CN109327330B - Chemical production abnormal slice management method based on data driving - Google Patents
Chemical production abnormal slice management method based on data driving Download PDFInfo
- Publication number
- CN109327330B CN109327330B CN201811058446.4A CN201811058446A CN109327330B CN 109327330 B CN109327330 B CN 109327330B CN 201811058446 A CN201811058446 A CN 201811058446A CN 109327330 B CN109327330 B CN 109327330B
- Authority
- CN
- China
- Prior art keywords
- historical
- matrix
- bit number
- incidence matrix
- alarm
- 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
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L41/00—Arrangements for maintenance, administration or management of data switching networks, e.g. of packet switching networks
- H04L41/06—Management of faults, events, alarms or notifications
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L41/00—Arrangements for maintenance, administration or management of data switching networks, e.g. of packet switching networks
- H04L41/06—Management of faults, events, alarms or notifications
- H04L41/0631—Management of faults, events, alarms or notifications using root cause analysis; using analysis of correlation between notifications, alarms or events based on decision criteria, e.g. hierarchy, tree or time analysis
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L41/00—Arrangements for maintenance, administration or management of data switching networks, e.g. of packet switching networks
- H04L41/08—Configuration management of networks or network elements
- H04L41/0893—Assignment of logical groups to network elements
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L67/00—Network arrangements or protocols for supporting network services or applications
- H04L67/50—Network services
- H04L67/55—Push-based network services
Landscapes
- Engineering & Computer Science (AREA)
- Computer Networks & Wireless Communication (AREA)
- Signal Processing (AREA)
- Testing And Monitoring For Control Systems (AREA)
- Alarm Systems (AREA)
Abstract
The invention relates to a data-driven chemical production abnormal slice management method, which comprises the following steps: step 1, collecting each bit number by using a sensor; step 2, carrying out periodic correlation analysis on the bit numbers to obtain correlation matrixes at different moments; step 3, at a certain moment, alarming occurs to the target position number, and an incidence matrix closest to the alarming moment is found out through time difference; step 4, calculating the similarity between the incidence matrix and the historical incidence matrix to obtain a group of historical incidence matrices with the highest similarity; step 5, calling out the historical operation corresponding to the most similar historical incidence matrix, and sending the historical operation to an operator; and 6, after the alarm is solved, recording the incidence matrix and the operation of an operator to form a knowledge slice, and putting the knowledge slice into a historical database.
Description
Technical Field
The invention relates to a slice management method, in particular to a data-driven chemical production abnormal slice management method, and belongs to the technical field of chemical production control.
Background
The chemical production process is a high-risk process, the parameters are numerous, an operator needs to reside in a control room and know the change of the production working condition in real time according to a DCS system interface, and the traditional management operation mode is limited by time and place and has strong passivity. The chemical production process needs to provide flexible technical operation schemes for operators due to multiple parameters, diverse flow and complex process so as to deal with the complex and variable chemical production process.
Currently, the main method for managing the position number is that an operator observes a DCS interface and receives corresponding single-parameter alarm information. The accident prevention management method for the chemical process is obviously lagged, the provided alarm information is single, and an accurate priori knowledge base is difficult to establish due to the fact that the chemical production process is multiple in parameters, diverse in flow and complex in process. More relevant parameter information and relevant operation basis are provided for operators, so that the possibility of shortening the accident judgment and accelerating the decision process of handling is provided, and therefore, a new scheme is urgently needed to solve the technical problem.
Disclosure of Invention
The invention provides a data-driven chemical production abnormal slice management method aiming at the technical problems in the prior art, and the technical scheme is that the incidence matrix of the target bit number is periodically calculated, and a group of parameters, usually 2-3 parameters, are determined. When the target bit number gives an alarm at a certain moment, similarity calculation is carried out on the incidence matrix at the alarm moment and the historical incidence matrix, the historical incidence matrix which is closest to the current incidence matrix is found, and then the alarm details and the related experience corresponding to the historical matrix are pushed to an operator through the mobile terminal. After the fault is removed, the accident details and experience are recorded, and the incidence matrix is stored in the database as a historical incidence matrix and is used as a judgment basis for the next similar fault.
In order to achieve the purpose, the technical scheme of the invention is as follows: a data-driven chemical production abnormal slice management method comprises the following steps:
step 1, collecting each bit number by using a sensor;
step 2, carrying out periodic correlation analysis on the bit numbers to obtain correlation matrixes at different moments;
step 3, at a certain moment, alarming occurs to the target position number, and an incidence matrix closest to the alarming moment is found out through time difference;
step 4, calculating the similarity between the incidence matrix and the historical incidence matrix to obtain a group of historical incidence matrices with the highest similarity;
step 5, calling out the historical operation corresponding to the most similar historical incidence matrix, and sending the historical operation to an operator;
and 6, after the alarm is solved, recording the incidence matrix and the operation of an operator to form a knowledge slice, and putting the knowledge slice into a historical database to serve as comparison information of the parameter alarm in the future.
As an improvement of the present invention, the specific process of step 2 is as follows: calculating the relevance between the alarm bit number and which bit numbers by adopting a relevance calculation method; the specific calculation method comprises the following steps:
in the normal production process, the bit number strongly correlated with the target bit number is calculated once every 12 hours (or other time periods) and a correlation matrix is formed,
as an improvement of the present invention, the step 3 is specifically as follows, when the target bit number generates an alarm, sending the alarm information to the operator, and calling out the correlation matrix closest to the alarm time from the database.
As an improvement of the present invention, in step 4, the mahalanobis distance is used to calculate which historical association matrix the association matrix of the target bit number is closest to.
As an improvement of the present invention, in step 5, the accident description and experience record corresponding to the historical association matrix are presented to the operator as operation reference information of the operator.
As a modification of the present invention, the step 6: after the fault is relieved, the alarm parameter incidence matrix and the corresponding accident description and experience record are put into a historical database to be used as comparison information of the incidence matrix and corresponding operation guidance when the next target bit number gives an alarm.
As an improvement of the present invention, the message push of step 3 and step 5 is displayed to the operator through the mobile terminal, and the mobile terminal has a personalized operation function, that is: whether to open a switch for receiving the alarm push information.
As an improvement of the present invention, in the step 2), when the bit number matrix strongly correlated with the target bit number is calculated, another parameter having a large correlation with the change of the target bit number is calculated by using a correlation coefficient algorithm, the target bit number is represented by x, y denotes another k bit numbers participating in the correlation calculation, and first, a correlation coefficient x between x and y is calculated as [ x1, x2, x3, … xn ═]，yk＝[yk1，yk2，yk3，...ykn](k is 1, … h), and mixing x, yknWith equation 1, the correlation coefficient of the two is calculated,
in statistics, the degree of correlation between two variables is quantitatively described by a "correlation coefficient", which is denoted by the symbol "r". The range of the correlation coefficient is limited as follows: -1. ltoreq. r.ltoreq.1;
the meaning of the correlation coefficient representation;
| correlation coefficient | 0.00 | 0.00…±0.3 | ±0.30...±0.50 | ±0.50...±0.80 | ±0.30...±1.00 |
| Degree of correlation | Without correlation | Micro positive and negative correlation | Real positive-negative correlation | Significant positive and negative correlation | High positive-negative correlation |
Here, a coefficient in which the absolute value of r is larger than 0.8 is taken as a strong correlation coefficient.
As an improvement of the present invention, at a certain time in step 4, when the target bit number alarms, the correlation matrix of the bit number at the time is calculated, then similarity calculation is performed with the historical correlation matrix of the bit number, a group of historical correlation matrices with the maximum similarity is found, and the similarity calculation between the correlation matrix and the historical correlation matrix at this time adopts a mahalanobis distance algorithm:
the target bit number is denoted by x, and the strong correlation matrix when an alarm occurs is A ═ x, a1，a2...am]TA certain historical association matrix P ═ x, P1p2...pm]And E is the covariance matrix of the A, P matrix, then the distance of A from P is:
the historical correlation matrix corresponding to the small value D is most similar to the correlation matrix A, and the accident description and the experience record corresponding to the historical correlation matrix are presented to the operator. And the incidence matrix is stored as a slice to become a historical incidence matrix which is the basis for the next operation similar to the fault.
Compared with the prior art, the method has the advantages that 1) the acquired parameters are subjected to relevance analysis in the step 2 of the technical scheme, and then the relevance parameter matrix is calculated, the method gets rid of the defect of the traditional single alarm information, the relevance parameters of the alarm parameters are found out through the relevance matrix calculation, and an operator can obtain more fault decision basis and operation basis according to the relevance parameters; 2) step 3 of the scheme is to periodically calculate the incidence matrix of the target bit number and record the corresponding time generated by the incidence matrix. The chemical process parameters are numerous, the flow is diversified, the incidence matrix of the target position number is periodically calculated, the timeliness of the incidence matrix is guaranteed, the production actual situation of the chemical process is met, and the incidence matrix closest to the alarm time of the target position number can be conveniently and quickly found out in the step 4; 3) in the scheme, the step 4 finds out which historical incidence matrix is most similar to the incidence matrix, because parameter dimensions of liquid, enterprises, solids, pressure and the like in the chemical process are numerous, the mahalanobis distance is not influenced by the dimensions, and the mahalanobis distance between two points is irrelevant to the measurement unit of the original data. The technical scheme adopts a big data method and a thinking mode to fully mine and utilize the value of historical data, finds out the historical incidence matrix which is most similar to the incidence matrix from a plurality of historical matrixes, and applies the corresponding operation and suggestion thereof to the current production management and control; 4) the operation of calling the most similar historical incidence matrix in the step 5 of the scheme occurs to the administrator, the administrator has a set of operation detailed rules according to professional knowledge and experience of the administrator according to the alarm parameter information, sufficient information is provided for alarm fault processing according to the corresponding historical operation detailed rules called out by the historical incidence matrix, and the administrator has richer and more reliable operation basis. Alarm information and experience operation records of historical similar faults are sent to an administrator through a mobile terminal, connection between a chemical process and a person is established, and through mobile interconnection, abnormal conditions can be rapidly and effectively issued, so that the management and control efficiency of the chemical process is improved; 5) step 6 of the scheme is that finally, after the fault is solved by an operator and an administrator, the operation of the administrator is recorded and is put into a historical database together with an alarm correlation matrix to form a knowledge slice, and through similarity judgment of abnormal slices, the value of historical data is fully mined and utilized, so that more operation bases are provided for alarming of a certain number in the future; 6) according to the technical scheme, the incidence matrix and the corresponding operation rule are stored in the database, the historical data amount is enriched, the accuracy of the next matrix similarity calculation is improved, and the operation effect of the technology is enhanced. The traditional DCS system only gives alarm information of a single parameter, the alarm information is single, an administrator usually needs to accumulate through solid professional knowledge and rich experience, and the acquisition of the alarm information is bound to a control room. The alarm push information of the technical scheme can be acquired at the mobile terminal, and the associated bit number information of the alarm bit number and the operation suggestion of historical similar alarms are given, so that the judgment basis of the accident type is enriched. The significance of the scheme is that historical operation information is recorded through mobile interconnection to form a knowledge slice, and operation suggestions are provided for future identical or similar alarm faults, so that the decision efficiency is improved, the accident judgment and processing efficiency in chemical production is reduced, the negative influence degree of accidents is reduced to the maximum extent, and secondary disasters are avoided.
Drawings
FIG. 1 is a schematic view of the working process of the present invention.
The specific implementation mode is as follows:
for the purpose of enhancing an understanding of the present invention, the present embodiment will be described in detail below with reference to the accompanying drawings.
Example 1: referring to fig. 1, a data-driven chemical production abnormal slice management method includes the following steps:
step 1: the acquisition of each parameter is realized by using a sensor;
step 2: correlation calculation, namely performing periodic correlation analysis on the bit numbers to obtain correlation matrixes at different moments; a specific calculation method is as follows,
in the normal production process, bit numbers which are strongly related to target bit numbers are periodically calculated and related bit number groups are formed, for example, the bit numbers which are strongly related to the A.WL.MT.LT901A bit numbers are calculated, other bit numbers which have larger change correlation with the A.WL.MT.LT901A are calculated through a correlation coefficient algorithm, the A.WL.MT.LT901A is represented by x, and y refers to other k bit numbers which participate in correlation calculation. First, the correlation coefficient x of x and y is calculated as [ x1, x2, x3, … xn [ ]]，yk＝[yk1，yk2，yk3...ykn](k is 1, … h), and mixing x, yknThe correlation coefficient of the two is calculated by an substituting formula,
in statistics, the degree of correlation between two variables is quantitatively described by a "correlation coefficient", and is represented by a symbol "r"; the range of the correlation coefficient is limited as follows: -1. ltoreq. r.ltoreq.1;
significance of the representation of the correlation coefficient
| Correlation coefficient | 0.00 | 0.00…±0.3 | ±0.30...±0.50 | ±0.50...±0.80 | ±0.80...±1.00 |
| Degree of correlation | Without correlation | Micro positive and negative correlation | Real positive-negative correlation | Significant positive and negative correlation | High positive-negative correlation |
Taking a coefficient with the absolute value of r larger than 0.8 as a strong correlation coefficient;
such as the correlation matrix of target bit numbers [ A.WL.MT.LT901D, A.WL.MT.LT-1110A, A.LY.YP.FI1501.PV …]It means that the several bit numbers have strong correlation with a.wl.mt.lt901a, and for convenience, a ═ x, a is used hereinafter1，a2...am]TA correlation matrix is represented.
And step 3: at a certain moment, the target bit number occursAlarming, namely finding out the incidence matrix closest to the alarming moment by making a difference in time; repeating the steps 1 and 2 according to the working condition decision at intervals of 12 hours, two days or one week, recalculating the bit number with strong correlation with the target bit number, and substituting the data at that time to form a correlation matrix, namely A ═ x, a1，a2...am]TWherein T is the time of generation of the correlation matrix;
and 4, step 4: calculating the similarity between the incidence matrix and the historical incidence matrix to obtain a group of historical incidence matrices with the highest similarity; at a certain time T1When the target bit number A.WL.MT.LT901A gives an alarm, the target bit number T is compared with the target bit number T1Similarity calculation is carried out on the correlation matrix corresponding to the moment with the minimum difference value and the historical correlation matrix, a group of historical correlation matrices with the closest similarity is found out, accident description and experience records corresponding to the historical correlation matrices are presented to an operator, and a Mahalanobis distance algorithm is adopted for similarity calculation of the correlation matrices and the historical correlation matrices:
the incidence matrix when alarm occurs is [ x, a ]1，a2...am]TA certain historical association matrix P ═ x, P1，p2...pm]And E is the covariance matrix of the A, P matrix, then the distance of A from P is:
the corresponding historical incidence matrix is most similar to the incidence matrix a when the value of D is the smallest,
step 5, sending the accident description and the experience operation record corresponding to the historical incidence matrix to an operator;
and 6, after the alarm is solved, recording the incidence matrix and the operation of an operator to form a knowledge slice, and putting the knowledge slice into a historical database to serve as comparison information of the parameter alarm in the future.
Application example:
in 2018, parameters of a catalytic operation device of a certain chemical plant are as follows: the period of the correlation matrix calculation is two days, and the data of 2018.5.25 # 00:00 to 11:50 collected by the sensor are shown in the following table:
and calculating the correlation between the A.WL.MT.LT901A bit number and other bit numbers by using the correlation coefficient algorithm in the second step.
Wherein h is 35, n is 72,
the correlation coefficient r of the a.wl.mt.lt901a parameter with other bit numbers is as follows:
TABLE 2
If the absolute value of the correlation coefficient is greater than 0.9, the correlation is strong, and the 0.9 is summarized according to practical experience of working conditions, the parameters which are strongly correlated with the parameter A.WL.MT.LT901A are shown in the following table.
| Parameter name | y | x |
| External pressure | A.LY.YP.LI08.PV | A.WL.MT.LT901A |
| External heat-taking material level | A.LY.YP.LI8410.PV | A.WL.MT.LT901A |
| Regeneration temperature | A.LY.YP.LXH12408S.U | A.WL.MT.LT901A |
As known from the above table, three parameters of the take-out pressure, the take-out hot material level and the regeneration temperature are strongly related to a.wl.mt.lt901a, and are carried into 2018.5.25 # 00:00 to 11:50, constituting a correlation matrix a, namely: a ═ a.wl.mt.lt901a, a.ly.yp.li08.pv, a.ly.yp.li8410.pv, a.ly.yp.lxh12408s.u2018.5.25.00.00.002018.5.25 # 16:40A. WL. MT. LT901A, comparing the time difference, calling out the correlation matrix and bringing the current alarm data when the calculation time of the correlation matrix is closest to the alarm occurrence time
P1, P2.. Pn is a historical correlation matrix calculated according to the above correlation coefficient algorithm when a historical alarm is given,
And finally, traversing the historical incidence matrix to calculate that the distance between A and P4 is minimum, 8.3298. Therefore, the historical experience record corresponding to P4 is called from the historical database, that is: and (3) conducting external heat taking and tube explosion, reporting to a workshop leader and a production place by telephone, immediately arriving at an external heat taking inlet and outlet header for external operation outside a team to prepare for finding a leaking tube bundle, leading a workshop leader to organize on duty managers to arrive at an operation room for commanding accident treatment at the first time, leading the external operation by a reaction unit technician to cut the 1# tube bundle leaked from the external heat taking device, then recovering the production to be stable, and recovering all parameters to be normal.
The operation records provide operation suggestions for operators, and managers decide to adopt the suggestions according to own professional knowledge and experience accumulation, so that all parameters are restored to be stable finally.
Because the pressure of a plurality of tube bundles changes slowly when the tube bundles are cut out one by one, the judgment of leakage points is influenced to a certain extent. The tube bundle starts to leak a small amount, and the team and team can find, judge and process leakage accidents timely, accurately and properly, so that the situation is prevented from being enlarged, and normal production is ensured. If the accident is not found in time, the judgment is not accurate, the safe and stable operation of the device and the pressure of a 3.5MPa steam pipe network of the whole plant are directly endangered, and the stability of other devices is further influenced.
It should be noted that the above-mentioned embodiments are not intended to limit the scope of the present invention, and all equivalent modifications and substitutions based on the above-mentioned technical solutions are within the scope of the present invention as defined in the claims.
Claims (6)
1. A data-driven chemical production abnormal slice management method is characterized by comprising the following steps:
step 1, acquiring each bit number by using a sensor, wherein the bit number is a parameter acquired by the sensor;
step 2, carrying out periodic correlation analysis on the bit numbers to obtain correlation matrixes at different moments;
in the course of normal production processes,periodically calculating bit numbers which are strongly related to a target bit number to form a related bit number group, calculating other bit numbers with large change correlation with the target bit number through a correlation coefficient algorithm, expressing the target bit number by x, designating y as other k bit numbers participating in correlation calculation, and firstly calculating the correlation coefficient x of x and y to be [ x1, x2, x3, … xn],yk＝[yk1,yk2,yk3,…ykn](k is 1, … h), and mixing x, yknWith equation 1, the correlation coefficient of the two is calculated,
calculating the correlation coefficients of the two, taking the bit number of which the phase relation number r is greater than a preset value, and forming a strong correlation matrix together with the target bit number;
step 3, at a certain moment, alarming occurs to the target position number, and an incidence matrix closest to the alarming moment is found out through time difference;
step 4, calculating the similarity between the incidence matrix and the historical incidence matrix to obtain a group of historical incidence matrices with the highest similarity; at a certain moment in the step 4, when the target bit number gives an alarm, calculating the incidence matrix of the bit number at the moment, then performing similarity calculation with the historical incidence matrix of the bit number, and finding out a group of historical incidence matrices with the maximum similarity, wherein the similarity calculation between the incidence matrix and the historical incidence matrix at the moment adopts a mahalanobis distance algorithm:
the target bit number is denoted by x, and the strong correlation matrix when an alarm occurs is A ═ x, a1,a2…am]TA certain historical association matrix P ═ x, P1,p2…pm]E is the covariance matrix of the A, P matrix, then the distance of a and P is:
the historical incidence matrix corresponding to the small value D is most similar to the incidence matrix A;
step 5, calling out the historical operation corresponding to the most similar historical incidence matrix, and sending the historical operation to an operator;
and 6, after the alarm is solved, recording the incidence matrix and the operation of an operator to form a knowledge slice, and putting the knowledge slice into a historical database to serve as comparison information of future parameter alarm.
2. The data-driven chemical production abnormal slice management method based on claim 1, wherein the specific process of the step 2 is as follows: calculating the relevance between the alarm bit number and which bit numbers by adopting a relevance calculation method; the specific calculation method comprises the following steps: in the normal production process, the bit number strongly correlated with the target bit number is calculated once every 12 hours and a correlation matrix is formed.
3. The data-driven chemical production abnormal slice management method based on the claim 2, wherein the step 3 is specifically as follows, when the target position number generates an alarm, an alarm message is sent to an operator, and the incidence matrix closest to the alarm time is called from the database.
4. The data-driven chemical production abnormal slice management method based on the claim 3, wherein the step 5 presents the accident description and experience record corresponding to the historical association matrix to the operator as the operation reference information of the operator.
5. The data-driven chemical production abnormal slice management method according to claim 4, wherein the step 6: after the fault is relieved, the alarm parameter incidence matrix and the corresponding accident description and experience record are put into a historical database to be used as comparison information of the incidence matrix and corresponding operation guidance when the next target bit number gives an alarm.
6. The data-driven chemical production abnormal slice management method based on the claim 5, wherein the message push of the step 3 and the message push of the step 5 are both displayed to an operator through a mobile terminal, and the mobile terminal has personalized operation functions: whether to open a switch for receiving the alarm push information.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201811058446.4A CN109327330B (en) | 2018-09-11 | 2018-09-11 | Chemical production abnormal slice management method based on data driving |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201811058446.4A CN109327330B (en) | 2018-09-11 | 2018-09-11 | Chemical production abnormal slice management method based on data driving |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN109327330A CN109327330A (en) | 2019-02-12 |
| CN109327330B true CN109327330B (en) | 2022-02-22 |
Family
ID=65265033
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201811058446.4A Active CN109327330B (en) | 2018-09-11 | 2018-09-11 | Chemical production abnormal slice management method based on data driving |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN109327330B (en) |
Families Citing this family (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111899119B (en) * | 2020-06-16 | 2022-12-09 | 西安交通大学 | Intelligent alarm suppression method and system for coal-fired power plant |
Citations (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN201222183Y (en) * | 2008-07-02 | 2009-04-15 | 南京金炼科技有限公司 | Optimized control system for heating stove |
| CN102053613A (en) * | 2010-12-31 | 2011-05-11 | 北京三博中自科技有限公司 | Energy real-time monitoring system of industrial process equipment and monitoring method thereof |
| CN102176221A (en) * | 2011-03-16 | 2011-09-07 | 中南大学 | Coke furnace temperature predicting method based on dynamic working conditions in coke furnace heating and burning process |
| CN104750072A (en) * | 2015-03-12 | 2015-07-01 | 用友网络科技股份有限公司 | Production process monitoring method and system, monitoring device and monitoring system |
| CN105893700A (en) * | 2016-04-26 | 2016-08-24 | 陆新建 | Chemical production on-line fault detection and diagnosis technique based on physical-large data hybrid model |
| CN106020154A (en) * | 2016-07-12 | 2016-10-12 | 中国石油化工股份有限公司 | Safe dynamic health assessment method and assessment system for ethylene production |
| CN107390646A (en) * | 2016-05-17 | 2017-11-24 | 通用电气公司 | Intelligent management system for factory's balance |
| CN107479518A (en) * | 2017-08-16 | 2017-12-15 | 南京联成科技发展股份有限公司 | A kind of method and system for automatically generating alarm association rule |
| CN107958089A (en) * | 2017-12-19 | 2018-04-24 | 新疆金风科技股份有限公司 | Build the method and apparatus of model and the detection method and device of abnormal data |
Family Cites Families (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US10474684B2 (en) * | 2012-02-08 | 2019-11-12 | Telefonaktiebolaget Lm Ericsson (Publ) | Method, computer program, computer program product and system for handling sensor data |
| CN105518432B (en) * | 2013-09-06 | 2018-03-30 | 国际商业机器公司 | Detection device, detection method and computer-readable medium |
| CN103699118B (en) * | 2013-12-18 | 2016-08-31 | 北京航天测控技术有限公司 | A kind of method and device for analyzing abnormal state of spacecraft in operating process |
| EP2919124A1 (en) * | 2014-03-12 | 2015-09-16 | Haltian Oy | Relevance determination of sensor event |
| US10503156B2 (en) * | 2015-09-18 | 2019-12-10 | Fisher-Rosemount Systems, Inc. | Methods and apparatus to define stages for multi-variate batch control analytics |
| US10289109B2 (en) * | 2015-10-01 | 2019-05-14 | Globalfoundries Inc. | Methods of error detection in fabrication processes |
-
2018
- 2018-09-11 CN CN201811058446.4A patent/CN109327330B/en active Active
Patent Citations (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN201222183Y (en) * | 2008-07-02 | 2009-04-15 | 南京金炼科技有限公司 | Optimized control system for heating stove |
| CN102053613A (en) * | 2010-12-31 | 2011-05-11 | 北京三博中自科技有限公司 | Energy real-time monitoring system of industrial process equipment and monitoring method thereof |
| CN102176221A (en) * | 2011-03-16 | 2011-09-07 | 中南大学 | Coke furnace temperature predicting method based on dynamic working conditions in coke furnace heating and burning process |
| CN104750072A (en) * | 2015-03-12 | 2015-07-01 | 用友网络科技股份有限公司 | Production process monitoring method and system, monitoring device and monitoring system |
| CN105893700A (en) * | 2016-04-26 | 2016-08-24 | 陆新建 | Chemical production on-line fault detection and diagnosis technique based on physical-large data hybrid model |
| CN107390646A (en) * | 2016-05-17 | 2017-11-24 | 通用电气公司 | Intelligent management system for factory's balance |
| CN106020154A (en) * | 2016-07-12 | 2016-10-12 | 中国石油化工股份有限公司 | Safe dynamic health assessment method and assessment system for ethylene production |
| CN107479518A (en) * | 2017-08-16 | 2017-12-15 | 南京联成科技发展股份有限公司 | A kind of method and system for automatically generating alarm association rule |
| CN107958089A (en) * | 2017-12-19 | 2018-04-24 | 新疆金风科技股份有限公司 | Build the method and apparatus of model and the detection method and device of abnormal data |
Non-Patent Citations (2)
| Title |
|---|
| DCS系统过程控制功能的设计与实现;孙勇;《中国优秀硕士学位论文全文数据库》;20140515;全文 * |
| 朱群雄 ; 高慧慧 ; 徐圆.工业过程报警管理研究进展.《自动化学报》.2017, * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN109327330A (en) | 2019-02-12 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN105955219B (en) | Distributed dynamic procedure failure testing method based on mutual information | |
| CN117014472B (en) | Cloud side end cooperation-based intelligent power plant equipment inspection method and system | |
| US5799148A (en) | System and method for estimating a measure of confidence in a match generated from a case-based reasoning system | |
| CN104636450B (en) | A kind of database table space monitoring method | |
| CN102055604B (en) | Fault location method and system thereof | |
| CN109327330B (en) | Chemical production abnormal slice management method based on data driving | |
| CN101523316A (en) | Method and system for detecting faults in a process plant | |
| CN106932670A (en) | A kind of distribution power automation terminal method for diagnosing status based on D S evidence theories | |
| CN115858303B (en) | Zabbix-based server performance monitoring method and system | |
| CN104240779B (en) | Supervising method and device for quadrant incline of reactor core power of nuclear power station | |
| CN111722976A (en) | Fault flow analysis method, device and medium based on intelligent operation and maintenance | |
| CN104965978A (en) | Diagnosis failure probability calculation method and device | |
| CN105681128A (en) | Method and device for monitoring big data system state | |
| CN108536130A (en) | A kind of Fault Diagnosis in Chemical Process system of colony intelligence optimizing | |
| CN112729723A (en) | Intelligent gas remote monitoring and management method and system based on Internet of things and artificial intelligence, electronic equipment and computer storage medium | |
| CN104361242B (en) | Based on data-driven and the Method for Bearing Fault Diagnosis of random intuitionistic fuzzy strategy | |
| CN113159503B (en) | Remote control intelligent safety evaluation system and method | |
| CN115829192B (en) | Digital management system and method for realizing engineering information security supervision | |
| EP3945383A1 (en) | Management of a filter in continuous flow engine | |
| CN116843530A (en) | Carbon neutralization management system | |
| CN109617740A (en) | A kind of method and device that application failure quickly positions | |
| CN111637369B (en) | Water supply pipe network anti-seepage maintenance method and maintenance system | |
| CN104965652A (en) | Human error probability calculation method and device | |
| CN117590793B (en) | Water treatment integrated monitoring system and method for water plant based on Internet of things | |
| CN108956888B (en) | Monitoring method for humidity abnormity of intelligent industrial control equipment |
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 |