CN109299156A - Electronic device, the electric power data predicting abnormality method based on XGBoost and storage medium - Google Patents
Electronic device, the electric power data predicting abnormality method based on XGBoost and storage medium Download PDFInfo
- Publication number
- CN109299156A CN109299156A CN201810954347.8A CN201810954347A CN109299156A CN 109299156 A CN109299156 A CN 109299156A CN 201810954347 A CN201810954347 A CN 201810954347A CN 109299156 A CN109299156 A CN 109299156A
- Authority
- CN
- China
- Prior art keywords
- data
- electric power
- power data
- abnormal
- model
- 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.)
- Withdrawn
Links
- 238000000034 method Methods 0.000 title claims abstract description 75
- 230000005856 abnormality Effects 0.000 title claims abstract description 53
- 230000002159 abnormal effect Effects 0.000 claims abstract description 157
- 238000012549 training Methods 0.000 claims abstract description 83
- 238000013480 data collection Methods 0.000 claims abstract description 40
- 238000010276 construction Methods 0.000 claims abstract description 13
- 238000012360 testing method Methods 0.000 claims description 56
- 230000008569 process Effects 0.000 claims description 26
- 238000012986 modification Methods 0.000 claims description 20
- 230000004048 modification Effects 0.000 claims description 20
- 238000012545 processing Methods 0.000 claims description 5
- 239000012141 concentrate Substances 0.000 claims description 3
- 238000012423 maintenance Methods 0.000 claims 2
- 230000005611 electricity Effects 0.000 description 23
- PEDCQBHIVMGVHV-UHFFFAOYSA-N Glycerine Chemical compound OCC(O)CO PEDCQBHIVMGVHV-UHFFFAOYSA-N 0.000 description 12
- 230000002354 daily effect Effects 0.000 description 11
- 230000003203 everyday effect Effects 0.000 description 10
- 230000006870 function Effects 0.000 description 9
- 238000004891 communication Methods 0.000 description 6
- 230000008901 benefit Effects 0.000 description 4
- 238000010586 diagram Methods 0.000 description 4
- 230000003442 weekly effect Effects 0.000 description 4
- 230000005540 biological transmission Effects 0.000 description 2
- 238000013500 data storage Methods 0.000 description 2
- 238000003066 decision tree Methods 0.000 description 2
- 230000001419 dependent effect Effects 0.000 description 2
- 230000001771 impaired effect Effects 0.000 description 2
- 230000036651 mood Effects 0.000 description 2
- 230000005236 sound signal Effects 0.000 description 2
- 241000208340 Araliaceae Species 0.000 description 1
- 235000005035 Panax pseudoginseng ssp. pseudoginseng Nutrition 0.000 description 1
- 235000003140 Panax quinquefolius Nutrition 0.000 description 1
- 230000004888 barrier function Effects 0.000 description 1
- 238000013523 data management Methods 0.000 description 1
- 235000008434 ginseng Nutrition 0.000 description 1
- 239000004973 liquid crystal related substance Substances 0.000 description 1
- 230000001737 promoting effect Effects 0.000 description 1
- 230000009466 transformation Effects 0.000 description 1
- 230000000007 visual effect Effects 0.000 description 1
Classifications
-
- 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
- G06Q10/00—Administration; Management
- G06Q10/06—Resources, workflows, human or project management; Enterprise or organisation planning; Enterprise or organisation modelling
- G06Q10/063—Operations research, analysis or management
- G06Q10/0635—Risk analysis of enterprise or organisation activities
-
- 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
-
- 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
- Business, Economics & Management (AREA)
- Human Resources & Organizations (AREA)
- Engineering & Computer Science (AREA)
- Economics (AREA)
- Strategic Management (AREA)
- Entrepreneurship & Innovation (AREA)
- Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Health & Medical Sciences (AREA)
- Marketing (AREA)
- General Physics & Mathematics (AREA)
- General Business, Economics & Management (AREA)
- Tourism & Hospitality (AREA)
- Educational Administration (AREA)
- Quality & Reliability (AREA)
- Operations Research (AREA)
- Game Theory and Decision Science (AREA)
- Development Economics (AREA)
- Public Health (AREA)
- Water Supply & Treatment (AREA)
- General Health & Medical Sciences (AREA)
- Primary Health Care (AREA)
- Supply And Distribution Of Alternating Current (AREA)
Abstract
The present invention proposes electronic device, electric power data predicting abnormality method and storage medium based on XGBoost, which comprises obtains electric power data of the predetermined building in the first time predefined section；The electric power data that method analysis obtains is determined using predetermined exceptional value, to obtain the first electric power data collection without exception and the first abnormal electric power data collection；The first abnormal power data set is substituted into the abnormal value prediction model that training is completed in advance and carries out exceptional value prediction, to predict the abnormal power data in the first abnormal power data set.It can be improved the accuracy that abnormal data is predicted in electric power data, provide reference data for the construction and popularization of smart grid.
Description
Technical field
The present invention relates to electric power data process field more particularly to a kind of electronic device, based on the electric power data of XGBoost
Predicting abnormality method and storage medium.
Background technique
Electric power data (the electricity consumption historical record for referring to all kinds of buildings collected in electric system) be smart grid construction and
Key reference factor in popularization and application, and electric power data mainly carries out daily meter reading by the meter reading personnel of power supply company at present
And then by professional's logging data management system, on the one hand due in entire Data Input Process, by means of a large amount of
Manpower complete, electric power data exception is easy to appear, on the other hand, in Data Input Process, it is possible to there is system therefore
The logging data as caused by equipment faults itself reason such as barrier, line interruption is abnormal；And electric power data is abnormal in smart grid
It will cause subjectively unimaginable accident in construction and application, for example, causing to miss certain building electricity consumption due to data exception
It is disconnected, and so that the building circut breaking is occurred or the problems such as circuit burnout occur, in some instances it may even be possible to it will lead to the generation of accident.Cause
This, how to predict the abnormal data in electric power data like clockwork is current urgent problem to be solved.
Summary of the invention
In view of this, the present invention proposes a kind of electronic device, the electric power data predicting abnormality method based on XGBoost and deposits
Storage media can be improved the accuracy that abnormal data is predicted in electric power data, provide ginseng for the construction and popularization of smart grid
Examine data.
Firstly, to achieve the above object, the present invention proposes a kind of electronic device, the electronic device include memory and
The processor connecting with the memory, the processor is for executing the electricity based on XGBoost stored on the memory
Force data predicting abnormality program, it is real when the electric power data predicting abnormality program based on XGBoost is executed by the processor
Existing following steps:
A1, electric power data of the predetermined building in the first time predefined section is obtained；
A2, the electric power data that method analysis obtains is determined using predetermined exceptional value, it is without exception to obtain first
Electric power data collection and the first abnormal electric power data collection；
A3, the first abnormal power data set is substituted into the abnormal value prediction model progress exceptional value that training is completed in advance
Prediction, to predict the abnormal power data in the first abnormal power data set.
Preferably, in the step A2, predetermined exceptional value determines that method is that box figure identifies exceptional value method,
The step A2 includes the following steps:
Determine the unit time in the described first predefined period；
The electric power data in determining each unit time is traversed respectively, is inquired in the electric power data in each unit time
With the presence or absence of the data for being greater than the first predefined thresholds and with the presence or absence of the data less than the second predefined thresholds；
If having the data existed in the electric power data in the unit time greater than first predefined thresholds, or there are small
In the data of second predefined thresholds, it is determined that there are abnormal power data in the electric power data in the unit time；
If having in the electric power data in the unit time, there is no the data greater than first predefined thresholds, and are not present
Less than the data of second predefined thresholds, it is determined that abnormal power number is not present in the electric power data in the unit time
According to；
The electric power data there are in the unit time of abnormal power data constitutes the described first abnormal electric power data
Collection, the electric power data there is no in the unit time of abnormal power data constitute the first electric power data collection without exception.
Preferably, first predefined thresholds are as follows:
QU+1.5IQRQU+1.5QR
Second predefined thresholds are as follows:
QL-1.5IQRQL-1.5IQR
Wherein, QU is the upper quartile that statistics obtains, and indicates have 1/4 data bigger than him in total data, QL is system
Obtained lower quartile is counted, indicates there are 1/4 data smaller than him in total data, it is QU and QL that IQR, which is quartile spacing,
Difference.
Preferably, in the step A3, the abnormal value prediction model that the preparatory training is completed is XGBoost model,
The exceptional value and model include the training process of model and the test process of model, and the training process of the model includes:
E1, electric power data sample set of the predetermined target construction in the second time predefined section is obtained,
The abnormality of the electric power data obtained is analyzed, to obtain the second data set without exception and the second abnormal data set；
F1, the training sample set and test sample that the data in the described second data set without exception are divided into preset ratio
Collection modifies in the training sample set partial data at random to obtain abnormality test sample set；
G1, label is added to each sample data that the training sample is concentrated, the label is preset each sample
Whether data are exceptional value；
H1, the electric power data that the training sample with the label is concentrated using the XGboost that pre-establishes touch type into
Row supervised learning, to obtain abnormal value prediction model；
J1, model accuracy test is carried out to the obtained abnormal value prediction model, if test passes through, model training
Terminate, if test does not pass through, increase the sample data that the training sample is concentrated, and repeat above-mentioned steps E1, F1,
G1、H1。
Preferably, the test process of the model includes:
Exceptional value mark is carried out to the data in the abnormality test sample set according to obtained abnormal value prediction model, with
It obtains with the electric power data collection marked extremely；
The data concentrated with the training sample after the electric power data collection that marks extremely and random modification that will be obtained into
Row compares, if obtained after the abnormal data with the electric power data marked extremely concentration mark and the random modification
Electric power data is compared, and the abnormal data of mark is that the probability value of the electric power data of modification is more than or equal to preset probability threshold
Value, it is determined that the accuracy test of model is not passed through；
Alternatively, if after abnormal data and the random modification for concentrating mark with the electric power data marked extremely
Obtained electric power data is compared, and the abnormal data of mark is that the probability value of the electric power data of modification is less than preset probability threshold value,
Then pass through for the accuracy test of model.
In addition, in order to solve the above technical problems, the present invention also proposes a kind of electric power data predicting abnormality based on XGBoost
Method, described method includes following steps:
S1, electric power data of the predetermined building in the first time predefined section is obtained；
S2, the electric power data that method analysis obtains is determined using predetermined exceptional value, it is without exception to obtain first
Electric power data collection and the first abnormal electric power data collection；
S3, the first abnormal power data set is substituted into the abnormal value prediction model progress exceptional value that training is completed in advance
Prediction, to predict the abnormal power data in the first abnormal power data set.
Preferably, in the step S2, predetermined exceptional value determines that method is that box figure identifies exceptional value method,
The step S2 includes the following steps:
Determine the unit time in the described first predefined period；
The electric power data in determining each unit time is traversed respectively, is inquired in the electric power data in each unit time
With the presence or absence of the data for being greater than the first predefined thresholds and with the presence or absence of the data less than the second predefined thresholds；
If having the data existed in the electric power data in the unit time greater than first predefined thresholds, or there are small
In the data of second predefined thresholds, it is determined that there are abnormal power data in the electric power data in the unit time；
If having in the electric power data in the unit time, there is no the data greater than first predefined thresholds, and are not present
Less than the data of second predefined thresholds, it is determined that abnormal power number is not present in the electric power data in the unit time
According to；
The electric power data there are in the unit time of abnormal power data constitutes the described first abnormal electric power data
Collection, the electric power data there is no in the unit time of abnormal power data constitute the first electric power data collection without exception.
Preferably, first predefined thresholds are as follows:
QU+1.5IQRQU+1.5QR
Second predefined thresholds are as follows:
QL-1.5IQRQL-1.5IQR
Wherein, QU is the upper quartile that statistics obtains, and indicates have 1/4 data bigger than him in total data, QL is system
Obtained lower quartile is counted, indicates there are 1/4 data smaller than him in total data, it is QU and QL that IQR, which is quartile spacing,
Difference.
Preferably, in the step S3, the abnormal value prediction model that the preparatory training is completed is XGBoost model,
The exceptional value and model include the training process of model and the test process of model, and the training process of the model includes:
E2, electric power data sample set of the predetermined target construction in the second time predefined section is obtained,
The abnormality of the electric power data obtained is analyzed, to obtain the second data set without exception and the second abnormal data set；
F2, the training sample set and test sample that the data in the described second data set without exception are divided into preset ratio
Collection modifies in the training sample set partial data at random to obtain abnormality test sample set；
G2, label is added to each sample data that the training sample is concentrated, the label is preset each sample
Whether data are exceptional value；
H2, the electric power data that the training sample with the label is concentrated using the XGboost that pre-establishes touch type into
Row supervised learning, to obtain abnormal value prediction model；
J2, model accuracy test is carried out to the obtained abnormal value prediction model, if test passes through, model training
Terminate, if test does not pass through, increase the sample data that the training sample is concentrated, and repeat above-mentioned steps E2, F2,
G2、H2。
In addition, in order to solve the above-mentioned technical problem, the present invention also proposes a kind of computer readable storage medium, the calculating
Machine readable storage medium storing program for executing is stored with the electric power data predicting abnormality program based on XGBoost, the electric power number based on XGBoost
It can be executed by least one processor according to predicting abnormality program, so that the execution of at least one described processor is based on as described above
The step of electric power data predicting abnormality method of XGBoost.
Electronic device proposed by the invention, the electric power data predicting abnormality method based on XGBoost and storage medium, it is first
First obtain electric power data of the predetermined building in the first time predefined section；Then predetermined exceptional value is utilized
The electric power data that method analysis obtains is determined, to obtain the first electric power data collection without exception and the first abnormal electric power data
Collection；It is pre- that the first abnormal power data set is finally substituted into the abnormal value prediction model progress exceptional value that training is completed in advance
It surveys, to predict the abnormal power data in the first abnormal power data set.It can be improved abnormal data in electric power data
The accuracy of prediction provides reference data for the construction and popularization of smart grid.
Detailed description of the invention
Fig. 1 is the schematic diagram of the optional hardware structure of electronic device one proposed by the present invention；
Fig. 2 is the program of the electric power data predicting abnormality program in one embodiment of electronic device of the present invention based on XGBoost
Module diagram；
Fig. 3 is the implementation flow chart of the electric power data predicting abnormality method preferred embodiment the present invention is based on XGBoost.
The embodiments will be further described with reference to the accompanying drawings for the realization, the function and the advantages of the object of the present invention.
Specific embodiment
In order to make the objectives, technical solutions, and advantages of the present invention clearer, with reference to the accompanying drawings and embodiments, right
The present invention is further elaborated.It should be appreciated that described herein, specific examples are only used to explain the present invention, not
For limiting the present invention.Based on the embodiments of the present invention, those of ordinary skill in the art are not before making creative work
Every other embodiment obtained is put, shall fall within the protection scope of the present invention.
It should be noted that the description for being related to " first ", " second " etc. in the present invention is used for description purposes only, and cannot
It is interpreted as its relative importance of indication or suggestion or implicitly indicates the quantity of indicated technical characteristic.Define as a result, " the
One ", the feature of " second " can explicitly or implicitly include at least one of the features.In addition, the skill between each embodiment
Art scheme can be combined with each other, but must be based on can be realized by those of ordinary skill in the art, when technical solution
Will be understood that the combination of this technical solution is not present in conjunction with there is conflicting or cannot achieve when, also not the present invention claims
Protection scope within.
As shown in fig.1, being the optional hardware structure schematic diagram of electronic device one proposed by the present invention.In the present embodiment,
Electronic device 10 may include, but be not limited only to, and connection memory 11, processor 12, net can be in communication with each other by communication bus 14
Network interface 13.It should be pointed out that Fig. 1 illustrates only the electronic device 10 with component 11-14, it should be understood that simultaneously
All components shown realistic are not applied, the implementation that can be substituted is more or less component.
Wherein, memory 11 includes at least a type of computer readable storage medium, computer readable storage medium
Including flash memory, hard disk, multimedia card, card-type memory (for example, SD or DX memory etc.), random access storage device (RAM), quiet
State random access storage device (SRAM), electrically erasable programmable read-only memory (EEPROM), can be compiled read-only memory (ROM)
Journey read-only memory (PROM), magnetic storage, disk, CD etc..In some embodiments, memory 11 can be electronics dress
Set 10 internal storage unit, such as the hard disk or memory of electronic device 10.In further embodiments, memory 11 can also be with
It is the outer packet storage device of electronic device 10, such as the plug-in type hard disk being equipped on electronic device 10, intelligent memory card (Smart
Media Card, SMC), secure digital (Secure Digital, SD) card, flash card (Flash Card) etc..Certainly, it stores
Device 11 can also both including electronic device 10 internal storage unit and also including its outer packet storage device.In the present embodiment, storage
Device 11 is installed on the operating system and types of applications software of electronic device 10, such as the electricity based on XGBoost commonly used in storage
Force data predicting abnormality program etc..It has exported or will export in addition, memory 11 can be also used for temporarily storing
Various types of data.
Processor 12 can be in some embodiments central processing unit (Central Processing Unit, CPU),
Controller, microcontroller, microprocessor or other data processing chips.Processor 12 is commonly used in control electronic device 10
Overall operation.In the present embodiment, program code or processing data of the processor 12 for being stored in run memory 11, such as
The electric power data predicting abnormality program based on XGBoost etc. of operation.
Network interface 13 may include radio network interface or wired network interface, and network interface 13 is commonly used in filling in electronics
It sets and establishes communication connection between 10 and other electronic equipments.
Communication bus 14 is for realizing the communication connection between component 11-13.
Fig. 1 illustrates only the electronics dress of the electric power data predicting abnormality program with component 11-14 and based on XGBoost
10 are set, it should be understood that it is not required for implementing all components shown, more or less groups of the implementation that can be substituted
Part.
Optionally, electronic device 10 can also include user interface (not shown in figure 1), and user interface may include display
Device, input unit such as keyboard, wherein user interface can also be including standard wireline interface and wireless interface etc..
Optionally, in some embodiments, display can be light-emitting diode display, liquid crystal display, touch control type LCD and show
Device and OLED touch device etc..Further, display is alternatively referred to as display screen or display unit, for being shown in electronic device
Information is handled in 10 and for showing visual user interface.
Optionally, in some embodiments, electronic device 10 can also include that audio unit (does not show in audio unit Fig. 1
Out), audio unit can be in call signal reception pattern, call mode, logging mode, speech recognition mould in electronic device 10
When under the isotypes such as formula, broadcast reception mode, received or storage audio data is converted into audio signal；Further
Ground, electronic device 10 can also include audio output unit, and the audio signal that audio output unit converts audio unit exports,
And audio output unit can also provide the relevant audio output of specific function that executes to electronic device 10 (such as calling is believed
Number receive sound, message sink sound etc.), audio output unit may include loudspeaker, buzzer etc..
Optionally, in some embodiments, electronic device 10 can also include alarm unit (not shown), alarm list
Member can provide output and the generation of event is notified electron device 10.Typical event may include calling reception, message
Reception, key signals input, touch input etc..Other than audio or video export, alarm unit can be with different sides
Formula provides output with the generation of notification event.For example, alarm unit can provide output in the form of vibration, exhaled when receiving
Cry, message or it is some other can make electronic device 10 enter communication pattern when, alarm unit can provide tactile output (that is,
Vibration) to notify to user.
In one embodiment, the electric power data predicting abnormality program based on XGBoost stored in memory 11 is processed
When device 12 executes, following operation is realized:
A1, electric power data of the predetermined building in the first time predefined section is obtained；
It, can also be with it is to be appreciated that the predetermined building can be inpatient building in hospital, patient terminal
It is the building in the places such as office block, family dependents' building, school instruction building, market, the first predefined period can be with
The moon is unit, can also be unit in week, can also be as unit of day；If the electric power data obtained is described as unit of the moon
First time predefined section, such as the Power system load data of the electricity consumption monthly every day in half a year；If being obtained as unit of week
Electric power data be the first time predefined section, such as nearest two months in electricity consumption every day weekly electric load number
According to；If the electric power data obtained is each daily in the first time predefined section, such as nearest 10 days as unit of day
The Power system load data of hour electricity consumption.It is understood that the first predefined period can also as unit of hour,
And the minimum unit of the electric power data obtained can be specific to need to use according to the predetermined building is practical to second grade
The variation of electricity and determine, be not specifically limited in the present embodiment.
A2, the electric power data that method analysis obtains is determined using predetermined exceptional value, it is without exception to obtain first
Electric power data collection and the first abnormal electric power data collection；
It is to be appreciated that since the electric power data got is manually to carry out meter reading, and typing by relevant staff
To the electric power data of electric power data storage system, therefore, the electric power data is typically due to human factor, such as staff's mood
Electric power data caused by bad exist it is abnormal, personnel relieve a sentry join it is indefinite caused by meter reading data it is abnormal etc., or can also be with
The reason of by equipment itself, such as line interruption, data transmission link is impaired etc., and to cause electric power data to exist abnormal；Further,
In the present embodiment, the predetermined exceptional value determine method be box figure identify exceptional value method, specifically include as
Lower step: the unit time in the described first predefined period is determined；It is traversed in determining each unit time respectively
Electric power data with the presence or absence of the data greater than the first predefined thresholds and is in the electric power data inquired in each unit time
The no data existed less than the second predefined thresholds；Make a reservation for if having and existing in the electric power data in the unit time greater than described first
The data of adopted threshold value, or there are the data for being less than second predefined thresholds, it is determined that the electric power number in the unit time
There are abnormal power data in, and the electric power data in the unit time is the sample that the described first abnormal electric power data is concentrated
Collection；If having in the electric power data in the unit time, there is no the data greater than first predefined thresholds, and there is no be less than
The data of second predefined thresholds, it is determined that abnormal power data are not present in the electric power data in the unit time, it should
Electric power data in unit time is the sample set that the described first electric power data without exception is concentrated；Wherein, described first is predetermined
Adopted threshold value is QU+ 1.5IQRQU+1.5QR, second predefined thresholds are QL-1.5IQRQL-1.5IQR, wherein QU is system
Obtained upper quartile is counted, indicates there are 1/4 data bigger than him in total data, QL is the lower quartile that statistics obtains,
Indicate there are 1/4 data smaller than him in total data, IQR is quartile spacing, is the difference of QU and QL.
Specifically, in this example, it is assumed that the first predefined period is the then institute with the period of moon unit
Every day that the predefined unit time is the middle of each month in the predefined first time period is stated, if having one day use in certain middle of the month
There are exceptional values for the Power system load data of electricity, then the power load charge values of the daily electricity consumption of this month are the first abnormal power data
The sample data of concentration, similarly, if exceptional value is not present in the Power system load data in the daily electricity consumption in certain middle of the month, the middle of the month
The power load charge values of daily electricity consumption be sample data that the described first electric power data without exception is concentrated.It is to be appreciated that false
If the first predefined period with unit around, then when the predefined unit time is described predefined first
In weekly in every day, if the first predefined period, as unit of day, institute's predefined unit time is
In the predefined first time daily in each hour.
A3, the first abnormal power data set is substituted into the abnormal value prediction model progress exceptional value that training is completed in advance
Prediction, to predict the abnormal power data in the first abnormal power data set.
Specifically, the abnormal value prediction model that the preparatory training is completed is XGBoost model, the exceptional value and model
The test process of training process and model including model, the training process of the model include:
E1, electric power data sample set of the predetermined target construction in the second time predefined section is obtained,
The abnormality of the electric power data obtained is analyzed, to obtain the second data set without exception and the second abnormal data set；
It is to be appreciated that the first time predefined section can be identical with the second time predefined section, it is preferable that
The second time predefined section is greater than the first time predefined section, and the first time predefined section is pre- with described second
The chronomere for defining the period must be identical, if the first time predefined Duan Yiyue is unit, then described second is pre-
Defining the period must be also as unit of the moon.It should be noted that during the abnormality for the electric power data that analysis obtains,
Need to consider the ratio for the electric power data exceptional value for including in the predefined period obtained, if such as described first predefined
Period and the second time predefined Duan Jun then need to consider as unit of the moon predefined described second as unit of the moon
In period, the exceptional value ratio of electric power data in the Power system load data of electricity consumption every day monthly, if pre- described second
There is the exceptional value ratio of certain month electric power data more than preset exceptional value proportion threshold value, such as 10% in the definition period, then
It needs to carry out error condition analysis to the electric power data of this month based on experience value, alternatively, need to abandon the electric power data of this month, and
The electric power data in more times is obtained as sample set.
F1, the training sample set and test sample that the data in the described second data set without exception are divided into preset ratio
Collection modifies in the training sample set partial data at random to obtain abnormality test sample set；
Specifically, the ratio of the training sample set is typically larger than the ratio of the test sample collection, such as preferably, institute
The ratio for stating training sample set and the test sample collection is 7:3.
G1, label is added to each sample data that the training sample is concentrated, the label is preset each sample
Whether data are exceptional value；
Specifically, in this example, it is assumed that being exceptional value, then preset label is 1, it is assumed that is not exceptional value, then in advance
If label be 0.
H1, the electric power data that the training sample with the label is concentrated using the XGboost that pre-establishes touch type into
Row supervised learning, to obtain abnormal value prediction model；
Specifically, it is a kind of boost model based on decision tree structure that the XGboost, which touches type, similar with GBDT, but by
Type is touched in XGboost to have carried out second level Taylor expansion to objective function and joined regularization term so that model have it is stronger
Generalization ability can evade the risk of fitting, strong antijamming capability.Specifically, to the training sample set benefit with the label
It being exercised supervision study, is inputted as sample and label with XGboost, the sample is the electric power data that the training sample is concentrated,
The label is whether the electric power data that the training sample is concentrated is exceptional value, exports the training sample concentration for prediction
Each electric power data be exceptional value probability.
F1, model accuracy test is carried out to the obtained abnormal value prediction model, if test passes through, model training
Terminate, if test does not pass through, increase the sample data that the training sample is concentrated, and repeat above-mentioned steps E1, F1,
G1、H1。
Specifically, the test process of the model includes: according to obtained abnormal value prediction model to the abnormality test
Data in sample set carry out exceptional value mark, to obtain with the electric power data collection marked extremely；It will obtain with abnormal
The electric power data collection of mark is compared with the data that the training sample after random modification is concentrated, if described with abnormal mark
Electric power data concentrate the abnormal data of mark to obtain after the random modification electric power data compared with, the abnormal number of mark
It is more than or equal to preset probability threshold value according to the probability value of the electric power data for modification, it is determined that test the accuracy of model
Do not pass through；Alternatively, if after abnormal data and the random modification for concentrating mark with the electric power data marked extremely
Obtained electric power data is compared, and the abnormal data of mark is that the probability value of the electric power data of modification is less than preset probability threshold value,
Then pass through for the accuracy test of model.
By above-mentioned thing embodiment it is found that electronic device proposed by the present invention, obtains predetermined building first
Electric power data in one time predefined section；Then the electric power number of method analysis acquisition is determined using predetermined exceptional value
According to obtain the first electric power data collection without exception and the first abnormal electric power data collection；Finally by first abnormal power
Data set substitutes into the abnormal value prediction model that training is completed in advance and carries out exceptional value prediction, to predict first abnormal power
Abnormal power data in data set.It can be improved the accuracy that abnormal data is predicted in electric power data, for building for smart grid
And if promoting and providing reference data.
In addition, the function that the electric power data predicting abnormality program of the invention based on XGBoost is realized according to its each section
Can be different, it can be described with program module with the same function.It please refers to shown in Fig. 2, is that electronic device one of the present invention is real
Apply the program module schematic diagram of the electric power data predicting abnormality program in example based on XGBoost.In the present embodiment, it is based on
The difference for the function that the electric power data predicting abnormality program of XGBoost is realized according to its each section, can be divided into acquisition
Module 201, analysis module 202 and prediction module 203.By above description it is found that the so-called program module of the present invention refers to
The series of computation machine program instruction section that can complete specific function, than program more suitable for describing the electric power based on XGBoost
Implementation procedure of the data exception Prediction program in electronic device 10.The functions or operations step that the module 201-203 is realized
Rapid similar as above, and will not be described here in detail, illustratively, such as wherein:
Module 201 is obtained for obtaining electric power data of the predetermined building in the first time predefined section；
Analysis module 202 is used to determine the electric power data that method analysis obtains using predetermined exceptional value, to obtain
First electric power data collection without exception and the first abnormal electric power data collection；
Prediction module 203, which is used to substituting into the first abnormal power data set into the exceptional value that training is completed in advance, predicts mould
Type carries out exceptional value prediction, to predict the abnormal power data in the first abnormal power data set.
In addition, the present invention also proposes a kind of electric power data predicting abnormality method based on XGBoost, please refer to shown in Fig. 3,
The electric power data predicting abnormality method based on XGBoost includes the following steps:
S301, electric power data of the predetermined building in the first time predefined section is obtained；
It, can also be with it is to be appreciated that the predetermined building can be inpatient building in hospital, patient terminal
It is the building in the places such as office block, family dependents' building, school instruction building, market, the first predefined period can be with
The moon is unit, can also be unit in week, can also be as unit of day；If the electric power data obtained is described as unit of the moon
First time predefined section, such as the Power system load data of the electricity consumption monthly every day in half a year；If being obtained as unit of week
Electric power data be the first time predefined section, such as nearest two months in electricity consumption every day weekly electric load number
According to；If the electric power data obtained is each daily in the first time predefined section, such as nearest 10 days as unit of day
The Power system load data of hour electricity consumption.It is understood that the first predefined period can also as unit of hour,
And the minimum unit of the electric power data obtained can be specific to need to use according to the predetermined building is practical to second grade
The variation of electricity and determine, be not specifically limited in the present embodiment.
S302, the electric power data that method analysis obtains is determined using predetermined exceptional value, it is without exception to obtain first
Electric power data collection and the first abnormal electric power data collection；
It is to be appreciated that since the electric power data got is manually to carry out meter reading, and typing by relevant staff
To the electric power data of electric power data storage system, therefore, the electric power data is typically due to human factor, such as staff's mood
Electric power data caused by bad exist it is abnormal, personnel relieve a sentry join it is indefinite caused by meter reading data it is abnormal etc., or can also be with
The reason of by equipment itself, such as line interruption, data transmission link is impaired etc., and to cause electric power data to exist abnormal；Further,
In the present embodiment, the predetermined exceptional value determine method be box figure identify exceptional value method, specifically include as
Lower step: the unit time in the described first predefined period is determined；It is traversed in determining each unit time respectively
Electric power data with the presence or absence of the data greater than the first predefined thresholds and is in the electric power data inquired in each unit time
The no data existed less than the second predefined thresholds；Make a reservation for if having and existing in the electric power data in the unit time greater than described first
The data of adopted threshold value, or there are the data for being less than second predefined thresholds, it is determined that the electric power number in the unit time
There are abnormal power data in, and the electric power data in the unit time is the sample that the described first abnormal electric power data is concentrated
Collection；If having in the electric power data in the unit time, there is no the data greater than first predefined thresholds, and there is no be less than
The data of second predefined thresholds, it is determined that abnormal power data are not present in the electric power data in the unit time, it should
Electric power data in unit time is the sample set that the described first electric power data without exception is concentrated；Wherein, described first is predetermined
Adopted threshold value is QU+1.5IQRQU+1.5QR, and second predefined thresholds are QL-1.5IQRQL-1.5IQR, wherein QU is system
Obtained upper quartile is counted, indicates there are 1/4 data bigger than him in total data, QL is the lower quartile that statistics obtains,
Indicate there are 1/4 data smaller than him in total data, IQR is quartile spacing, is the difference of QU and QL.
Specifically, in this example, it is assumed that the first predefined period is the then institute with the period of moon unit
Every day that the predefined unit time is the middle of each month in the predefined first time period is stated, if having one day use in certain middle of the month
There are exceptional values for the Power system load data of electricity, then the power load charge values of the daily electricity consumption of this month are the first abnormal power data
The sample data of concentration, similarly, if exceptional value is not present in the Power system load data in the daily electricity consumption in certain middle of the month, the middle of the month
The power load charge values of daily electricity consumption be sample data that the described first electric power data without exception is concentrated.It is to be appreciated that false
If the first predefined period with unit around, then when the predefined unit time is described predefined first
In weekly in every day, if the first predefined period, as unit of day, institute's predefined unit time is
In the predefined first time daily in each hour.
S303, the abnormal value prediction model progress that the first abnormal power data set is substituted into training completion in advance are abnormal
Value prediction, to predict the abnormal power data in the first abnormal power data set.
Specifically, the abnormal value prediction model that the preparatory training is completed is XGBoost model, the exceptional value and model
The test process of training process and model including model, the training process of the model include:
E2, electric power data sample set of the predetermined target construction in the second time predefined section is obtained,
The abnormality of the electric power data obtained is analyzed, to obtain the second data set without exception and the second abnormal data set；
It is to be appreciated that the first time predefined section can be identical with the second time predefined section, it is preferable that
The second time predefined section is greater than the first time predefined section, and the first time predefined section is pre- with described second
The chronomere for defining the period must be identical, if the first time predefined Duan Yiyue is unit, then described second is pre-
Defining the period must be also as unit of the moon.It should be noted that during the abnormality for the electric power data that analysis obtains,
Need to consider the ratio for the electric power data exceptional value for including in the predefined period obtained, if such as described first predefined
Period and the second time predefined Duan Jun then need to consider as unit of the moon predefined described second as unit of the moon
In period, the exceptional value ratio of electric power data in the Power system load data of electricity consumption every day monthly, if pre- described second
There is the exceptional value ratio of certain month electric power data more than preset exceptional value proportion threshold value, such as 10% in the definition period, then
It needs to carry out error condition analysis to the electric power data of this month based on experience value, alternatively, need to abandon the electric power data of this month, and
The electric power data in more times is obtained as sample set.
F2, the training sample set and test sample that the data in the described second data set without exception are divided into preset ratio
Collection modifies in the training sample set partial data at random to obtain abnormality test sample set；
Specifically, the ratio of the training sample set is typically larger than the ratio of the test sample collection, such as preferably, institute
The ratio for stating training sample set and the test sample collection is 7:3.
G2, label is added to each sample data that the training sample is concentrated, the label is preset each sample
Whether data are exceptional value；
Specifically, in this example, it is assumed that being exceptional value, then preset label is 1, it is assumed that is not exceptional value, then in advance
If label be 0.
H2, the electric power data that the training sample with the label is concentrated using the XGboost that pre-establishes touch type into
Row supervised learning, to obtain abnormal value prediction model；
Specifically, it is a kind of boost model based on decision tree structure that the XGboost, which touches type, similar with GBDT, but by
Type is touched in XGboost to have carried out second level Taylor expansion to objective function and joined regularization term so that model have it is stronger
Generalization ability can evade the risk of fitting, strong antijamming capability.Specifically, to the training sample set benefit with the label
It being exercised supervision study, is inputted as sample and label with XGboost, the sample is the electric power data that the training sample is concentrated,
The label is whether the electric power data that the training sample is concentrated is exceptional value, exports the training sample concentration for prediction
Each electric power data be exceptional value probability.
F2, model accuracy test is carried out to the obtained abnormal value prediction model, if test passes through, model training
Terminate, if test does not pass through, increase the sample data that the training sample is concentrated, and repeat above-mentioned steps E2, F2,
G2、H2。
Specifically, the test process of the model includes: according to obtained abnormal value prediction model to the abnormality test
Data in sample set carry out exceptional value mark, to obtain with the electric power data collection marked extremely；It will obtain with abnormal
The electric power data collection of mark is compared with the data that the training sample after random modification is concentrated, if described with abnormal mark
Electric power data concentrate the abnormal data of mark to obtain after the random modification electric power data compared with, the abnormal number of mark
It is more than or equal to preset probability threshold value according to the probability value of the electric power data for modification, it is determined that test the accuracy of model
Do not pass through；Alternatively, if after abnormal data and the random modification for concentrating mark with the electric power data marked extremely
Obtained electric power data is compared, and the abnormal data of mark is that the probability value of the electric power data of modification is less than preset probability threshold value,
Then pass through for the accuracy test of model.
By above-mentioned thing embodiment it is found that the electric power data predicting abnormality method proposed by the present invention based on XGBoost, first
Obtain electric power data of the predetermined building in the first time predefined section；Then true using predetermined exceptional value
The electric power data that method analysis obtains is determined, to obtain the first electric power data collection without exception and the first abnormal electric power data collection；
The first abnormal power data set is finally substituted into the abnormal value prediction model that training is completed in advance and carries out exceptional value prediction, with
Predict the abnormal power data in the first abnormal power data set.It can be improved what abnormal data in electric power data was predicted
Accuracy provides reference data for the construction and popularization of smart grid.
In addition, the present invention also proposes a kind of computer readable storage medium, stored on the computer readable storage medium
There is the electric power data predicting abnormality program based on XGBoost, the electric power data predicting abnormality program based on XGBoost is located
It manages and realizes following operation when device executes:
Obtain electric power data of the predetermined building in the first time predefined section；
The electric power data that method analysis obtains is determined using predetermined exceptional value, to obtain the first electric power without exception
Data set and the first abnormal electric power data collection；
It is pre- that the first abnormal power data set is substituted into the abnormal value prediction model progress exceptional value that training is completed in advance
It surveys, to predict the abnormal power data in the first abnormal power data set.
Computer readable storage medium of the present invention, specific implementation process and electronic device and the electricity based on XGBoost
Force data predicting abnormality method is similar, and details are not described herein.
By above-mentioned analysis it is found that computer readable storage medium of the invention, obtains predetermined building first
Electric power data in the first time predefined section；Then the electric power of method analysis acquisition is determined using predetermined exceptional value
Data, to obtain the first electric power data collection without exception and the first abnormal electric power data collection；It is finally abnormal electric by described first
Force data collection substitutes into the abnormal value prediction model that training is completed in advance and carries out exceptional value prediction, abnormal electric to predict described first
The abnormal power data that force data is concentrated.It can be improved the accuracy that abnormal data is predicted in electric power data, be smart grid
Construction and popularization provide reference data.
Through the above description of the embodiments, those skilled in the art can be understood that above-described embodiment side
Method can be realized by means of software and necessary general hardware platform, naturally it is also possible to by hardware, but in many cases
The former is more preferably embodiment.Based on this understanding, technical solution of the present invention substantially in other words does the prior art
The part contributed out can be embodied in the form of software products, which is stored in a storage medium
In (such as ROM/RAM, magnetic disk, CD), including some instructions are used so that a terminal device (can be mobile phone, computer, clothes
Business device, air conditioner or the network equipment etc.) execute method described in each embodiment of the present invention.The above is only of the invention excellent
Embodiment is selected, is not intended to limit the scope of the invention, it is all using made by description of the invention and accompanying drawing content etc.
Structure or equivalent process transformation are imitated, is applied directly or indirectly in other relevant technical fields, and is similarly included in the present invention
Scope of patent protection in.
Claims (10)
1. a kind of electronic device, which is characterized in that the electronic device includes memory and the processing that connect with the memory
Device, the processor are described for executing the electric power data predicting abnormality program based on XGBoost stored on the memory
Electric power data predicting abnormality program based on XGBoost realizes following steps when being executed by the processor:
A1, electric power data of the predetermined building in the first time predefined section is obtained；
A2, the electric power data that method analysis obtains is determined using predetermined exceptional value, to obtain the first electric power without exception
Data set and the first abnormal electric power data collection；
A3, the abnormal value prediction model progress exceptional value that the first abnormal power data set is substituted into training completion in advance are pre-
It surveys, to predict the abnormal power data in the first abnormal power data set.
2. electronic device as described in claim 1, which is characterized in that in the step A2, predetermined exceptional value is true
Determining method is that box figure identifies exceptional value method, and the step A2 includes the following steps:
Determine the unit time in the described first predefined period；
Traverse the electric power data in determining each unit time respectively, inquire in the electric power data in each unit time whether
In the presence of the data for being greater than the first predefined thresholds and with the presence or absence of the data less than the second predefined thresholds；
If there are the data existed in the electric power data in the unit time greater than first predefined thresholds, or exists and be less than institute
State the data of the second predefined thresholds, it is determined that there are abnormal power data in the electric power data in the unit time；
If having in the electric power data in the unit time, there is no the data greater than first predefined thresholds, and there is no be less than
The data of second predefined thresholds, it is determined that abnormal power data are not present in the electric power data in the unit time；
The electric power data there are in the unit time of abnormal power data constitutes the described first abnormal electric power data collection, institute
The electric power data stated in the unit time there is no abnormal power data constitutes the first electric power data collection without exception.
3. electronic device as claimed in claim 2, which is characterized in that first predefined thresholds are as follows:
QU+1.5IQRQU+1.5QR
Second predefined thresholds are as follows:
QL-1.5IQRQL-1.5IQR
Wherein, QU is the upper quartile that statistics obtains, and indicates have 1/4 data bigger than him in total data, QL is to count
The lower quartile arrived indicates have 1/4 data smaller than him in total data, and IQR is quartile spacing, is the difference of QU and QL
Value.
4. electronic device as claimed in claim 3, which is characterized in that in the step A3, what the preparatory training was completed
Abnormal value prediction model is XGBoost model, and the exceptional value and model include the training process of model and the test of model
The training process of journey, the model includes:
E1, electric power data sample set of the predetermined target construction in the second time predefined section, analysis are obtained
The abnormality of the electric power data of acquisition, to obtain the second data set without exception and the second abnormal data set；
F1, the training sample set and test sample collection that the data in the described second data set without exception are divided into preset ratio, with
Machine maintenance changes in the training sample set partial data to obtain abnormality test sample set；
G1, label is added to each sample data that the training sample is concentrated, the label is preset each sample data
It whether is exceptional value；
H1, the electric power data concentrated to the training sample with the label are touched type using the XGboost pre-established and are supervised
Educational inspector practises, to obtain abnormal value prediction model；
J1, model accuracy test is carried out to the obtained abnormal value prediction model, if test passes through, model training knot
Beam increases the sample data that the training sample is concentrated if test does not pass through, and repeat above-mentioned steps E1, F1, G1,
H1。
5. electronic device as claimed in claim 4, which is characterized in that the test process of the model includes:
Exceptional value mark is carried out to the data in the abnormality test sample set according to obtained abnormal value prediction model, to obtain
With the electric power data collection marked extremely；
The obtained data concentrated with the electric power data collection marked extremely with the training sample after random modification are compared
Compared with if the electric power for concentrating the abnormal data of mark and the random modification to obtain later with the electric power data marked extremely
Data are compared, and the abnormal data of mark is that the probability value of the electric power data of modification is more than or equal to preset probability threshold value, then
It determines and the accuracy test of model is not passed through；
Alternatively, if described concentrate the abnormal data of mark and the random modification to obtain later with the electric power data marked extremely
Electric power data compare, the abnormal data of mark is that the probability value of the electric power data of modification is less than preset probability threshold value, then needle
The accuracy test of model is passed through.
6. a kind of electric power data predicting abnormality method based on XGBoost, which is characterized in that described method includes following steps:
S1, electric power data of the predetermined building in the first time predefined section is obtained；
S2, the electric power data that method analysis obtains is determined using predetermined exceptional value, to obtain the first electric power without exception
Data set and the first abnormal electric power data collection；
S3, the abnormal value prediction model progress exceptional value that the first abnormal power data set is substituted into training completion in advance are pre-
It surveys, to predict the abnormal power data in the first abnormal power data set.
7. the electric power data predicting abnormality method based on XGBoost as claimed in claim 6, which is characterized in that in the step
In rapid S2, predetermined exceptional value determines that method is that box figure identifies exceptional value method, and the step S2 includes the following steps:
Determine the unit time in the described first predefined period；
Traverse the electric power data in determining each unit time respectively, inquire in the electric power data in each unit time whether
In the presence of the data for being greater than the first predefined thresholds and with the presence or absence of the data less than the second predefined thresholds；
If there are the data existed in the electric power data in the unit time greater than first predefined thresholds, or exists and be less than institute
State the data of the second predefined thresholds, it is determined that there are abnormal power data in the electric power data in the unit time；
If having in the electric power data in the unit time, there is no the data greater than first predefined thresholds, and there is no be less than
The data of second predefined thresholds, it is determined that abnormal power data are not present in the electric power data in the unit time；
The electric power data there are in the unit time of abnormal power data constitutes the described first abnormal electric power data collection, institute
The electric power data stated in the unit time there is no abnormal power data constitutes the first electric power data collection without exception.
8. the electric power data predicting abnormality method based on XGBoost as claimed in claim 7, which is characterized in that described first
Predefined thresholds are as follows:
QU+1.5IQRQU+1.5QR
Second predefined thresholds are as follows:
QL-1.5IQRQL-1.5IQR
Wherein, QU is the upper quartile that statistics obtains, and indicates have 1/4 data bigger than him in total data, QL is to count
The lower quartile arrived indicates have 1/4 data smaller than him in total data, and IQR is quartile spacing, is the difference of QU and QL
Value.
9. the electric power data predicting abnormality method based on XGBoost as claimed in claim 8, which is characterized in that in the step
In rapid S3, the abnormal value prediction model that the preparatory training is completed is XGBoost model, and the exceptional value and model include model
Training process and model test process, the training process of the model includes:
E2, electric power data sample set of the predetermined target construction in the second time predefined section, analysis are obtained
The abnormality of the electric power data of acquisition, to obtain the second data set without exception and the second abnormal data set；
F2, the training sample set and test sample collection that the data in the described second data set without exception are divided into preset ratio, with
Machine maintenance changes in the training sample set partial data to obtain abnormality test sample set；
G2, label is added to each sample data that the training sample is concentrated, the label is preset each sample data
It whether is exceptional value；
H2, the electric power data concentrated to the training sample with the label are touched type using the XGboost pre-established and are supervised
Educational inspector practises, to obtain abnormal value prediction model；
J2, model accuracy test is carried out to the obtained abnormal value prediction model, if test passes through, model training knot
Beam increases the sample data that the training sample is concentrated if test does not pass through, and repeat above-mentioned steps E2, F2, G2,
H2。
10. a kind of computer readable storage medium, the computer-readable recording medium storage has the electric power number based on XGBoost
According to predicting abnormality program, the electric power data predicting abnormality program based on XGBoost can be executed by least one processor, with
The electric power data based on XGBoost for executing at least one described processor as described in any one of claim 6-9 is abnormal
The step of prediction technique.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810954347.8A CN109299156A (en) | 2018-08-21 | 2018-08-21 | Electronic device, the electric power data predicting abnormality method based on XGBoost and storage medium |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810954347.8A CN109299156A (en) | 2018-08-21 | 2018-08-21 | Electronic device, the electric power data predicting abnormality method based on XGBoost and storage medium |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN109299156A true CN109299156A (en) | 2019-02-01 |
Family
ID=65165284
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201810954347.8A Withdrawn CN109299156A (en) | 2018-08-21 | 2018-08-21 | Electronic device, the electric power data predicting abnormality method based on XGBoost and storage medium |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN109299156A (en) |
Cited By (12)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110866366A (en) * | 2019-11-26 | 2020-03-06 | 南京工程学院 | XGboost algorithm-based island detection method for photovoltaic microgrid containing PHEV |
| CN111275288A (en) * | 2019-12-31 | 2020-06-12 | 华电国际电力股份有限公司十里泉发电厂 | XGboost-based multi-dimensional data anomaly detection method and device |
| CN112988725A (en) * | 2021-03-15 | 2021-06-18 | 广东电网有限责任公司清远供电局 | Power transmission line data cleaning method and system, electronic equipment and storage medium |
| CN113052195A (en) * | 2019-12-27 | 2021-06-29 | 北京国双科技有限公司 | Logging curve abnormal value marking method and device, computer equipment and storage medium |
| CN113128612A (en) * | 2021-04-26 | 2021-07-16 | 国网河北省电力有限公司营销服务中心 | Processing method of abnormal value in power data and terminal equipment |
| CN113222245A (en) * | 2021-05-11 | 2021-08-06 | 深圳供电局有限公司 | Method and system for checking monthly electric quantity and electricity charge abnormity of residential user and storage medium |
| US11288602B2 (en) | 2019-09-18 | 2022-03-29 | Hartford Steam Boiler Inspection And Insurance Company | Computer-based systems, computing components and computing objects configured to implement dynamic outlier bias reduction in machine learning models |
| US11328177B2 (en) | 2019-09-18 | 2022-05-10 | Hartford Steam Boiler Inspection And Insurance Company | Computer-based systems, computing components and computing objects configured to implement dynamic outlier bias reduction in machine learning models |
| US11334645B2 (en) | 2011-08-19 | 2022-05-17 | Hartford Steam Boiler Inspection And Insurance Company | Dynamic outlier bias reduction system and method |
| US11550874B2 (en) | 2014-04-11 | 2023-01-10 | Hartford Steam Boiler Inspection And Insurance Company | Future reliability prediction based on system operational and performance data modelling |
| US11615348B2 (en) | 2019-09-18 | 2023-03-28 | Hartford Steam Boiler Inspection And Insurance Company | Computer-based systems, computing components and computing objects configured to implement dynamic outlier bias reduction in machine learning models |
| US11636292B2 (en) | 2018-09-28 | 2023-04-25 | Hartford Steam Boiler Inspection And Insurance Company | Dynamic outlier bias reduction system and method |
-
2018
- 2018-08-21 CN CN201810954347.8A patent/CN109299156A/en not_active Withdrawn
Cited By (15)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US11334645B2 (en) | 2011-08-19 | 2022-05-17 | Hartford Steam Boiler Inspection And Insurance Company | Dynamic outlier bias reduction system and method |
| US11868425B2 (en) | 2011-08-19 | 2024-01-09 | Hartford Steam Boiler Inspection And Insurance Company | Dynamic outlier bias reduction system and method |
| US11550874B2 (en) | 2014-04-11 | 2023-01-10 | Hartford Steam Boiler Inspection And Insurance Company | Future reliability prediction based on system operational and performance data modelling |
| US11803612B2 (en) | 2018-09-28 | 2023-10-31 | Hartford Steam Boiler Inspection And Insurance Company | Systems and methods of dynamic outlier bias reduction in facility operating data |
| US11636292B2 (en) | 2018-09-28 | 2023-04-25 | Hartford Steam Boiler Inspection And Insurance Company | Dynamic outlier bias reduction system and method |
| US11288602B2 (en) | 2019-09-18 | 2022-03-29 | Hartford Steam Boiler Inspection And Insurance Company | Computer-based systems, computing components and computing objects configured to implement dynamic outlier bias reduction in machine learning models |
| US11328177B2 (en) | 2019-09-18 | 2022-05-10 | Hartford Steam Boiler Inspection And Insurance Company | Computer-based systems, computing components and computing objects configured to implement dynamic outlier bias reduction in machine learning models |
| US11615348B2 (en) | 2019-09-18 | 2023-03-28 | Hartford Steam Boiler Inspection And Insurance Company | Computer-based systems, computing components and computing objects configured to implement dynamic outlier bias reduction in machine learning models |
| CN110866366A (en) * | 2019-11-26 | 2020-03-06 | 南京工程学院 | XGboost algorithm-based island detection method for photovoltaic microgrid containing PHEV |
| CN113052195A (en) * | 2019-12-27 | 2021-06-29 | 北京国双科技有限公司 | Logging curve abnormal value marking method and device, computer equipment and storage medium |
| CN111275288B (en) * | 2019-12-31 | 2023-12-26 | 华电国际电力股份有限公司十里泉发电厂 | XGBoost-based multidimensional data anomaly detection method and device |
| CN111275288A (en) * | 2019-12-31 | 2020-06-12 | 华电国际电力股份有限公司十里泉发电厂 | XGboost-based multi-dimensional data anomaly detection method and device |
| CN112988725A (en) * | 2021-03-15 | 2021-06-18 | 广东电网有限责任公司清远供电局 | Power transmission line data cleaning method and system, electronic equipment and storage medium |
| CN113128612A (en) * | 2021-04-26 | 2021-07-16 | 国网河北省电力有限公司营销服务中心 | Processing method of abnormal value in power data and terminal equipment |
| CN113222245A (en) * | 2021-05-11 | 2021-08-06 | 深圳供电局有限公司 | Method and system for checking monthly electric quantity and electricity charge abnormity of residential user and storage medium |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN109299156A (en) | Electronic device, the electric power data predicting abnormality method based on XGBoost and storage medium | |
| CN109410068A (en) | Electronic device, the electric power data Missing Data Filling method based on XGBoost and storage medium | |
| CN109543992A (en) | Intelligent polling method, device, intelligent terminal and server | |
| CN111639885A (en) | Agricultural sewage management system and management method | |
| CN109472550A (en) | The anti-human-equation error method, apparatus in million kilowatt nuclear power station and terminal device | |
| CN101364291A (en) | Method of calibrating data accuracy, device and labour hour data verification system | |
| CN106501760A (en) | A kind of electrical energy meter fault management-control method and system | |
| CN114519524A (en) | Enterprise risk early warning method and device based on knowledge graph and storage medium | |
| CN111158806B (en) | Interface display method, device, computer equipment and storage medium | |
| CN109976823A (en) | A kind of application program launching method, device and terminal device | |
| CN109064104A (en) | Material equipment management method, system and terminal device | |
| CN110232450A (en) | A kind of security control method of station, system and terminal device | |
| CN110163651A (en) | A kind of monitoring administration method, device and terminal device | |
| CN117193975A (en) | Task scheduling method, device, equipment and storage medium | |
| Park et al. | Comparisons of human reliability data between analog and digital environments | |
| CN109214578A (en) | Electronic device, the building electro-load forecast method based on decision-tree model and storage medium | |
| CN110400187A (en) | A kind of billing method, billing device and terminal device | |
| CN112365156B (en) | Data processing method, data processing device, terminal and storage medium | |
| CN107291514A (en) | Method and device for unloading application program and terminal equipment | |
| CN111861021A (en) | Business risk prediction method, device, equipment and computer readable storage medium | |
| CN116843395A (en) | Alarm classification method, device, equipment and storage medium of service system | |
| CN115190045B (en) | Monitoring method and device for express cabinet system service, electronic equipment and storage medium | |
| CN109902486A (en) | Electronic device, abnormal user processing strategie Intelligent Decision-making Method and storage medium | |
| CN116128696A (en) | Emergency flow detection method, system, electronic equipment and storage medium | |
| CN108614685A (en) | Distribution security control method and system |
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 | ||
| WW01 | Invention patent application withdrawn after publication |
Application publication date: 20190201 |
|
| WW01 | Invention patent application withdrawn after publication |