CN113487252B - Enterprise safety early warning, prevention and control method and system based on big data analysis - Google Patents
Enterprise safety early warning, prevention and control method and system based on big data analysis Download PDFInfo
- Publication number
- CN113487252B CN113487252B CN202111046027.0A CN202111046027A CN113487252B CN 113487252 B CN113487252 B CN 113487252B CN 202111046027 A CN202111046027 A CN 202111046027A CN 113487252 B CN113487252 B CN 113487252B
- Authority
- CN
- China
- Prior art keywords
- equipment
- danger
- devices
- sample track
- fault
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q10/00—Administration; Management
- G06Q10/06—Resources, workflows, human or project management; Enterprise or organisation planning; Enterprise or organisation modelling
- G06Q10/063—Operations research, analysis or management
- G06Q10/0635—Risk analysis of enterprise or organisation activities
Landscapes
- Business, Economics & Management (AREA)
- Human Resources & Organizations (AREA)
- Engineering & Computer Science (AREA)
- Strategic Management (AREA)
- Entrepreneurship & Innovation (AREA)
- Economics (AREA)
- Operations Research (AREA)
- Game Theory and Decision Science (AREA)
- Development Economics (AREA)
- Marketing (AREA)
- Educational Administration (AREA)
- Quality & Reliability (AREA)
- Tourism & Hospitality (AREA)
- Physics & Mathematics (AREA)
- General Business, Economics & Management (AREA)
- General Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The invention provides an enterprise safety early warning, prevention and control method and system based on big data analysis, wherein the method comprises the following steps: constructing a device induced relation matrix according to fault induced relations among devices in the enterprise production environment, wherein the device induced relation matrix comprises a first dimension inducing faults of other devices and a second dimension induced by the other devices; acquiring the ratio of the first dimension to the second dimension of each device to obtain a device state distribution sequence; generating a sample track by utilizing an equipment evoked relation matrix and an equipment state distribution sequence according to a Markov chain model; and obtaining the dangerous value of each sample track according to the probability value of each sample track and the dangerous value of each device on each sample track, and carrying out early warning, prevention and control on the devices in the enterprise production environment according to the dangerous value of each sample track. The method can quickly and accurately position the root cause of the danger in the production area of the enterprise, and is beneficial to quickly and accurately making relevant measures to reduce the safety risk in the production area of the enterprise.
Description
Technical Field
The application relates to the field of big data, in particular to an enterprise safety early warning, prevention and control method and system based on big data analysis.
Background
With social development and technological progress, enterprise security problems are more and more prominent. Production, storage equipment, required materials and the like of enterprises are potential dangers in the production environment of the enterprises, so that enterprise safety early warning and prevention and control have important significance for enterprise development.
In order to reduce potential risks in a production environment and reduce the probability of accidents, an enterprise needs to overhaul equipment or adjust related production parameters. In order to solve the problem of enterprise safety, in the prior art, the running state of equipment is analyzed, and then a proper method is adopted to reduce the probability of equipment failure, so that the risk of a production area is reduced. However, the method is inefficient in analyzing the fault source, and the enterprise suffers risks and losses.
Disclosure of Invention
In order to solve the problems, the invention provides an enterprise safety early warning and prevention and control method based on big data analysis, which comprises the following steps:
constructing a device induced relation matrix according to fault induced relations among devices in the enterprise production environment, wherein the device induced relation matrix comprises a first dimension inducing faults of other devices and a second dimension induced by the other devices, and elements in the device induced relation matrix represent fault induced probabilities;
acquiring the ratio of the first dimension to the second dimension of each device as a state value of the device, wherein the state values of all the devices form a device state distribution sequence;
generating a sample track by utilizing an equipment evoked relation matrix and an equipment state distribution sequence according to a Markov chain model;
obtaining a danger value of each sample track according to the probability value of each sample track and the danger values of all devices on the sample track; acquiring the fault induction probability of adjacent equipment and the state value of first equipment in a sample track, and accumulating and multiplying the fault induction probability of the adjacent equipment and the state value of the first equipment to obtain the probability value of the sample track; the danger values of all the devices on the sample track are the modular length of the sum of the danger generation vectors of all the devices on the sample track; the risk generation vector of the device is specifically: obtaining a plurality of danger sources according to the comprehensive danger distribution map by utilizing a neural network; acquiring a danger source associated equipment set according to a danger source; the current analysis equipment is equipment to be analyzed, and a correlation risk value of the equipment to be analyzed and the hazard source is obtained according to the overlapping area of the danger areas of the equipment to be analyzed and equipment in the hazard source correlation equipment set; forming a danger generating vector of the equipment to be analyzed by the equipment to be analyzed and the associated danger values of all the danger sources;
and early warning, prevention and control are carried out on equipment in the enterprise production environment according to the dangerous values of the sample tracks. Preferably, the fault inducing relationship between the devices is determined according to the occurrence time of the device fault and the correlation between the fault degrees of any two devices. Preferably, the area with the equipment as the center is obtained as the dangerous area of the equipment, the dangerous distribution of the equipment is obtained by assigning values to the dangerous area of the equipment according to the fault degree of the equipment, and the dangerous distributions of all the equipment are superposed to obtain the comprehensive dangerous distribution map.
The invention also provides an enterprise safety early warning and prevention and control system based on big data analysis, which comprises:
the data acquisition module is used for constructing a device induced relation matrix according to fault induced relations among devices in the enterprise production environment, the device induced relation matrix comprises a first dimension inducing faults of other devices and a second dimension induced by the other devices, and elements in the device induced relation matrix represent fault induced probabilities; acquiring the ratio of the first dimension to the second dimension of each device as a state value of the device, wherein the state values of all the devices form a device state distribution sequence;
the sample track generation module is used for generating a sample track by utilizing the equipment evoked relation matrix and the equipment state distribution sequence according to the Markov chain model;
the early warning analysis module is used for obtaining the danger value of each sample track according to the probability value of each sample track and the danger values of all devices on the sample track; acquiring the fault induction probability of adjacent equipment and the state value of first equipment in a sample track, and accumulating and multiplying the fault induction probability of the adjacent equipment and the state value of the first equipment to obtain the probability value of the sample track; the danger values of all the devices on the sample track are the modular length of the sum of the danger generation vectors of all the devices on the sample track; the risk generation vector of the device is specifically: obtaining a plurality of danger sources according to the comprehensive danger distribution map by utilizing a neural network; acquiring a danger source associated equipment set according to a danger source; the current analysis equipment is equipment to be analyzed, and a correlation risk value of the equipment to be analyzed and the hazard source is obtained according to the overlapping area of the danger areas of the equipment to be analyzed and equipment in the hazard source correlation equipment set; forming a danger generating vector of the equipment to be analyzed by the equipment to be analyzed and the associated danger values of all the danger sources; and early warning, prevention and control are carried out on equipment in the enterprise production environment according to the dangerous values of the sample tracks.
The invention has the following beneficial effects:
according to the method, the equipment fault time sequence is obtained through the sequence of equipment faults in historical data, so that the fault induction relationship between the equipment is obtained, finally, a sample track is generated through the fault induction relationship matrix between the equipment and the equipment state distribution sequence based on a Markov chain model and is used for representing which equipment faults are mainly generated by superposition and induction of a hazard source, and the root cause of the enterprise production area hazard can be quickly and accurately positioned according to the structure, so that the problem is pertinently solved, and related measures can be quickly and accurately formulated to reduce the safety risk of the enterprise production area.
Drawings
FIG. 1 is a process flow diagram.
Detailed Description
The invention provides an enterprise safety early warning, prevention and control method and system based on big data analysis, and the invention is further described in detail by combining the attached drawings and specific embodiments.
The first embodiment is as follows:
the embodiment provides an enterprise security early warning and prevention and control method based on big data analysis, and a flow chart of the method is shown in fig. 1.
Firstly, acquiring an area with equipment as a center as a dangerous area of the equipment, assigning values to the dangerous area of the equipment according to the fault degree of the equipment to obtain the dangerous distribution of the equipment, and superposing the dangerous distributions of all the equipment to obtain a comprehensive dangerous distribution map.
And installing cameras in a production area, enabling the visual angles of the cameras to be downward in an oblique view, enabling adjacent cameras to have partially overlapped visual fields, acquiring images acquired by all the cameras, splicing and fusing the images, and obtaining a whole ground overlook spliced graph in the area.
Acquiring state data of each device on the production line, such as the temperature, pressure and the like of the chemical reaction device, the raw material concentration and the like of the raw material storage device, the catalyst residual amount and the like of the catalyst storage device, the output temperature and the output pressure of the device providing high temperature and high pressure, the output power of the power device and the like. And analyzing the state data of the equipment according to classification models such as a neural network and the like to obtain the fault degree of the equipment. The failure degree is divided into ten grades: 0.0, 0.1, … …, 0.9. When the degree of the failure is greater than 0, the device is considered to be failed.
The positions of the devices are fixed and unchangeable, in the embodiment, the positions of the devices are artificially marked on a ground overlook splicing map, a circular dangerous area is divided by taking the devices as the center and taking the preset radius as each device for representing the accident influence range of each device after a dangerous accident happens, and the ranges of the dangerous accidents generated are different due to different action types of the devices, so that the radiuses of the dangerous areas of each device are different.
On a ground overlook splicing diagram, generating a Gaussian hot spot with the same size as a corresponding dangerous area for one equipment position, wherein the generation method of the Gaussian hot spot comprises the following steps: and taking the position of the equipment as a center and the diameter of the dangerous area as a size to generate a Gaussian convolution kernel, wherein the Gaussian convolution kernel is the Gaussian hot spot. Different positions on the Gaussian hot spot correspond to a heat value, and the heat values of all the positions except the Gaussian hot spot are 0.
The heat value at each location is multiplied by the degree of failure of the equipment, and the multiplication result represents the magnitude of the risk at each location. So far, for one device, the danger size of each location is obtained, and the danger sizes of all the locations are referred to as the danger distribution of one device. And superposing the danger distributions of all the devices to obtain a comprehensive danger distribution map, and representing the result on a ground overlook splicing map.
Then, obtaining a device fault time sequence according to the time sequence of the device faults in any time period; and processing the equipment fault time sequence by using the neural network to obtain the fault characteristic point of the equipment.
And acquiring the time when each device fails according to the historical data, and acquiring a sequence of the devices according to the sequence of the failures, wherein the sequence is called a comprehensive failure time sequence and is used for representing the sequence of the failures of the devices. If the time interval of the two devices with faults in the comprehensive fault time sequence is larger than the preset threshold value, the sequence is disconnected from the middle of the two devices, and therefore the comprehensive fault time sequence is divided into a plurality of device fault time sequences. Thus, a plurality of equipment fault time sequence sequences are obtained according to the sequence of equipment faults. Each equipment failure time sequence represents the sequence of equipment failure. The sequence of the occurrence of the equipment faults comprises fault inducing characteristics among the equipment, wherein the fault inducing characteristics among the equipment refer to the probability that the fault occurring in one equipment induces the fault occurring in the other equipment.
In this embodiment, the above-mentioned one equipment failure time sequence is regarded as one text statement, one equipment element in the sequence represents one word, a large number of equipment failure time sequence sequences are obtained, and a word vector is allocated to each equipment by adopting a word2vec method according to the sequences, which is called an equipment failure vector.
In practice, the data volume of the equipment fault time sequence is insufficient, so that a chemical production line needs to be simulated by a computer, some production faults are artificially set, and equipment with faults in sequence is obtained through computer simulation, so that a large number of equipment fault time sequence sequences are obtained and used as training data of the word2 vec.
An equipment fault vector is a point in a high dimensional space called an equipment fault feature point, and if two equipment fault feature points are close in distance, the two equipment fault feature points are more similar. For example, if a device a has a greater similarity to a device b, then when a faulty device induces a fault in the device a, the faulty device may also induce a fault in the device b.
Then, obtaining a plurality of danger sources according to the comprehensive danger distribution map by utilizing a neural network; and determining the fault induction relationship between the devices according to the occurrence time of the devices and the correlation relationship between the fault degrees of any two devices.
And inputting the comprehensive danger distribution map into the DNN network to acquire the danger source in the production area. The danger source is a communicated area, a plurality of positions with high danger degree are concentrated in the area, and if safety accidents occur in the danger source, the interlocking effect of the accidents can be generated, so that large-area damage or destruction is caused. The task of the DNN network is the same as the conventional semantic segmentation task, training is carried out through artificial marking, and a cross entropy loss function is adopted as a loss function. In addition, the DNN network of the present embodiment may output a plurality of hazard sources.
In the embodiment, the dangerous source is obtained in real time, when the total area of the dangerous source reaches a certain preset threshold value at the current moment, the time when different equipment fails within a period of time before the current moment is obtained, and the equipment failure time sequence is obtained according to the time sequence when the equipment fails
N represents a sequence
The length of (a) of (b),
representing a sequence
The ith device.
The meaning of the characterization is that the equipment
Post-failure device
A fault also occurs. Experience has shown that it is possible to have a device
Fault causing device of
But due to the complexity of the production equipment, equipment is also possible
Fault causing device of
Fault of
。
Sequence of device failure
In obtaining any two devices
And
. Then the equipment
And apparatus
The fault induction relationship of
. In the formula
For describing the apparatus
The occurring fault induces the device
Probability of failure.
in the sequence
Middle equipment
Is and the equipment
Adjacent, possibly equipment
Failure of (2) induces a device
Malfunction, possibly of the apparatus
Any of the previous devices induced a device
Failure of (2). For the same reason, for the equipment
May be a device
Any of the previous devices induced a device
Failure of (2). Device
To the equipment
Can be obtained by analyzing the distance between the fault characteristic points of the two devices and the device-to-device relationship between the two devices
The distance of (2) obtains the association relationship between the two devices. The analysis equipment is required in the embodiment
Fault inducing device of
Possibility of failure, on the one hand by analysis
、
The distance between the two can obtain the probability of directly inducing the fault; on the one hand if
Inducing a failure of the device between the two devices if the device is out of order
And apparatus
Are similar, then the device
Is likely to induce the device
A failure occurs; if the equipment
And apparatus
Not similar, then the device
May not induce the device
A failure occurs; if the equipment
And apparatus
The similarity is very high, then the device
Is likely to induce a device
A failure occurs; device
And apparatus
The more similar the two devices are, the closer the fault signature points are, and the further away they are otherwise.
Device
The equipment failure characteristic point is
Device
The equipment failure characteristic point is
Device
K other devices nearest in time (excluding
) Is a set of equipment fault feature points
One device in the set is
. Association relation
The second embodiment of (1):
when in use
When the temperature of the water is higher than the set temperature,
。
wherein
，
To represent
The elements (A) and (B) in (B),
indicating the distance of the two equipment failure characteristic points,
the smaller the ratio, the equipment is illustrated
Fault characteristic point and
the distance between the fault characteristic points of the medium equipment is far longer than that between the fault characteristic points of other equipment and the equipment
The distance between the fault feature points is small, which shows
Device and equipment
Strong substitutability of the compound
The larger.
When in use
When the temperature of the water is higher than the set temperature,
this means that a device that fails later does not induce a device that fails earlier, and is a cause and effect relationship.
Further, a device induced relation matrix is constructed according to fault induced relations among devices in the enterprise production environment, the device induced relation matrix comprises a first dimension inducing faults of other devices and a second dimension inducing faults of other devices, and elements in the device induced relation matrix represent fault induced probabilities; acquiring the ratio of the first dimension to the second dimension of each device as a state value of the device, wherein the state values of all the devices form a device state distribution sequence; and generating a sample track by utilizing the equipment evoked relation matrix and the equipment state distribution sequence according to the Markov chain model.
Obtaining a matrix based on evoked relationships between devices
The number of rows and columns of the matrix being
Length of (2)
In a matrix of
The value of the coordinate is taken as the equipment
And apparatus
In this embodiment, a first dimension of the device evoked relationship matrix is a row, each row represents a probability that one device induces a fault of another device, and a second dimension is a column, each column represents a probability that one device induces a fault of another device.
Acquisition device
In a first dimension, i.e. second
Mean value of all non-zero elements of a row and second dimension, i.e. second
Device for obtaining ratio of mean values of all non-zero elements in column
State value of (2), representing a device
Probability of inducing other devices to malfunction and other devices inducing devices
Ratio of probability of failure. And acquiring state values of all the devices, and normalizing the state values to obtain a device state distribution sequence.
The device evoked relationship matrix and the device state distribution sequence can be used as parameters of a Markov chain model, and a plurality of sample tracks can be generated according to the parameters, wherein the sample tracks are chain diagram structure data, and each node on each sample track represents a device. Each sample track characterizes an induced relationship between devices, i.e. a failure of a device represented by each node on the sample track can induce a failure of a device represented by the next node. The edge weights of two node devices on the sample track are: and setting the device corresponding to the previous node as a and the device corresponding to the next adjacent node as b, acquiring an element corresponding to a first dimension corresponding to the device a, namely a second dimension corresponding to the device b, namely a row where the device a is located and a column where the device b is located on the fault inducing relationship matrix, wherein the probability value represented by the element is the edge weight value between the two node devices.
Then, the current analysis equipment is equipment to be analyzed, and the associated risk value of the equipment to be analyzed and the risk source is obtained according to the overlapping area of the dangerous area of the equipment to be analyzed and the equipment in the risk source associated equipment set; and the associated danger values of the equipment to be analyzed and all the danger sources form a danger generating vector of the equipment to be analyzed.
In this embodiment, it is assumed that K hazard sources are obtained, and hazard source associated equipment is obtainedA set S, wherein the set S meets the condition that the dangerous area of each device in the set S has intersection with the kth dangerous source; device for installing
The sum of the summation and the intersection ratio of the dangerous areas of all the devices in the set S is recorded as the result
Then equipment
Associated hazard value with the kth hazard source of
. Device
Associated hazard values with all hazard sources forming device
Generates a vector.
And finally, obtaining the dangerous value of each sample track according to the probability value of each sample track and the dangerous values of all the devices on the sample track, and carrying out early warning, prevention and control on the devices in the enterprise production environment according to the dangerous values of each sample track. The method comprises the steps of obtaining fault induction probability of adjacent equipment in a sample track and a state value of first equipment, and accumulating and multiplying the fault induction probability of the adjacent equipment and the state value of the first equipment to obtain a probability value of the sample track. The hazard value for all devices on the sample track is the modulo length of the sum of the hazard generation vectors for all devices on the sample track.
Each sample track corresponds to a risk value, and the calculation method of the risk value of the sample track is as follows: the state value of the device represented by the first node in a sample track is
The product of all edge weights for that sample track (i.e., allThe product of the fault-inducing probabilities of neighboring devices) is
Then the probability value of the sample track is
. Obtaining the modular length of the danger generating vector of each device on the sample track to obtain the danger value of each device on the sample track, summing the danger generating vectors of all the devices, calculating the modular length of the vectors and recording the modular length of the vectors as the danger value
. The hazard value of the sample track is
The risk sources within the current production area are characterized by the probability of being due to fault stacking and fault induction of the devices represented by the sample track.
Calculating the danger values of a plurality of sample tracks, and acquiring the sample track with the largest danger value, wherein the sample track is used for representing that the danger source in the current production area is generated due to fault superposition and fault induction of the devices represented by the sample track, so that the devices must be firstly subjected to priority treatment to prevent the current danger source from generating danger or reduce the danger degree of the danger source, and the device at the first node of the sample track induces the other devices to generate faults, so that the device is considered to be the root source of the danger, and further, the devices in the enterprise production environment are subjected to early warning, prevention and control, and the problem of the faults of the devices is solved in time, so that the danger of the enterprise production environment is rapidly reduced at the first time.
Example two:
this embodiment provides an enterprise safety precaution and prevention and control system based on big data analysis, and this system includes:
the data acquisition module is used for constructing a device induced relation matrix according to fault induced relations among devices in the enterprise production environment, the device induced relation matrix comprises a first dimension inducing faults of other devices and a second dimension induced by the other devices, and elements in the device induced relation matrix represent fault induced probabilities; acquiring the ratio of the first dimension to the second dimension of each device as a state value of the device, wherein the state values of all the devices form a device state distribution sequence;
the sample track generation module is used for generating a sample track by utilizing the equipment evoked relation matrix and the equipment state distribution sequence according to the Markov chain model;
the early warning analysis module is used for obtaining the danger value of each sample track according to the probability value of each sample track and the danger values of all devices on the sample track; acquiring the fault induction probability of adjacent equipment and the state value of first equipment in a sample track, and accumulating and multiplying the fault induction probability of the adjacent equipment and the state value of the first equipment to obtain the probability value of the sample track; the danger values of all the devices on the sample track are the modular length of the sum of the danger generation vectors of all the devices on the sample track; the risk generation vector of the device is specifically: obtaining a plurality of danger sources according to the comprehensive danger distribution map by utilizing a neural network; acquiring a danger source associated equipment set according to a danger source; the current analysis equipment is equipment to be analyzed, and a correlation risk value of the equipment to be analyzed and the hazard source is obtained according to the overlapping area of the danger areas of the equipment to be analyzed and equipment in the hazard source correlation equipment set; forming a danger generating vector of the equipment to be analyzed by the equipment to be analyzed and the associated danger values of all the danger sources; and early warning, prevention and control are carried out on equipment in the enterprise production environment according to the dangerous values of the sample tracks.
The above description is only for the purpose of illustrating the preferred embodiments of the present invention and is not to be construed as limiting the invention, and any modifications, equivalents, improvements and the like that fall within the spirit and principle of the present invention are intended to be included therein.
Claims (4)
1. An enterprise safety early warning and prevention and control method based on big data analysis is characterized by comprising the following steps:
determining a fault induction relation between the devices according to the occurrence time of the device faults in the enterprise production environment and the correlation relation between the fault degrees of any two devices; constructing a device induced relation matrix according to fault induced relations among devices, wherein the device induced relation matrix comprises a first dimension inducing faults of other devices and a second dimension induced by the other devices, and elements in the device induced relation matrix represent fault induced probabilities;
acquiring the ratio of the first dimension to the second dimension of each device as a state value of the device, wherein the state values of all the devices form a device state distribution sequence;
generating a sample track by utilizing an equipment evoked relation matrix and an equipment state distribution sequence according to a Markov chain model; obtaining a danger value of each sample track according to the probability value of each sample track and the danger value of each device on each sample track; acquiring the fault induction probability of adjacent equipment and the state value of first equipment in a sample track, and accumulating and multiplying the fault induction probability of the adjacent equipment and the state value of the first equipment to obtain the probability value of the sample track; the hazard value of each device on the sample track is specifically: obtaining the modular length of the danger generating vector of each device on the sample track to obtain the danger value of each device on the sample track; the risk generation vector of the device is specifically: obtaining a plurality of danger sources according to the comprehensive danger distribution map by utilizing a neural network; acquiring a danger source associated equipment set according to a danger source; the current analysis equipment is equipment to be analyzed, and a correlation risk value of the equipment to be analyzed and the hazard source is obtained according to the overlapping area of the danger areas of the equipment to be analyzed and equipment in the hazard source correlation equipment set; forming a danger generating vector of the equipment to be analyzed by the equipment to be analyzed and the associated danger values of all the danger sources;
and early warning, prevention and control are carried out on equipment in the enterprise production environment according to the dangerous values of the sample tracks.
2. The method of claim 1, further comprising:
and acquiring an area taking the equipment as a center as a dangerous area of the equipment, assigning values to the dangerous area of the equipment according to the fault degree of the equipment to obtain the dangerous distribution of the equipment, and superposing the dangerous distributions of all the equipment to obtain a comprehensive dangerous distribution map.
3. The utility model provides an enterprise safety precaution and prevention and control system based on big data analysis which characterized in that, the system includes:
the data acquisition module is used for determining the fault induction relationship between the devices according to the occurrence time of the device faults in the enterprise production environment and the correlation relationship between the fault degrees of any two devices; constructing a device induced relation matrix according to fault induced relations among devices, wherein the device induced relation matrix comprises a first dimension inducing faults of other devices and a second dimension induced by the other devices, and elements in the device induced relation matrix represent fault induced probabilities; acquiring the ratio of the first dimension to the second dimension of each device as a state value of the device, wherein the state values of all the devices form a device state distribution sequence;
the sample track generation module is used for generating a sample track by utilizing the equipment evoked relation matrix and the equipment state distribution sequence according to the Markov chain model;
the early warning analysis module is used for obtaining a danger value of each sample track according to the probability value of each sample track and the danger value of each device on each sample track; acquiring the fault induction probability of adjacent equipment and the state value of first equipment in a sample track, and accumulating and multiplying the fault induction probability of the adjacent equipment and the state value of the first equipment to obtain the probability value of the sample track; the hazard value of each device on the sample track is specifically: obtaining the modular length of the danger generating vector of each device on the sample track to obtain the danger value of each device on the sample track; the risk generation vector of the device is specifically: obtaining a plurality of danger sources according to the comprehensive danger distribution map by utilizing a neural network; acquiring a danger source associated equipment set according to a danger source; the current analysis equipment is equipment to be analyzed, and a correlation risk value of the equipment to be analyzed and the hazard source is obtained according to the overlapping area of the danger areas of the equipment to be analyzed and equipment in the hazard source correlation equipment set; forming a danger generating vector of the equipment to be analyzed by the equipment to be analyzed and the associated danger values of all the danger sources; and early warning, prevention and control are carried out on equipment in the enterprise production environment according to the dangerous values of the sample tracks.
4. The system of claim 3, wherein the system is further configured to:
and acquiring an area taking the equipment as a center as a dangerous area of the equipment, assigning values to the dangerous area of the equipment according to the fault degree of the equipment to obtain the dangerous distribution of the equipment, and superposing the dangerous distributions of all the equipment to obtain the comprehensive dangerous distribution map.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202111046027.0A CN113487252B (en) | 2021-09-07 | 2021-09-07 | Enterprise safety early warning, prevention and control method and system based on big data analysis |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202111046027.0A CN113487252B (en) | 2021-09-07 | 2021-09-07 | Enterprise safety early warning, prevention and control method and system based on big data analysis |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN113487252A CN113487252A (en) | 2021-10-08 |
| CN113487252B true CN113487252B (en) | 2021-11-26 |
Family
ID=77947316
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202111046027.0A Active CN113487252B (en) | 2021-09-07 | 2021-09-07 | Enterprise safety early warning, prevention and control method and system based on big data analysis |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN113487252B (en) |
Family Cites Families (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US9158303B2 (en) * | 2012-03-27 | 2015-10-13 | General Electric Company | Systems and methods for improved reliability operations |
| CN103714401A (en) * | 2014-01-21 | 2014-04-09 | 国家电网公司 | Method for predicting distribution and transformation fault frequency of electric network |
| CN106327033B (en) * | 2015-06-18 | 2020-08-28 | 中国电力科学研究院 | Power system cascading failure analysis method based on Markov process |
| CN106503923B (en) * | 2016-11-15 | 2019-07-16 | 国网四川省电力公司电力科学研究院 | A kind of method and system of Risk assessment of cascading failure in power system |
| CN110336280B (en) * | 2019-07-31 | 2022-07-01 | 国网河南省电力公司经济技术研究院 | Power system cascading failure analysis method based on dictionary set acceleration |
-
2021
- 2021-09-07 CN CN202111046027.0A patent/CN113487252B/en active Active
Also Published As
| Publication number | Publication date |
|---|---|
| CN113487252A (en) | 2021-10-08 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Saeed et al. | Novel fault diagnosis scheme utilizing deep learning networks | |
| Wang et al. | Fault identification and diagnosis based on KPCA and similarity clustering for nuclear power plants | |
| Saeed et al. | Online fault monitoring based on deep neural network & sliding window technique | |
| US11487266B2 (en) | Method for recognizing contingencies in a power supply network | |
| CN107704992A (en) | The method and device of transmission line lightning stroke risk assessment | |
| Taylor et al. | Verification and validation of neural networks: a sampling of research in progress | |
| CN110334478A (en) | Machinery equipment abnormality detection model building method, detection method and model | |
| Ma et al. | Adequate and precise evaluation of quality models in software engineering studies | |
| Dong et al. | Attention-based time series analysis for data-driven anomaly detection in nuclear power plants | |
| Choi et al. | Unsupervised learning algorithm for signal validation in emergency situations at nuclear power plants | |
| Kumar et al. | Advanced analytics on IV curves and electroluminescence images of photovoltaic modules using machine learning algorithms | |
| Ogar et al. | The use of artificial neural network for low latency of fault detection and localisation in transmission line | |
| US20210270797A1 (en) | Systems and methods for dissolved gas analysis | |
| Racheal et al. | Evaluation of optimized machine learning models for nuclear reactor accident prediction | |
| Hou et al. | An artificial neural network based anomaly detection algorithm for nuclear power plants | |
| CN110378358A (en) | A kind of power distribution network isomeric data integration method and system | |
| Kollias et al. | Machine learning for analysis of real nuclear plant data in the frequency domain | |
| CN113487252B (en) | Enterprise safety early warning, prevention and control method and system based on big data analysis | |
| Chen et al. | Peripheric sensors-based leaking source tracking in a chemical industrial park with complex obstacles | |
| Yang et al. | Prediction of hydrogen leakage location and intensity in hydrogen refueling stations based on deep learning | |
| Firos et al. | Fault detection in power transmission lines using AI model | |
| US20180067831A1 (en) | Fine-Grained Causal Anomaly Inference for Complex System Fault Diagnosis | |
| Liu et al. | A new classification method of infrasound events using Hilbert‐Huang transform and support vector machine | |
| Mengash et al. | Search and Rescue Optimization with Machine Learning Enabled Cybersecurity Model. | |
| CN116032726A (en) | Fault root cause positioning model training method, device, equipment and readable storage medium |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant | ||
| TR01 | Transfer of patent right |
Effective date of registration: 20220309 Address after: 226100 Fuxing Road, Yuet Lai Town, Haimen, Nantong, Jiangsu, 20 Patentee after: NANTONGYOUYUAN ART DESIGN Co.,Ltd. Address before: 226131 group 25, Puxin village, Yuelai Town, Haimen District, Nantong City, Jiangsu Province Patentee before: Jiangsu Changcun Copper Co.,Ltd. |
|
| TR01 | Transfer of patent right |