CN107862468A - The method and device that equipment Risk identification model is established - Google Patents
The method and device that equipment Risk identification model is established Download PDFInfo
- Publication number
- CN107862468A CN107862468A CN201711184571.5A CN201711184571A CN107862468A CN 107862468 A CN107862468 A CN 107862468A CN 201711184571 A CN201711184571 A CN 201711184571A CN 107862468 A CN107862468 A CN 107862468A
- Authority
- CN
- China
- Prior art keywords
- risk identification
- identification model
- sample data
- data
- equipment
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
Classifications
-
- G—PHYSICS
- G06—COMPUTING OR CALCULATING; 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 OR CALCULATING; COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/21—Design or setup of recognition systems or techniques; Extraction of features in feature space; Blind source separation
- G06F18/214—Generating training patterns; Bootstrap methods, e.g. bagging or boosting
-
- G—PHYSICS
- G06—COMPUTING OR CALCULATING; COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/24—Classification techniques
Landscapes
- Engineering & Computer Science (AREA)
- Business, Economics & Management (AREA)
- Theoretical Computer Science (AREA)
- Data Mining & Analysis (AREA)
- Human Resources & Organizations (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Economics (AREA)
- Evolutionary Biology (AREA)
- General Engineering & Computer Science (AREA)
- Bioinformatics & Computational Biology (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Artificial Intelligence (AREA)
- Life Sciences & Earth Sciences (AREA)
- Strategic Management (AREA)
- Evolutionary Computation (AREA)
- Entrepreneurship & Innovation (AREA)
- Educational Administration (AREA)
- Game Theory and Decision Science (AREA)
- Development Economics (AREA)
- Marketing (AREA)
- Operations Research (AREA)
- Quality & Reliability (AREA)
- Tourism & Hospitality (AREA)
- General Business, Economics & Management (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The embodiment of the present application discloses a kind of method of equipment Risk identification model foundation and equipment Risk identification model establishes device.The embodiment of the present application method includes：First sample data are obtained from Internet of Things cloud platform, first sample data are contained in total number of samples evidence, and total number of samples uploads to Internet of Things cloud platform according to for things-internet gateway；First sample data are handled, the first sample data after processing are divided into training set data and test set data；Equipment Risk identification model is trained using training set data, obtains target device risk identification model, equipment Risk identification model is the equipment Risk identification model to be trained pre-established；Using test set data test target device risk identification model, the accuracy rate of target device risk identification model is obtained；Whether judging nicety rate is higher than predetermined threshold value；If accuracy rate is higher than predetermined threshold value, it is determined that the confirmation rate of target device risk identification model is up to standard.
Description
Technical field
The invention relates to computer realm, more particularly to a kind of equipment Risk identification model establish method, set
Standby risk identification model establishes device and computer-readable recording medium.
Background technology
With the development of science and technology, equipment is used among many production fields, but because the running environment of equipment is answered
It is miscellaneous, the long aging for causing equipment component of operation hours, equipment fault is become certainty and contingency,
Therefore, in order to ensure the steady Effec-tive Function of equipment, it is necessary to which risk existing for equipment is identified, timely discovering device is present
Risk, reduce the fault rate of equipment.
In order to reduce influence of the failure of equipment to user's production operation, set by establishing equipment Risk identification model pair
Standby existing risk is identified, and is the equipment Risk identification model established by statistic law in the prior art, statistic law is built
Vertical equipment Risk identification model is the sample data by collecting device, and then the equipment Risk identification model is to the sample number
According to being learnt, the equipment Risk identification model is set to obtain the statistical law of sample data, needing to carry out setting for risk identification
Standby data input device risk identification model, obtains risk existing for the equipment.
But the quality of sample data is to rely on based on the equipment Risk identification model that statistic law is established, if without height
The sample data of quality, high feature, the accuracy of the data of equipment Risk identification model prediction can be caused relatively low.
The content of the invention
The embodiment of the present application provides the method and equipment Risk identification model that a kind of equipment Risk identification model is established
Device is established, for establishing equipment Risk identification model, improves the accuracy rate of equipment Risk identification model identification equipment risk.
The embodiment of the present application first aspect provides a kind of method that equipment Risk identification model is established, including：
First sample data are obtained from Internet of Things cloud platform, the first sample packet contains device-dependent ginseng
Number, the first sample data are contained in total number of samples evidence, and the total number of samples uploads to the Internet of Things according to for things-internet gateway
The all devices data of net cloud platform；
The first sample data are handled, by the first sample data after processing be divided into training set data and
Test set data；
Equipment Risk identification model is trained using the training set data, obtains target device risk identification model, it is described
Equipment Risk identification model is the equipment Risk identification model to be trained pre-established；
Using target device risk identification model described in the test set data test, obtain the target device risk and know
The accuracy rate of other model；
Judge whether the accuracy rate is higher than predetermined threshold value；
If the accuracy rate is higher than the predetermined threshold value, it is determined that the confirmation rate of the target device risk identification model reaches
Mark.
Alternatively, after the confirmation rate of the determination target device risk identification model is up to standard, methods described is also wrapped
Include：
The second sample data is obtained from the Internet of Things cloud platform, second sample data is contained in the total number of samples
According to, and second sample data and the first sample data are not overlapping；
The target device risk identification model is updated using second sample data.
Alternatively, described to judge whether the accuracy rate is higher than after predetermined threshold value, methods described also includes：
If the accuracy rate is less than the predetermined threshold value, the 3rd sample data, institute are obtained from the Internet of Things cloud platform
State the 3rd sample data and be contained in the total number of samples evidence, and the 3rd sample data and the first sample data and described
Second sample data is not overlapping；
The target device risk identification model is trained using the 3rd sample data.
Alternatively, before the training equipment Risk identification model using the training set data, methods described also includes：
The training set data is converted into fisrt feature matrix；
It is corresponding, equipment Risk identification model is trained using the training set data, including：
The equipment Risk identification model is trained using the fisrt feature matrix.
Alternatively, before target device risk model described in the use test set data test, methods described is also
Including：
The test set data are converted into second characteristic matrix；
It is corresponding, using target device risk model described in the test set data test, including：
The target device risk identification model is tested using the second characteristic matrix.
Alternatively, it is described processing is carried out to the first sample data to include：
The abnormal data and wrong data in the first sample data are removed, with the first sample number after being handled
According to.
The embodiment of the present application second aspect provides a kind of equipment Risk identification model and establishes device, equipment Risk identification mould
Type, which establishes device, has the function of realizing that equipment Risk identification model in above-mentioned first aspect establishes device behavior.The function can be with
Realized by hardware, corresponding software can also be performed by hardware and is realized.The hardware or software include it is one or more with it is upper
State module corresponding to function phase.
The embodiment of the present application third aspect provides a kind of computer-readable storage medium, and the computer-readable storage medium is used to store
Establish computer software instructions used in device for the equipment Risk identification model of above-mentioned first aspect, it include be for execution
The equipment Risk identification model of first aspect establishes the program designed by device.
The embodiment of the present application fourth aspect provides a kind of computer program product, and the computer program product includes calculating
Machine software instruction, the computer software instructions can be loaded by processor to realize the method stream in above-mentioned first aspect
Journey.
As can be seen from the above technical solutions, the embodiment of the present application has advantages below：By being obtained from Internet of Things cloud platform
The first sample data of taking equipment, it is then training set data and survey according to certain ratio cut partition by the first sample data
Examination collection data, the equipment Risk identification model pre-established using training set data training, obtain target device risk identification mould
Type, then using the test set data test target device risk identification model, determine the target device risk identification model
Accuracy rate, deep learning Algorithm Learning sample data is used in the present embodiment, set for target of the accuracy rate higher than predetermined threshold value
Standby risk identification model, can constantly obtain the second sample data and continue to update, continue to optimize the target device risk identification mould
The accuracy rate of formula.It is less than the target device risk identification model of predetermined threshold value for accuracy rate, can constantly obtains the 3rd sample
Data re -training target device risk identification model, until the rate of accuracy reached of target device risk identification model is to default threshold
Value, so as to improve the accuracy rate of equipment Risk identification model prediction equipment Risk.
Brief description of the drawings
Fig. 1 is that equipment Risk identification model establishes system framework schematic diagram in the embodiment of the present application；
Fig. 2 is one embodiment schematic diagram of the method that equipment Risk identification model is established in the embodiment of the present application；
Fig. 3 is another embodiment schematic diagram of the method that equipment Risk identification model is established in the embodiment of the present application；
Fig. 4 is another embodiment schematic diagram of the method that equipment Risk identification model is established in the embodiment of the present application；
Fig. 5 is another embodiment schematic diagram of the method that equipment Risk identification model is established in the embodiment of the present application；
Fig. 6 is one embodiment schematic diagram that equipment Risk identification model establishes device in the embodiment of the present application；
Fig. 7 is another embodiment schematic diagram that equipment Risk identification model establishes device in the embodiment of the present application.
Embodiment
The embodiment of the present application provides the method and equipment Risk identification model that a kind of equipment Risk identification model is established
Device is established, for establishing equipment Risk identification model, improves the accuracy rate of equipment Risk identification model identification equipment risk.
The embodiment of the present application can be applied to equipment Risk identification model as shown in Figure 1 and establish system framework figure figure, and this is
System frame diagram includes：Sample data 101, sample data 102, sample data 103, things-internet gateway 104, Internet of Things cloud platform
105 and machine learning system 106.It should be noted that the sample data in the embodiment of the present application includes the parameter of equipment, this
Place only lists sample data 101, sample data 102 and sample data 103 as shown in Figure 1, can also be more, herein not
It is limited.In addition, the sample data in the embodiment of the present application in Fig. 1 can be obtained by same things-internet gateway, can also
It is that the corresponding things-internet gateway of each sample data obtains, only enumerates sample data acquisition modes as shown in Figure 1 herein, it is right
Do not limited in the acquisition modes of sample data.
Sample data 101, sample data 102 and sample data 103 can be the operational factor of industrial equipment, equipment life
Produce the use duration of ambient parameter and equipment.
The device data that things-internet gateway 104 gathers from industrial equipment, obtains is as sample data.Things-internet gateway with
Internet of Things cloud platform 105 is communicated to connect, and the device data acquired is combed, classified, and by the device data of acquisition
Report to Internet of Things cloud platform 105.
Specifically, the things-internet gateway 104 in the embodiment of the present invention, it is a intelligent network for being directed to industrial Internet of Things
Close, including data acquisition module, communication module, locating module, data processing chip module etc., multiple industry can be docked simultaneously to be set
Standby or sensor, support Ethernet interface (Ethernet), RS485 serial ports, RS232 serial ports, the uplink mode such as be wirelessly transferred, or
The wireless transmission methods such as GPRS, 433MHZ, 2.4GHZ, WI-FI.Different communication protocol and multiple servers are supported to exchange number
According to.Collect the work(such as data acquisition, data classification, data transfer, communication management, data receiver, protocol conversion, data processing forwarding
Energy.
It is understood that the device data of industrial equipment corresponding to being obtained in the present embodiment only with above-mentioned description of contents
Concrete mode, in actual applications, other manner can also be used, as long as the number of devices of industrial equipment corresponding to can obtaining
According to being not specifically limited herein.
In the present embodiment, industrial equipment can include but is not limited to compressor, generating set, diesel engine, Industrial Boiler, vapour
Turbine, water treatment facilities and packaging facilities, are not specifically limited herein.
Internet of Things cloud platform 105 can carry out a series of processing to the device data that things-internet gateway uploads, including：Connect
Receipts, store, manage, organize, associate, contrast and trigger.
Specifically, the Internet of Things cloud platform 105 in the embodiment of the present invention, the data processing being made up of multiple server zones
Maincenter, each cluster are made up of more physical servers, and its overall capacity can be carried on all separate unit physical servers
The summation of oncurrent processing ability, it assures that can establish redundancy backup center on multiple ground, data, services are under any circumstance all
Do not interrupt.And possess powerful Data Concurrent disposal ability, it is personalized to possess hundred million grades of high concurrent disposal ability and Millisecond
Event triggering ability, therefore million grades of the things-internet gateway can be supported to connect well, efficient transceiving data.
The specific effect of Internet of Things cloud platform is also embodied in the corresponding data that can receive things-internet gateway transmission, and preserves
These data；A series of logic rules are safeguarded, such as：Incidence relation, Early-warning Model, threshold value control, boundary condition setting etc.
Deng；Data are arranged, organized, associated, analyzed；According to logic rules, a series of trigger mechanism is formed；Give Internet of Things net
Close lower photos and sending messages (data)；The corresponding data received from things-internet gateway, there is provided to other platforms etc..
Machine learning system 106 is mainly to obtain the information that environment provides, and knowledge base is changed using these information, to promote
System executable portion completes the efficiency of task, and executable portion completes task according to knowledge base, while is carried out using the information obtained
Deep learning, and use the continuous repetition training of these information.
Based on above-mentioned Fig. 1 system framework figure, Fig. 2 is refer to, equipment Risk identifies the side established in the embodiment of the present application
One embodiment of method includes：
201st, first sample data are obtained from Internet of Things cloud platform.
After the first sample data obtained in slave unit are uploaded to Internet of Things cloud platform by things-internet gateway, Ke Yicong
First sample data are obtained in Internet of Things cloud platform.It should be noted that the first sample data are the sample data of tape label,
The first sample data are corresponding to label, and label includes the operational factor of equipment；In addition, the first sample data include and equipment
Related equipment operational factor, equipment production environment parameter and equipment use duration.
202nd, first sample data are handled, by the first sample data after processing be divided into training set data and
Test set data.
After first sample data are got, after handling first sample data, then by after processing
One sample data is divided into training set data and test set data.
Specifically, in the present embodiment, first sample data can be divided according to certain ratio.For example, can be with
First sample data according to training set data are accounted for into 70%, test set data account for 30% ratio and divided, can also be according to
Training set data, which accounts for 80%, test set data and accounts for 20% ratio, to be divided, and is not limited herein.
203rd, equipment Risk identification model is trained using training set data.
After first sample data after by processing are divided into training set data, equipment wind is trained using training set data
Dangerous identification model, the accuracy rate of the target device risk identification model is obtained, specifically, this is set using back-propagation algorithm
Standby risk identification model is trained.Declined by continuing on gradient to try to achieve the artificial of the target device risk identification model
The minimum parameter value of neutral net error, so as to obtain the target device risk of the artificial neural network comprising a local optimum
Identification model.
It should be noted that in the present embodiment, the equipment Risk identification model trained using training set data can be pre-
The equipment Risk identification model established first with random forest scheduling algorithm.
204th, using test set data test target device risk identification model.
, can be further to using test set data to the target device after target device risk identification model is obtained
Risk identification is tested, and obtains the accuracy rate of the target device risk identification model, specifically, should by test set data input
Target risk identification model, compare the target device risk identification model output result and the error of physical tags, and calculate defeated
Go out result and physical tags variance obtains the accuracy rate of the target device risk identification model.
205th, whether judging nicety rate is higher than predetermined threshold value, if then performing step 206, if otherwise performing step 207.
When obtaining the accuracy rate of the target device risk identification model, the accuracy rate can be verified, pass through judgement
Whether the accuracy rate can detect whether the target device risk identification model can be accurately judged to set higher than default accuracy rate
Standby existing risk.
206th, determine that the confirmation rate of target device risk identification model is up to standard.
If the accuracy rate of the target device risk identification model is higher than predetermined threshold value, illustrate that the target device risk is known
Other model can be accurately judged to risk existing for equipment, it may be determined that and the confirmation rate of target device risk identification model is up to standard,
It is believed that the equipment Risk identification model can use.
207th, other are operated.
If the accuracy rate of the target device risk identification model is less than predetermined threshold value, illustrate the target device risk identification
The accuracy rate of model is relatively low, it is impossible to which risk existing for Accurate Prediction equipment to the equipment Risk identification model, it is necessary to carry out other
Operation, such as correct or reconfigure parameter.
From above scheme, it can be seen that the present embodiment has the advantage that：By obtaining equipment from Internet of Things cloud platform
First sample data, it is then training set data and test set number according to certain ratio cut partition by the first sample data
According to using the equipment Risk identification model that pre-establishes of training set data training, obtaining target device risk identification model, then
Using the test set data test target device risk identification model, the accuracy rate of the target device risk identification model is determined,
The target device risk identification model that accuracy rate is higher than predetermined threshold value is exported, is by the first sample data of equipment in the present embodiment
It is training set parameter and test set parameter according to certain weight distribution, uses the deep learning Algorithm Learning training set number
According to obtaining the higher target device risk identification model of accuracy rate.
The method established above to equipment Risk identification model in the embodiment of the present application is described, below to equipment Risk
The method that identification model is established is described in actual applications, specifically refer to Fig. 3, and equipment Risk is known in the embodiment of the present application
Another embodiment includes in the method that other model is established：
301st, things-internet gateway obtains total number of samples evidence.
Things-internet gateway obtains the total number of samples evidence of equipment, and the total number of samples is according to including first sample data, the second sample
Data and the 3rd sample data, specifically, the total number of samples that things-internet gateway obtains is according to operational factor, the equipment for including equipment
Production environment parameter and equipment use duration, can also be other, do not limit herein.
It should be noted that the operational factor of above-mentioned equipment can be electric current, voltage；The production environment parameter of equipment can
To be temperature, pneumatic air humidity etc., specifically determined by the actual parameter of specific equipment.
302nd, total number of samples evidence is uploaded to Internet of Things cloud platform by things-internet gateway.
After things-internet gateway gets total number of samples evidence, total number of samples evidence is uploaded to Internet of Things cloud by things-internet gateway
Platform.
303rd, equipment Risk identification model establishes device and obtains first sample data.
After total number of samples evidence is uploaded to Internet of Things cloud platform by things-internet gateway, equipment Risk identification model device can
First sample data are obtained from Internet of Things cloud platform to obtain, specifically, operational factor of the first sample data including equipment,
The production environment parameter of equipment and the use duration of equipment, can also be other, do not limit herein.
304th, equipment Risk identification model establishes device and first sample data is handled.
After first sample data are got, the data of the abnormal data and mistake in first sample data are removed,
With the data standardized；Specifically, removing the abnormal data in first sample data and wrong data is included beyond the
The data of one sample data scope, the data that go beyond the scope are abnormal data or wrong data, including nonnegative value for negative etc..
305th, equipment Risk identification model establish device by the first sample data after processing be divided into training set data with
And test set data.
After handling first sample data, then equipment Risk identification model establishes device by first after processing
Sample data is divided into training set data and test set data.
Specifically, in the present embodiment, first sample data can be divided according to certain ratio；For example, can be with
First sample data according to training set data are accounted for into 70%, test set data account for 30% ratio and divided, can also be according to
Training set data, which accounts for 80%, test set data and accounts for 20% ratio, to be divided, and is not limited herein.
306th, equipment Risk identification model establishes device and training set data is converted into fisrt feature matrix and test set number
According to being converted into second characteristic matrix.
After first sample data after by processing are divided into training set data, equipment Risk identification model establishes device
Training set data is converted into fisrt feature matrix, specifically, makes full use of its sequential relationship, training set number after treatment
According to plus previous training set data and the latter training set data, directly composition fisrt feature matrix.
After first sample data after by processing are divided into test set data, equipment Risk identification model establishes device
Test set data are converted into second characteristic matrix, specifically, make full use of its sequential relationship, test set number after treatment
According to plus previous test set data and the latter test set data, directly forming second characteristic matrix.
307th, equipment Risk identification model establishes device and uses fisrt feature matrix training equipment Risk identification model.
Equipment Risk identification model, which establishes device and advances with random forests algorithm and establish equipment, in the present embodiment waits to train
Equipment Risk identification model, then equipment Risk identification model to be trained is trained using fisrt feature matrix, obtained
To target device risk identification model.Specifically, the equipment Risk is known using fisrt feature matrix using back-propagation algorithm
Other model is trained.Declined by continuing on gradient to try to achieve the artificial neural network of the target device risk identification model
The minimum parameter value of error, so as to obtain the target risk identification model of the artificial neural network comprising a local optimum, need
It is noted that can also be trained in this implementation using other algorithms to equipment Risk identification model, do not limit herein.
308th, equipment Risk identification model establishes device and uses second characteristic matrix test target equipment Risk identification model.
When obtain training after target device risk identification model after, equipment Risk identification model establish device use by
The second characteristic matrix of test set data conversion is tested target device risk identification model, obtains the knowledge of target device risk
The accuracy rate of other model.Specifically, second characteristic matrix is inputted into the target device risk identification model, compares the target risk
The error of identification model output result and physical tags, and calculate output result and physical tags variance obtains target risk knowledge
The accuracy rate of other model.
309th, equipment Risk identification model establishes whether device judging nicety rate is higher than predetermined threshold value, if so, then performing step
Rapid 310 to step 312, if it is not, then performing step 313 to step 314.
Equipment Risk identification model, which establishes device, can determine whether that the accuracy rate of target device risk identification model is
No to be higher than predetermined threshold value, confirmation rate is higher, illustrates that risk existing for the target device risk identification model prediction equipment is more accurate
Really, specifically, the predetermined threshold value can be preset according to the usage time of equipment and the quality of equipment.
310th, equipment Risk identification model establishes device and determines that target risk identification model is up to standard.
Judge the accuracy rate of the target device risk identification model higher than pre- if equipment Risk identification model establishes device
If threshold value, illustrate that the target device risk identification model can more accurately predict equipment risk that may be present in production,
It is capable of the operation conditions of reliable assessment equipment, it is believed that the target device risk identification model can use.
311st, equipment Risk identification model establishes device and obtains the second sample data.
Equipment Risk identification model establishes device and obtains the second sample data, second sample data from Internet of Things cloud platform
Comprising with total number of samples evidence, and the second sample data is not overlapping with first sample data, and same second sample data includes
The use duration of the operational factor of equipment, the production environment parameter of equipment and equipment, can also be other, does not limit herein
It is fixed.
312nd, equipment Risk identification model is established device target device risk identification model is entered using the second sample data
Row renewal.
In order that risk existing for the more Accurate Prediction equipment of accuracy rate target risk identification model up to standard, equipment
Risk identification device constantly can obtain the second sample data from Internet of Things cloud platform, and model is entered using the second sample data
Row matching, updates target risk identification model, it is ensured that and the target risk model can upgrade in time according to the change of device data,
Risk existing for Accurate Prediction equipment.
313rd, equipment Risk identification model establishes device and obtains the 3rd sample data.
Equipment Risk identification model establishes device and obtains the 3rd sample data, the 3rd sample data from Internet of Things cloud platform
Comprising with total number of samples evidence, and not overlapping with first sample data and the second sample data, same 3rd sample data
The production environment parameter of operational factor, equipment comprising equipment and the use duration of equipment, can also be other, do not do herein
Limit.
314th, equipment Risk identification model is established device target device risk identification model is entered using the 3rd sample data
Row training.
Judge the accuracy rate of the target device risk identification model less than pre- if equipment Risk identification model establishes device
If threshold value, illustrate the target device risk identification model can not Accurate Prediction go out equipment risk that may be present in production,
Equipment Risk identification model, which establishes device, can constantly obtain the 3rd sample data, then using the 3rd sample data to target
Risk identification model carries out training again, until the rate of accuracy reached of the target device risk identification model is to predetermined threshold value, energy
It is enough more accurately to predict risk existing for equipment.
In the present embodiment, the target device risk identification model of predetermined threshold value is higher than for accuracy rate, can constantly be obtained
Second sample data continues to update, and continues to optimize the accuracy rate of the target device risk identification pattern, improves target device wind
Dangerous identification model judges that equipment has the accuracy rate of risk.It is less than the target device risk identification mould of predetermined threshold value for accuracy rate
Type, the 3rd sample data re -training target device risk identification model can be constantly obtained, until target device risk identification
The rate of accuracy reached of model can accurately identify risk existing for equipment to predetermined threshold value.
In the embodiment of the present application, whether predetermined threshold value can also be higher than according to the accuracy rate of target device risk identification model
Operation is performed to target device risk identification model, is described separately below：
First, the accuracy rate of target device risk identification model is higher than predetermined threshold value.
In the present embodiment, when the accuracy rate for judging target device risk identification model is higher than predetermined threshold value, target is determined
Equipment Risk identification model it is accurate determine rate it is up to standard after, can continue to obtain the second sample data from Internet of Things cloud platform,
Then it is updated using the second sample data equipment Risk identification model up to standard to accuracy rate, specifically refer to Fig. 4, this Shen
Another embodiment for the method that equipment Risk identification model is established it please include in embodiment：
401st, the second sample data is obtained from Internet of Things cloud platform.
After the first sample data obtained in slave unit are uploaded to Internet of Things cloud platform by things-internet gateway, Ke Yicong
The second sample data is obtained in Internet of Things cloud platform.It should be noted that second sample data is the sample data of tape label,
Second sample data is corresponding to label, and label includes the operational factor of equipment；In addition, the first sample data include and equipment
Related equipment operational factor, equipment production environment parameter and equipment use duration.
402nd, the second sample data is handled.
After the second sample data is got, the data of the abnormal data and mistake in the second sample data are removed,
With the data standardized；Specifically, removing the abnormal data in the second sample data and wrong data is included beyond the
The data of two sample data scopes, the data that go beyond the scope are abnormal data or wrong data, including nonnegative value for negative etc..
403rd, the second sample data is converted into third feature matrix.
After the second sample data is handled, training set data is converted into third feature matrix, specifically, filled
Divide and utilize its sequential relationship, the second sample data after treatment, plus previous data and the latter data, directly form
Third feature matrix.
404th, target device risk identification model is updated using third feature matrix.
After the second sample data will be converted into third feature matrix, third feature matrix can be used to accuracy rate
The target device risk identification model for reaching predetermined threshold value is constantly updated so that the target device risk identification model can be with
The database of itself is updated according to the second sample data, more accurately judges risk existing for equipment.
In the present embodiment, the target device risk identification model of predetermined threshold value is higher than for accuracy rate, can constantly be obtained
Second sample data continues to update, and continues to optimize the accuracy rate of the target risk recognition mode, improves target risk identification mould
Type judges that equipment has the accuracy rate of risk.
2nd, the accuracy rate of target device risk identification model is less than predetermined threshold value.
In the present embodiment, when the accuracy rate for judging target device risk identification model is less than predetermined threshold value, illustrate the mesh
Marking device risk identification model is unable to risk existing for Accurate Prediction equipment, then can continue to obtain from Internet of Things cloud platform
Three sample datas, then equipment Risk identification model of the accuracy rate less than predetermined threshold value is instructed using the 3rd sample data
Practice, specifically refer to Fig. 5, another embodiment for the method that equipment Risk identification model is established includes in the embodiment of the present application：
501st, the 3rd sample data is obtained from Internet of Things cloud platform.
502nd, the 3rd sample data is handled.
In the present embodiment, step 501 to step 502 is similar to step 402 with Fig. 4 steps 401, and here is omitted.
503rd, the 3rd sample data after processing is divided into training set data and test set data.
After handling the 3rd sample data, the 3rd sample data after processing is then divided into training set data
And test set data.
Specifically, in the present embodiment, the 3rd sample data can be divided according to certain ratio.For example, can be with
3rd sample data according to training set data is accounted for into 70%, test set data account for 30% ratio and divided, can also be according to
Training set data, which accounts for 80%, test set data and accounts for 20% ratio, to be divided, and is not limited herein.
504th, training set data is converted into fourth feature matrix and test set data is converted into fifth feature matrix.
After the 3rd sample data after by processing is divided into training set data, it is special that training set data is converted into the 4th
Matrix is levied, specifically, makes full use of its sequential relationship, training set data after treatment, plus previous training set data
With the latter training set data, fourth feature matrix is directly formed.
After the 3rd sample data after by processing is divided into test set data, establishes device and convert test set data
For fifth feature matrix, specifically, its sequential relationship is made full use of, test set data after treatment, plus previous survey
Examination collection data and the latter test set data, directly form fifth feature matrix.
505th, it is less than the target device risk identification model of predetermined threshold value using fourth feature matrix training accuracy rate.
Fourth feature matrix is used to target device risk identification model of the accuracy rate less than predetermined threshold value in the present embodiment
It is trained, specifically, target of the accuracy rate less than predetermined threshold value is set using fourth feature matrix using back-propagation algorithm
Standby risk identification model is trained.Declined by continuing on gradient to try to achieve the artificial of the target device risk identification model
The minimum parameter value of neutral net error, so as to obtain the identification of the target risk of the artificial neural network comprising a local optimum
Model can also be trained, herein not in this implementation, it is necessary to explanation using other algorithms to equipment Risk identification model
Limit.
506th, using fifth feature matrix test target equipment Risk identification model.
After being trained to accuracy rate less than the target device risk identification model of predetermined threshold value, using by test set
The fifth feature matrix of data conversion is tested, and obtains the accuracy rate of target device risk identification model again.Specifically, will
The fifth feature Input matrix target risk identification model, compare the target risk identification model output result and physical tags
Error, and calculate output result and physical tags variance obtains the accuracy rate of the target risk identification model.
507th, whether judging nicety rate is higher than predetermined threshold value, if so, step 508 is then performed, if it is not, then performing step 509.
Whether the accuracy rate for determining whether to be tested to obtain using fifth feature matrix is higher than predetermined threshold value, it is necessary to say
Bright, predetermined threshold value herein as default predetermined threshold value for the first time or can reset predetermined threshold value.
508th, determine that target risk identification model is up to standard.
If the accuracy rate of the target device risk identification model by the training of the 3rd sample data is higher than predetermined threshold value, say
The bright target device risk identification model can reach standard compared with accuracy rate, and can predict equipment may deposit in production
Risk, be capable of the operation conditions of reliable assessment equipment, it is believed that the target device risk identification model can use.
509th, other operations are performed.
If after being trained by the 3rd sample data to this, the accuracy rate of the target device risk identification model is still
So it is less than predetermined threshold value, then continues to obtain the 4th sample data from Internet of Things cloud platform to equipment Risk identification model progress
Training.
In the present embodiment, the target device risk identification model of predetermined threshold value is less than for accuracy rate, can constantly be obtained
3rd sample data re -training target device risk identification model, until the rate of accuracy reached of target device risk identification model arrives
Predetermined threshold value, risk existing for equipment can be accurately identified, by repetition training, lift target device risk identification model
Accuracy rate.
The method established above to the equipment Risk identification model in the embodiment of the present application is described, below to this Shen
Equipment Risk identification model that please be in embodiment is established device and is described：
Fig. 6 is refer to, one embodiment that the equipment Risk identification model in the application implementation establishes device includes：
First acquisition unit 601, for obtaining first sample data, the first sample data from Internet of Things cloud platform
Comprising device-dependent parameter, the first sample data are contained in total number of samples evidence, and the total number of samples evidence is Internet of Things
Gateway uploads to all devices data of the Internet of Things cloud platform；
Processing unit 602, for handling first sample data, the first sample data after processing are divided into instruction
Practice collection data and test set data；
First training unit 603, for using training set data training equipment Risk identification model, obtaining target device wind
Dangerous identification model, equipment Risk identification model are the equipment Risk identification models to be trained pre-established；
Test cell 604, for using test set data test target device risk identification model, obtain target device wind
The accuracy rate of dangerous identification model；
Judging unit 605, whether it is higher than predetermined threshold value for judging nicety rate；
Determining unit 606, for when accuracy rate is higher than predetermined threshold value, determining the confirmation of target device risk identification model
Rate is up to standard.
In the present embodiment, equipment Risk identification model, which establishes device, also to be included：
Second acquisition unit 607, for obtaining the second sample data from Internet of Things cloud platform, the second sample data is contained in
Total number of samples evidence, and the second sample data is not overlapping with first sample data；
Updating block 608, for being updated using the second sample data to target device risk identification model.
In the present embodiment, equipment Risk identification model, which establishes device, also to be included：
3rd acquiring unit 609, for when accuracy rate is less than predetermined threshold value, the 3rd sample to be obtained from Internet of Things cloud platform
Data, the 3rd sample data are contained in total number of samples evidence, and the 3rd sample data and first sample data and the second sample data
It is not overlapping；
Second training unit 610, for being trained using the 3rd sample data to target device risk identification model.
In the present embodiment, equipment Risk identification model, which establishes device, also to be included：
First conversion unit 611, for training set data to be converted into fisrt feature matrix；
Corresponding, the first training unit 603 is specifically used for using fisrt feature matrix training equipment Risk identification model.
In the present embodiment, equipment Risk identification model, which establishes device, also to be included：
Second conversion unit 612, for test set data to be converted into second characteristic matrix；
Corresponding, test cell 604 is specifically used for using second characteristic matrix test target equipment Risk identification model.
In the present embodiment, processing unit specific 602 is used to remove the abnormal data and wrong data in first sample data,
With the first sample data after being handled.
In the present embodiment, equipment Risk identification model establish flow in device performed by each unit and earlier figures 2 and
The method flow described in embodiment shown in Fig. 5 is similar, and here is omitted.
In the present embodiment, the target device risk identification model of predetermined threshold value, second acquisition unit are higher than for accuracy rate
607 can constantly obtain the second sample data continues to update the target device risk identification model by updating block 608, continues
Optimize the accuracy rate of the target risk recognition mode, improve target risk identification model and judge that equipment has the accurate of risk
Rate.It is less than the target device risk identification model of predetermined threshold value for accuracy rate, the 3rd acquiring unit 609 can constantly obtain the
Three sample datas are by the re -training target device risk identification model of the second training unit 610, until target device risk is known
The rate of accuracy reached of other model can accurately identify risk existing for equipment to predetermined threshold value.
It refer to Fig. 7, equipment Risk identification model is established another embodiment of device and included in the embodiment of the present application：
Fig. 7 is that a kind of equipment Risk identification model provided in an embodiment of the present invention establishes apparatus structure schematic diagram, the equipment
Risk identification model, which establishes device 700, to produce bigger difference because configuration or performance are different, can include one or one
Individual above central processing unit (central processing units, CPU) 701 (for example, one or more processors)
With memory 705, one or more application program or data are stored with the memory 705.
Wherein, memory 705 can be volatile storage or persistently storage.Being stored in the program of memory 705 can wrap
One or more modules are included, each module can include the series of instructions established to equipment Risk identification model in device
Operation.Further, central processing unit 701 could be arranged to communicate with memory 705, be established in equipment Risk identification model
The series of instructions operation in memory 705 is performed on device 700.
Equipment Risk identification model, which establishes device 700, can also include one or more power supplys 702, one or one
Above wired or wireless network interface 703, one or more input/output interfaces 704, and/or, one or more
Operating system, such as Windows ServerTM, Mac OS XTM, UnixTM, LinuxTM, FreeBSDTM etc..
In above-described embodiment as shown in the step performed by equipment Risk identification model establishes device can be based on the Fig. 7
Equipment Risk identification model establishes apparatus structure.
Equipment Risk identification model establishes device in the present embodiment, the flow performed by central processing unit 701 and earlier figures 2
And the method flow described in the embodiment shown in Fig. 5 is similar, here is omitted.
The embodiment of the present application also provides a kind of computer-readable storage medium, and the computer-readable storage medium is used to save as foregoing set
Standby risk identification model establishes the computer software instructions used in device, and it includes being used to perform building for equipment Risk identification model
Program designed by vertical device.
The embodiment of the present application also provides a kind of computer program product, and the computer program product refers to including computer software
Order, the computer software instructions can be loaded by processor to realize the method stream in the embodiment shown in earlier figures 2 to 5
Journey.
It is apparent to those skilled in the art that for convenience and simplicity of description, the system of foregoing description,
The specific work process of device and unit, the corresponding process in preceding method embodiment is may be referred to, will not be repeated here.
In several embodiments provided herein, it should be understood that disclosed system, apparatus and method can be with
Realize by another way.For example, device embodiment described above is only schematical, for example, the unit
Division, only a kind of division of logic function, can there is other dividing mode, such as multiple units or component when actually realizing
Another system can be combined or be desirably integrated into, or some features can be ignored, or do not perform.It is another, it is shown or
The mutual coupling discussed or direct-coupling or communication connection can be the indirect couplings by some interfaces, device or unit
Close or communicate to connect, can be electrical, mechanical or other forms.
The unit illustrated as separating component can be or may not be physically separate, show as unit
The part shown can be or may not be physical location, you can with positioned at a place, or can also be distributed to multiple
On NE.Some or all of unit therein can be selected to realize the mesh of this embodiment scheme according to the actual needs
's.
In addition, each functional unit in each embodiment of the application can be integrated in a processing unit, can also
That unit is individually physically present, can also two or more units it is integrated in a unit.Above-mentioned integrated list
Member can both be realized in the form of hardware, can also be realized in the form of SFU software functional unit.
If the integrated unit is realized in the form of SFU software functional unit and is used as independent production marketing or use
When, it can be stored in a computer read/write memory medium.Based on such understanding, the technical scheme of the application is substantially
The part to be contributed in other words to prior art or all or part of the technical scheme can be in the form of software products
Embody, the computer software product is stored in a storage medium, including some instructions are causing a computer
Equipment (can be personal computer, server, or network equipment etc.) performs the complete of each embodiment methods described of the application
Portion or part steps.And foregoing storage medium includes：USB flash disk, mobile hard disk, read-only storage (ROM, Read-Only
Memory), random access memory (RAM, Random Access Memory), magnetic disc or CD etc. are various can store journey
The medium of sequence code.
Described above, above example is only to illustrate the technical scheme of the application, rather than its limitations；Although with reference to before
Embodiment is stated the application is described in detail, it will be understood by those within the art that：It still can be to preceding
State the technical scheme described in each embodiment to modify, or equivalent substitution is carried out to which part technical characteristic；And these
Modification is replaced, and the essence of appropriate technical solution is departed from the spirit and scope of each embodiment technical scheme of the application.
Claims (14)
1. a kind of method that equipment Risk identification model is established, it is characterised in that including：
First sample data are obtained from Internet of Things cloud platform, the first sample packet contains device-dependent parameter, institute
State first sample data and be contained in total number of samples evidence, the total number of samples evidence uploads to the Internet of Things cloud for things-internet gateway and put down
The all devices data of platform；
The first sample data are handled, the first sample data after processing are divided into training set data and test
Collect data；
Equipment Risk identification model is trained using the training set data, obtains target device risk identification model, the equipment
Risk identification model is the equipment Risk identification model to be trained pre-established；
Using target device risk identification model described in the test set data test, the target device risk identification mould is obtained
The accuracy rate of type；
Judge whether the accuracy rate is higher than predetermined threshold value；
If the accuracy rate is higher than the predetermined threshold value, it is determined that the confirmation rate of the target device risk identification model is up to standard.
2. according to the method for claim 1, it is characterised in that described to determine the target device risk identification model really
Recognize rate it is up to standard after, methods described also includes：
The second sample data is obtained from the Internet of Things cloud platform, second sample data is contained in the total number of samples evidence,
And second sample data and the first sample data are not overlapping；
The target device risk identification model is updated using second sample data.
3. according to the method for claim 2, it is characterised in that it is described judge the accuracy rate whether be higher than predetermined threshold value it
Afterwards, methods described also includes：
If the accuracy rate is less than the predetermined threshold value, the 3rd sample data is obtained from the Internet of Things cloud platform, described the
Three sample datas are contained in the total number of samples evidence, and the 3rd sample data and the first sample data and described second
Sample data is not overlapping；
The target device risk identification model is trained using the 3rd sample data.
4. according to the method in any one of claims 1 to 3, it is characterised in that described to be instructed using the training set data
Before practicing equipment Risk identification model, methods described also includes：
The training set data is converted into fisrt feature matrix；
It is corresponding, equipment Risk identification model is trained using the training set data, including：
The equipment Risk identification model is trained using the fisrt feature matrix.
5. according to the method in any one of claims 1 to 3, it is characterised in that described to be surveyed using the test set data
Before trying the target device risk model, methods described also includes：
The test set data are converted into second characteristic matrix；
It is corresponding, using target device risk model described in the test set data test, including：
The target device risk identification model is tested using the second characteristic matrix.
6. according to the method for claim 5, it is characterised in that described processing is carried out to the first sample data to include：
The abnormal data and wrong data in the first sample data are removed, with the first sample data after being handled.
7. a kind of equipment Risk identification model establishes device, it is characterised in that including：
First acquisition unit, for from Internet of Things cloud platform obtain first sample data, the first sample packet contain with
Device-dependent parameter, the first sample data are contained in total number of samples evidence, and the total number of samples is according to on things-internet gateway
Pass to all devices data of the Internet of Things cloud platform；
Processing unit, for handling the first sample data, the first sample data after processing are divided into training
Collect data and test set data；
First training unit, for using training set data training equipment Risk identification model, obtaining target device risk
Identification model, the equipment Risk identification model are the equipment Risk identification models to be trained pre-established；
Test cell, for using target device risk identification model described in the test set data test, obtaining the target
The accuracy rate of equipment Risk identification model；
Judging unit, for judging whether the accuracy rate is higher than predetermined threshold value；
Determining unit, for when the accuracy rate is higher than the predetermined threshold value, determining the target device risk identification model
Confirmation rate it is up to standard.
8. equipment Risk identification model according to claim 7 establishes device, it is characterised in that the equipment Risk identification
Model, which establishes device, also to be included：
Second acquisition unit, for obtaining the second sample data from the Internet of Things cloud platform, second sample data includes
In the total number of samples evidence, and second sample data and the first sample data are not overlapping；
Updating block, for being updated using second sample data to the target device risk identification model.
9. equipment Risk identification model according to claim 8 establishes device, it is characterised in that the equipment Risk identification
Model, which establishes device, also to be included：
3rd acquiring unit, for when the accuracy rate is less than the predetermined threshold value, the is obtained from the Internet of Things cloud platform
Three sample datas, the 3rd sample data are contained in the total number of samples evidence, and the 3rd sample data and described first
Sample data and second sample data be not overlapping；
Second training unit, for being trained using the 3rd sample data to the target device risk identification model.
10. the equipment Risk identification model according to any one of claim 7 to 9 establishes device, the equipment Risk is known
Other model, which establishes device, also to be included：
First conversion unit, for the training set data to be converted into fisrt feature matrix；
Corresponding, first training unit is specifically used for training the equipment Risk identification mould using the fisrt feature matrix
Type.
11. the equipment Risk identification model according to any one of claim 7 to 9 establishes device, it is characterised in that described
Equipment Risk identification model, which establishes device, also to be included：
Second conversion unit, for the test set data to be converted into second characteristic matrix；
Corresponding, the test cell is specifically used for testing the target device risk identification mould using the second characteristic matrix
Type.
12. equipment Risk identification model according to claim 11 establishes device, it is characterised in that the processing unit tool
Body is used to remove abnormal data and wrong data in the first sample data, with the first sample data after being handled.
13. a kind of equipment Risk identification model establishes device, it is characterised in that including：Memory, transceiver and at least one place
Device is managed, instruction is stored with the memory；The memory, the transceiver and at least one processor pass through circuit
Interconnection；
At least one processor calls the instruction, and perform claim requires to set described in the method described in 1 to 6 any one
Standby risk identification model establishes the Message Processing or control operation of device side progress.
A kind of 14. computer-readable recording medium, it is characterised in that including instructing, when the instruction is run on computers,
So that the method any one of computer perform claim requirement 1 to 6.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201711184571.5A CN107862468A (en) | 2017-11-23 | 2017-11-23 | The method and device that equipment Risk identification model is established |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201711184571.5A CN107862468A (en) | 2017-11-23 | 2017-11-23 | The method and device that equipment Risk identification model is established |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN107862468A true CN107862468A (en) | 2018-03-30 |
Family
ID=61702633
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201711184571.5A Pending CN107862468A (en) | 2017-11-23 | 2017-11-23 | The method and device that equipment Risk identification model is established |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN107862468A (en) |
Cited By (36)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108564376A (en) * | 2018-04-20 | 2018-09-21 | 阿里巴巴集团控股有限公司 | Risk control method, device, server and readable storage medium storing program for executing |
| CN109195154A (en) * | 2018-08-13 | 2019-01-11 | 中国联合网络通信集团有限公司 | Internet of Things alters card user recognition methods and device |
| CN109522304A (en) * | 2018-11-23 | 2019-03-26 | 中国联合网络通信集团有限公司 | Exception object recognition methods and device, storage medium |
| CN109784403A (en) * | 2019-01-16 | 2019-05-21 | 武汉斗鱼鱼乐网络科技有限公司 | A kind of method and relevant device identifying risk equipment |
| CN110322254A (en) * | 2019-07-04 | 2019-10-11 | 同盾控股有限公司 | Online fraud recognition methods, device, medium and electronic equipment |
| CN110958305A (en) * | 2019-11-15 | 2020-04-03 | 锐捷网络股份有限公司 | Method and device for identifying terminal equipment of Internet of things |
| CN111177585A (en) * | 2018-11-13 | 2020-05-19 | 北京四维图新科技股份有限公司 | Map POI feedback method and device |
| CN111275453A (en) * | 2018-12-03 | 2020-06-12 | 中国移动通信集团上海有限公司 | An industry identification method and system for Internet of Things equipment |
| CN111461892A (en) * | 2020-03-31 | 2020-07-28 | 支付宝(杭州)信息技术有限公司 | Method and device for selecting derived variables of risk identification model |
| CN111558226A (en) * | 2020-04-28 | 2020-08-21 | 腾讯科技(成都)有限公司 | Method, device, equipment and storage medium for detecting abnormal operation behaviors |
| CN111652284A (en) * | 2020-05-09 | 2020-09-11 | 杭州数梦工场科技有限公司 | Scanner identification method and device, electronic device, storage medium |
| CN112384924A (en) * | 2018-07-26 | 2021-02-19 | 西门子股份公司 | Method and device for establishing product performance prediction model, computer equipment, computer readable storage medium, product performance prediction method and prediction system |
| CN112436969A (en) * | 2020-11-24 | 2021-03-02 | 成都西加云杉科技有限公司 | Internet of things equipment management method, system, equipment and medium |
| CN112640381A (en) * | 2018-06-18 | 2021-04-09 | 帕洛阿尔托网络公司 | Pattern matching based detection in IOT security |
| CN112766398A (en) * | 2021-01-27 | 2021-05-07 | 无锡中车时代智能装备有限公司 | Generator rotor vent hole identification method and device |
| CN113239969A (en) * | 2021-04-15 | 2021-08-10 | 联合汽车电子有限公司 | Feature model establishing method, equipment fingerprint generating method and equipment identification method |
| CN113642805A (en) * | 2021-08-27 | 2021-11-12 | Oppo广东移动通信有限公司 | Algorithm optimization method of Internet of things equipment, electronic equipment and readable storage medium |
| CN113887609A (en) * | 2021-09-28 | 2022-01-04 | 广州绿怡信息科技有限公司 | Device screen aging detection model training method and device screen aging detection method |
| CN113901996A (en) * | 2021-09-29 | 2022-01-07 | 广州绿怡信息科技有限公司 | Device screen perspective detection model training method and device screen perspective detection method |
| CN114298204A (en) * | 2021-12-24 | 2022-04-08 | 广州绿怡信息科技有限公司 | Device screen scratch detection model training method and device screen scratch detection method |
| CN114705444A (en) * | 2022-03-29 | 2022-07-05 | 华北电力科学研究院有限责任公司 | Method and device for monitoring vibration of steam turbine |
| CN115018656A (en) * | 2022-08-08 | 2022-09-06 | 太平金融科技服务(上海)有限公司深圳分公司 | Risk identification method, and training method, device and equipment of risk identification model |
| CN115130713A (en) * | 2021-03-29 | 2022-09-30 | 淘宝（中国）软件有限公司 | Risk data identification, object detection and model optimization method, device and system |
| US11552954B2 (en) | 2015-01-16 | 2023-01-10 | Palo Alto Networks, Inc. | Private cloud control |
| US11552975B1 (en) | 2021-10-26 | 2023-01-10 | Palo Alto Networks, Inc. | IoT device identification with packet flow behavior machine learning model |
| US11671327B2 (en) | 2017-10-27 | 2023-06-06 | Palo Alto Networks, Inc. | IoT device grouping and labeling |
| US11683328B2 (en) | 2017-09-27 | 2023-06-20 | Palo Alto Networks, Inc. | IoT device management visualization |
| US11681812B2 (en) | 2016-11-21 | 2023-06-20 | Palo Alto Networks, Inc. | IoT device risk assessment |
| US11689573B2 (en) | 2018-12-31 | 2023-06-27 | Palo Alto Networks, Inc. | Multi-layered policy management |
| US11706246B2 (en) | 2018-12-12 | 2023-07-18 | Palo Alto Networks, Inc. | IOT device risk assessment and scoring |
| US11722875B2 (en) | 2020-06-01 | 2023-08-08 | Palo Alto Networks, Inc. | IoT device discovery and identification |
| US12289328B2 (en) | 2018-10-15 | 2025-04-29 | Palo Alto Networks, Inc. | Multi-dimensional periodicity detection of IOT device behavior |
| US12289329B2 (en) | 2015-04-07 | 2025-04-29 | Palo Alto Networks, Inc. | Packet analysis based IOT management |
| US12294482B2 (en) | 2018-09-04 | 2025-05-06 | Palo Alto Networks, Inc. | IoT application learning |
| US12301600B2 (en) | 2022-01-18 | 2025-05-13 | Palo Alto Networks, Inc. | IoT device identification by machine learning with time series behavioral and statistical features |
| US12302451B2 (en) | 2020-06-01 | 2025-05-13 | Palo Alto Networks, Inc. | IoT security policy on a firewall |
Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102567391A (en) * | 2010-12-20 | 2012-07-11 | 中国移动通信集团广东有限公司 | Method and device for building classification forecasting mixed model |
| CN103824092A (en) * | 2014-03-04 | 2014-05-28 | 国家电网公司 | Image classification method for monitoring state of electric transmission and transformation equipment on line |
| CN105182161A (en) * | 2015-09-23 | 2015-12-23 | 国网山东莒县供电公司 | Transformer monitoring system and method |
| CN105550700A (en) * | 2015-12-08 | 2016-05-04 | 国网山东省电力公司电力科学研究院 | Time series data cleaning method based on correlation analysis and principal component analysis |
| CN105595990A (en) * | 2016-01-27 | 2016-05-25 | 浙江铭众科技有限公司 | Intelligent terminal device for evaluating and distinguishing quality of electrocardiosignal |
| CN107133118A (en) * | 2016-02-26 | 2017-09-05 | 华为技术有限公司 | A kind of fault diagnosis model training method, method for diagnosing faults and relevant apparatus |
-
2017
- 2017-11-23 CN CN201711184571.5A patent/CN107862468A/en active Pending
Patent Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102567391A (en) * | 2010-12-20 | 2012-07-11 | 中国移动通信集团广东有限公司 | Method and device for building classification forecasting mixed model |
| CN103824092A (en) * | 2014-03-04 | 2014-05-28 | 国家电网公司 | Image classification method for monitoring state of electric transmission and transformation equipment on line |
| CN105182161A (en) * | 2015-09-23 | 2015-12-23 | 国网山东莒县供电公司 | Transformer monitoring system and method |
| CN105550700A (en) * | 2015-12-08 | 2016-05-04 | 国网山东省电力公司电力科学研究院 | Time series data cleaning method based on correlation analysis and principal component analysis |
| CN105595990A (en) * | 2016-01-27 | 2016-05-25 | 浙江铭众科技有限公司 | Intelligent terminal device for evaluating and distinguishing quality of electrocardiosignal |
| CN107133118A (en) * | 2016-02-26 | 2017-09-05 | 华为技术有限公司 | A kind of fault diagnosis model training method, method for diagnosing faults and relevant apparatus |
Non-Patent Citations (1)
| Title |
|---|
| 龚文引 等: "《智能算法在高光谱遥感数据处理中的应用》", 30 November 2014 * |
Cited By (53)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US12244599B2 (en) | 2015-01-16 | 2025-03-04 | Palo Alto Networks, Inc. | Private cloud control |
| US11552954B2 (en) | 2015-01-16 | 2023-01-10 | Palo Alto Networks, Inc. | Private cloud control |
| US12289329B2 (en) | 2015-04-07 | 2025-04-29 | Palo Alto Networks, Inc. | Packet analysis based IOT management |
| US12399999B2 (en) | 2016-11-21 | 2025-08-26 | Palo Alto Networks, Inc. | IoT device risk assessment |
| US11681812B2 (en) | 2016-11-21 | 2023-06-20 | Palo Alto Networks, Inc. | IoT device risk assessment |
| US11683328B2 (en) | 2017-09-27 | 2023-06-20 | Palo Alto Networks, Inc. | IoT device management visualization |
| US11671327B2 (en) | 2017-10-27 | 2023-06-06 | Palo Alto Networks, Inc. | IoT device grouping and labeling |
| US12021697B2 (en) | 2017-10-27 | 2024-06-25 | Palo Alto Networks, Inc. | IoT device grouping and labeling |
| CN108564376A (en) * | 2018-04-20 | 2018-09-21 | 阿里巴巴集团控股有限公司 | Risk control method, device, server and readable storage medium storing program for executing |
| CN112640381B (en) * | 2018-06-18 | 2024-03-08 | 帕洛阿尔托网络公司 | Methods and systems for detecting undesirable behavior of Internet of Things devices |
| US11777965B2 (en) | 2018-06-18 | 2023-10-03 | Palo Alto Networks, Inc. | Pattern match-based detection in IoT security |
| US12381902B2 (en) | 2018-06-18 | 2025-08-05 | Palo Alto Networks, Inc. | Pattern match-based detection in IOT security |
| CN112640381A (en) * | 2018-06-18 | 2021-04-09 | 帕洛阿尔托网络公司 | Pattern matching based detection in IOT security |
| CN112384924A (en) * | 2018-07-26 | 2021-02-19 | 西门子股份公司 | Method and device for establishing product performance prediction model, computer equipment, computer readable storage medium, product performance prediction method and prediction system |
| CN112384924B (en) * | 2018-07-26 | 2024-07-05 | 西门子股份公司 | Method and device for establishing product performance prediction model, computer equipment, computer readable storage medium, product performance prediction method and prediction system |
| US11940782B2 (en) | 2018-07-26 | 2024-03-26 | Siemens Aktiengesellschaft | Product performance prediction modeling to predict final product performance in case of device exception |
| CN109195154A (en) * | 2018-08-13 | 2019-01-11 | 中国联合网络通信集团有限公司 | Internet of Things alters card user recognition methods and device |
| US12294482B2 (en) | 2018-09-04 | 2025-05-06 | Palo Alto Networks, Inc. | IoT application learning |
| US12289328B2 (en) | 2018-10-15 | 2025-04-29 | Palo Alto Networks, Inc. | Multi-dimensional periodicity detection of IOT device behavior |
| CN111177585A (en) * | 2018-11-13 | 2020-05-19 | 北京四维图新科技股份有限公司 | Map POI feedback method and device |
| CN109522304A (en) * | 2018-11-23 | 2019-03-26 | 中国联合网络通信集团有限公司 | Exception object recognition methods and device, storage medium |
| CN111275453A (en) * | 2018-12-03 | 2020-06-12 | 中国移动通信集团上海有限公司 | An industry identification method and system for Internet of Things equipment |
| US11706246B2 (en) | 2018-12-12 | 2023-07-18 | Palo Alto Networks, Inc. | IOT device risk assessment and scoring |
| US11689573B2 (en) | 2018-12-31 | 2023-06-27 | Palo Alto Networks, Inc. | Multi-layered policy management |
| US12438774B2 (en) | 2018-12-31 | 2025-10-07 | Palo Alto Networks, Inc. | Multi-layered policy management |
| CN109784403B (en) * | 2019-01-16 | 2022-07-05 | 武汉斗鱼鱼乐网络科技有限公司 | A method of identifying risk equipment and related equipment |
| CN109784403A (en) * | 2019-01-16 | 2019-05-21 | 武汉斗鱼鱼乐网络科技有限公司 | A kind of method and relevant device identifying risk equipment |
| CN110322254B (en) * | 2019-07-04 | 2022-12-16 | 同盾控股有限公司 | Online fraud identification method, device, medium and electronic equipment |
| CN110322254A (en) * | 2019-07-04 | 2019-10-11 | 同盾控股有限公司 | Online fraud recognition methods, device, medium and electronic equipment |
| CN110958305A (en) * | 2019-11-15 | 2020-04-03 | 锐捷网络股份有限公司 | Method and device for identifying terminal equipment of Internet of things |
| CN111461892B (en) * | 2020-03-31 | 2021-07-06 | 支付宝(杭州)信息技术有限公司 | Derivative variable selection method and apparatus for risk identification model |
| CN111461892A (en) * | 2020-03-31 | 2020-07-28 | 支付宝(杭州)信息技术有限公司 | Method and device for selecting derived variables of risk identification model |
| CN111558226A (en) * | 2020-04-28 | 2020-08-21 | 腾讯科技(成都)有限公司 | Method, device, equipment and storage medium for detecting abnormal operation behaviors |
| CN111558226B (en) * | 2020-04-28 | 2023-04-18 | 腾讯科技(成都)有限公司 | Method, device, equipment and storage medium for detecting abnormal operation behaviors |
| CN111652284B (en) * | 2020-05-09 | 2025-01-10 | 杭州数梦工场科技有限公司 | Scanner identification method and device, electronic device, and storage medium |
| CN111652284A (en) * | 2020-05-09 | 2020-09-11 | 杭州数梦工场科技有限公司 | Scanner identification method and device, electronic device, storage medium |
| US11722875B2 (en) | 2020-06-01 | 2023-08-08 | Palo Alto Networks, Inc. | IoT device discovery and identification |
| US12302451B2 (en) | 2020-06-01 | 2025-05-13 | Palo Alto Networks, Inc. | IoT security policy on a firewall |
| CN112436969A (en) * | 2020-11-24 | 2021-03-02 | 成都西加云杉科技有限公司 | Internet of things equipment management method, system, equipment and medium |
| CN112766398B (en) * | 2021-01-27 | 2022-09-16 | 无锡中车时代智能装备研究院有限公司 | Generator rotor vent hole identification method and device |
| CN112766398A (en) * | 2021-01-27 | 2021-05-07 | 无锡中车时代智能装备有限公司 | Generator rotor vent hole identification method and device |
| CN115130713A (en) * | 2021-03-29 | 2022-09-30 | 淘宝（中国）软件有限公司 | Risk data identification, object detection and model optimization method, device and system |
| CN113239969A (en) * | 2021-04-15 | 2021-08-10 | 联合汽车电子有限公司 | Feature model establishing method, equipment fingerprint generating method and equipment identification method |
| CN113239969B (en) * | 2021-04-15 | 2024-04-30 | 联合汽车电子有限公司 | Feature model establishment method, equipment fingerprint generation method and equipment identification method |
| CN113642805A (en) * | 2021-08-27 | 2021-11-12 | Oppo广东移动通信有限公司 | Algorithm optimization method of Internet of things equipment, electronic equipment and readable storage medium |
| CN113887609A (en) * | 2021-09-28 | 2022-01-04 | 广州绿怡信息科技有限公司 | Device screen aging detection model training method and device screen aging detection method |
| CN113901996A (en) * | 2021-09-29 | 2022-01-07 | 广州绿怡信息科技有限公司 | Device screen perspective detection model training method and device screen perspective detection method |
| US11552975B1 (en) | 2021-10-26 | 2023-01-10 | Palo Alto Networks, Inc. | IoT device identification with packet flow behavior machine learning model |
| US12255906B2 (en) | 2021-10-26 | 2025-03-18 | Palo Alto Networks, Inc. | IoT device identification with packet flow behavior machine learning model |
| CN114298204A (en) * | 2021-12-24 | 2022-04-08 | 广州绿怡信息科技有限公司 | Device screen scratch detection model training method and device screen scratch detection method |
| US12301600B2 (en) | 2022-01-18 | 2025-05-13 | Palo Alto Networks, Inc. | IoT device identification by machine learning with time series behavioral and statistical features |
| CN114705444A (en) * | 2022-03-29 | 2022-07-05 | 华北电力科学研究院有限责任公司 | Method and device for monitoring vibration of steam turbine |
| CN115018656A (en) * | 2022-08-08 | 2022-09-06 | 太平金融科技服务(上海)有限公司深圳分公司 | Risk identification method, and training method, device and equipment of risk identification model |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN107862468A (en) | The method and device that equipment Risk identification model is established | |
| CN107992888A (en) | The recognition methods of operation of industrial installation and server | |
| CN104732276B (en) | One kind metering production facility on-line fault diagnosis method | |
| CN111241154A (en) | Storage battery fault early warning method and system based on big data | |
| CN102923538A (en) | Elevator health management and maintenance system based on Internet of things and collection and assessment method | |
| CN107358388A (en) | A kind of WMS based on Internet of Things and the storage quality risk appraisal procedure based on the system | |
| CN110189070B (en) | Intelligent logistics Internet of things system | |
| CN114297935A (en) | Airport terminal building departure optimization operation simulation system and method based on digital twin | |
| CN118396510A (en) | A logistics solution intelligent design system and method | |
| CN106772205A (en) | A kind of automatic power-measuring system terminal unit exception monitoring method and device | |
| CN109905885A (en) | A kind of method and inspection device of determining inspection station list | |
| CN113516244A (en) | Intelligent operation and maintenance method and device, electronic equipment and storage medium | |
| CN110458351A (en) | Area management method, device, equipment and readable storage medium storing program for executing based on flow of the people | |
| CN117391464A (en) | Project progress prediction method and system based on RBF neural network | |
| CN111882074A (en) | Data preprocessing system, method, computer device and readable storage medium | |
| CN113807462A (en) | AI-based network equipment fault reason positioning method and system | |
| CN110378586B (en) | Power transformation equipment defect early warning method and system based on dynamic closed-loop knowledge management | |
| CN118536910A (en) | Retired power battery intelligent storage method and device, intelligent storage system and storage medium | |
| CN114279554A (en) | Multi-place synchronous self-adaptive performance testing method and system of low-temperature flutter sensor | |
| CN114118678A (en) | Iron works management system based on edge internet of things and architecture method thereof | |
| CN105447518B (en) | One kind being based on K-means telemetry interpreting system | |
| CN120725344A (en) | Smart factory management methods, systems, equipment and media based on the Industrial Internet of Things | |
| CN119961821A (en) | Vehicle fault handling method, device, electronic equipment and storage medium | |
| CN111062827B (en) | Engineering supervision method based on artificial intelligence mode | |
| CN117684243B (en) | Intelligent electroplating control system and control method |
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 | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20180330 |
|
| RJ01 | Rejection of invention patent application after publication |