CN110097259A - A kind of concentrating type electrical safety hidden danger pre-judging method based on artificial neural network - Google Patents
A kind of concentrating type electrical safety hidden danger pre-judging method based on artificial neural network Download PDFInfo
- Publication number
- CN110097259A CN110097259A CN201910299825.0A CN201910299825A CN110097259A CN 110097259 A CN110097259 A CN 110097259A CN 201910299825 A CN201910299825 A CN 201910299825A CN 110097259 A CN110097259 A CN 110097259A
- Authority
- CN
- China
- Prior art keywords
- neural network
- artificial neural
- output
- electrical safety
- data
- 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
- 238000013528 artificial neural network Methods 0.000 title claims abstract description 61
- 238000000034 method Methods 0.000 title claims abstract description 26
- 238000012544 monitoring process Methods 0.000 claims abstract description 48
- 210000002569 neuron Anatomy 0.000 claims abstract description 44
- 238000012549 training Methods 0.000 claims abstract description 39
- 230000006870 function Effects 0.000 claims description 17
- 238000001914 filtration Methods 0.000 claims description 9
- 238000007781 pre-processing Methods 0.000 claims description 9
- 230000004913 activation Effects 0.000 claims description 3
- 230000001788 irregular Effects 0.000 claims description 3
- 230000000694 effects Effects 0.000 abstract description 5
- 230000002159 abnormal effect Effects 0.000 description 7
- 238000004364 calculation method Methods 0.000 description 2
- 125000004122 cyclic group Chemical group 0.000 description 2
- 238000004519 manufacturing process Methods 0.000 description 2
- 238000012545 processing Methods 0.000 description 2
- 238000012937 correction Methods 0.000 description 1
- 230000005611 electricity Effects 0.000 description 1
- 230000001537 neural effect Effects 0.000 description 1
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/04—Architecture, e.g. interconnection topology
- G06N3/045—Combinations of networks
-
- 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
Landscapes
- Business, Economics & Management (AREA)
- Engineering & Computer Science (AREA)
- Human Resources & Organizations (AREA)
- Theoretical Computer Science (AREA)
- Economics (AREA)
- Physics & Mathematics (AREA)
- Health & Medical Sciences (AREA)
- General Physics & Mathematics (AREA)
- Strategic Management (AREA)
- Entrepreneurship & Innovation (AREA)
- General Health & Medical Sciences (AREA)
- Marketing (AREA)
- Tourism & Hospitality (AREA)
- General Business, Economics & Management (AREA)
- Development Economics (AREA)
- Biophysics (AREA)
- Water Supply & Treatment (AREA)
- Educational Administration (AREA)
- Public Health (AREA)
- Game Theory and Decision Science (AREA)
- Operations Research (AREA)
- Quality & Reliability (AREA)
- Life Sciences & Earth Sciences (AREA)
- Artificial Intelligence (AREA)
- Biomedical Technology (AREA)
- Primary Health Care (AREA)
- Computational Linguistics (AREA)
- Data Mining & Analysis (AREA)
- Evolutionary Computation (AREA)
- Molecular Biology (AREA)
- Computing Systems (AREA)
- General Engineering & Computer Science (AREA)
- Mathematical Physics (AREA)
- Software Systems (AREA)
- Emergency Alarm Devices (AREA)
Abstract
The present invention relates to a kind of concentrating type electrical safety hidden danger pre-judging method based on artificial neural network, the monitoring data of electrical safety monitoring point are collected as training data, pretreatment obtains sample training data, using electrical safety monitoring point as the neuron of artificial neural network, artificial neural network is inputted with pretreated sample training data, obtain the reality output of desired output vector Yu current manual's neural network, error function is calculated to the partial derivative of each neuron of output layer, connection weight and threshold value are corrected using the output of partial derivative and input layer, judge whether to continue iteration with global error；With adopting data judges whether there is electrical safety hidden danger in fact after the completion of training.The present invention is based on the relevance of existing power grid; the early warning effect for realizing electrical safety under concentrating type monitoring mode increases with sample size and self study time, and prejudging result and early warning will be more acurrate effective; electrical safety hidden danger, effective protection the people's lives and property safety are found early.
Description
Technical Field
The present invention relates to computing; calculating; the technical field of counting, in particular to a cluster type electric potential safety hazard prejudging method based on an artificial neural network.
Background
Because the electric energy application is extensive, the electric safety also has the universality, no matter in the production field, still the life field, can not leave the electricity, can all meet various electric safety problems, for the safety of better guarantee people's production, life, early warning system is applied to in a large amount of power consumption occasions, through reporting to the police in order to guarantee electric safety.
However, most of the methods related to electrical safety precaution in the existing market use "false early warning" in a sensor monitoring manner, parameters monitored in this manner include current, voltage, residual current, temperature, etc., more professional parameters also include fault arc, active power, reactive power, fundamental power factor, harmonic distortion rate, etc., but the working mode of this type of monitoring generally sends out alarm information when abnormal data is monitored, so as to achieve the so-called "early warning" effect, but in fact, real early warning should give out a warning before an abnormal condition occurs, otherwise, time for processing electrical safety fault cannot be won, and a large loss may be formed.
Disclosure of Invention
The invention solves the problem that the real early warning effect cannot be achieved because the electric safety monitoring starts to feed back information and warn only when the abnormity is found in the prior art, and provides an optimized cluster type electric safety hidden danger pre-judging method based on an artificial neural network.
The invention adopts the technical scheme that a cluster type electric potential safety hazard prejudging method based on an artificial neural network comprises the following steps:
step 1: collecting monitoring data of the electrical safety monitoring points as training data;
step 2: preprocessing the training data to obtain k sample training data;
and step 3: setting an error function, an error precision epsilon and a maximum iteration number as M, and enabling M to be 0; initializing a connection weight value omega and a threshold value h of a neuron, wherein omega is rand (0,1), and h is rand (0, 1);
and 4, step 4: taking an electrical safety monitoring point as a neuron of an artificial neural network, and training data of any preprocessed sampleInputting an artificial neural network, wherein i is the serial number of the sample training data, and i is more than or equal to 1 and less than or equal to k;
and 5: obtaining sample training dataExpected output vector after input into artificial neural networkActual output z from the current artificial neural networkoutput(i) Calculating partial derivative delta of error function to each neuron of output layero(i)；
Step 6: using partial derivatives delta of neurons in the output layero(i) And output x of input layer neuronsoi(i) Correcting a connection weight omega and a threshold h corresponding to any neuron;
and 7: calculating a global error E, if E is less than epsilon or M +1 is equal to M, carrying out the next step, otherwise, if M is equal to M +1, and returning to the step 4;
and 8: and (4) taking the electrical safety monitoring points as the neurons of the trained artificial neural network, acquiring monitoring data by the electrical safety monitoring points in real time, alarming if the output result of the artificial neural network shows that the electrical safety hidden danger exists, and otherwise, repeating the step 8.
Preferably, in step 1, the monitoring data includes corresponding key data sets a [ n ]]And an auxiliary data set b [ n ]]The training dataWherein, a [ n ]]＝[a1，a2，a3，...an]，b[n]＝[b1，b2，b3，...bn]C is the geographical position information carried by each monitoring point, and n is a positive integer.
Preferably, the critical data includes one or more of cable temperature, residual current and fault arc, and the auxiliary data includes one or more of voltage, frequency, waveform distortion rate and power factor.
Preferably, in the step 2, the preprocessing includes filtering irregular data in the key data set a [ n ] and filtering discrete values in the monitoring data; and (4) inheriting the parent level position information of the electrical safety monitoring point with the c being 0, updating the c, and filtering if the c is 0 and no parent level position information exists.
Preferably, in the step 4, the artificial neural network includes an input layer and an output layer, and the sample training dataThe input is from the input layer, and the output is from the output layer after passing through the artificial neural network.
Preferably, the neurons in the artificial neural network are electrical safety monitoring points, and the vector of any neuron comprises an input vector zinput(i) And an output vector zoutput(i)， Where f is the activation function.
Preferably, in the step 5, the error function isPartial derivative delta of error function to each neuron of output layero(i)＝(d(i)-zoutput(i))zoutput(i)(1-zoutput(i))。
Preferably, in step 6, the connection weight ω' ═ ω + η δ corresponding to any neuron is correctedo(i)zoutput(i) The threshold h' corresponding to any neuron is corrected to h + η δo(i) Wherein η represents a learning rate, η ∈ (0, 1).
Preferably, in the step 7,
preferably, in said step 7, ε ∈ [0.3,0.8 ].
The invention provides an optimized cluster type electric safety hidden danger prejudging method based on an artificial neural network, which comprises the steps of collecting monitoring data of electric safety monitoring points as training data and preprocessing the training data to obtain k sample training data, using the electric safety monitoring points as neurons of the artificial neural network, inputting any preprocessed sample training data into the artificial neural network to obtain an expected output vector and actual output of the current artificial neural network, calculating a partial derivative of an error function on each neuron of an output layer, correcting a connection weight and a threshold value by using the partial derivative and the output of the neurons of the input layer, judging whether to continue iteration by using a global error, and using the maximum iteration number as the termination of the iteration; and after the artificial neural network training is finished, the real data is used for operation, and whether the electrical potential safety hazard exists at that time is judged.
The invention solves the problem that the abnormal data is monitored and then becomes more serious accidents without adopting an automatic intervention mode in the prior art, simultaneously reduces the cost investment, realizes the early warning effect of the electrical safety in a cluster monitoring mode on the basis of the relevance of the existing power grid, and can more accurately and effectively predict the electrical safety hidden danger and discover the electrical safety hidden danger as soon as possible along with the increase of the sample volume and the increase of the self-learning time, thereby effectively protecting the life and property safety of people.
Detailed Description
The present invention is described in further detail with reference to the following examples, but the scope of the present invention is not limited thereto.
The invention relates to a cluster type electric safety hidden danger prejudging method based on an artificial neural network.
In the present invention, these data parameters are divided into electrical safety parameters and electrical quality parameters. Anomalies in electrical safety parameters often indicate that a safety hazard has occurred at the monitoring point, which can directly lead to safety problems, such as: the cable temperature, the residual current, the fault arc and the like can be regarded as potential hazard index parameters of electrical safety, and the fractional weight can be given according to the importance degree of the parameters; the electrical quality parameters mainly acquire data within a certain time to obtain the change characteristics of the data, and analyze the quality of the power supply quality of the power grid according to the change factors to prevent some electrical safety parameters from being abnormal, wherein the electrical quality parameters comprise voltage, frequency, waveform distortion rate, power factor and the like.
In the invention, two types of parameters with electrical safety parameters as main parameters and electrical quality parameters as auxiliary parameters are used as 'forward propagation input quantity' of an artificial neural network pre-judging algorithm, feedback of a pre-judging result is used as 'error reverse propagation quantity' of the neural network pre-judging algorithm, and the aim of continuously correcting the pre-judging precision and giving an accurate pre-judging result of the electrical safety hidden danger is achieved by utilizing the self-learning characteristic of the neural network, and the result is embodied in a safety value (namely an electrical safety score value of 1-100, and 100 is optimal), so that a user can make early warning work according to the difference of the safety values of each region.
The method comprises the following steps.
Step 1: and collecting monitoring data of the electrical safety monitoring points as training data.
In step 1, the monitoring data includes a corresponding key data set a [ n ]]And an auxiliary data set b [ n ]]The training dataWherein, a [ n ]]＝[a1，a2，a3，...an]，b[n]＝[b1，b2，b3，...bn]C is the geographical position information carried by each monitoring point, and n is a positive integer.
The critical data includes one or more of cable temperature, residual current and fault arc, and the auxiliary data includes one or more of voltage, frequency, waveform distortion rate and power factor.
In the present invention, critical data includes, but is not limited to, cable temperature, residual current, and fault arcing; the auxiliary data includes, but is not limited to, voltage, frequency, waveform distortion rate, and power factor.
Step 2: and preprocessing the training data to obtain k sample training data.
In the step 2, the preprocessing comprises filtering irregular data in the key data set a [ n ] and filtering discrete values in the monitoring data; and (4) inheriting the parent level position information of the electrical safety monitoring point with the c being 0, updating the c, and filtering if the c is 0 and no parent level position information exists.
In the invention, the amount of the collected original data is very large, and the original data is often provided with a small amount of abnormal data, but even a small amount of abnormal data can bring great interference to the calculation result, so that preprocessing is required to be executed.
In the invention, the parameter of the key data set a [ n ] has a large referential meaning, so that the unconventional data needs to be strictly filtered, for example, when the sensor is damaged to cause data exception, the returned value is often 65535 of the maximum byte of the data, and the unconventional data needs to be filtered.
In the invention, in the process of multiple acquisition, if a numerical value with a great change amplitude value appears, a discrete point is formed and is filtered; the skilled person can set the threshold value for this amplitude value according to the actual requirements, such as the type of parameter.
In the invention, the geographical position information can be manually set, and if a certain building or point location position information is empty, the current geographical position information inherits the parent-level position information, thereby ensuring that the current geographical position information is not empty.
And step 3: setting an error function, an error precision epsilon and a maximum iteration number as M, and enabling M to be 0; initializing a connection weight ω and a neuron threshold h, ω ═ rand (0,1) and h ═ rand (0, 1).
And 4, step 4: taking an electrical safety monitoring point as a neuron of an artificial neural network, and training data of any preprocessed sampleAnd (4) inputting the artificial neural network, wherein i is the serial number of the sample training data, and i is more than or equal to 1 and less than or equal to k.
In the step 4, the artificial neural network comprises an input layer and an output layer, and the sample training dataInput from input layer, output layer after passing through artificial neural networkAnd (6) outputting.
The neurons in the artificial neural network are electrical safety monitoring points, and the vector of any neuron comprises an input vector zinput(i) And an output vector zoutput(i)， Where f is the activation function.
In the present invention, the network input to each neuron isWherein,and In is the network input of each most original neuron, which is the corresponding connection weight of the neuron.
In the invention, the artificial neural network in general conditions contains hidden layers which can be regarded as special input layers, so that only the input layer and the output layer are defaulted in the invention, and in practical application, a person skilled in the art can set the number of the hidden layers according to requirements and carry out cyclic calculation and cyclic deviation correction according to the number of the hidden layers.
And 5: obtaining sample training dataExpected output vector after input into artificial neural networkActual output z from the current artificial neural networkoutput(i) Calculating partial derivative delta of error function to each neuron of output layero(i)。
In said step 5, the error function isPartial derivative delta of error function to each neuron of output layero(i)＝(d(i)-zoutput(i))zoutput(i)(1-zoutput(i))。
Step 6: using partial derivatives delta of neurons in the output layero(i) And output x of input layer neuronsoi(i) And correcting the connection weight omega and the threshold h corresponding to any neuron.
In step 6, the connection weight ω' ═ ω + η δ corresponding to any neuron is correctedo(i)zoutput(i) The threshold h' corresponding to any neuron is corrected to h + η δo(i) Wherein η represents a learning rate, η ∈ (0, 1).
And 7: and calculating a global error E, if E is less than epsilon or M +1 is equal to M, carrying out the next step, otherwise, if M is equal to M +1, and returning to the step 4.
In the step 7, the process is carried out,
in said step 7,. epsilon. [0.3,0.8 ].
In the invention, after the pre-judgment result is output each time, personnel can confirm and investigate the real situation according to the situation and process the real situation, so that a real value can be obtained, and an error amount epsilon can be formed between the real value and the pre-judgment result each time.
In the present invention, in general, ∈ is 0.6.
And 8: and (4) taking the electrical safety monitoring points as the neurons of the trained artificial neural network, acquiring monitoring data by the electrical safety monitoring points in real time, alarming if the output result of the artificial neural network shows that the electrical safety hidden danger exists, and otherwise, repeating the step 8.
In the invention, after the training of the artificial neural network is finished, real-time acquisition data is input in real time, the artificial neural network is used for operation, the output value is monitored, actually, the output value is compared with an expected output value, and if the output value has deviation, the existence of electrical potential safety hazard can be judged, and alarm processing is carried out.
The method comprises the steps of collecting monitoring data of electrical safety monitoring points as training data and preprocessing the training data to obtain k sample training data, using the electrical safety monitoring points as neurons of an artificial neural network, inputting any preprocessed sample training data into the artificial neural network to obtain an expected output vector and actual output of the current artificial neural network, calculating partial derivatives of error functions on the neurons of an output layer, correcting connection weights and thresholds by using the partial derivatives and the output of the neurons of the input layer, judging whether iteration is continued by using global errors, and using the maximum iteration number as the termination of the iteration; and after the artificial neural network training is finished, the real data is used for operation, and whether the electrical potential safety hazard exists at that time is judged.
The invention solves the problem that the abnormal data is monitored and then becomes more serious accidents without adopting an automatic intervention mode in the prior art, simultaneously reduces the cost investment, realizes the early warning effect of the electrical safety in a cluster monitoring mode on the basis of the relevance of the existing power grid, and can more accurately and effectively predict the electrical safety hidden danger and discover the electrical safety hidden danger as soon as possible along with the increase of the sample volume and the increase of the self-learning time, thereby effectively protecting the life and property safety of people.
Claims (10)
1. A cluster type electric potential safety hazard pre-judging method based on an artificial neural network is characterized by comprising the following steps: the method comprises the following steps:
step 1: collecting monitoring data of the electrical safety monitoring points as training data;
step 2: preprocessing the training data to obtain k sample training data;
and step 3: setting an error function, an error precision epsilon and a maximum iteration number as M, and enabling M to be 0;
initializing a connection weight ω and a neuron threshold h, ω ═ rand (0,1),
h＝rand(0,1)；
and 4, step 4: taking an electrical safety monitoring point as a neuron of an artificial neural network, and training data of any preprocessed sampleInputting an artificial neural network, wherein i is the serial number of the sample training data, and i is more than or equal to 1 and less than or equal to k;
and 5: obtaining sample training dataExpected output vector after input into artificial neural networkActual output z from the current artificial neural networkoutput(i) Calculating partial derivative delta of error function to each neuron of output layero(i)；
Step 6: using partial derivatives delta of neurons in the output layero(i) And output x of input layer neuronsoi(i) Correcting a connection weight omega and a threshold h corresponding to any neuron;
and 7: calculating a global error E, if E is less than epsilon or M +1 is equal to M, carrying out the next step, otherwise, if M is equal to M +1, and returning to the step 4;
and 8: and (4) taking the electrical safety monitoring points as the neurons of the trained artificial neural network, acquiring monitoring data by the electrical safety monitoring points in real time, alarming if the output result of the artificial neural network shows that the electrical safety hidden danger exists, and otherwise, repeating the step 8.
2. The cluster type electric potential safety hazard prejudging method based on the artificial neural network as claimed in claim 1, characterized in that: in step 1, the monitoring data includes a corresponding key data set a [ n ]]And an auxiliary data set b [ n ]]The training data Wherein, a [ n ]]＝[a1,a2,a3,…an]，b[n]＝[b1,b2,b3,...bn]C is the geographical position information carried by each monitoring point, and n is a positive integer.
3. The cluster type electric potential safety hazard pre-judging method based on the artificial neural network as claimed in claim 2, characterized in that: the critical data includes one or more of cable temperature, residual current and fault arc, and the auxiliary data includes one or more of voltage, frequency, waveform distortion rate and power factor.
4. The cluster type electric potential safety hazard prejudging method based on the artificial neural network as claimed in claim 1, characterized in that: in the step 2, the preprocessing comprises filtering irregular data in the key data set a [ n ] and filtering discrete values in the monitoring data; and (4) inheriting the parent level position information of the electrical safety monitoring point with the c being 0, updating the c, and filtering if the c is 0 and no parent level position information exists.
5. The cluster type electric potential safety hazard prejudging method based on the artificial neural network as claimed in claim 1, characterized in that: in the step 4, the artificial neural network comprises an input layer and an output layer, and the sample training dataThe input is from the input layer, and the output is from the output layer after passing through the artificial neural network.
6. The cluster type electric potential safety hazard pre-judging method based on the artificial neural network as claimed in claim 5, wherein: the neurons in the artificial neural network are electrically safeMonitor points, the vector for any neuron comprising an input vector zinput(i) And an output vector zoutput(i)，Where f is the activation function.
7. The cluster type electric potential safety hazard prejudging method based on the artificial neural network as claimed in claim 1, characterized in that: in said step 5, the error function isPartial derivative delta of error function to each neuron of output layero(i)＝(d(i)-zoutput(i))zoutput(i)(1-zoutput(i))。
8. The method for predicting the electrical safety hazards in a cluster based on the artificial neural network as claimed in claim 7, wherein in step 6, the connection weight ω' ═ ω + η δ corresponding to any neuron is correctedo(i)zoutput(i) The threshold h' corresponding to any neuron is corrected to h + η δo(i) Wherein η represents a learning rate, η ∈ (0, 1).
9. The cluster type electric potential safety hazard prejudging method based on the artificial neural network as claimed in claim 1, characterized in that: in the step 7, the process is carried out,
10. the cluster type electric potential safety hazard prejudging method based on the artificial neural network as claimed in claim 1, characterized in that: in said step 7,. epsilon. [0.3,0.8 ].
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910299825.0A CN110097259A (en) | 2019-04-15 | 2019-04-15 | A kind of concentrating type electrical safety hidden danger pre-judging method based on artificial neural network |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910299825.0A CN110097259A (en) | 2019-04-15 | 2019-04-15 | A kind of concentrating type electrical safety hidden danger pre-judging method based on artificial neural network |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN110097259A true CN110097259A (en) | 2019-08-06 |
Family
ID=67444887
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201910299825.0A Pending CN110097259A (en) | 2019-04-15 | 2019-04-15 | A kind of concentrating type electrical safety hidden danger pre-judging method based on artificial neural network |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN110097259A (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN115659243A (en) * | 2022-12-22 | 2023-01-31 | 四川九通智路科技有限公司 | Infrastructure risk monitoring method and monitoring system based on MEMS |
Citations (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104200266A (en) * | 2014-08-22 | 2014-12-10 | 浙江工业大学 | Drying room heating system fault diagnosis method based on extension neural network |
| CN104951836A (en) * | 2014-03-25 | 2015-09-30 | 上海市玻森数据科技有限公司 | Posting predication system based on nerual network technique |
| CN107655690A (en) * | 2017-10-30 | 2018-02-02 | 王才旺 | Motor bearings method for diagnosing faults based on artificial neural network |
| CN108427986A (en) * | 2018-02-26 | 2018-08-21 | 中车青岛四方机车车辆股份有限公司 | A kind of production line electrical fault prediction technique and device |
| CN108562835A (en) * | 2018-03-19 | 2018-09-21 | 杭州拓深科技有限公司 | A kind of fault arc detection method based on BP neural network |
| CN108761377A (en) * | 2018-05-28 | 2018-11-06 | 国网福建省电力有限公司电力科学研究院 | A kind of electric energy metering device method for detecting abnormality based on long memory models in short-term |
| CN108892014A (en) * | 2018-09-19 | 2018-11-27 | 歌拉瑞电梯股份有限公司 | A kind of elevator internal contracting brake fault early warning method based on Elman neural network |
| CN109615070A (en) * | 2018-12-06 | 2019-04-12 | 浙江巨磁智能技术有限公司 | Electric power artificial intelligence chip and power failure recognition methods |
-
2019
- 2019-04-15 CN CN201910299825.0A patent/CN110097259A/en active Pending
Patent Citations (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104951836A (en) * | 2014-03-25 | 2015-09-30 | 上海市玻森数据科技有限公司 | Posting predication system based on nerual network technique |
| CN104200266A (en) * | 2014-08-22 | 2014-12-10 | 浙江工业大学 | Drying room heating system fault diagnosis method based on extension neural network |
| CN107655690A (en) * | 2017-10-30 | 2018-02-02 | 王才旺 | Motor bearings method for diagnosing faults based on artificial neural network |
| CN108427986A (en) * | 2018-02-26 | 2018-08-21 | 中车青岛四方机车车辆股份有限公司 | A kind of production line electrical fault prediction technique and device |
| CN108562835A (en) * | 2018-03-19 | 2018-09-21 | 杭州拓深科技有限公司 | A kind of fault arc detection method based on BP neural network |
| CN108761377A (en) * | 2018-05-28 | 2018-11-06 | 国网福建省电力有限公司电力科学研究院 | A kind of electric energy metering device method for detecting abnormality based on long memory models in short-term |
| CN108892014A (en) * | 2018-09-19 | 2018-11-27 | 歌拉瑞电梯股份有限公司 | A kind of elevator internal contracting brake fault early warning method based on Elman neural network |
| CN109615070A (en) * | 2018-12-06 | 2019-04-12 | 浙江巨磁智能技术有限公司 | Electric power artificial intelligence chip and power failure recognition methods |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN115659243A (en) * | 2022-12-22 | 2023-01-31 | 四川九通智路科技有限公司 | Infrastructure risk monitoring method and monitoring system based on MEMS |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN109359894B (en) | RPN-based risk evaluation method and device for electric power metering equipment | |
| CN111523785A (en) | Power system dynamic security assessment method based on generation countermeasure network | |
| CN107330573B (en) | State evaluation method and device for photovoltaic system key equipment | |
| CN105893757B (en) | Comprehensive evaluation method for severity of voltage sag | |
| CN108847686B (en) | Photovoltaic inverter fault prediction method | |
| CN113887846B (en) | Out-of-tolerance risk early warning method for capacitive voltage transformer | |
| CN106768262B (en) | A kind of transformer online monitoring method based on surface vibration signals analysis | |
| CN109861206B (en) | Transient power angle stability fault screening system and method based on support vector machine | |
| CN103699668B (en) | Electric distribution network electrical equipment assembled state appraisal procedure based on data section uniformity | |
| CN111582542B (en) | Power load prediction method and system based on anomaly repair | |
| CN113935440A (en) | Iterative evaluation method and system for error state of voltage transformer | |
| CN114369849A (en) | Method and device for monitoring health degree of electrolytic cell and electrolytic cell monitoring system | |
| CN111046327A (en) | Prony analysis method suitable for low-frequency oscillation and subsynchronous oscillation identification | |
| CN113533910A (en) | Method and system suitable for converter transformer partial discharge early warning | |
| CN110097259A (en) | A kind of concentrating type electrical safety hidden danger pre-judging method based on artificial neural network | |
| CN112649642A (en) | Electricity stealing position judging method, device, equipment and storage medium | |
| CN115081597A (en) | Method, system, equipment and medium for performing carbon check by using electricity based on machine learning | |
| CN112436542A (en) | Steady-state safety emergency control online pre-decision method considering stability control strategy | |
| CN102095953B (en) | A kind of performance of accumulator charger online test method | |
| CN104617575A (en) | Method for using static security analysis to assess flow insolubility of anticipated faults | |
| CN106325258B (en) | Relay protection device state evaluation method based on online monitoring information | |
| CN103928923B (en) | A kind of network stationary power quality method for early warning based on sensitivity analysis | |
| CN117334032A (en) | Multi-feature fusion photovoltaic fire multistage early warning system and method | |
| CN112182960A (en) | Power transformer state risk assessment method based on Bayesian network | |
| CN108844662B (en) | Method for evaluating state of electrical cabinet of numerical control machine tool |
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 | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20190806 |