CN110322368A - A kind of harmonic data method for detecting abnormality, terminal device and storage medium - Google Patents
A kind of harmonic data method for detecting abnormality, terminal device and storage medium Download PDFInfo
- Publication number
- CN110322368A CN110322368A CN201910594035.5A CN201910594035A CN110322368A CN 110322368 A CN110322368 A CN 110322368A CN 201910594035 A CN201910594035 A CN 201910594035A CN 110322368 A CN110322368 A CN 110322368A
- Authority
- CN
- China
- Prior art keywords
- data
- harmonic
- isolated
- data set
- tree
- 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.)
- Pending
Links
- 230000005856 abnormality Effects 0.000 title claims abstract description 61
- 238000000034 method Methods 0.000 title claims abstract description 28
- 238000003860 storage Methods 0.000 title claims abstract description 16
- 238000001514 detection method Methods 0.000 claims abstract description 47
- 238000004590 computer program Methods 0.000 claims description 21
- 241001269238 Data Species 0.000 claims description 8
- 238000004140 cleaning Methods 0.000 claims description 3
- 238000004422 calculation algorithm Methods 0.000 abstract description 35
- 239000000463 material Substances 0.000 abstract description 3
- 230000002159 abnormal effect Effects 0.000 description 22
- 238000012549 training Methods 0.000 description 17
- 238000013528 artificial neural network Methods 0.000 description 11
- 238000010586 diagram Methods 0.000 description 11
- 230000006870 function Effects 0.000 description 6
- 238000010276 construction Methods 0.000 description 5
- 238000002474 experimental method Methods 0.000 description 5
- 230000009466 transformation Effects 0.000 description 5
- 238000004458 analytical method Methods 0.000 description 4
- 230000008569 process Effects 0.000 description 4
- 238000012545 processing Methods 0.000 description 4
- 238000005070 sampling Methods 0.000 description 4
- 238000003491 array Methods 0.000 description 3
- 230000008901 benefit Effects 0.000 description 3
- 238000004364 calculation method Methods 0.000 description 3
- 238000004378 air conditioning Methods 0.000 description 2
- 230000005540 biological transmission Effects 0.000 description 2
- 238000013480 data collection Methods 0.000 description 2
- 238000009826 distribution Methods 0.000 description 2
- 230000000694 effects Effects 0.000 description 2
- FNMKZDDKPDBYJM-UHFFFAOYSA-N 3-(1,3-benzodioxol-5-yl)-7-(3-methylbut-2-enoxy)chromen-4-one Chemical compound C1=C2OCOC2=CC(C2=COC=3C(C2=O)=CC=C(C=3)OCC=C(C)C)=C1 FNMKZDDKPDBYJM-UHFFFAOYSA-N 0.000 description 1
- 241000208340 Araliaceae Species 0.000 description 1
- 208000025274 Lightning injury Diseases 0.000 description 1
- 238000000342 Monte Carlo simulation Methods 0.000 description 1
- 235000005035 Panax pseudoginseng ssp. pseudoginseng Nutrition 0.000 description 1
- 235000003140 Panax quinquefolius Nutrition 0.000 description 1
- 230000008859 change Effects 0.000 description 1
- 230000000052 comparative effect Effects 0.000 description 1
- 238000005520 cutting process Methods 0.000 description 1
- 230000003247 decreasing effect Effects 0.000 description 1
- 230000002708 enhancing effect Effects 0.000 description 1
- 235000008434 ginseng Nutrition 0.000 description 1
- 230000008676 import Effects 0.000 description 1
- 230000006872 improvement Effects 0.000 description 1
- 238000007689 inspection Methods 0.000 description 1
- 230000014759 maintenance of location Effects 0.000 description 1
- 238000013178 mathematical model Methods 0.000 description 1
- 238000005259 measurement Methods 0.000 description 1
- 238000012544 monitoring process Methods 0.000 description 1
- 230000035772 mutation Effects 0.000 description 1
- 235000001968 nicotinic acid Nutrition 0.000 description 1
- 238000007781 pre-processing Methods 0.000 description 1
- 238000011160 research Methods 0.000 description 1
- 238000004088 simulation Methods 0.000 description 1
- 238000001228 spectrum Methods 0.000 description 1
- 238000012360 testing method Methods 0.000 description 1
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F17/00—Digital computing or data processing equipment or methods, specially adapted for specific functions
- G06F17/10—Complex mathematical operations
- G06F17/15—Correlation function computation including computation of convolution operations
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/21—Design or setup of recognition systems or techniques; Extraction of features in feature space; Blind source separation
- G06F18/214—Generating training patterns; Bootstrap methods, e.g. bagging or boosting
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q50/00—Information and communication technology [ICT] specially adapted for implementation of business processes of specific business sectors, e.g. utilities or tourism
- G06Q50/06—Energy or water supply
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Data Mining & Analysis (AREA)
- Business, Economics & Management (AREA)
- Economics (AREA)
- Pure & Applied Mathematics (AREA)
- Mathematical Physics (AREA)
- Mathematical Optimization (AREA)
- Mathematical Analysis (AREA)
- Computational Mathematics (AREA)
- General Engineering & Computer Science (AREA)
- Health & Medical Sciences (AREA)
- Marketing (AREA)
- Software Systems (AREA)
- Tourism & Hospitality (AREA)
- Computing Systems (AREA)
- Strategic Management (AREA)
- Primary Health Care (AREA)
- Human Resources & Organizations (AREA)
- Algebra (AREA)
- General Health & Medical Sciences (AREA)
- Water Supply & Treatment (AREA)
- Databases & Information Systems (AREA)
- General Business, Economics & Management (AREA)
- Public Health (AREA)
- Life Sciences & Earth Sciences (AREA)
- Artificial Intelligence (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Bioinformatics & Computational Biology (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Evolutionary Biology (AREA)
- Evolutionary Computation (AREA)
- Remote Monitoring And Control Of Power-Distribution Networks (AREA)
Abstract
Include: S1 in this method the present invention relates to a kind of harmonic data method for detecting abnormality, terminal device and storage medium: the harmonic current data that acquisition equipment generates in a period of time sequence is as data set;S2: collection is built with the forest of more isolated trees composition according to the collected data;S3: the path length of each harmonic data in data set is calculated;S4: the average path length of all data in data set is calculated;S5: the abnormality score of each harmonic data is calculated;S6: exception is determined if according to the abnormality score of each harmonic data.The present invention not only reduces the expenditure of manpower and material resources, increases economic efficiency and service quality, increase administration of power networks level also promote electric power netting safe running ability by isolating forest algorithm realization to the abnormality detection of harmonic data.
Description
Technical field
The present invention relates to harmonic data abnormality detection technical field more particularly to a kind of sides of harmonic data abnormality detection
Method, terminal device and storage medium.
Background technique
It is run under open-air atmosphere since electric power networks transmission is generally lain in, voltage and current can during power transmission
The influence of the incidents such as lightning stroke can be will receive.In the harmonic data of generation, it may appear that differ larger with daily monitoring harmonic wave
And the data that the duration is very short, and these abnormal harmonic datas will affect the accuracy of harmonic wave normality analysis.Study harmonic wave
Data exception detection, grasps the harmonic wave actual state of electric system, prevents abnormal Harmfulness Caused by Harmonics power grid, to promotion electric power networks peace
Row for the national games promotes economic benefit, and enhancing administration of power networks is horizontal, and improving grid service quality has important influence.
It is most commonly used that abnormal harmonics based on Fourier transformation, which are detected in current practical application, but Fourier becomes
Change there are the shortcomings that spectrum leakage effect cause based on its harmonic detecting method reliability it is bad, and Fourier calculate work
Amount is very big, and actual requirement is also not achieved in accuracy as a result.The theoretical basis of wavelet transformation is Fourier transformation, is provided simultaneously with
The characteristics of different from Fourier transformation, but wavelet transformation is there are disadvantage, structural model are built than less easily, reliability
It is not high.
Artificial neural network (ArtificialNeuralNetworks) theory belongs to the subject of an edge crossing, hair
Open up swift and violent and stylish, theoretical basis is bionics, have powerful independent learning ability, but when due to abnormal harmonics detection it
It needs a large amount of training sample, at present the artificial neural network construction method without standard and calculation has training sample
Serious dependence.
Summary of the invention
To solve the above-mentioned problems, the invention proposes a kind of harmonic data method for detecting abnormality, terminal device and storages
Medium.
Concrete scheme is as follows:
A kind of harmonic data method for detecting abnormality, comprising the following steps:
S1: the harmonic data that acquisition equipment generates in a period of time sequence is as data set;
S2: collection is built with the forest of more isolated trees composition according to the collected data;
S3: calculating the path length h (x) of each harmonic data in data set, and x indicates harmonic data;
S4: the average path length C (n) of all harmonic datas in data set is calculated:
H (n-1)=Ln (n-1)+γ
Wherein, n indicates the data count of harmonic data in data set, and γ is Euler's constant;
S5: the abnormality score S (x, n) of each harmonic data is calculated:
Wherein, E (h (x)) indicates the average value of the path length of all harmonic datas;
S6: exception is determined if according to the abnormality score of each harmonic data.
Further, step S1 further includes carrying out cleaning to the harmonic current data of acquisition to reject junk data and delete superfluous
Remainder evidence.
Further, step S2 specifically includes the following steps:
S201: Ψ data are randomly choosed from data set as Sub Data Set;
S202: a data are randomly selected from Sub Data Set and are put into the root node of isolated tree, and choose harmonic data
Classification Index of one index as isolated tree;
S203: the divide value of Classification Index is set;
S204: the data point of divide value of the divide value of index to data and more than or equal to index will be less than in Sub Data Set
It is not set as the left branch and right branch of node;
S205: repeating step S203 and S204, and looping construct isolates the left and right branch of tree node, until the height of isolated tree
Reach only one data in given threshold or child node.
Further, the maximum height of every isolated tree is h=log2Ψ。
It further, include 256 data in each Sub Data Set, the number of isolated tree is 100.
Further, step S6 specifically: as S (x, n) → 1, be determined as exception;As S (x, n) → 0, determine to be positive
Often.
A kind of harmonic data abnormality detection terminal device, including processor, memory and storage are in the memory
And the computer program that can be run on the processor, the processor realize that the present invention is real when executing the computer program
The step of applying example above-mentioned method.
A kind of computer readable storage medium, the computer-readable recording medium storage have computer program, feature
The step of being, above-mentioned method of the embodiment of the present invention realized when the computer program is executed by processor.
The present invention uses technical solution as above, by isolating forest algorithm realization to the abnormality detection of harmonic data, not only
The expenditure for reducing manpower and material resources, increases economic efficiency and service quality, increase administration of power networks level also promote electric power netting safe running
Ability.
Detailed description of the invention
Fig. 1 show the harmonic current data schematic diagram of the acquisition of the embodiment of the present invention one.
Fig. 2 show rejecting outliers process schematic in the embodiment.
Fig. 3 show the construction flow chart of the isolated tree in the embodiment.
Fig. 4 show the abnormal harmonics overhaul flow chart based on isolated forest in the embodiment.
Fig. 5 show the building forest flow chart in the embodiment.
Fig. 6 show the building flow chart of the isolated tree in the embodiment.
Fig. 7 show the path length calculation flow chart in the embodiment.
Fig. 8 show the schematic diagram of random data group abnormality detection in the embodiment.
Fig. 9 show the schematic diagram of harmonic voltage value data in the embodiment.
Figure 10 show the schematic diagram of harmonic voltage abnormality detection in the embodiment.
Figure 11 show the schematic diagram of equipment A phase triple harmonic current virtual value abnormality detection in the embodiment.
Figure 12 show the schematic diagram of equipment B phase triple harmonic current virtual value abnormality detection in the embodiment.
Figure 13 show the schematic diagram of equipment C phase triple harmonic current virtual value abnormality detection in the embodiment.
Figure 14 show the schematic diagram of equipment A phase triple harmonic current abnormality detection in the embodiment.
Figure 15 show the schematic diagram of equipment B phase triple harmonic current abnormality detection in the embodiment.
Figure 16 show the schematic diagram of equipment C phase triple harmonic current abnormality detection in the embodiment.
Specific embodiment
To further illustrate that each embodiment, the present invention are provided with attached drawing.These attached drawings are that the invention discloses one of content
Point, mainly to illustrate embodiment, and the associated description of specification can be cooperated to explain the operation principles of embodiment.Cooperation ginseng
These contents are examined, those of ordinary skill in the art will be understood that other possible embodiments and advantages of the present invention.
Now in conjunction with the drawings and specific embodiments, the present invention is further described.
Embodiment one:
The embodiment of the present invention one provides a kind of harmonic data method for detecting abnormality, comprising the following steps:
Step 1: the harmonic data of equipment is acquired.
The harmonic current data for acquiring air-conditioning in laboratory in the embodiment using power quality analyzer, due to air-conditioning
It is that power grid three-phase accesses, when installing power quality analyzer, A, B, C three-phase of utility network are wanted and power quality analyzer
Three-phase measurement current transform er it is corresponding.Harmonic current data acquires the duration close to five hours, laboratory environment temperature
Degree is 24 DEG C, and the operating temperature of equipment setting is 23 DEG C.
Collected harmonic current data is exported using software PQ Diffractor, and it is humorous three times therefrom to extract A, B, C three-phase
Wave current effective value.Since there are a large amount of manifest error data in the initial data of acquisition, by pretreatment, rubbish is rejected in cleaning
Rubbish data, after deleting redundant data, it is as shown in Figure 1 that last Effective selection obtains 156 harmonic current datas.Horizontal seat in Fig. 1
Mark represents Beijing time when sampling, and ordinate represents three-phase triple harmonic current virtual value.By Fig. 1 it can be concluded that B phase three times
Primary acutely mutation occurs when equipment starts (about 1:48) soon for harmonic current, then falls back to acceptable region at once
Value illustrates that B phase triple-frequency harmonics data value 1.8297 is an abnormal data, manually can be concluded that.After this, A, B, C tri-
Phase triple harmonic current degree of fluctuation is small, and situation is substantially similar, can not the abnormal judgement of direct labor's progress.
Step 2: isolated forest model is constructed, determines the exceptional value in the harmonic data of acquisition.
There are two types of characteristics for abnormal data tool, few and unusual.Isolated forest algorithm is suitable for the exception of continuous data
Exception definition can be " being easy the outlier isolated " by detection, popular to be interpreted as being distributed sparse and high density group farther out
Point.Inside the big region of data, the data probability that densely distributed region is interpreted as falling in this region is very high, thus is fallen
Data in these regions are normal;Otherwise the sparse probability of distribution is extremely low.
Isolated forest algorithm does not both have to define mathematical model or does not need markd training set, and uses a kind of very high
The strategy of effect.Assuming that dividing data space with a random planar, divides and two sub-regions once can be generated.Each sub-district
Domain continuation is subdivided, and circulation is gone down always, until only existing a data point inside each subregion.The high number of closeness
It is that can divide region many times can just stop, but the extremely low data point of closeness is easy to and stops at one very early according to cluster
In sub-regions, thus the realization of the isolated forest algorithm isolated forest that include building be made of t isolated tree and to data into
Row abnormality detection.
1, the forest being made of t isolated tree is constructed.
The isolated tree of isolated forest is binary tree structure.In binary tree, N is that the node of T if N is leaf node claims it
For external node;If N is the node with two child nodes, it is called internal node.Due to being random division, need to use
Monte carlo method obtains a convergency value, and repetition is divided since node, then the average result divided every time.
The construction process of isolated tree:
(1) Ψ data are randomly choosed from data set as Sub Data Set;
(2) data point is randomly selected from Sub Data Set and is put into isolated root vertex, and root node is randomly assigned dimension
Generate divide value P again (i.e. divide value results from Sub Data Set between the maximin of specified dimension);
(3) a division plane is generated with divide value P, data space is then divided into two sub-regions, in specified dimension
Using the data less than divide value P as the left branch of node in degree, using the big data equal to divide value P as the right side of node point
Branch;
(4) recursion step (2) and (3) in child node, constantly construct new child node, until there was only one in child node
A data (can not be further continued for cutting) or child node have reached the height log of restriction2Ψ。
2, detected data is carried out abnormality detection.
Path length refers to that in an isolated tree, the number on undergone side, is denoted as h (x) from root node to external node.
For a training dataset X, it is enabled to traverse each isolated tree, then calculates each training data x in training dataset X
Which layer (i.e. final height of the training data x in isolated tree) of isolated tree finally is fallen in, then calculates training data x in isolated tree
Path length h (x).The schematic diagram that test data traverses isolated tree is as shown in Figure 2.
The height of every isolated tree is normalized again, calculates the average path length C of all training datas
(n), as shown in formula (1).
Wherein, n indicate training data sum, H (i)=Ln (i)+γ, γ be Euler's constant (i.e. γ=
0.5772156649)。
The abnormality score of each training data x is finally calculated, so that whether training of judgement data x is abnormal.Training data x's
Shown in the calculation formula of abnormality score S (x, n) such as formula (2).
Wherein E (h (x)) is the average value of the path length h (x) of all training datas.
When the value of E (h (x)) level off to 0 when, then the value of S (x, n) levels off to 1, illustrates that training data x is exceptional value;Work as E
The value of (h (x)) level off to n-1 when, then the value of S (x, n) levels off to 0, illustrates that training data x is normal value;When the value of E (h (x))
Level off to C (n) when, then the value of S (x, n) levels off to 0.5, illustrate training data x cannot distinguish between whether exceptional value.
3, based on the harmonic data abnormality detection of isolated forest algorithm.
When processing using isolated forest algorithm harmonic data, the construction of isolated tree is as shown in Figure 3.
Since the harmonic data generated when equipment operation is all continuous variable, then isolated tree realizes that process is as follows:
(1) index in 2 harmonic data indexs is randomly selected;
(2) a value p of the harmonic wave index is randomly choosed;
(3) it is less than the left side for being placed on tree node of p, the right for being placed on tree node greater than p in index;
(4) the left and right branch of looping construct tree node, when the data record of loading is constant or the height of isolated tree reaches
Setting value then stops constructing.
Using the flow chart of isolated forest algorithm construction harmonic data method for detecting abnormality, as shown in Figure 4.Wherein every orphan
The maximum height of vertical tree is h=log2Ψ.Harmonic data is analyzed by isolated forest algorithm, according to abnormality score S's
Size picks out the abnormal data in harmonic wave.
Step 3: isolated forest model software realization.
Isolated forest model is realized in the embodiment using MATLAB software.Software program design is by four subprogram groups
At, wherein three are used to describe isolated forest algorithm for subfunction program, and it is in addition that main program is used to call subfunction program, it is real
Now isolate the operation of forest algorithm, the program thread of three sub- function programs introduced below.
First sub- function program is to construct the forest being made of t isolated tree, as shown in Figure 5.
Second sub- function program is building isolated tree, as shown in Figure 6.
The sub- function program of third is to calculate path length, as shown in Figure 7.
Step 4: experiment simulation.
By the analysis to isolated forest algorithm, it is known that with the increase and number of sampled samples of isolated tree quantity
Increase, the time of algorithm operation can also increase therewith, and precision improvement is limited after the quantity of isolated tree reaches certain value, and isolated
The quantity of tree is excessive, and model performance is caused to be decreased obviously.So taking the sample number Ψ of isolated tree in the embodiment is 256
It is a, the tree t=100 of isolated tree.
Emulation experiment is divided into three phases progress: the first stage is random data group's abnormality detection, and second stage is last year
Data exception detection and phase III are experiment acquisition data exception detections.Circle encloses the point come in Fig. 8,10,11,12,13
Data exception is represented, it is normal that remaining point represents data.
(1) random data group abnormality detection.
Setting generate three random data groups, one of them be generate by between (0,1) uniform random number form
Ten rows, two columns group, the other two are generate by between (0,1) standardized normal distribution random number form 300 rows two column
Array, then basic arithmetic is carried out to three arrays of generation and generates new overall data group, collectively generate 610 arrays.
Data set is imported in isolated forest algorithm model and is analyzed, as a result as shown in Figure 8.
(2) given data collection abnormality detection.
Given data collection derives from the harmonic voltage data of last year acquisition, and the harmonic voltage data volume of acquisition is 450, will
Harmonic voltage data set line chart 9 shows.It observes harmonic voltage and fluctuates situation, import in isolated forest algorithm model and divide
Analysis, the results are shown in Figure 10, and the abnormal point table 1 in Figure 10 is enumerated.The abscissa of line chart represents time, ordinate generation
Table harmonic voltage.
Table 1
(3) experiment acquisition data exception detection
Using the harmonic current virtual value that equipment generates as research object, the harmonic wave sampling interval 1 minute, 300 are obtained altogether
Sampled point.Then after pre-processing to harmonic current virtual value, effective sampling points 156 are obtained.By collected equipment
A, B, C three-phase triple harmonic current virtual value are loaded into isolated forest algorithm model and obtain as shown in Figure 11,12,13, and will figure
11, abnormal point table 2, table 3 and the table 4 in 12,13 are enumerated.
Table 2
Table 3
Table 4
Ordinate respectively represents harmonic voltage and current effective value in Figure 10,11,12,13, and abscissa represents when sampling
Beijing time.Abnormality degree refers to that abnormal current data point accounts for the percentage of total size.
It observes in Fig. 8 it can be found that distinguishing biggish random array can be identified compared with other arrays.In conjunction with
It is identified in Fig. 9, Figure 10 and table 1 it can be seen that deviateing apparent harmonic voltage point, illustrates that isolated forest algorithm can answer
It is detected for abnormal harmonics.
According to the reasonable artificial judgment of Fig. 1 as a result, the current effective value that B phase triple-frequency harmonics occurs 48 points on one point is agreed
It surely is exceptional value, this manually can be concluded that, and carry out abnormal mark to this obvious abnormal harmonic data in Figure 12, say
Bright isolated forest algorithm is effective to harmonic current abnormality detection.
Under identical judgment threshold, A, B, C three-phase triple harmonic current virtual value abnormality degree is 3.21% respectively,
2.56% and 2.56%, when illustrating equipment operation, there is a possibility that abnormal greater than B phase and C in A phase triple harmonic current virtual value
Phase, therefore A phase triple harmonic current virtual value can cause harmonic current data abnormal vulnerable to the influence of fluctuation.According to table 2,3
And 4 it is found that there is abnormal data simultaneously in A, B, C three-phase triple harmonic current virtual value at 4 points between 15 minutes to 30 minutes,
Illustrate in this time, fluctuation is violent when equipment is run.
In conclusion isolated forest algorithm has the ability for quick and precisely identifying harmonic current abnormal data, can push away
It is wide to arrive detection harmonic voltage.And fluctuation can be oriented as ordinate using the real time and occurred at some time point.Also
According to prevailing circumstances situation, it is inferred to that factor causes mains by harmonics unusual fluctuations occur.
Step 5: algorithm Accuracy Analysis.
In order to analyze the accuracy of isolated forest algorithm, therefore carried out using based on the detection of neural network algorithm abnormal harmonics
Comparative result.It is mainly characterized by based on neural network algorithm: directly detection harmonic current data line chart, according to neural network
The setting of algorithm threshold value, the data higher than threshold line are all exceptional values.Experimental result picture based on neural network algorithm such as Figure 14,
15, shown in 16, and the fault point table 5,6,7 in Figure 14,15,16 is enumerated.Abscissa represents the time in Figure 14,15,16, indulges
Coordinate represents three-phase triple harmonic current virtual value.
Table 5
Table 6
Table 7
Contrast table 2,3,4 analyzes two kinds of detection methods to A phase triple harmonic current virtual value abnormality detection with table 5,6,7
As a result almost the same;To B, C phase triple harmonic current virtual value abnormality detection, the harmonic current data processing of forest algorithm is isolated
Can be more more accurate than neural network algorithm, it can detect the abnormal data lower than threshold value.But neural network algorithm can be to minority
Data are more sensitive.In conclusion based on isolated forest algorithm abnormality detection result with to be based on neural network almost the same.It is used to
The accuracy of the neural network algorithm of comparison is up to 90%, therefore the accuracy of isolated forest algorithm can also be up to 90%.
The embodiment of the present invention one, which is proposed, carries out abnormality detection harmonic data using isolated forest algorithm, constructs isolated
The harmonic data abnormality detection model of forest algorithm.By carrying out emulation experiment to the model of foundation, to collected harmonic number
According to being analyzed, as a result demonstrating isolated forest algorithm has the energy that can quick and precisely identify harmonic data appearance exception
Power, and can also orient in the generation of some period, the inspection of harmonic data exception is realized in verifying using isolated forest algorithm
Measuring tool has practical application value, not only reduces the expenditure of manpower and material resources, increases economic efficiency and service quality, increases administration of power networks
Level also promotes electric power netting safe running ability.
Embodiment two:
The present invention also provides a kind of harmonic data abnormality detection terminal device, including memory, processor and it is stored in
In the memory and the computer program that can run on the processor, when the processor executes the computer program
Realize the step in the above method embodiment of the embodiment of the present invention one.
Further, as an executable scheme, the harmonic data abnormality detection terminal device can be desktop
Computer, notebook, palm PC and cloud server etc. calculate equipment.The harmonic data abnormality detection terminal device can wrap
It includes, but is not limited only to, processor, memory.It will be understood by those skilled in the art that above-mentioned harmonic data abnormality detection terminal is set
Standby composed structure is only the example of harmonic data abnormality detection terminal device, is not constituted whole to harmonic data abnormality detection
The restriction of end equipment may include perhaps combining certain components or different components than above-mentioned more or fewer components,
Such as the harmonic data abnormality detection terminal device can also be including input-output equipment, network access equipment, bus etc., this
Inventive embodiments do not limit this.
Further, as an executable scheme, alleged processor can be central processing unit (Central
Processing Unit, CPU), it can also be other general processors, digital signal processor (Digital Signal
Processor, DSP), it is specific integrated circuit (Application Specific Integrated Circuit, ASIC), existing
At programmable gate array (Field-Programmable Gate Array, FPGA) or other programmable logic device, discrete
Door or transistor logic, discrete hardware components etc..General processor can be microprocessor or the processor can also
To be any conventional processor etc., the processor is the control centre of the harmonic data abnormality detection terminal device, benefit
With the various pieces of various interfaces and the entire harmonic data abnormality detection terminal device of connection.
The memory can be used for storing the computer program and/or module, and the processor is by operation or executes
Computer program in the memory and/or module are stored, and calls the data being stored in memory, described in realization
The various functions of harmonic data abnormality detection terminal device.The memory can mainly include storing program area and storing data
Area, wherein storing program area can application program needed for storage program area, at least one function;Storage data area can store
Created data etc. are used according to mobile phone.In addition, memory may include high-speed random access memory, can also include
Nonvolatile memory, such as hard disk, memory, plug-in type hard disk, intelligent memory card (Smart Media Card, SMC), safety
Digital (Secure Digital, SD) card, flash card (Flash Card), at least one disk memory, flush memory device or
Other volatile solid-state parts.
The present invention also provides a kind of computer readable storage medium, the computer-readable recording medium storage has computer
Program, when the computer program is executed by processor the step of the realization above method of the embodiment of the present invention.
If the integrated module/unit of the harmonic data abnormality detection terminal device is real in the form of SFU software functional unit
Now and when sold or used as an independent product, it can store in a computer readable storage medium.Based in this way
Understanding, the present invention realize above-described embodiment method in all or part of the process, can also be instructed by computer program
Relevant hardware is completed, and the computer program can be stored in a computer readable storage medium, the computer program
When being executed by processor, it can be achieved that the step of above-mentioned each embodiment of the method.Wherein, the computer program includes computer
Program code, the computer program code can be source code form, object identification code form, executable file or certain centres
Form etc..The computer-readable medium may include: can carry the computer program code any entity or device,
Recording medium, USB flash disk, mobile hard disk, magnetic disk, CD, computer storage, read-only memory (ROM, ROM, Read-Only
Memory), random access memory (RAM, Random Access Memory) and software distribution medium etc..
Although specifically showing and describing the present invention in conjunction with preferred embodiment, those skilled in the art should be bright
It is white, it is not departing from the spirit and scope of the present invention defined by the appended claims, it in the form and details can be right
The present invention makes a variety of changes, and is protection scope of the present invention.
Claims (8)
1. a kind of harmonic data method for detecting abnormality, which comprises the following steps:
S1: the harmonic data that acquisition equipment generates in a period of time sequence is as data set;
S2: collection is built with the forest of more isolated trees composition according to the collected data;
S3: calculating the path length h (x) of each harmonic data in data set, and x indicates harmonic data;
S4: the average path length C (n) of all harmonic datas in data set is calculated:
H (n-1)=Ln (n-1)+γ
Wherein, n indicates the data count of harmonic data in data set, and γ is Euler's constant;
S5: the abnormality score S (x, n) of each harmonic data is calculated:
Wherein, E (h (x)) indicates the average value of the path length of all harmonic datas;
S6: exception is determined if according to the abnormality score of each harmonic data.
2. harmonic data method for detecting abnormality according to claim 1, it is characterised in that: step S1 further includes to acquisition
Harmonic current data carries out cleaning and rejects junk data and delete redundant data.
3. harmonic data method for detecting abnormality according to claim 1, it is characterised in that: step S2 specifically includes following step
It is rapid:
S201: Ψ data are randomly choosed from data set as Sub Data Set;
S202: a data are randomly selected from Sub Data Set and are put into the root node of isolated tree, and choose one of harmonic data
Classification Index of the index as isolated tree;
S203: the divide value of Classification Index is set;
S204: the divide value of index will be less than in Sub Data Set and set respectively to data and more than or equal to the data of the divide value of index
It is set to the left branch and right branch of node;
S205: repeating step S203 and S204, and looping construct isolates the left and right branch of tree node, until the height of isolated tree reaches
Only one data in given threshold or child node.
4. harmonic data method for detecting abnormality according to claim 3, it is characterised in that: the maximum height of every isolated tree
For h=log2Ψ。
5. harmonic data method for detecting abnormality according to claim 3, it is characterised in that: include in each Sub Data Set
256 data, the number of isolated tree are 100.
6. harmonic data method for detecting abnormality according to claim 1, it is characterised in that: step S6 specifically: when S (x,
When n) → 1, it is determined as exception;As S (x, n) → 0, it is determined as normal.
7. a kind of harmonic data abnormality detection terminal device, it is characterised in that: including processor, memory and be stored in described
The computer program run in memory and on the processor, the processor are realized such as when executing the computer program
In claim 1~6 the step of any the method.
8. a kind of computer readable storage medium, the computer-readable recording medium storage has computer program, and feature exists
In when the computer program is executed by processor in realization such as claim 1~6 the step of any the method.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201910594035.5A CN110322368A (en) | 2019-07-03 | 2019-07-03 | A kind of harmonic data method for detecting abnormality, terminal device and storage medium |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201910594035.5A CN110322368A (en) | 2019-07-03 | 2019-07-03 | A kind of harmonic data method for detecting abnormality, terminal device and storage medium |
Publications (1)
Publication Number | Publication Date |
---|---|
CN110322368A true CN110322368A (en) | 2019-10-11 |
Family
ID=68122508
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201910594035.5A Pending CN110322368A (en) | 2019-07-03 | 2019-07-03 | A kind of harmonic data method for detecting abnormality, terminal device and storage medium |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN110322368A (en) |
Cited By (6)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN111669368A (en) * | 2020-05-07 | 2020-09-15 | 宜通世纪科技股份有限公司 | End-to-end network sensing abnormity detection and analysis method, system, device and medium |
CN113109642A (en) * | 2021-03-10 | 2021-07-13 | 中国电力科学研究院有限公司 | Method and system for processing power quality signal data |
CN113484662A (en) * | 2021-06-02 | 2021-10-08 | 青岛鼎信通讯股份有限公司 | Fault indicator starting wave recording method |
CN114609546A (en) * | 2022-03-10 | 2022-06-10 | 东南大学 | Modularized multi-level converter open-circuit fault diagnosis method based on isolated forest |
CN116756597A (en) * | 2023-08-16 | 2023-09-15 | 山东泰开电力电子有限公司 | Wind turbine generator harmonic data real-time monitoring method based on artificial intelligence |
WO2024131358A1 (en) * | 2022-12-19 | 2024-06-27 | 蔚来电池科技(安徽)有限公司 | Battery cell consistency evaluation method, apparatus, system and medium |
Citations (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN102066956A (en) * | 2008-04-14 | 2011-05-18 | 努沃尔特股份有限公司 | Electrical anomaly detection method and system |
CN202330555U (en) * | 2011-09-06 | 2012-07-11 | 常州帕威尔测控技术有限公司 | Harmonic detection circuit device |
CN109859029A (en) * | 2019-01-04 | 2019-06-07 | 深圳壹账通智能科技有限公司 | Abnormal application detection method, device, computer equipment and storage medium |
-
2019
- 2019-07-03 CN CN201910594035.5A patent/CN110322368A/en active Pending
Patent Citations (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN102066956A (en) * | 2008-04-14 | 2011-05-18 | 努沃尔特股份有限公司 | Electrical anomaly detection method and system |
CN202330555U (en) * | 2011-09-06 | 2012-07-11 | 常州帕威尔测控技术有限公司 | Harmonic detection circuit device |
CN109859029A (en) * | 2019-01-04 | 2019-06-07 | 深圳壹账通智能科技有限公司 | Abnormal application detection method, device, computer equipment and storage medium |
Non-Patent Citations (2)
Title |
---|
余翔等: "基于孤立森林算法的用电数据异常检测研究", 《信息技术》 * |
李新鹏等: "基于孤立森林算法的电力调度流", 《电网技术》 * |
Cited By (9)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN111669368A (en) * | 2020-05-07 | 2020-09-15 | 宜通世纪科技股份有限公司 | End-to-end network sensing abnormity detection and analysis method, system, device and medium |
CN111669368B (en) * | 2020-05-07 | 2022-12-06 | 宜通世纪科技股份有限公司 | End-to-end network sensing abnormity detection and analysis method, system, device and medium |
CN113109642A (en) * | 2021-03-10 | 2021-07-13 | 中国电力科学研究院有限公司 | Method and system for processing power quality signal data |
CN113109642B (en) * | 2021-03-10 | 2023-12-05 | 中国电力科学研究院有限公司 | Method and system for processing power quality signal data |
CN113484662A (en) * | 2021-06-02 | 2021-10-08 | 青岛鼎信通讯股份有限公司 | Fault indicator starting wave recording method |
CN114609546A (en) * | 2022-03-10 | 2022-06-10 | 东南大学 | Modularized multi-level converter open-circuit fault diagnosis method based on isolated forest |
WO2024131358A1 (en) * | 2022-12-19 | 2024-06-27 | 蔚来电池科技(安徽)有限公司 | Battery cell consistency evaluation method, apparatus, system and medium |
CN116756597A (en) * | 2023-08-16 | 2023-09-15 | 山东泰开电力电子有限公司 | Wind turbine generator harmonic data real-time monitoring method based on artificial intelligence |
CN116756597B (en) * | 2023-08-16 | 2023-11-14 | 山东泰开电力电子有限公司 | Wind turbine generator harmonic data real-time monitoring method based on artificial intelligence |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN110322368A (en) | A kind of harmonic data method for detecting abnormality, terminal device and storage medium | |
CN112149873B (en) | Low-voltage station line loss reasonable interval prediction method based on deep learning | |
CN103605016B (en) | A kind of power quality data treating method and apparatus | |
CN108802535B (en) | Screening method, main interference source identification method and device, server and storage medium | |
CN104504508A (en) | Area closed-loop data analysis method on basis of analytical hierarchy process and wavelet regression | |
CN108009730A (en) | A kind of photovoltaic power station system health status analysis method | |
CN115563477B (en) | Harmonic data identification method, device, computer equipment and storage medium | |
CN104156775B (en) | Meteorological disaster Forecasting Methodology based on arithmetic of linearity regression | |
CN110287762A (en) | A kind of non-intrusion type load discrimination method and device based on data mining technology | |
CN110244185A (en) | A kind of multi-source harmonic contributions division methods, terminal device and storage medium | |
CN108399505B (en) | Distributed photovoltaic power supply access capacity planning method and terminal equipment | |
CN115239105A (en) | Method and device for evaluating wind resources of in-service wind power plant | |
CN115184674A (en) | Insulation test method and device, electronic terminal and storage medium | |
CN114240041A (en) | Lean line loss analysis method and system for distribution network distribution area | |
CN103617356A (en) | Self-adaptive on-line monitoring data trend extraction method | |
CN113538063A (en) | Electricity charge abnormal data analysis method, device, equipment and medium based on decision tree | |
CN112668225A (en) | Distribution network grid planning method and device, computer equipment and storage medium | |
CN109033564A (en) | A kind of aircraft airborne equipment testability demonstration index evaluating method | |
CN116780518A (en) | Theoretical power generation amount acquisition method, device, equipment and medium | |
CN110033236A (en) | A kind of project duplicate checking method and system based on concurrent tasks | |
CN115292202A (en) | Product test analysis method and device, electronic equipment and storage medium | |
CN113793049B (en) | Method, device, equipment and medium for positioning bad root cause in production process of product | |
CN112967154B (en) | Assessment method and device for Well-rolling of power system | |
CN110896218B (en) | Harmonic modeling method and system for establishing collective residential load | |
CN116973695A (en) | Partial discharge mode identification method, device and equipment for power cable and storage medium |
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 | ||
WD01 | Invention patent application deemed withdrawn after publication | ||
WD01 | Invention patent application deemed withdrawn after publication |
Application publication date: 20191011 |