CN108564193A - The prediction technique of electric power personal injury, misoperation fault and the accident of operation against rules against regulations based on neural network - Google Patents
The prediction technique of electric power personal injury, misoperation fault and the accident of operation against rules against regulations based on neural network Download PDFInfo
- Publication number
- CN108564193A CN108564193A CN201810012435.6A CN201810012435A CN108564193A CN 108564193 A CN108564193 A CN 108564193A CN 201810012435 A CN201810012435 A CN 201810012435A CN 108564193 A CN108564193 A CN 108564193A
- Authority
- CN
- China
- Prior art keywords
- neural network
- electric power
- personal injury
- accident
- following
- 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; 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
- G06Q10/04—Forecasting or optimisation specially adapted for administrative or management purposes, e.g. linear programming or "cutting stock problem"
-
- 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
- G06N3/08—Learning methods
- G06N3/084—Backpropagation, e.g. using gradient descent
-
- 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—Systems or methods specially adapted for specific business sectors, e.g. utilities or tourism
- G06Q50/06—Electricity, gas or water supply
-
- 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—Systems or methods specially adapted for specific business sectors, e.g. utilities or tourism
- G06Q50/10—Services
- G06Q50/26—Government or public services
- G06Q50/265—Personal security, identity or safety
Abstract
The present invention relates to a kind of electric power personal injury, misoperation fault and the prediction technique for operating accident against regulations based on neural network, include the following steps：S1. various electric power personal injuries, misoperation fault, the historical data for operating accident against regulations are acquired；S2. neural network is built, neural network is trained using the historical data of various electric power personal injuries, misoperation fault, operation against rules accident；S3. test data is input in trained neural network, obtains prediction result.
Description
Technical field
The present invention relates to electric power safety management and control technical fields, more particularly, to the electric power person thing based on neural network
Therefore the prediction technique of misoperation fault and the accident of operation against rules against regulations.
Background technology
The development of big data is to rise in this world, and also becomes the hot issue of current information industry, and so-called big data counts
It is huge to can not be counted by main software tool at this stage according to measuring, and reach acquisition within reasonable time, management,
It handles and arranges as the more valuable more effective information of enterprise management decision-making is helped.
From the point of view of the data kept the safety in production in recent years from electric power enterprise, human casualty accident number and death toll are in rise year by year
Trend, the serious accidents such as personal injury, especially big and substantial equipment accident also happen occasionally in power generation, in people because of secure side
Still the situation is tense in face.Meanwhile problem violating the regulations is also always a weight difficult point of electric power enterprise production safety management.Correlative study
Show that act of violating regulations can cause the effect of the subjective sensation of personal injury, act of violating regulations can be by the effect that regulation is punished, violating the regulations
Behavior meets the difference etc. of Physiological Psychology needs effect with Zhang Hangwei is abided by；Idea of leaving things to chance, careless psychology, saves the energy heart at group psychology
Reason, antagonistic psychology, careless and sloppy psychology etc. are also to generate the reason of operating against regulations.In addition, electric misoperation problem is constantly subjected to height
Pay attention to.The reason of operator maloperation behavior includes human factor, factor of natural environment, apparatus factor, human-equation error
And cognition and technical ability factor, environmental factor, physiology/psychology/nature factor etc..
However electric power personal injury, maloperation and research method violating the regulations and its achievement are mostly close at present, mostly with qualitative
Based on analysis.Such as：Relevance presenting levels are formed by each target factor according to the size of importance using analytic hierarchy process (AHP), the row of foundation
Sequence judgment matrix, and as the foundation of decision, there are consistency and it is computationally intensive the problems such as.Using the level based on ambiguity function
Analytic approach can effectively solve the problem that the above problem, but be difficult to disclose the rule and feature of the electric power person and misoperation fault event.It adopts
With the analytic hierarchy process (AHP) based on triangle ambiguity function, the previous simple deficiency for using analytic hierarchy process (AHP) is improved, to a certain degree
On contribute to disclose electric power human-initiated accident event genesis mechanism, but lack basic data.
Invention content
The present invention is to solve the defect of the above prior art, provide a kind of electric power personal injury based on neural network,
The prediction technique of misoperation fault and the accident of operation against rules against regulations.
To realize the above goal of the invention, the technical solution adopted is that：
The prediction technique of electric power personal injury, misoperation fault and the accident of operation against rules against regulations based on neural network, including with
Lower step：
S1. various electric power personal injuries, misoperation fault, the historical data for operating accident against regulations are acquired；
S2. neural network is built, various electric power personal injuries, misoperation fault, the history number for operating accident against regulations are utilized
It is trained according to neural network；
S3. test data is input in trained neural network, obtains prediction result.
Compared with prior art, the beneficial effects of the invention are as follows：
The present invention utilizes historical data, the personal injury of research electric power, misoperation fault, the feature for operating accident against regulations to adopt
Probe into accident pests occurrence rule with neural network, analysis detects various accident conditions, the personal injury of research electric power, misoperation fault,
The big data analysis method of the characteristic elements such as operation against rules accident and people, machine, material, method, ring relevance and sensitivity discloses electric power
Personal injury, misoperation fault, the pests occurrence rule for operating accident against regulations specify electric power personal injury, misoperation fault, behaviour violating the regulations
Make the key message needed for crash analysis, determining leads to electric power personal injury, misoperation fault, the high risk for operating accident against regulations
Factor formulates electric power personal injury, misoperation fault, the pre-control strategy for operating accident against regulations.
Description of the drawings
Fig. 1 is the flow chart schematic diagram of method.
Fig. 2 is the schematic diagram of training neural network.
Specific implementation mode
The attached figures are only used for illustrative purposes and cannot be understood as limitating the patent；
Below in conjunction with drawings and examples, the present invention is further elaborated.
Embodiment 1
As shown in Figure 1, method provided by the invention includes following steps：
S1. various electric power personal injuries, misoperation fault, the historical data for operating accident against regulations are acquired；
S2. neural network is built, various electric power personal injuries, misoperation fault, the history number for operating accident against regulations are utilized
It is trained according to neural network；
S3. test data is input in trained neural network, obtains prediction result.
In specific implementation process, as shown in Fig. 2, described when being trained to neural network, electric power people occurs for acquisition
Power generation work, power construction work when body accident, misoperation fault, operation against rules accident, condition flag, personnel characteristics
Input value as neural network；And acquire the accident etc. when electric power personal injury, misoperation fault, operation against rules accident occurs
Grade, personal injury type, electric power personal injury grade, the output of personal injury event hierarchy, responsibility type as neural network
Value, is trained neural network by input value, output valve；
Power generation work can be any one of following：Transmission of electricity, distribution, power supply, electricity consumption, scheduling, runs, patrols power transformation
Depending on, overhaul, safeguard, test, meter reading examination, power utility check, repairing, technological transformation, industry expand；
Power construction work can be any one of following：The construction of power transmission and transformation distribution engineering, installation, debugging and management；
Condition flag can be any one of following：It gets an electric shock, bar, falling from high altitude, fire, explosion explosion；
Personnel characteristics can be any one of following：Gender, age, educational background, recruitment form, the length of service, work post, this work post work
Age, educational training, safety examination, credentials, occupation taboo；
Incident classification can be any one of following：It is especially big, great, larger, general；
Personal injury type can be any one of following：Personal dead, personal injury；
Electric power personal injury grade can be any one of following：Level-one, two level, three-level, level Four, Pyatyi；
Personal injury event hierarchy can be any one of following：500kV、220kV、110kV、35kV、10kV、380V、
220V；
Responsibility type can be any one of following：Leadership responsibility, management responsibility execute responsibility or direct liability, indirectly
Responsibility.
The detailed process being wherein trained to neural network is as follows：
1) data prediction
The data of separate sources, format, property are subjected to organic combination, are stored according to data characteristics, when necessary into
Row association storage or distributed storage.Data screening is carried out, data incomplete, inconsistent in data are removed, removes noise
Data, will be available, and the Data Integration of completion is data set.Data are converted, smoothly, standardization, the processing such as normalization are protected
Demonstrate,proving it becomes the input data format of neural network.
2) network is built
Neural network structure is built as follows：Electric power personal injury, misoperation fault, the data volume Pang for operating accident against regulations
Greatly, ready power generation is worked, power construction work, condition flag, personnel characteristics as input signal X, by pair
The analysis of these signals can predict the result type and grade of electric power accident, these states are as output signal Y, the present invention
Using 4 layers of neural network, it is 7 and 5 to hide node layer, that is, builds the electric power accident BP god of double hidden layers by network, simultaneously
It is using hyperbolic tangent function, function in view of hyperbolic tangent function has better effect, activation primitive in the problem：
F (x)=tanh (x)
3) network is initialized, weight and biasing are initialized
Weight and biasing are initialized using random value
4) forward-propagating
(4.1) input layer is calculated to hidden layer
It needs to be trained after network establishment is complete, for neural network, training is the learning process of parameter, and just
It is constituted to propagating with backpropagation, in forward-propagating, the communication process of input layer to hidden layer is shown in formula (2), by data
It is supplied to input layer one by one, and calculates the output of each nodal value of hidden layer and concealed nodes.Wherein yhkFor hidden layer neuron
Value, ωhiFor weights, xiFor input value, bhkFor biasing, f is activation primitive.
(4.2) hidden layer is calculated to output layer
The communication process of hidden layer to output layer is identical to that this is identical, and formula is as follows, and network is calculated using formula (3)
Output valve
5) error is calculated
Output layer obtains theoretical output, by it compared with reality output, calculates error and carries out backpropagation according to error,
Mean square deviation is to calculate the usual way of error.Error, wherein y are calculated using formula (4)iFor the output valve of output layer neuron, Yi
For actual value.
6) backpropagation
Backpropagation is carried out according to error, updates weights and the biasing of hidden layer, the newer formula of weight is as follows, wherein the
One modified weight item for conventional algorithm,Indicate the gradient direction of error, η is learning rate, and Section 2 is momentum term, and α is
Factor of momentum
7) learning rate self study
Learning rate self study is carried out according to formula (6) to (10), adjusts the learning rate η of network, wherein gtFor gradient, mt, vt
Respectively to the single order moments estimation of gradient and second order moments estimation,It is to mtAnd vtCorrection, β1, β2For parameter, can set respectively
For 0.9 and 0.999, ∈ may be configured as 10-8, it is 0, m that can prevent divisor0, v0Initial value is 0
mt=β1mt-1+(1-β1)gt (6)
8) factor of momentum self study
Factor of momentum self study is carried out according to formula (11), adjusts the factor of momentum α of network, wherein λ is normal number, is used for
Control the size of factor of momentum, it is clear that the value of α is between 0 and 1, and the gradient norm with error about weight vector
Change and changes.
9) iteration
Return to step (5) carries out iteration, until the error of network is less than preset value or iterations are more than
Given value
10) terminate training
Or then terminate to train to restrain when not up to presetting number.
Obviously, the above embodiment of the present invention be only to clearly illustrate example of the present invention, and not be pair
The restriction of embodiments of the present invention.For those of ordinary skill in the art, may be used also on the basis of the above description
To make other variations or changes in different ways.There is no necessity and possibility to exhaust all the enbodiments.It is all this
All any modification, equivalent and improvement etc., should be included in the claims in the present invention made by within the spirit and principle of invention
Protection domain within.
Claims (4)
1. the prediction technique of electric power personal injury, misoperation fault and the accident of operation against rules against regulations based on neural network, feature exist
In：Include the following steps：
S1. various electric power personal injuries, misoperation fault, the historical data for operating accident against regulations are acquired；
S2. neural network is built, various electric power personal injuries, misoperation fault, the historical data pair for operating accident against regulations are utilized
Neural network is trained；
S3. test data is input in trained neural network, obtains prediction result.
2. electric power personal injury, misoperation fault and the accident of operation against rules against regulations according to claim 1 based on neural network
Prediction technique, it is characterised in that：Described when being trained to neural network, electric power personal injury, maloperation thing occur for acquisition
Therefore the power generation work, power construction work, condition flag, personnel characteristics when operating accident against regulations are as neural network
Input value；And acquire incident classification when electric power personal injury, misoperation fault, operation against rules accident occurs, personal injury class
Type, electric power personal injury grade, the output valve of personal injury event hierarchy, responsibility type as neural network, by input value,
Output valve is trained neural network；
Power generation work can be any one of following：Transmission of electricity, power transformation, distribution, power supply, electricity consumption, scheduling, operation, tour, inspection
Repair, safeguard, testing, meter reading examination, power utility check, repairing, technological transformation, industry expand；
Power construction work can be any one of following：The construction of power transmission and transformation distribution engineering, installation, debugging and management；
Condition flag can be any one of following：It gets an electric shock, bar, falling from high altitude, fire, explosion explosion；
Personnel characteristics can be any one of following：Gender, the age, educational background, recruitment form, the length of service, work post, this work post length of service,
Educational training, safety examination, credentials, occupation taboo；
Incident classification can be any one of following：It is especially big, great, larger, general；
Personal injury type can be any one of following：Personal dead, personal injury；
Electric power personal injury grade can be any one of following：Level-one, two level, three-level, level Four, Pyatyi；
Personal injury event hierarchy can be any one of following：500kV、220kV、110kV、35kV、10kV、380V、220V；
Responsibility type can be any one of following：Leadership responsibility, executes responsibility or direct liability, indirect liability at management responsibility.
3. electric power personal injury, misoperation fault and the accident of operation against rules against regulations according to claim 2 based on neural network
Prediction technique, it is characterised in that：After the step S1 collects historical data, historical data is pre-processed.
4. electric power personal injury, misoperation fault and the accident of operation against rules against regulations according to claim 2 based on neural network
Prediction technique, it is characterised in that：The neural network includes input layer, two layers of hidden layer and output layer.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810012435.6A CN108564193A (en) | 2018-01-05 | 2018-01-05 | The prediction technique of electric power personal injury, misoperation fault and the accident of operation against rules against regulations based on neural network |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810012435.6A CN108564193A (en) | 2018-01-05 | 2018-01-05 | The prediction technique of electric power personal injury, misoperation fault and the accident of operation against rules against regulations based on neural network |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN108564193A true CN108564193A (en) | 2018-09-21 |
Family
ID=63530630
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201810012435.6A Pending CN108564193A (en) | 2018-01-05 | 2018-01-05 | The prediction technique of electric power personal injury, misoperation fault and the accident of operation against rules against regulations based on neural network |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN108564193A (en) |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111723970A (en) * | 2020-05-06 | 2020-09-29 | 国网浙江省电力有限公司衢州供电公司 | Power operation hidden danger prediction method |
| CN111862557A (en) * | 2019-04-24 | 2020-10-30 | 广州煜煊信息科技有限公司 | Household accident handling method based on Internet of things monitoring mode |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2014093670A1 (en) * | 2012-12-12 | 2014-06-19 | University Of North Dakota | Analyzing flight data using predictive models |
| CN104462727A (en) * | 2014-12-31 | 2015-03-25 | 中国科学院遥感与数字地球研究所 | Oil spilling simulation parameter optimization method based on dynamic remote sensing data driving |
| CN104992299A (en) * | 2015-07-23 | 2015-10-21 | 河南行知专利服务有限公司 | Power grid risk analysis and early warning method |
| KR101830295B1 (en) * | 2017-06-23 | 2018-03-29 | 주식회사 주빅스 | INTELLIGENT ACCIDENT PREVENTION TRAINING SELECTION DECISION SYSTEM OF TOXIC SUBSTANCE BUSINESS USING IoT SENSOR |
-
2018
- 2018-01-05 CN CN201810012435.6A patent/CN108564193A/en active Pending
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2014093670A1 (en) * | 2012-12-12 | 2014-06-19 | University Of North Dakota | Analyzing flight data using predictive models |
| CN104462727A (en) * | 2014-12-31 | 2015-03-25 | 中国科学院遥感与数字地球研究所 | Oil spilling simulation parameter optimization method based on dynamic remote sensing data driving |
| CN104992299A (en) * | 2015-07-23 | 2015-10-21 | 河南行知专利服务有限公司 | Power grid risk analysis and early warning method |
| KR101830295B1 (en) * | 2017-06-23 | 2018-03-29 | 주식회사 주빅스 | INTELLIGENT ACCIDENT PREVENTION TRAINING SELECTION DECISION SYSTEM OF TOXIC SUBSTANCE BUSINESS USING IoT SENSOR |
Non-Patent Citations (7)
| Title |
|---|
| 唐开元等: "《小波网络理论及其在经济建模中的应用》", 31 December 2008 * |
| 徐荣贞等: "《金融生态视角下系统性风险研究》", 31 December 2017 * |
| 湛从昌: "《液压可靠性最优化与智能故障诊断》", 31 December 2015 * |
| 赵林明等: "《多层前向人工神经网络》", 31 December 1999 * |
| 郭艳玲: "《线切割机床虚拟制造系统》", 31 December 2001 * |
| 陈丽华等: "《人工神经网络及其在水质信息检测中的应用》", 31 December 2011 * |
| 高晓旭: "《构建本质安全型煤矿理论与实践》", 31 December 2016 * |
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111862557A (en) * | 2019-04-24 | 2020-10-30 | 广州煜煊信息科技有限公司 | Household accident handling method based on Internet of things monitoring mode |
| CN111862557B (en) * | 2019-04-24 | 2022-08-16 | 广州煜煊信息科技有限公司 | Household accident handling method based on Internet of things monitoring mode |
| CN111723970A (en) * | 2020-05-06 | 2020-09-29 | 国网浙江省电力有限公司衢州供电公司 | Power operation hidden danger prediction method |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN102496069B (en) | Cable multimode safe operation evaluation method based on fuzzy analytic hierarchy process (FAHP) | |
| Wang et al. | Coal mine safety production forewarning based on improved BP neural network | |
| CN106972481A (en) | Scale electrically-charging equipment accesses the security quantitative estimation method of active power distribution network | |
| CN106022596A (en) | Urban gas pipeline system danger forecast and evaluation method | |
| CN104951588A (en) | Aided design method for mine ventilation systems | |
| CN108921230A (en) | Method for diagnosing faults based on class mean value core pivot element analysis and BP neural network | |
| CN107491694A (en) | Method for quantitative evaluation SCADA system information security fragility | |
| CN108053121A (en) | A kind of safe big data health degree appraisal procedure of structural fire protection based on AHP | |
| CN105868884A (en) | Evaluation method for domino accident prevention of petrochemical storage tank area | |
| CN108564193A (en) | The prediction technique of electric power personal injury, misoperation fault and the accident of operation against rules against regulations based on neural network | |
| CN104037760A (en) | Anticipated accident selection method for electric power system with uncertain injection power | |
| Ren et al. | A universal defense strategy for data-driven power system stability assessment models under adversarial examples | |
| CN102682212B (en) | Reliability measurement method for mechanical and electrical product | |
| CN109918791A (en) | A kind of nuclear plant digital master control room operator human reliability analysis method | |
| CN103400213A (en) | Backbone net rack survivability assessment method based on LDA (Linear Discriminant Analysis) and PCA (Principal Component Analysis) | |
| CN106874607A (en) | Power network self_organized criticla quantitative evaluating method based on multi-layer variable-weight theory | |
| CN103488911A (en) | Investment risk assessment method of photovoltaic power generation project | |
| CN108183499A (en) | A kind of static security analysis method based on Latin Hypercube Sampling Probabilistic Load Flow | |
| CN103577700B (en) | Boat firefighting system interlock failure prediction method | |
| Xu et al. | Power Quality Indices Online Prediction Based on VMD-LSTM Residual Analysis | |
| CN110414047A (en) | A method of it is evaluated for telecommunication transmission equipment health status | |
| Bian et al. | A Novel Dynamic Weight Allocation Method for Assessing the Health Status of Remote Terminal Unit in Distribution Automation System Based on AHM and GRA | |
| Zhou et al. | The integrated safety assessment on chemical industry park | |
| CN112132397A (en) | Large-scale business district security risk fuzzy comprehensive evaluation method and device based on network analysis | |
| CN108108909A (en) | Data analysing method towards electric power accident, misoperation fault with operating accident against regulations |
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: 20180921 |
|
| RJ01 | Rejection of invention patent application after publication |