CN106909490B - Monitoring equipment data flow evaluation and noise elimination method - Google Patents

Monitoring equipment data flow evaluation and noise elimination method Download PDF

Info

Publication number
CN106909490B
CN106909490B CN201710111079.9A CN201710111079A CN106909490B CN 106909490 B CN106909490 B CN 106909490B CN 201710111079 A CN201710111079 A CN 201710111079A CN 106909490 B CN106909490 B CN 106909490B
Authority
CN
China
Prior art keywords
point
data
correction value
mutation
detection
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
Application number
CN201710111079.9A
Other languages
Chinese (zh)
Other versions
CN106909490A (en
Inventor
蔡宇翔
肖琦敏
李霆
周晓东
付婷
王雪晶
蔡力军
董衍旭
陈锐
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
State Grid Corp of China SGCC
State Grid Fujian Electric Power Co Ltd
Information and Telecommunication Branch of State Grid Fujian Electric Power Co Ltd
Original Assignee
State Grid Corp of China SGCC
State Grid Fujian Electric Power Co Ltd
Information and Telecommunication Branch of State Grid Fujian Electric Power Co Ltd
Priority date (The priority date 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 date listed.)
Filing date
Publication date
Application filed by State Grid Corp of China SGCC, State Grid Fujian Electric Power Co Ltd, Information and Telecommunication Branch of State Grid Fujian Electric Power Co Ltd filed Critical State Grid Corp of China SGCC
Priority to CN201710111079.9A priority Critical patent/CN106909490B/en
Publication of CN106909490A publication Critical patent/CN106909490A/en
Application granted granted Critical
Publication of CN106909490B publication Critical patent/CN106909490B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Images

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F11/00—Error detection; Error correction; Monitoring
    • G06F11/30—Monitoring
    • G06F11/3055—Monitoring arrangements for monitoring the status of the computing system or of the computing system component, e.g. monitoring if the computing system is on, off, available, not available
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F11/00—Error detection; Error correction; Monitoring
    • G06F11/30—Monitoring
    • G06F11/34—Recording or statistical evaluation of computer activity, e.g. of down time, of input/output operation ; Recording or statistical evaluation of user activity, e.g. usability assessment
    • G06F11/3409—Recording or statistical evaluation of computer activity, e.g. of down time, of input/output operation ; Recording or statistical evaluation of user activity, e.g. usability assessment for performance assessment
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F18/00—Pattern recognition
    • G06F18/20—Analysing
    • G06F18/23—Clustering techniques

Landscapes

  • Engineering & Computer Science (AREA)
  • Theoretical Computer Science (AREA)
  • General Engineering & Computer Science (AREA)
  • Physics & Mathematics (AREA)
  • Data Mining & Analysis (AREA)
  • General Physics & Mathematics (AREA)
  • Quality & Reliability (AREA)
  • Bioinformatics & Cheminformatics (AREA)
  • Evolutionary Computation (AREA)
  • Evolutionary Biology (AREA)
  • Computer Vision & Pattern Recognition (AREA)
  • Bioinformatics & Computational Biology (AREA)
  • Artificial Intelligence (AREA)
  • Computing Systems (AREA)
  • Life Sciences & Earth Sciences (AREA)
  • Computer Hardware Design (AREA)
  • Complex Calculations (AREA)

Abstract

The invention relates to a monitoring equipment data flow evaluation and noise elimination method. The method comprises the following steps: firstly, inputting data, and obtaining the relevance between parameters to be analyzed according to a relevance analysis method; adopting different countermeasures according to different data stream forms and types; for two sequences with strong correlation: if the two sequences generate mutation at the same time, the mutation amount is considered not to be error data according to the principle of the association rule, and the detection points are not processed; if one sequence generates mutation, the mutation is regarded as error data, and the detection point is corrected; for sequences that are not strongly related to other sequences: the detection point is corrected by considering the mutation amount or the separated data as error data. The invention realizes the data flow evaluation and noise elimination of the monitoring equipment, is convenient to master the state of the monitoring equipment and takes the corresponding measures in time.

Description

Monitoring equipment data flow evaluation and noise elimination method
Technical Field
The invention relates to a method for evaluating and eliminating noise of data stream of monitoring equipment
Background
With the continuous progress of information communication technology, the rapid development of technologies such as internet of things and cloud computing, the digitalization and informatization degree of production and life is higher and higher, the global data volume is about doubled every two years, not only is the data volume generated every day rapidly increased, but also the data structure becomes more and more complex, including various unstructured data, the traditional data processing mode is not free from worry, and the big data era comes, and becomes the focus of the discussion of the information technology industry. In this context, the processing of streaming data is becoming increasingly important. The data flow is a real-time sequence, wherein the data expresses the state of the current time object, the data has timeliness, and the data change has high speed and non-determinacy along with the continuous change of time, so that the limited data arriving in the data flow needs to be analyzed and processed in time, the future change and trend of the data flow are predicted, the state of the object is convenient to master, and measures are taken in time.
Disclosure of Invention
The invention aims to provide a monitoring equipment data flow evaluation and noise elimination method, which realizes the data flow evaluation and noise elimination of the monitoring equipment, is convenient to master the state of the monitoring equipment and takes corresponding measures in time.
In order to achieve the purpose, the technical scheme of the invention is as follows: a monitoring device data stream evaluation and noise cancellation method includes the steps of,
s1: inputting data, and obtaining the relevance between parameters to be analyzed according to a relevance analysis method; adopting different countermeasures according to different data stream forms and types;
s2: two sequences with strong correlation: if the two sequences generate mutation at the same time, the mutation amount is considered not to be error data according to the principle of the association rule, and the detection points are not processed; if one sequence generates mutation, the mutation is regarded as error data, and the detection point is corrected;
s3: sequences that are not strongly related to other sequences: the detection point is corrected by considering the mutation amount or the separated data as error data.
In an embodiment of the present invention, the modification method in step S2 is: when detecting that a detection point needs to be corrected, firstly, calculating the data value of the point to be zero, and then calculating the correction value of the point by using a backward interpolation method and a forward interpolation method, specifically: let X be the sequence to be cleaned, and X be a segment of data in the sequencei-5,xi-4,xi-3,xi-2,xi-1,xi,xi+1,xi+2,xi+3,xi+4,xi+5In which the detected point to be corrected is xi；
(1) Firstly, a correction value of the point is obtained by using a forward interpolation method:
polynomial of structure prefix
Figure BDA0001234131170000021
Wherein Ij＝{1,2,...,5}；
Constructing a pre-difference function
Figure BDA0001234131170000022
j＝1,2,...,5；
Thus obtaining forward correction values of the detection points as
Figure BDA0001234131170000023
(2) Then, a correction value of the point is obtained by using a backward interpolation method:
polynomial with postpositional structure
Figure BDA0001234131170000024
Wherein Ij＝{1,2,...,5}；
Constructing a post-difference function
Figure BDA0001234131170000025
j＝1,2,...,5；
Thus obtaining a backward correction value of the detection point of
Figure BDA0001234131170000026
(3) Obtaining a final correction value of the detection point as
Figure BDA0001234131170000027
In an embodiment of the present invention, the modification method in step S3 is: when detecting that a detection point needs to be corrected, firstly, calculating the data value of the point to be zero, and then calculating the correction value of the point by using a backward interpolation method and a forward interpolation method, specifically: let X be the sequence to be cleaned, and X be a segment of data in the sequencei-5,xi-4,xi-3,xi-2,xi-1,xi,xi+1,xi+2,xi+3,xi+4,xi+5In which the detected point to be corrected is xi；
(1) Firstly, a correction value of the point is obtained by using a forward interpolation method:
polynomial of structure prefix
Figure BDA0001234131170000028
Wherein Ij＝{1,2,...,5}；
Constructing a pre-difference function
Figure BDA0001234131170000029
j＝1,2,...,5；
Thus obtaining forward correction values of the detection points as
Figure BDA00012341311700000210
(2) Then, a correction value of the point is obtained by using a backward interpolation method:
polynomial with postpositional structure
Figure BDA00012341311700000211
Wherein Ij＝{1,2,...,5}；
Constructing a post-difference function
Figure BDA0001234131170000031
j＝1,2,...,5；
Thus obtaining a backward correction value of the detection point of
Figure BDA0001234131170000032
(3) Obtaining a final correction value of the detection point as
Figure BDA0001234131170000033
In an embodiment of the present invention, the association analysis method adopted in step S1 is a DBSCAN clustering algorithm.
In an embodiment of the present invention, in step S1, before the correlation analysis method is adopted, missing value detection is further performed on the input data, and if a missing value is detected, an average value is adopted instead.
Compared with the prior art, the invention has the following beneficial effects: the invention realizes the data flow evaluation and noise elimination of the monitoring equipment, is convenient to master the state of the monitoring equipment and takes the corresponding measures in time.
Drawings
FIG. 1 is a flow chart of the method of the present invention.
FIG. 2 is an example DBSCAN.
Fig. 3 is a diagram of the three-phase total active power and three-phase total current cleaning results (strong correlation) according to an embodiment of the present invention.
Fig. 4 is a diagram of the three-phase total active power and three-phase total current cleaning results (strong correlation and noise) according to an embodiment of the present invention.
Detailed Description
The technical scheme of the invention is specifically explained below with reference to the accompanying drawings.
As shown in fig. 1, a monitoring device data stream evaluation and noise cancellation method of the present invention includes the steps of,
s1: inputting data, and obtaining the relevance between parameters to be analyzed according to a relevance analysis method; adopting different countermeasures according to different data stream forms and types;
s2: two sequences with strong correlation: if the two sequences generate mutation at the same time, the mutation amount is considered not to be error data according to the principle of the association rule, and the detection points are not processed; if one sequence generates mutation, the mutation is regarded as error data, and the detection point is corrected;
s3: sequences that are not strongly related to other sequences: the detection point is corrected by considering the mutation amount or the separated data as error data.
The correction method in steps S2 and S3 is: when detecting that a detection point needs to be corrected, firstly, calculating the data value of the point to be zero, and then calculating the correction value of the point by using a backward interpolation method and a forward interpolation method, specifically: let X be the sequence to be cleaned, and X be a segment of data in the sequencei-5,xi-4,xi-3,xi-2,xi-1,xi,xi+1,xi+2,xi+3,xi+4,xi+5In which the detected point to be corrected is xi；
(1) Firstly, a correction value of the point is obtained by using a forward interpolation method:
polynomial of structure prefix
Figure BDA0001234131170000041
Wherein Ij＝{1,2,...,5}；
Constructing a pre-difference function
Figure BDA0001234131170000042
j＝1,2,...,5；
Thus obtaining forward correction values of the detection points as
Figure BDA0001234131170000043
(2) Then, a correction value of the point is obtained by using a backward interpolation method:
polynomial with postpositional structure
Figure BDA0001234131170000044
Wherein Ij＝{1,2,...,5}；
Constructing a post-difference function
Figure BDA0001234131170000045
j＝1,2,...,5；
Thus obtaining a backward correction value of the detection point of
Figure BDA0001234131170000046
(3) Obtaining a final correction value of the detection point as
Figure BDA0001234131170000047
The correlation analysis method adopted in step S1 is a DBSCAN clustering algorithm. In step S1, before the correlation analysis method is used, missing value detection is performed on the input data, and if a missing value is detected, an average value is used instead.
The following is a specific embodiment of the present invention.
The DBSCAN (Density-based Spatial Clustering of Application with Noise) algorithm belongs to a Spatial data Clustering method based on a Density mode, and is originally an algorithm proposed by Ester Martin et al. The algorithm can divide regions originally having a high density into different clusters, and can find arbitrary-shaped clustering patterns and similar clusters in spatial data having "noise". The most central idea of the DBSCAN algorithm is: for each analysis object in each cluster, the number of data objects in the neighborhood (neighbor) of a given radius (often in Eps) must be greater than the initially set given value. I.e. the neighborhood density must be larger than a certain threshold (often denoted MinPts). The DBSCAN algorithm can obtain the final result by searching for all data only once by searching for adjacent points, so the operation speed is very high. And the DBSCAN has a great advantage in that it can handle the clustering property of an arbitrary shape, is not interfered by noise, and can also remove the contained noise according to the threshold value MinPts.
The definition involved in the DBSCAN algorithm is given below:
define Eps-neighborhood of 1 point: the Eps neighborhood of any point p in the space is a set of points contained in a region with p as the center and Eps as the radius.
Definition 2 density: the density of any point p in space is the number of points included in a circular region having the point p as the center and the radius of Eps as the center.
Define 3 core points and boundary points: a point in space is said to be a core point if its density is greater than a given threshold value MinPts. Otherwise, the point is called a boundary point.
Definition 4 direct density is achievable: point p is directly density-reachable from point q if they satisfy the following two conditions: p is in the neighborhood; q is the core point.
Definition 5 the density can be reached: point p is reachable from the point q density if (p)1,p2,…,pnWherein p is1＝p，pnQ) and has piFrom pi+1The direct density can be reached.
Definition of 6 density connections: the point p and the point q are connected in density, and if for any o, both p and q are reachable from o density.
Defining 7 clusters: a non-empty set A of database D is a class if and only if A satisfies the following condition "
(1) For p and q, if p belongs to A and the density of p can reach q, then q belongs to A;
(2) for p and q, p and q are density-connected if p ∈ A and q ∈ A.
Define 8 noise: points in the database D that do not belong to any class are noise.
DBSCAN is different from other clustering algorithms, and the DBSCAN algorithm can find abnormal points which do not meet clustering. As shown in fig. 2, the cluster obtained by the clustering method using the radius neighborhood Eps as an input is shown in fig. 2. The DBSCAN algorithm also relies on another parameter, minimum number Minpts, and if the minimum number Minpts is set equal to 4, cluster 3 and cluster 5 may be considered isolated outliers.
The DBSCAN algorithm can be classified as one of the most representative algorithms in the clustering method. It defines the class to be analyzed as the largest aggregate of densely connected points, while having high noise immunity. The DBSCAN algorithm is often used in finding isolated points for data mining. But DBSCAN itself has certain limitations. Firstly, the DBSCAN depends on two input parameters, that is, the parameter radius neighborhood Eps and the minimum number Minpts need to be determined in advance, and for the selection of different parameters, the clustering results of final data are different. Secondly, in the DBSCAN algorithm, since the variable radius neighborhood Eps and the minimum number Minpts are set as global variables and have uniqueness, the clustering effect is general when the data distribution is not uniform. However, the DBSCAN algorithm is applied to the monitoring data of the power transmission and transformation equipment of the power system, and the characteristics of the monitored and collected data are considered, so the latter problem does not need to be considered.
According to the advantages of the DBSCAN algorithm, as shown in fig. 1, the monitoring device data stream evaluation and noise elimination method of the present invention includes the following steps,
s1: inputting data, and obtaining the relevance between parameters to be analyzed according to a relevance analysis method; adopting different countermeasures according to different data stream forms and types; the adopted correlation analysis method is a DBSCAN clustering algorithm, in addition, before the correlation analysis method is adopted, missing value detection needs to be carried out on input data, and if the missing value is detected, an average value is adopted for replacement.
S2: two sequences with strong correlation: if the two sequences generate mutation at the same time, the mutation amount is considered not to be error data according to the principle of the association rule, and the detection points are not processed; if one sequence generates mutation, the mutation is regarded as error data, and the detection point is corrected;
s3: sequences that are not strongly related to other sequences: the detection point is corrected by considering the mutation amount or the separated data as error data.
The correction method comprises the following steps: when detecting that a detection point needs to be corrected, firstly, calculating the data value of the point to be zero, and then calculating the correction value of the point by using a backward interpolation method and a forward interpolation method, specifically: let X be the sequence to be cleaned, and X be a segment of data in the sequencei-5,xi-4,xi-3,xi-2,xi-1,xi,xi+1,xi+2,xi+3,xi+4,xi+5In which the detected point to be corrected is xi；
(1) Firstly, a correction value of the point is obtained by using a forward interpolation method:
polynomial of structure prefix
Figure BDA0001234131170000061
Wherein Ij＝{1,2,...,5}；
Constructing a pre-difference function
Figure BDA0001234131170000062
j＝1,2,...,5；
Thus obtaining forward correction values of the detection points as
Figure BDA0001234131170000063
(2) Then, a correction value of the point is obtained by using a backward interpolation method:
polynomial with postpositional structure
Figure BDA0001234131170000064
Wherein Ij＝{1,2,...,5}；
Constructing a post-difference function
Figure BDA0001234131170000065
j＝1,2,...,5；
Thus obtaining a backward correction value of the detection point of
Figure BDA0001234131170000066
(3) Obtaining a final correction value of the detection point as
Figure BDA0001234131170000067
The following are specific examples of the present invention.
Taking the ambient temperature data of the power grid as an example. The total length of the original data is 480, the sequence number n in the FCM analysis is 10, and q is 48.
As shown in fig. 3, the first sub-graph is the input raw data, where the blue line represents the total active power of the three-phase circuit and the green line represents the total current of the three-phase circuit. The second sub-graph is the modified graph, where the black represents the total current of the three-phase circuit and the pink represents the total active power of the three-phase circuit. The blue circles mark isolated points of the total active power obtained by the DBSCAN algorithm, corresponding to times t-28 and t-84. The red circles mark isolated points of total current, corresponding to times t 28 and t 205. It can be seen that the positions where both occur at the first isolated point are the same. And because there is strong correlation between the total active power and the total current, the data at this time point are not processed. However, for an isolated point where the three-phase power occurs at the time t-84, the three-phase current is not abnormal, so that it can be considered that the point is an abnormal point, and the data at the point needs to be corrected and replaced by data cleaning. Similarly, the point at which the three-phase current occurs at time t 205 also needs to be cleaned and corrected. The two corrected images are shown in the second sub-diagram, and it can be seen that the first point is retained for further analysis and processing, and the rest of the points are already processed.
As shown in fig. 4, the first sub-graph is the input raw data, where the blue line represents the total active power of the three-phase circuit and the green line represents the total current of the three-phase circuit. The second sub-graph is the modified graph, where the black represents the total current of the three-phase circuit and the pink represents the total active power of the three-phase circuit. Wherein a noise is superimposed near 300 total active power t for the three phases. A large number of abnormal points exist in the part through the DBSCAN algorithm, and then the part is cleaned initially through the cleaning algorithm, so that the influence of noise is reduced. But there is still some noise interference that requires further processing and analysis.
The above are preferred embodiments of the present invention, and all changes made according to the technical scheme of the present invention that produce functional effects do not exceed the scope of the technical scheme of the present invention belong to the protection scope of the present invention.

Claims (3)

1. A method for monitoring device data stream assessment and noise cancellation, characterized by: comprises the following steps of (a) carrying out,
s1: inputting data, and obtaining the relevance between parameters to be analyzed according to a relevance analysis method; adopting different countermeasures according to different data stream forms and types;
s2: two sequences with strong correlation: if the two sequences generate mutation at the same time, the mutation amount is considered not to be error data according to the principle of the association rule, and the detection points are not processed; if one sequence generates mutation, the mutation is regarded as error data, and the detection point is corrected;
s3: sequences that are not strongly related to other sequences: regarding the mutation amount or the separated data as error data, and correcting the detection point;
the correction method in step S2 is: when detecting that a detection point needs to be corrected, firstly, calculating the data value of the point to be zero, and then calculating the correction value of the point by using a backward interpolation method and a forward interpolation method, specifically: let X be the sequence to be cleaned, and X be a segment of data in the sequencei-5,xi-4,xi-3,xi-2,xi-1,xi,xi+1,xi+2,xi+3,xi+4,xi+5In which the detected point to be corrected is xi；
(1) Firstly, a correction value of the point is obtained by using a forward interpolation method:
polynomial of structure prefix
Figure FDA0002344643060000011
Wherein Ij＝{1,2,...,5}；
Constructing a pre-difference function
Figure FDA0002344643060000012
Thus obtaining forward correction values of the detection points as
Figure FDA0002344643060000013
(2) Then, a correction value of the point is obtained by using a backward interpolation method:
polynomial with postpositional structure
Figure FDA0002344643060000014
Wherein Ij＝{1,2,...,5}；
Constructing a post-difference function
Figure FDA0002344643060000015
Thus obtaining a backward correction value of the detection point of
Figure FDA0002344643060000016
(3) Obtaining a final correction value of the detection point as
Figure FDA0002344643060000017
The correction method in step S3 is: when detecting that the detection point needs to be corrected, firstly, the data value of the point is counted as zeroThen, calculating a correction value of the point by using a backward interpolation method and a forward interpolation method, specifically: let X be the sequence to be cleaned, and X be a segment of data in the sequencei-5,xi-4,xi-3,xi-2,xi-1,xi,xi+1,xi+2,xi+3,xi+4,xi+5In which the detected point to be corrected is xi；
(1) Firstly, a correction value of the point is obtained by using a forward interpolation method:
polynomial of structure prefix
Figure FDA0002344643060000021
Wherein Ij＝{1,2,...,5}；
Constructing a pre-difference function
Figure FDA0002344643060000022
Thus obtaining forward correction values of the detection points as
Figure FDA0002344643060000023
(2) Then, a correction value of the point is obtained by using a backward interpolation method:
polynomial with postpositional structure
Figure FDA0002344643060000024
Wherein Ij＝{1,2,...,5}；
Constructing a post-difference function
Figure FDA0002344643060000025
Thus obtaining a backward correction value of the detection point of
Figure FDA0002344643060000026
(3) Obtaining a final correction value of the detection point as
Figure FDA0002344643060000027
2. The method of claim 1, wherein the method comprises: the correlation analysis method adopted in step S1 is a DBSCAN clustering algorithm.
3. The method of claim 1, wherein the method comprises: in step S1, before the correlation analysis method is used, missing value detection is performed on the input data, and if a missing value is detected, an average value is used instead.
CN201710111079.9A 2017-02-28 2017-02-28 Monitoring equipment data flow evaluation and noise elimination method Active CN106909490B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN201710111079.9A CN106909490B (en) 2017-02-28 2017-02-28 Monitoring equipment data flow evaluation and noise elimination method

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN201710111079.9A CN106909490B (en) 2017-02-28 2017-02-28 Monitoring equipment data flow evaluation and noise elimination method

Publications (2)

Publication Number Publication Date
CN106909490A CN106909490A (en) 2017-06-30
CN106909490B true CN106909490B (en) 2020-05-05

Family

ID=59209189

Family Applications (1)

Application Number Title Priority Date Filing Date
CN201710111079.9A Active CN106909490B (en) 2017-02-28 2017-02-28 Monitoring equipment data flow evaluation and noise elimination method

Country Status (1)

Country Link
CN (1) CN106909490B (en)

Families Citing this family (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN110083593B (en) * 2019-04-23 2023-02-10 中国大唐集团科学技术研究院有限公司华东电力试验研究院 Power station operation parameter cleaning and repairing method and repairing system
CN110188810A (en) * 2019-05-23 2019-08-30 西北工业大学 A kind of detection of online outlier, identification and modification method based on Density Clustering
CN112699113B (en) * 2021-01-12 2022-08-05 上海交通大学 Industrial manufacturing process operation monitoring system driven by time sequence data stream

Citations (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN105488736A (en) * 2015-12-02 2016-04-13 国家电网公司 Data processing method for photovoltaic power station data acquisition system
CN106295850A (en) * 2016-07-15 2017-01-04 广东电网有限责任公司电力科学研究院 A kind of bad data recognition method and system based on historical metrology

Patent Citations (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN105488736A (en) * 2015-12-02 2016-04-13 国家电网公司 Data processing method for photovoltaic power station data acquisition system
CN106295850A (en) * 2016-07-15 2017-01-04 广东电网有限责任公司电力科学研究院 A kind of bad data recognition method and system based on historical metrology

Non-Patent Citations (3)

* Cited by examiner, † Cited by third party
Title
多变量时间序列滑动窗口异常点的检测;戴慧等;《南京信息工程大学学报：自然科学版》;20141231;第06卷(第06期);第515-519页 *
戴慧等.多变量时间序列滑动窗口异常点的检测.《南京信息工程大学学报：自然科学版》.2014,第06卷(第06期),第515-519页. *
数据挖掘方法在短期负荷预测中的应用研究;王志勇;《中国博士学位论文全文数据库工程科技II辑》;20080415(第04期);摘要，第二章 *

Also Published As

Publication number Publication date
CN106909490A (en) 2017-06-30

Similar Documents

Publication Publication Date Title
CN110839016B (en) Abnormal flow monitoring method, device, equipment and storage medium
Cheng et al. MS-LSTM: A multi-scale LSTM model for BGP anomaly detection
CN109086720B (en) Face clustering method, face clustering device and storage medium
CN106909490B (en) Monitoring equipment data flow evaluation and noise elimination method
Wang et al. An efficient switching median filter based on local outlier factor
CN110995153B (en) Abnormal data detection method and device for photovoltaic power station and electronic equipment
CN111046968B (en) Road network track clustering analysis method based on improved DPC algorithm
CN111179276B (en) Image processing method and device
CN115905990A (en) Transformer oil temperature abnormity monitoring method based on density aggregation algorithm
CN113660237B (en) Industrial Internet data flow abnormity detection method based on dynamic sliding window, memory and processor
CN105049286A (en) Cloud platform speed measurement data judging method based on hierarchical clustering
Nair et al. An efficient adaptive weighted switching median filter for removing high density impulse noise
Hou et al. Denoising point cloud
CN105282720B (en) A kind of method for filtering spam short messages and device
Janos et al. On finding all reducts of consistent decision tables
Qing et al. Device type identification via network traffic and lightweight convolutional neural network for Internet of Things
CN115175192A (en) Vehicle networking intrusion detection method based on graph neural network
CN106814608B (en) Predictive control adaptive filtering algorithm based on posterior probability distribution
Zhou et al. Network-wide anomaly detection based on router connection relationships
CN113255541A (en) Intrinsic mode function recombination signal relative entropy-based process parameter denoising method for adaptive process industrial process
Beliakov et al. Non-monotone averaging aggregation
Zola et al. Temporal graph-based approach for behavioural entity classification
Lee et al. Estimating the degrees of neighboring nodes in online social networks
US9729412B2 (en) Periodicity detection method, periodicity detection device, and periodicity detection program
CN105471639B (en) Network flow entropy evaluation method based on median and device

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