteristics | |
| CN112288597A (en) | Energy consumption online anomaly detection method based on hierarchical clustering and histogram algorithm | |
| CN116821729A (en) | Data processing system for driving measurement calibration | |
| CN114119289A (en) | Method and device for processing comprehensive energy monitoring data | |
| CN114241425A (en) | Training method and device of garbage detection model, storage medium and equipment | |
| CN107835025A (en) | A kind of electric load historical data self-adapting compressing method and system | |
| CN116706992A (en) | Self-adaptive power prediction method, device and equipment for distributed photovoltaic cluster | |
| US10338547B2 (en) | System, method and apparatuses for determining parameter settings for a power generation system and a tangible computer readable medium | |
| CN111539519A (en) | Convolutional neural network training engine method and system for mass data | |
| CN116016115A (en) | Method, device, equipment, medium and program product for monitoring flow of network line | |
| CN115408860A (en) | Abnormal value monitoring and correcting method for wind speed-power data of wind power plant | |
| CN115936156A (en) | Partition error correction method based on rotation empirical orthogonal function decomposition | |
| CN111625525B (en) | Environment data repairing/filling method and system | |
| CN114398348A (en) | Smart grid abnormal data processing method, gateway, electronic device and storage medium | |
| CN113627885A (en) | Power grid power supply abnormity monitoring system and monitoring method thereof | |
| CN114969017B (en) | Wind power data cleaning method, cleaning device and prediction method | |
| CN110942469B (en) | Dual-channel dual-mode background modeling method for high-definition high-speed video | |
| CN113268646B (en) | Abnormal user data determination method, device, server and storage medium | |
| CN113742326B (en) | Power optimizer and power missing value filling method and device thereof | |
| CN116067433A (en) | Vibration wire data acquisition method and acquisition instrument thereof |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |