CN101634851A - Method based on cause-and-effect relation of variables for diagnosing failures in process industry - Google Patents
Method based on cause-and-effect relation of variables for diagnosing failures in process industry Download PDFInfo
- Publication number
- CN101634851A CN101634851A CN200910023676A CN200910023676A CN101634851A CN 101634851 A CN101634851 A CN 101634851A CN 200910023676 A CN200910023676 A CN 200910023676A CN 200910023676 A CN200910023676 A CN 200910023676A CN 101634851 A CN101634851 A CN 101634851A
- Authority
- CN
- China
- Prior art keywords
- variable
- cause
- effect
- node
- influence
- 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.)
- Granted
Links
Images
Classifications
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02P—CLIMATE CHANGE MITIGATION TECHNOLOGIES IN THE PRODUCTION OR PROCESSING OF GOODS
- Y02P90/00—Enabling technologies with a potential contribution to greenhouse gas [GHG] emissions mitigation
- Y02P90/02—Total factory control, e.g. smart factories, flexible manufacturing systems [FMS] or integrated manufacturing systems [IMS]
Abstract
The invention discloses a method based on the cause-and-effect relation of variables for diagnosing failures in the process industry. The method comprises the following steps: a human-computer interaction module inputs the description equations and type of the production process into a system; a variable cause-and-effect relation generating module generates a process variable cause-and-effect relation model according to the inputted equations and variables of the system, generates the failure matching rules on the basis of the process variable cause-and-effect relation model and stores the rules in a failure rule library in a relational database form; a variable state information acquisition module extracts the variable state information in a production process DCS (distributed control system) or a production monitoring real-time database, calculates the variable qualitative state discriminant parameter and sends the calculation result to a failure searching module; and the failure searching module searches the failure rule library according to the inputted variable abnormal state information, finds out the failure cause matching the inputted variable abnormal state information, and outputs the result through the human-computer interaction module. The invention is capable of clearing the trouble and preventing the occurring of the production accidents.
Description
Technical field
The present invention relates to a kind of process industry fault diagnosis system, particularly a kind of process industry method for diagnosing faults based on variable cause and effect influence relation.
Prior art
The fault pilosity takes place frequently in the process industry production run, often causes serious damage sequence and bigger economic loss, and production status monitoring timely and effectively and fault diagnosis technology can reduce system failure occurrence frequency, and the assurance security of system is efficiently moved.Existing fault diagnosis technology is divided into based on quantitative model, based on qualutative model with based on operation history data three classes.Based on process mathematical model, can draw accurate diagnostic results, but process industry production system apparatus is many based on the fault diagnosis technology of quantitative model, complex structure, and follow various chemical reactions, the systematic quantification model is difficult for obtaining accurately.Fault diagnosis technology based on operation history data does not rely on system model, status signal in the general using production run, as the rotary axis vibrating signal, methods such as employing spectrum analysis are extracted the failure message that comprises in the signal and are realized fault diagnosis, these class methods generally are confined to the fault diagnosis of large-scale rotary axis device and equipment, and adopt offline mode, real-time is not enough more, and is powerless to the fault diagnosis of other equipment in the process industry simultaneously.The characteristics that are fit to the process industry production system based on the fault diagnosis technology of qualutative model, as fault tree (FTA), event tree (ETA), signed digraph (SDG) etc., these methods do not rely on system mathematic model, various devices, equipment and the process that can adapt to process industry simultaneously, but also therefore there is the diagnostic result out of true, the defective that deceptive information is more, and to large complicated process industry system, the fault search efficient of these class methods is low, and real-time is poor, has influenced the application in actual production process.On the whole, in the existing fault diagnosis technology, qualitative fault diagnosis technology meets process industry production run characteristics, but it is accurate inadequately to the production run portrayal to need to solve its qualutative model, comprise more deceptive information, the not high problem of simultaneous faults searching method efficient.
Summary of the invention
The present invention is directed to the problems referred to above, a kind of process industry method for diagnosing faults based on variable cause and effect influence relation is provided, can be on the basis that accurately generates the qualitative cause and effect influence of variable relation, adopt diagnosis rule match search method, fault diagnosis under the quick realization flow industrial processes abnormality, production control system and operating personnel can be foundation with the native system diagnostic result, adjust processing parameter, or take maintenance measure targetedly, in time fix a breakdown, avoid taking place industrial accident.
Process industry method for diagnosing faults based on variable cause and effect influence relation of the present invention comprises:
1) operating personnel are by descriptive equation group and the variable of human-computer interaction module to system's input production run, and production run type, the production run type comprises equilibrium process, dynamic process and mixed process three classes, the process of describing with Algebraic Equation set is an equilibrium process, the process of describing with differential equation group is a dynamic process, and mixing the process of describing with the algebraic equation and the differential equation is mixed process;
2) variable cause and effect influence concern that generation module influence relational model according to the system equation and the variable generative process variable cause and effect of input, serves as basic generation fault matched rule with it then, and with these rale store in the diagnosis rule storehouse of relational database form;
3) variable status information capture module is extracted variable status information in production run DCS control system or the production monitoring real-time data base according to variable threshold, calculates the qualitative condition discrimination parameter of variable, and result of calculation is sent to the fault search module;
4) the fault search module is found out the failure cause that is complementary with input variable abnormality information according to the variable abnormality information search diagnosis rule storehouse of input, and the result is exported by human-computer interaction module;
Process industry fault diagnosis system of the present invention is a foundation with the production run system equation, it is accurately irredundant that its variable cause and effect influences relational model information, adopt the rule match way of search to carry out fault search, quick and precisely, real-time, adopt the process industry fault diagnosis system based on variable cause and effect influence relation of the present invention can diagnose out the failure cause of production system under the abnormality fast and accurately, diagnostic result can the guiding operation personnel in time be taked maintenance measure, the prosthetic appliance fault, or as the parameter input control system, with this adjusting process parameter, correcting system is unusual, the prevention industrial accident.
Description of drawings
Fig. 1 is a flow process graph of a relation of the present invention;
Fig. 2 influences relation model figure for the variable cause and effect;
Fig. 3 is equilibrium process variable causal ordering figure;
Fig. 4 is dynamic process variable incidence relation figure;
Fig. 5 variable cause and effect influences relational model branch road exploded view;
Below in conjunction with accompanying drawing content of the present invention is described in further detail.
Embodiment
With reference to shown in Figure 1, the process industry method for diagnosing faults that influence concerns based on the variable cause and effect is by constituting with the lower part: human-computer interaction module, the influence of variable cause and effect concern generation module, diagnosis rule storehouse, variable status information capture module, fault search module.The present invention is object with the production run, system equation and variable information by human-computer interaction module input production run, concern that by the influence of variable cause and effect generation module generates the variable cause and effect according to input information and influences relational model, on the model basis, generate the fault matched rule, and deposit it in diagnosis rule storehouse, variable status information capture module is responsible for gathering the variable status information in the production run DCS supervisory system (or real-time data base RTDB), calculate the qualitative condition discrimination parameter of variable, and result of calculation is sent to fault diagnosis module, fault diagnosis module compares the fault matched rule in abnormality variable and state and the diagnosis rule storehouse, search out failure cause, and diagnostic result is exported by human-computer interaction module;
With reference to shown in Figure 2, the tape label letter is the production process variable, represent (being illustrated as the circle form) with node, the oriented line between the node is represented cause and effect influence relation, points to result node by the reason node, solid line is represented just to influence, promptly when reason node variable value raises (or reduction), cause the result node variate-value to raise equally (or reduction), dotted line is represented negatively influencing, when being reason node variable value rising (or reduction), cause the result node variate-value to reduce (or rising) on the contrary;
With reference to shown in Figure 3, tape label letter representation production run variable, directed line segment are represented the causal sequence relation between the variable, point to outcome variable by causal variable, among the figure, and variable V
3, V
5, V
6, V
7Be variable V
8Causal variable, variable V
4Be variable V
5Causal variable, variable V
2Be variable V
3Causal variable, variable V
1Be variable V
2Causal variable;
With reference to shown in Figure 4, constitute by the interrelated relation between dynamic process variable, variable derivative, variable and variable derivative, dotted line is represented the causalnexus relation between variable and the variable derivative, and the solid line that has the i sign is the integration chain, and the integral relation of this variable itself is pointed in expression by the variable derivative;
With reference to shown in Figure 5, expression influences relational model is decomposed into no closed-loop path from the form that has the closed-loop path decomposition branch road form process with the variable cause and effect, the cause and effect of variable shown in the legend influences relational model and is made up of A, B, C, D, six variablees of E, F and mutual cause and effect influence relation thereof, the same Fig. 2 of method for expressing, this model connects the limit branch road and converges at node D and F place, form the closed-loop path, model can be disassembled from this two node and be the decomposition branch road form of four kinds of no closed-loop paths as shown in the figure;
Equilibrium process, dynamic process, mixed process, the variable cause and effect that relates among the present invention influenced relational model, exogenous variable, the qualitative condition discrimination parameter of variable notion to be done as giving a definition:
Equilibrium process: refer to adopt Algebraic Equation set
Definite production run of formal description is represented with S, and the Algebraic Equation set of this process can be separated and separate unique, comprises n variable, a n equation, and satisfies following condition:
(a) k is individual arbitrarily (in the system of equations that 0＜k≤n) equation is formed, has at least and contains k above nonzero coefficient variable in the equation;
(b) in the system of equations of any k equation composition, if m 〉=k nonzero coefficient variable occur, get wherein the value of (m-k) individual variable arbitrarily, a remaining k variate-value can come definite by the group of solving an equation;
Dynamic process: refer to adopt the differential equation of first order group
(i=1 ..., n) formal description (the differential equation of higher order form can be converted into the differential equation of first order form), the process of separating is determined in qualification within reason and can drawing, and represents that with D the differential equation group of this class process can be separated and separate unique, comprise n variable, a n equation, and satisfy following condition:
(a) any k (in the system of equations that 0＜k≤n) equation is formed, the first order derivative of k different variablees occurs at least;
(b) r occurs (in the system of equations that k equation of the individual first order derivative of r 〉=k) formed, appoint and get wherein the value of (r-k) individual first order derivative arbitrarily, a remaining k derivative value can be used as the function of n variable in the system of equations and unique definite any;
Mixed process: mix definite process of describing by the algebraic equation and the differential equation, represent with M, the subclass that whole differential equations are formed among the M is designated as D (M), the subclass (V is that derivative appears at the variable subset among the D (M)) that the constant equation of whole balance equations and variable V constitutes among the M is designated as S (M), such process variable value can be separated and separate unique, comprise n variable, a n equation, and satisfy following condition:
(a) one or more in n equation are differential equation of first orders, and remaining is an algebraic equation;
(b) in the k of D (M) the unit subclass, the first order derivative of k different variablees appears at least;
(c) (among the r 〉=k), get (r-k) individual first order derivative arbitrarily, k remaining variable can uniquely be determined in the k of any D (M) that r first order derivative occur unit subclass;
(d) have the first order derivative of d different variablees to appear among the D (M) just, wherein d is the number of equation among the set D (M);
(e) S (M) is a balanced system.
The variable cause and effect influences relational model: but a kind of graphic model of representing cause and effect influence relation between the production run variable, adopt node to represent variable, adopt the oriented limit that connects to represent variable cause and effect dependence, the outcome variable node is pointed to by the causal variable node in the oriented limit that connects, the oriented limit that connects of full lines is represented just to influence, promptly when reason node variable value raises (or reduction), cause result node variate-value raise equally (or reduce), negatively influencing is represented on the oriented limit that connects of dashed line form, when being reason node variable value rising (or reduction), cause the result node variate-value to reduce (or rising) on the contrary;
Exogenous variable: the variable of determining by system's external factor in the balance portion system equation of equilibrium process or mixed process, these variablees are not subjected to the influence of internal system operation, its value variation is determined by external environment condition, it is the reason of inner other variable change of decision systems, be called exogenous variable, as the material input of procedures system, environment temperature etc.
The qualitative condition discrimination parameter of variable: discriminatory variable value is in the parameter of higher, normal, on the low side these three kinds of qualitative states, three kinds of qualitative states that variable has, higher, normal, on the low side, represent that respectively variate-value is higher than upper threshold (with+1 expression), variate-value (with 0 expression), variate-value within threshold range and is lower than threshold value lower limit (with-1 expression), variable higher and on the low side is in abnormality;
Native system is an object with the process industry production run, according to system equation type difference, the process industry production run is divided three classes: equilibrium process, dynamic process and mixed process, operating personnel are input to system equation and variable information in the system by human-computer interaction module, to the balance portion that comprises in equilibrium process and the mixed process, determine its exogenous variable,, determine the intermediate variable of its balance portion and dynamic part mixed process;
The influence of variable cause and effect concerns that generation module receives the system equation and the variable information of human-computer interaction module input, and according to the difference of production run type, generating the variable cause and effect respectively influences relational model.
To equilibrium process S, its variable cause and effect influences the relational model generative process and is:
1) determine all subclass S ' of S, S ' is made up of the part equation among the S, and S ' is for meeting the equilibrium process of above-mentioned definition, and its inside no longer comprises the balance subprocess, claims that S ' is the minimum complete subclass of S;
2) establish S
0Be the union of all minimum complete subclass of S, claim S
0Be the complete subclass in 0 rank, find the solution S
0System of equations draws its variate-value;
3) bring the variate-value that solves into system of equations (S-S
0) in, obtain a new equilibrium process, be called the derivation process, repeating step 1), find out all minimum complete subclass of this derivation process; If S
1Be the union of the minimum complete subclass of this derivation process, be called the complete subclass of single order;
4) repeating step 3), find out each rank and derive process and minimum complete subclass thereof, derive process up to high-order, the balance subprocess is no longer contained in this process inside;
5) minimum complete subclass variable cause and effect depends on the minimum complete subclass variable of next rank process in the high-order process, and to each equation among the S, depend on its dependent variable the equation from the variable cause and effect of eliminating at last, determine that with this causal ordering between variable concerns that equilibrium process variable causal ordering representation is with reference to Fig. 3;
6) setting up on all variable cause and effect dependence bases, will have the residing equation of cause and effect dependence variable and be transformed to
Form, coefficient a
IjSymbol promptly represent cause and effect dependence x
j→ x
iThe cause and effect impact effect, positive sign is for just influencing (promote influence), negative sign is negatively influencing (suppressing influence);
7) adopt as shown in Figure 2 the variable cause and effect to influence relational model and represent balanced system variable cause and effect influence relation.
To dynamic process D, its variable cause and effect influences the relational model generative process and is:
1) the given dynamic process that contains n variable carries out solving n derivative after the conversion, draw shape as
(i=1 ..., n n) canonical form equation;
2) with the derivative dx of each equation equation left side variable
iBe outcome variable, the non-vanishing variable of equation the right coefficient is a causal variable, generates causalnexus relation between variable and variable derivative;
3) with the variable x on each equation equation left side
iBe outcome variable, with its derivative dx
iBe causal variable, generate cause and effect integration chain relation, through step 2) and 3) the dynamic process variable incidence relation form of back generation is with reference to Fig. 4;
4) yojan variable cause and effect integration chain is removed variable derivative dx
i, generate dynamic system variable cause and effect dependence;
5) on the basis of the above each differential equation variable is asked local derviation, ask
It is worth positive and negative definite cause and effect impact effect, and positive sign is for just influencing (promoting influence), and negative sign is negatively influencing (suppressing influence);
6) adopt as shown in Figure 2 the variable cause and effect to influence relational model and represent balanced system variable cause and effect influence relation.
To mixed process M, its variable cause and effect influences the relational model generative process and is:
1) concerns that according to above-mentioned dynamic process variable cause and effect influence generative process generates the variable cause and effect influence relation of its dynamic part D (M);
2) concern that according to above-mentioned equilibrium process cause and effect influence generative process generates the system variable cause and effect influence relation of its balance portion S (M);
3) the intermediate variable V in merging D (M) subclass variable cause and effect influence relation and S (M) the subclass variable cause and effect influence relation, drawing mixed process variable cause and effect influences relation;
4) adopt as shown in Figure 2 the variable cause and effect to influence relational model and represent mixed process variable cause and effect influence relation.
According to diagnosis production run object type difference, the influence of variable cause and effect concerns that generation module generates corresponding variable cause and effect and influences relational model, on this model basis, generates the fault matched rule by following steps:
1) finding out the variable cause and effect influences that all converge branch road in the relational model, with reference to Fig. 5 variables D and F, from converging the branch road variable node model is decomposed into no closed branch road form, with reference to Fig. 5, from node D and F model is decomposed into four no closed branch road models;
2) choosing the variable cause and effect influences a certain node of relational model as start node, and with reference to Fig. 5, choosing variables A is start node;
3) from this node, influence model along the variable cause and effect and connect the limit and arrive its connected node, just influencing as Lian Bianwei and connect the limit, then should connect limit with the p prefix designates, with reference to Fig. 5, A, B two nodes connect the limit for just influencing, and generate the pAB rule, even A, the B jack per line (be all+1, or be all-1), pAB=TRUE then, otherwise pAB=FALSE, otherwise, Ruo Lianbian is a negatively influencing, then should connect the limit with the m prefix designates, with reference to Fig. 5, B, it is negatively influencing that C two nodes connect the limit, generate the mBC rule, B even, (B is+1 to the C contrary sign, C be-1 or B be-1, C is 1), mBC=TRUE then, otherwise mBC=FALSE;
4) adopt traversal mode based on breadth First, generate one by one according to step 3) and respectively connect the limit rule, decompose branch road rules up to generating all, each rule links to each other with the and conjunction; With reference to Fig. 5, its four decomposition branch road rules are followed successively by:
(pAB)and(pBD)and(pDF)and(mAC)and(pCE)、
(pAB)and(mAC)and(pCD)and(mCE)and(pEF)、
(pAB)and(pBD)and(pEF)and(mAC)and(pCE)、
(pAB)and(pBD)and(pDF)and(mAC)and(pCE)；
5) the branch road rule is respectively decomposed in merging, and the branch road meet links to each other with the or conjunction, sets up the fault matched rule, and start node is a source of trouble node, and with reference to Fig. 5, the fault matched rule that above-mentioned decomposition branch road rule merges back foundation is:
IF[(pAB) and and (mAC) and (pCE)] and[(pBD) or (pCD)] and[(pDF) or (pEF)]=TRUETHEN A is the source of trouble };
6) choosing the variable cause and effect influences that another node is a start node in the relational model, repeating step 1)～5), next bar fault matched rule set up;
According to above-mentioned steps, generate all fault matched rules that the variable cause and effect influences model, it is deposited in the diagnosis rule storehouse;
Variable status information capture module is gathered the variable status information by input interface from production run DCS control system (or real-time data base RTDB system), this information comprises the current measured value of variable and variable operate as normal value and working range threshold value thereof, and establishing the current measured value of variable is V
C, variable operate as normal value V
N, variable threshold is V
T, variable state value S
VBe calculated as follows:
The qualitative condition discrimination parameter of variable J
VBe calculated as follows:
The variable information acquisition module calculates the qualitative condition discrimination parameter value of variable, and with its input fault search module;
The fault search module according to production run with the variable abnormality information of the variable state acquisition module input fault matched rule in the driving malfunction rule base one by one, finding out all operation result values is the rule of TRUE, the corresponding source node of these rules is the failure cause node that meets under the current production run abnormality, after diagnosis is finished this diagnostic result is outputed to display interface by human-computer interaction module, so that operating personnel take maintenance measure, the prosthetic appliance fault, or be input in the production control system as parameter, with this as reference adjusting process parameter, repair the production run abnormality, prevent industrial accident.
Claims (4)
1. the process industry method for diagnosing faults based on variable cause and effect influence relation is characterized in that, may further comprise the steps:
1) the descriptive equation group by human-computer interaction module input production run
(i=1 ..., n), x
iThe expression process variable, a
IjExpression variation coefficient, and the type of production run comprise equilibrium process, dynamic process and mixed process;
2) variable cause and effect influence concern that generation module generates the variable cause and effect according to the descriptive equation group of input and variable and influence relational model, serves as basic generation fault matched rule with it then, and with these rale store in the diagnosis rule storehouse of relational database form;
3) variable status information capture module is gathered actual production process variable status signal, calculates the qualitative condition discrimination parameter of variable, and is transported to the fault search module;
4) the fault search module is found out failure cause, and it is exported by human-computer interaction module according to the variable abnormality information search diagnosis rule storehouse of input.
2. method according to claim 1 is characterized in that: described generation variable cause and effect influences the relational model process and may further comprise the steps:
To equilibrium process (being designated as S), its variable cause and effect influences the relational model generative process and is:
1) determine all subclass S ' of S, S ' is made up of the part equation among the S, and S ' is an equilibrium process, and its inside no longer comprises the balance subprocess, claims that S ' is the minimum complete subclass of S;
2) establish S
0Be the union of all minimum complete subclass of S, claim S
0Be the complete subclass in 0 rank, find the solution S
0System of equations draws its variate-value;
3) bring the variate-value that solves into system of equations (S-S
0) in, obtain a new equilibrium process, be called the derivation process, repeating step 1) find out all minimum complete subclass of this derivation process; If S
1Be the union of the minimum complete subclass of this derivation process, be called the complete subclass of single order;
4) repeating step 3), find out each rank and derive process and minimum complete subclass thereof, derive process up to high-order, the balance subprocess is no longer contained in this process inside;
5) minimum complete subclass variable cause and effect depends on the minimum complete subclass variable of next rank process in the high-order process, and to each equation among the S, depends on its dependent variable the equation from the variable cause and effect of eliminating at last, determines that with this causal ordering between variable concerns;
6) setting up on all variable cause and effect dependence bases, will have the residing equation of cause and effect dependence variable and be transformed to
Form, coefficient a
IjSymbol promptly represent cause and effect dependence x
j→ x
iThe cause and effect impact effect, positive sign is for just influencing (promote influence), negative sign is negatively influencing (suppressing influence);
7) represent variable with node, the oriented line between the node is represented cause and effect influence relation, points to result node by the reason node, represents just to influence with solid line, and dotted line is represented negatively influencing;
Dynamic process is designated as D, and its variable cause and effect influences the relational model generative process and is:
1) the given dynamic process that contains n variable carries out solving n derivative after the conversion, draw shape as
(i=1 ..., n n) canonical form equation;
2) with the derivative dx of each equation equation left side variable
iBe outcome variable, the non-vanishing variable of equation the right coefficient is a causal variable, generates causalnexus relation between variable and variable derivative;
3) with the variable x on each equation equation left side
iBe outcome variable, with its derivative dx
iBe causal variable, generate cause and effect integration chain relation, pass through step 2) and 3) the back dynamic process variable incidence relation that generates;
4) yojan variable cause and effect integration chain is removed variable derivative dx
i, generate dynamic system variable cause and effect dependence;
5) on the basis of the above each differential equation variable is asked local derviation, ask
It is worth positive and negative definite cause and effect impact effect, and positive sign is for just influencing (promoting influence), and negative sign is negatively influencing (suppressing influence);
6) represent variable with node, the oriented line between the node is represented cause and effect influence relation, points to result node by the reason node, represents just to influence with solid line, and dotted line is represented negatively influencing;
To mixed process (be designated as M, its dynamic ingredient is designated as D (M), and equilibrium composition partly is designated as S (M)), its variable cause and effect influences the relational model generative process and is:
1) concerns that according to above-mentioned dynamic process variable cause and effect influence generative process generates the variable cause and effect influence relation of its dynamic part D (M);
2) concern that according to above-mentioned equilibrium process variable cause and effect influence generative process generates the variable cause and effect influence relation of its balance portion S (M);
3) the intermediate variable V in merging D (M) subclass variable cause and effect influence relation and S (M) the subclass variable cause and effect influence relation, drawing mixed process variable cause and effect influences relation;
4) represent variable with node, the oriented line between the node is represented cause and effect influence relation, points to result node by the reason node, represents just to influence with solid line, and dotted line is represented negatively influencing;
3. method according to claim 1 is characterized in that: described fault matched rule generative process may further comprise the steps:
1) finding out the variable cause and effect influences that all converge branch road in the relational model, from converging the branch road variable node model is decomposed into no closed branch road form;
2) choose the variable cause and effect and influence a certain node of relational model as start node;
3) from this node, influencing model along the variable cause and effect connects the limit and arrives its connected node, just influencing even limit as Lian Bianwei, then should connect the limit with the p prefix designates, generate pAB rule (A, B represent two node variables that link to each other), if A, B jack per line (be all+1, or be all-1), pAB=TRUE then, otherwise pAB=FALSE, otherwise Ruo Lianbian is a negatively influencing, then should connect the limit with the m prefix designates, generate the mAB rule, even A, B contrary sign (A is+1, B for-1 or A for-1, B is _ 1), mAB=TRUE then, otherwise mAB=FALSE;
4) adopt traversal mode based on breadth First, generate one by one according to step 3) and respectively connect the limit rule, decompose branch road rules up to generating all, each rule links to each other with the and conjunction;
5) the branch road rule is respectively decomposed in merging, and the branch road meet links to each other with the or conjunction, sets up the fault matched rule, and start node is a source of trouble node;
6) choosing the variable cause and effect influences that another node is a start node in the relational model, repeating step 1)～5), next bar fault matched rule set up, up to generating all variable node fault matched rules;
4. method according to claim 1 is characterized in that: the qualitative condition discrimination parameter calculation procedure of described variable is as follows:
1) gather the variable status information by input interface from production run DCS control system (or real-time data base RTDB system), this information comprises the current measured value of variable and variable operate as normal value and working range threshold value thereof, and the current measured value of note variable is V
C, variable operate as normal value V
N, variable threshold is V
T
2) variable state value S
VBe calculated as follows:
3) the qualitative condition discrimination parameter of variable J
VBe calculated as follows:
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN2009100236761A CN101634851B (en) | 2009-08-25 | 2009-08-25 | Method based on cause-and-effect relation of variables for diagnosing failures in process industry |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN2009100236761A CN101634851B (en) | 2009-08-25 | 2009-08-25 | Method based on cause-and-effect relation of variables for diagnosing failures in process industry |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN101634851A true CN101634851A (en) | 2010-01-27 |
| CN101634851B CN101634851B (en) | 2011-09-21 |
Family
ID=41594064
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN2009100236761A Expired - Fee Related CN101634851B (en) | 2009-08-25 | 2009-08-25 | Method based on cause-and-effect relation of variables for diagnosing failures in process industry |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN101634851B (en) |
Cited By (11)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102103715A (en) * | 2010-11-18 | 2011-06-22 | 上海海事大学 | Negative binomial regression-based maritime traffic accident investigation analysis and prediction method |
| WO2011131079A1 (en) * | 2010-04-20 | 2011-10-27 | 杭州和利时自动化有限公司 | Event processing method and system for distributed control system |
| CN102707712A (en) * | 2012-06-06 | 2012-10-03 | 广州山锋测控技术有限公司 | Electronic equipment fault diagnosis method and system |
| CN102736546A (en) * | 2012-06-28 | 2012-10-17 | 西安交通大学 | State monitoring device of complex electromechanical system for flow industry and method |
| CN104503434A (en) * | 2014-12-01 | 2015-04-08 | 北京航天试验技术研究所 | Fault diagnosis method based on active fault symptom pushing |
| CN105955241A (en) * | 2016-06-03 | 2016-09-21 | 北京科技大学 | Quality fault locating method based on federated data driven production process |
| CN109816940A (en) * | 2019-03-21 | 2019-05-28 | 北京天诚同创电气有限公司 | The fault alarm method and device of sewage treatment plant |
| CN111413491A (en) * | 2020-04-11 | 2020-07-14 | 深圳市资通科技有限公司 | Oil-filled equipment insulation aging evaluation platform |
| CN112114564A (en) * | 2019-06-19 | 2020-12-22 | 恩格尔奥地利有限公司 | Device for monitoring a production facility |
| CN113467326A (en) * | 2015-10-09 | 2021-10-01 | 费希尔-罗斯蒙特系统公司 | System and method for configuring separate monitor and result blocks of a process control system |
| CN114091710A (en) * | 2022-01-20 | 2022-02-25 | 广东智修互联大数据有限公司 | National maintenance technology supporting method and system |
-
2009
- 2009-08-25 CN CN2009100236761A patent/CN101634851B/en not_active Expired - Fee Related
Cited By (17)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2011131079A1 (en) * | 2010-04-20 | 2011-10-27 | 杭州和利时自动化有限公司 | Event processing method and system for distributed control system |
| CN102103715A (en) * | 2010-11-18 | 2011-06-22 | 上海海事大学 | Negative binomial regression-based maritime traffic accident investigation analysis and prediction method |
| CN102707712A (en) * | 2012-06-06 | 2012-10-03 | 广州山锋测控技术有限公司 | Electronic equipment fault diagnosis method and system |
| CN102707712B (en) * | 2012-06-06 | 2014-06-18 | 广州山锋测控技术有限公司 | Electronic equipment fault diagnosis method and system |
| CN102736546A (en) * | 2012-06-28 | 2012-10-17 | 西安交通大学 | State monitoring device of complex electromechanical system for flow industry and method |
| CN102736546B (en) * | 2012-06-28 | 2014-06-04 | 西安交通大学 | State monitoring device of complex electromechanical system for flow industry and method |
| CN104503434B (en) * | 2014-12-01 | 2017-05-03 | 北京航天试验技术研究所 | Fault diagnosis method based on active fault symptom pushing |
| CN104503434A (en) * | 2014-12-01 | 2015-04-08 | 北京航天试验技术研究所 | Fault diagnosis method based on active fault symptom pushing |
| CN113467326A (en) * | 2015-10-09 | 2021-10-01 | 费希尔-罗斯蒙特系统公司 | System and method for configuring separate monitor and result blocks of a process control system |
| CN105955241A (en) * | 2016-06-03 | 2016-09-21 | 北京科技大学 | Quality fault locating method based on federated data driven production process |
| CN105955241B (en) * | 2016-06-03 | 2018-09-14 | 北京科技大学 | A kind of quality fault localization method based on joint data-driven production process |
| CN109816940A (en) * | 2019-03-21 | 2019-05-28 | 北京天诚同创电气有限公司 | The fault alarm method and device of sewage treatment plant |
| CN109816940B (en) * | 2019-03-21 | 2023-05-09 | 北京天诚同创电气有限公司 | Fault alarm method and device for sewage treatment plant |
| CN112114564A (en) * | 2019-06-19 | 2020-12-22 | 恩格尔奥地利有限公司 | Device for monitoring a production facility |
| CN111413491A (en) * | 2020-04-11 | 2020-07-14 | 深圳市资通科技有限公司 | Oil-filled equipment insulation aging evaluation platform |
| CN114091710A (en) * | 2022-01-20 | 2022-02-25 | 广东智修互联大数据有限公司 | National maintenance technology supporting method and system |
| CN114091710B (en) * | 2022-01-20 | 2022-04-12 | 广东立升数字技术有限公司 | National maintenance technology supporting method and system |
Also Published As
| Publication number | Publication date |
|---|---|
| CN101634851B (en) | 2011-09-21 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN101634851B (en) | Method based on cause-and-effect relation of variables for diagnosing failures in process industry | |
| CN104298225B (en) | Chemical process unusual service condition causality inference pattern is modeled and graphical representation method | |
| CN104573850B (en) | A kind of Power Plant Equipment state evaluating method | |
| US8868985B2 (en) | Supervised fault learning using rule-generated samples for machine condition monitoring | |
| CN109501834B (en) | Method and device for predicting fault of turnout switch machine | |
| CN106932670B (en) | A kind of distribution power automation terminal method for diagnosing status based on D-S evidence theory | |
| DE102009025855A1 (en) | Systems and methods for modern condition monitoring of an asset system | |
| DE19808197C2 (en) | System and method for diagnosing engine conditions | |
| CN107423414A (en) | A kind of process industry complex electromechanical systems fault source tracing method based on information transmission model | |
| EP0768584B1 (en) | Method for monitoring a machine or plant | |
| CN109933905A (en) | A kind of mechanical equipment health state evaluation method based on multidimensional early warning analysis | |
| CN101446831A (en) | Decentralized process monitoring method | |
| CN103901880A (en) | Industrial process fault detection method based on multiple classifiers and D-S evidence fusion | |
| CN104267668A (en) | Bayes-method-based spaceflight valve part fault diagnosis method in machining process | |
| WO2022171788A1 (en) | Prediction model for predicting product quality parameter values | |
| CN104569666A (en) | Power transformer fault prediction method based on electricity-graph model | |
| CN104268338B (en) | Complex product failure effect transfer relation model analysis and evaluation method | |
| Samhouri et al. | An intelligent machine condition monitoring system using time-based analysis: neuro-fuzzy versus neural network | |
| CN115081647A (en) | Industrial intelligent instrument fault pre-diagnosis method based on Bayesian network model | |
| CN106933097A (en) | A kind of Fault Diagnosis for Chemical Process method based on multi-level optimization PCC SDG | |
| CN103745261A (en) | Method for dynamic fault diagnosis by constructing three-dimensional DUCG intelligent system | |
| CN106647274A (en) | Steady-state discrimination method for operation conditions in continuous production process | |
| EP4193227A1 (en) | Device and method for identifying anomalies in an industrial system for carrying out a production process | |
| Hussain et al. | Tuning of fault semantic network using Bayesian theory for probabilistic fault diagnosis in process industry | |
| M’halla et al. | Monitoring of a milk manufacturing workshop using chronicle and fault tree approaches |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| C14 | Grant of patent or utility model | ||
| GR01 | Patent grant | ||
| CF01 | Termination of patent right due to non-payment of annual fee |
Granted publication date: 20110921 Termination date: 20150825 |
|
| EXPY | Termination of patent right or utility model |