CA2406899A1 - Process and automation of an industrial process in steps, with mastery of an uncertain stress chain, and its application for control of noise and of the risk (value-at-risk, var) of a clearing house - Google Patents
Process and automation of an industrial process in steps, with mastery of an uncertain stress chain, and its application for control of noise and of the risk (value-at-risk, var) of a clearing house Download PDFInfo
- Publication number
- CA2406899A1 CA2406899A1 CA002406899A CA2406899A CA2406899A1 CA 2406899 A1 CA2406899 A1 CA 2406899A1 CA 002406899 A CA002406899 A CA 002406899A CA 2406899 A CA2406899 A CA 2406899A CA 2406899 A1 CA2406899 A1 CA 2406899A1
- Authority
- CA
- Canada
- Prior art keywords
- production
- uncertain
- abovementioned
- industrial
- electronically
- 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
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
- G06Q40/00—Finance; Insurance; Tax strategies; Processing of corporate or income taxes
- G06Q40/02—Banking, e.g. interest calculation or account maintenance
-
- G—PHYSICS
- G05—CONTROLLING; REGULATING
- G05B—CONTROL OR REGULATING SYSTEMS IN GENERAL; FUNCTIONAL ELEMENTS OF SUCH SYSTEMS; MONITORING OR TESTING ARRANGEMENTS FOR SUCH SYSTEMS OR ELEMENTS
- G05B19/00—Programme-control systems
- G05B19/02—Programme-control systems electric
- G05B19/418—Total factory control, i.e. centrally controlling a plurality of machines, e.g. direct or distributed numerical control [DNC], flexible manufacturing systems [FMS], integrated manufacturing systems [IMS], computer integrated manufacturing [CIM]
- G05B19/41865—Total factory control, i.e. centrally controlling a plurality of machines, e.g. direct or distributed numerical control [DNC], flexible manufacturing systems [FMS], integrated manufacturing systems [IMS], computer integrated manufacturing [CIM] characterised by job scheduling, process planning, material flow
-
- 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
- G06Q10/0631—Resource planning, allocation, distributing or scheduling for enterprises or organisations
- G06Q10/06312—Adjustment or analysis of established resource schedule, e.g. resource or task levelling, or dynamic rescheduling
-
- 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
- G06Q10/0637—Strategic management or analysis, e.g. setting a goal or target of an organisation; Planning actions based on goals; Analysis or evaluation of effectiveness of goals
- G06Q10/06375—Prediction of business process outcome or impact based on a proposed change
-
- 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 process comprises regulating production by means of a probabilistic automatic control (3) with action loop (5) and feedback loop (6). The action loop (5) of the automatism (3) comprises an inductive probabilistic simulator (11) evaluating the chaining of random stresses in the production chain leading to a probabilistic measurement of the industrial impact I (r) resulting, on the basis of the adjustable level of an industrial stock parameter (r). The industrial action parameter (r) is adjusted over time to an extremal value, maintaining the estimator of VaR effect (p,T(r)) below an authorised nuisance level M.
Claims (4)
1) Process to regulate flow F, the flow being multi-stage and multi-linked aiming to optimise this production flow (F) by acting on an industrial action parameter (r), while mastering an industrial impact (I) resulting from an uncertain chained stress at the different production steps, this regulation process being specifically applicable to a multi-stage or multi-linked production, i.e. a production that is:
- composed of several production steps or stages called E1,...,E m, - where each production step E i is composed of productive subsystems Si,j, - receiving one or more production subflows F i,j,k from one or more subsystems S i-1,k from the preceding step E i-1 (except at the first step where i=1).
- transmitting one or more production subflows F i+1,j,k to one or more subsystems Si+1, in the following step E i+1 (except at the last step where i=m), - whose production subflows F i,j,k can be controlled with the aid of an industrial action parameter, possibly multivariable, (r)= (r1,...,r n) - and whose global industrial production flow F(r) results from a combination of the production subflows F i,j,k, this regulation process is specifically applied to a production of the type which has a chain of uncertain stresses, i.e. a production:
- which has an industrial impact parameter I(X, T) (r) (in general harmful), possibly multivariable (I) - (I1,...,I H).
- which needs to be mastered (especially for regulatory reasons) within a time horizon T, - and where the impact parameter is the result of a cascading chain of aggregated stresses Wi,j (measurable phenomena) suffered by the productive subsystems S i,j, and where the abovementioned industrial impact parameter I (X, T) (r) depends on - the globally monotone variation (increasing or decreasing) of the aggregated stresses W i,j of the subsystems S i,j at the different production steps E i, where at least one of the elementary components of the impact I h depends on the aggregated stresses W m,j of the subsystems S m,j at the last production step E m, - an environmental uncertain production multivariable (X) = (x1,...,X N), where each of the X i; is called a subfactor of the unknown production value, (X), as well as on the - the time horizon T
- and the globally monotone variation (increasing or decreasing, in the same sense as the relation to aggregated stresses W m,j) of the industrial action parameter (r) via the intermediary aggregated stresses W m,j, where each aggregated stress W i,j is, in the normal mode, contained at the system level S i,j, but which can, in failure mode, be partially or totally transmitted to the following step E i+l, that is where each aggregated stress W i,j is the result (sum) of:
- a "self" stress W i,j at the productive subsystem, depending - in a known manner and specific to each subsystem - on an uncertain multivariable environmental factor (X) as well as on the globally increasing function of the industrial impact parameter (r), and - the stress transmitted W i,j,k by certain productive subsystems S i-1,k in the previous step E i-1, where each of the transmitted stresses W i,j,k is the combination, actually the product, of:
- the aggregated stresses W i-1,k from the subsystems S i-1,k, - a linkage coefficient q i,j,k, constant or at least known, not uncertain, and an uncertain transmission coefficient d i-i,k from the productive subsystem S i,l,k, varying between 0 and 1, whose elementary transmission probability distribution Pr i,j(X,T,r,W i,j,a i,j) of the productive subsystem S,j i is known and depends on - the multivariable factor X of the uncertain production value, - the time interval T, - the industrial action parameter r, - the stress W i,j and - a confidence coefficient a i,j, possibly multivariable, pertaining to the productive subsystem S i.j, - and whose characteristics result in particular but exclusively, from a historical analysis of productive subsystem failure, Whose production configuration can be:
- either tree-like - for each subsystem S i,k at a step E i, one of the linkage coefficients q i+i,j,k=1 and the other q i,l,j,k are zero, which corresponds to the case where one productive subsystem at stage E i can transmit its stress only to a single productive subsystem at the following stage E i+1, - or it can be in the more general case, matrix-like, and in this case each of the linkage coefficient q i,j,k can take any value, where the abovementioned transmission coefficients d i,j may be - either binary, they take only the values 0 and 1, i.e. a productive subsystem fails or it does not fail, but there is no such thing as a partial failure -and, in this case, the elementary probability distribution for failure, Pr i,j(X,T,r,W i,j,a i,j) for each productive subsystem S i,j is reduced to a single number p i,j(X,T,r,W i,j,a i,j) which is the probability that a failure will take place here, (i.e. d ij=1), - or, in the more general case, any value -that is they can take all values between 0 and 1;
partial failures are possible and the elementary probability distribution for failure is Prig (X, T, r,W i,j, a i,j) for each productive subsystem S i,j and a positive function defined on the interval [0,1]
of the possible transmission coefficient values, this production control process is specifically implemented in the case:
- where the variations over the time interval T
of the uncertain sub-values X;, which constitute the abovementioned multivariable X for the uncertain environment - are quantifiable by a known probability rule Prob (x1,...,x N) , where x i is the generic state which the uncertain subvalue X i can take - and where, in general, the characteristics, especially the average, the variance and the correlations, as well as the extremes of behaviour, are the results of a statistical analysis of the historical record of the uncertain subvalues X1, - and where the elementary impact components I n (X, T) (r) of the abovementioned industrial impact I (X, T) (r) from the production flow F(r) must not violate upwards (type 1) or downwards (type 2) the impact levels Mh, and that at an industrial probabilistic confidence level Prob [Ih(X,T)(r)<Mh]>.RHO.h (type 1) or Prob [Ih(X,T)(r)>Mh]>.RHO.h (type 2) where the abovementioned harm level Mh and the probabilities ph are typically imposed by regulation;
this process is specifically suited, in a known manner - to electronically perform a sampling which is as true as possible, and which is a function of the industrial action parameter (r), in the time interval (T), of the industrial production state, consisting of:
- the values which the uncertain subfactors X i can take, - the stresses W i,j for the productive subsystems S i,j, - their transmission coefficients d i,j, - the elementary components Ih (X, T) (r), - the production subflows F i,j,k (r), - The production flow F(r), - to electronically generate, always for a time horizon T, and as a function of the possible values for the industrial impact parameter (r) - the action loop (5) for production control automation (3) - the multivariable estimator for maximal impact (p) -(P 1.....P m), defined as a vector VaR (p,T) (r) of the limits VaR h (p h,T) (r) such that the abovementioned elementary impact components are only violated (upwards or downwards according to the type of system) with a probability (1-ph) , that is Prob [Ih(X,T)(r)<M n]>p h (type 1) or Prob [Ih(X,T)(r)>Mh]>ph (type 2), - and to regulate over time - reaction loop (6) in the production control automation (3) - the level for the industrial action parameter (r) to the extremal multivariable value (r max) or (r min) while maintaining the abovementioned estimator for the effect VaR(p,T) (r) on the right side of its authorized harm level M, that is, for each index h, VaR h, (ph,T) (I max or rain) < M n (type 1) or VaRh (ph, T) (rmaX or rm2n) > Mh (type 2), to render the production level F(r) extremal, while nevertheless compatible with respect to the abovementioned regulatory condition that the harm be controlled such that Prob [I n(X,T)(r)<N n]>p n (type 1) or Prob [I n(X,T)(r)>M n]>P n (type 2), this process is additionally specifically of a type capable of electronically performing an abovementioned sampling of the "Monte-Carlo" type, whose the global number of random draws is Z, end to do this, proceeds generally in the following fashion:
- to make a choice (validated by adequate classical statistical tests) of a simultaneous behavior model for each of the uncertain production subfactors (it can be, in particular, normal, log-normal or more generally or more generally, a distribution of levels of uncertain production subfactors (X i) justified by observations of historical data for this uncertain value), - then, less critically, (with the goal of accelerating the control loop, and thus to improve the performance and reliability of the control automation (3), by reducing the time horizon T) to perform an analysis by principal components (ACP) of the different uncertain subfactors, that is to consider the relations which are expressed by the behavior of these uncertain production subfactors (X i) as a function of common uncertain factors, independent factors among them, which, in the form of an uncertain indexed multivariable factor for the production environment (Y) - (Y+, Y 2,...,Y G), where the (Y g) are called the indexed common uncertain subfactors of production, with, in general, G<<N, - after that, starting from a part of the abovementioned behavior modeller for the various uncertain production subfactors (X i) and the distribution parameters (Y i) for the indexed uncertain values, electronically construct according to the Monte-Carlo method:
- either a number Z of pseudo-random samples of the state vector, x z =(x z,1, s z,2,...,x z,N) z=1,...,Z for the possible values of the abovementioned uncertain production subfactors (X i), [this electronic construction is performed starting from the parameters that describe each of the abovementioned uncertain sub-factors X i taken individually, but also starting from the correlations linking them to each other, especially according to the known decomposition methods by Cholesky and the "singular values"], - or, when the ACP is used, a number of pseudo-random values Z giving the index state y z -(y z,1.y z,2,....y z,G), z=1,...,Z , the possible values of the abovementioned common uncertain index production subfactors (Y g), then electronically determine, for each pseudo-random sample of the abovementioned vector, the index state y z, the value corresponding to an uncertain multivariable factor (x z)=(x z,1,x z,2,...,x z,N) according to the coefficients from the analysis of principal components, to electronically determine, for each pseudo-random sample of the abovementioned specific state vector x z, and as a function of the abovementioned action parameter (r), the level corresponding to the multivariable industrial impact vector I (x z, T(r), - to electronically organize and bring together the Z results, and for each of the elementary impact components I n to electronically take into account for each value V which can affect the component I n the number Z n(V) of the abovementioned electronic samples for which this impact component I n(x z, T)(r) violates (upwards or downwards, depending on whether we have a control system of type 1 or 2) the value V and thus to electronically calculate the pseudo-probability p'h(V) - Z h(V)/Z of violating the value V for the abovementioned impact component I h, - to electronically deduce from that the variations of VaRh (p h,T)(r) for an imposed p h, and this as a function of the abovementioned action parameter (r)[defined by Pr[I h(X,T)(r)>(or<)VaR h] - (1-phi], - to determine the extremal multivalue, [r max]
or [r min] according to the type of system, for which the multivariable estimator VaR (p,T)(r) is exactly the regulatory value M and to specifically adjust, by a process using servo actuators or servo motors, the action variable (r) to this level, (possibly multiple), the abovementioned process having been characterised in that, to electronically generate the elementary impact components I h (x2,T)(r) corresponding to the each sample of the uncertain quantity x z, we electronically determine for each pseudo-random sample of the abovementioned uncertain multivariable (xZ), and this as a function of the industrial action parameter (r), the level of aggregated stress W z,j,j for each of the productive subsystems S i,j by a inductive method (11), starting at the first step E1 and working towards the last step Em, that is to say that:
- starting from the first step E1, we measure the level of the abovementioned "self" stress W1,j (x Z) for each of the subsystems S1,j at step E1 of the of the industrial production, a) we determine, only for the step under consideration, E1, the self stress W1,j (xz) for each of the subsystems 51,3 with the aggregated stress W Z,1,j, b) then, we electronically perform a pseudo-random production of the abovementioned transmission coefficient d Z,1,j for each of the productive subsystems S1,j at the step E1, using the abovementioned probability distribution for the elementary failures, Pr1,j (x z,T,r,W z,1,j,a1,j) for the productive subsystems s1,j, that is, we electronically generate a pseudo-random number uz,1,j in the interval [0,1), and we electronically apply it to the inverse distribution function ~1,j (x z,T,r,W z,1,j,a1,j) (u z,1,j) giving the probability of elementary failure Pr1,j (x z,T,r,WZ,1,j,a1,j) for the productive subsystems S1,j, c) we measure the level of the abovementioned self stress W2,k (x z) for each of the subsystems S2,k at step E2 for the industrial production, d) we electronically evaluate the abovementioned aggregated stress for each subsystem S2,k at step E2 using the formula:
W z,2,k = W2,k (S z) + .SIGMA. j W' z,2,j,k=
W2,k (x2) + .SIGMA.w2,1,j(X z) ~ d2,1,j ~ q2,j,k, - we iterate these operations (b through d) for each step until we get the aggregated stress W z,m,j for the productive subsystems S m,j at step E m, - and we deduce the multivariable industrial impact I (x Z, T,) (r) linked to the sample with index z;
in such a manner that for each Monte-Carlo sample which constitutes a multivariable environmental factor x z and the emitted transmission coefficients d z,j,j:
- we measure the industrial impact variable I
while keeping track of the chained stress cascade W z,i,j at each production step and of the uncertain character of the transmission coefficients d z,i,j, - from this we obtain a more precise measurement of the probability of violating a level given by the supplied value V by the impact I(r) and thus, of the maximal impact VaR (p, T) (r), in such a manner that - the effective violation frequencies for the authorized limits fox the elementary components of the impact, I n, are closer to the target values, (1-ph) /T, - this allows a reduction of the industrial safety margins to be applied at the resistance level M, and, as a consequence, increases the production flaw F(r) while still respecting the given regulations, - we can build, thanks to this control and automation process, a more efficient production control system.
- composed of several production steps or stages called E1,...,E m, - where each production step E i is composed of productive subsystems Si,j, - receiving one or more production subflows F i,j,k from one or more subsystems S i-1,k from the preceding step E i-1 (except at the first step where i=1).
- transmitting one or more production subflows F i+1,j,k to one or more subsystems Si+1, in the following step E i+1 (except at the last step where i=m), - whose production subflows F i,j,k can be controlled with the aid of an industrial action parameter, possibly multivariable, (r)= (r1,...,r n) - and whose global industrial production flow F(r) results from a combination of the production subflows F i,j,k, this regulation process is specifically applied to a production of the type which has a chain of uncertain stresses, i.e. a production:
- which has an industrial impact parameter I(X, T) (r) (in general harmful), possibly multivariable (I) - (I1,...,I H).
- which needs to be mastered (especially for regulatory reasons) within a time horizon T, - and where the impact parameter is the result of a cascading chain of aggregated stresses Wi,j (measurable phenomena) suffered by the productive subsystems S i,j, and where the abovementioned industrial impact parameter I (X, T) (r) depends on - the globally monotone variation (increasing or decreasing) of the aggregated stresses W i,j of the subsystems S i,j at the different production steps E i, where at least one of the elementary components of the impact I h depends on the aggregated stresses W m,j of the subsystems S m,j at the last production step E m, - an environmental uncertain production multivariable (X) = (x1,...,X N), where each of the X i; is called a subfactor of the unknown production value, (X), as well as on the - the time horizon T
- and the globally monotone variation (increasing or decreasing, in the same sense as the relation to aggregated stresses W m,j) of the industrial action parameter (r) via the intermediary aggregated stresses W m,j, where each aggregated stress W i,j is, in the normal mode, contained at the system level S i,j, but which can, in failure mode, be partially or totally transmitted to the following step E i+l, that is where each aggregated stress W i,j is the result (sum) of:
- a "self" stress W i,j at the productive subsystem, depending - in a known manner and specific to each subsystem - on an uncertain multivariable environmental factor (X) as well as on the globally increasing function of the industrial impact parameter (r), and - the stress transmitted W i,j,k by certain productive subsystems S i-1,k in the previous step E i-1, where each of the transmitted stresses W i,j,k is the combination, actually the product, of:
- the aggregated stresses W i-1,k from the subsystems S i-1,k, - a linkage coefficient q i,j,k, constant or at least known, not uncertain, and an uncertain transmission coefficient d i-i,k from the productive subsystem S i,l,k, varying between 0 and 1, whose elementary transmission probability distribution Pr i,j(X,T,r,W i,j,a i,j) of the productive subsystem S,j i is known and depends on - the multivariable factor X of the uncertain production value, - the time interval T, - the industrial action parameter r, - the stress W i,j and - a confidence coefficient a i,j, possibly multivariable, pertaining to the productive subsystem S i.j, - and whose characteristics result in particular but exclusively, from a historical analysis of productive subsystem failure, Whose production configuration can be:
- either tree-like - for each subsystem S i,k at a step E i, one of the linkage coefficients q i+i,j,k=1 and the other q i,l,j,k are zero, which corresponds to the case where one productive subsystem at stage E i can transmit its stress only to a single productive subsystem at the following stage E i+1, - or it can be in the more general case, matrix-like, and in this case each of the linkage coefficient q i,j,k can take any value, where the abovementioned transmission coefficients d i,j may be - either binary, they take only the values 0 and 1, i.e. a productive subsystem fails or it does not fail, but there is no such thing as a partial failure -and, in this case, the elementary probability distribution for failure, Pr i,j(X,T,r,W i,j,a i,j) for each productive subsystem S i,j is reduced to a single number p i,j(X,T,r,W i,j,a i,j) which is the probability that a failure will take place here, (i.e. d ij=1), - or, in the more general case, any value -that is they can take all values between 0 and 1;
partial failures are possible and the elementary probability distribution for failure is Prig (X, T, r,W i,j, a i,j) for each productive subsystem S i,j and a positive function defined on the interval [0,1]
of the possible transmission coefficient values, this production control process is specifically implemented in the case:
- where the variations over the time interval T
of the uncertain sub-values X;, which constitute the abovementioned multivariable X for the uncertain environment - are quantifiable by a known probability rule Prob (x1,...,x N) , where x i is the generic state which the uncertain subvalue X i can take - and where, in general, the characteristics, especially the average, the variance and the correlations, as well as the extremes of behaviour, are the results of a statistical analysis of the historical record of the uncertain subvalues X1, - and where the elementary impact components I n (X, T) (r) of the abovementioned industrial impact I (X, T) (r) from the production flow F(r) must not violate upwards (type 1) or downwards (type 2) the impact levels Mh, and that at an industrial probabilistic confidence level Prob [Ih(X,T)(r)<Mh]>.RHO.h (type 1) or Prob [Ih(X,T)(r)>Mh]>.RHO.h (type 2) where the abovementioned harm level Mh and the probabilities ph are typically imposed by regulation;
this process is specifically suited, in a known manner - to electronically perform a sampling which is as true as possible, and which is a function of the industrial action parameter (r), in the time interval (T), of the industrial production state, consisting of:
- the values which the uncertain subfactors X i can take, - the stresses W i,j for the productive subsystems S i,j, - their transmission coefficients d i,j, - the elementary components Ih (X, T) (r), - the production subflows F i,j,k (r), - The production flow F(r), - to electronically generate, always for a time horizon T, and as a function of the possible values for the industrial impact parameter (r) - the action loop (5) for production control automation (3) - the multivariable estimator for maximal impact (p) -(P 1.....P m), defined as a vector VaR (p,T) (r) of the limits VaR h (p h,T) (r) such that the abovementioned elementary impact components are only violated (upwards or downwards according to the type of system) with a probability (1-ph) , that is Prob [Ih(X,T)(r)<M n]>p h (type 1) or Prob [Ih(X,T)(r)>Mh]>ph (type 2), - and to regulate over time - reaction loop (6) in the production control automation (3) - the level for the industrial action parameter (r) to the extremal multivariable value (r max) or (r min) while maintaining the abovementioned estimator for the effect VaR(p,T) (r) on the right side of its authorized harm level M, that is, for each index h, VaR h, (ph,T) (I max or rain) < M n (type 1) or VaRh (ph, T) (rmaX or rm2n) > Mh (type 2), to render the production level F(r) extremal, while nevertheless compatible with respect to the abovementioned regulatory condition that the harm be controlled such that Prob [I n(X,T)(r)<N n]>p n (type 1) or Prob [I n(X,T)(r)>M n]>P n (type 2), this process is additionally specifically of a type capable of electronically performing an abovementioned sampling of the "Monte-Carlo" type, whose the global number of random draws is Z, end to do this, proceeds generally in the following fashion:
- to make a choice (validated by adequate classical statistical tests) of a simultaneous behavior model for each of the uncertain production subfactors (it can be, in particular, normal, log-normal or more generally or more generally, a distribution of levels of uncertain production subfactors (X i) justified by observations of historical data for this uncertain value), - then, less critically, (with the goal of accelerating the control loop, and thus to improve the performance and reliability of the control automation (3), by reducing the time horizon T) to perform an analysis by principal components (ACP) of the different uncertain subfactors, that is to consider the relations which are expressed by the behavior of these uncertain production subfactors (X i) as a function of common uncertain factors, independent factors among them, which, in the form of an uncertain indexed multivariable factor for the production environment (Y) - (Y+, Y 2,...,Y G), where the (Y g) are called the indexed common uncertain subfactors of production, with, in general, G<<N, - after that, starting from a part of the abovementioned behavior modeller for the various uncertain production subfactors (X i) and the distribution parameters (Y i) for the indexed uncertain values, electronically construct according to the Monte-Carlo method:
- either a number Z of pseudo-random samples of the state vector, x z =(x z,1, s z,2,...,x z,N) z=1,...,Z for the possible values of the abovementioned uncertain production subfactors (X i), [this electronic construction is performed starting from the parameters that describe each of the abovementioned uncertain sub-factors X i taken individually, but also starting from the correlations linking them to each other, especially according to the known decomposition methods by Cholesky and the "singular values"], - or, when the ACP is used, a number of pseudo-random values Z giving the index state y z -(y z,1.y z,2,....y z,G), z=1,...,Z , the possible values of the abovementioned common uncertain index production subfactors (Y g), then electronically determine, for each pseudo-random sample of the abovementioned vector, the index state y z, the value corresponding to an uncertain multivariable factor (x z)=(x z,1,x z,2,...,x z,N) according to the coefficients from the analysis of principal components, to electronically determine, for each pseudo-random sample of the abovementioned specific state vector x z, and as a function of the abovementioned action parameter (r), the level corresponding to the multivariable industrial impact vector I (x z, T(r), - to electronically organize and bring together the Z results, and for each of the elementary impact components I n to electronically take into account for each value V which can affect the component I n the number Z n(V) of the abovementioned electronic samples for which this impact component I n(x z, T)(r) violates (upwards or downwards, depending on whether we have a control system of type 1 or 2) the value V and thus to electronically calculate the pseudo-probability p'h(V) - Z h(V)/Z of violating the value V for the abovementioned impact component I h, - to electronically deduce from that the variations of VaRh (p h,T)(r) for an imposed p h, and this as a function of the abovementioned action parameter (r)[defined by Pr[I h(X,T)(r)>(or<)VaR h] - (1-phi], - to determine the extremal multivalue, [r max]
or [r min] according to the type of system, for which the multivariable estimator VaR (p,T)(r) is exactly the regulatory value M and to specifically adjust, by a process using servo actuators or servo motors, the action variable (r) to this level, (possibly multiple), the abovementioned process having been characterised in that, to electronically generate the elementary impact components I h (x2,T)(r) corresponding to the each sample of the uncertain quantity x z, we electronically determine for each pseudo-random sample of the abovementioned uncertain multivariable (xZ), and this as a function of the industrial action parameter (r), the level of aggregated stress W z,j,j for each of the productive subsystems S i,j by a inductive method (11), starting at the first step E1 and working towards the last step Em, that is to say that:
- starting from the first step E1, we measure the level of the abovementioned "self" stress W1,j (x Z) for each of the subsystems S1,j at step E1 of the of the industrial production, a) we determine, only for the step under consideration, E1, the self stress W1,j (xz) for each of the subsystems 51,3 with the aggregated stress W Z,1,j, b) then, we electronically perform a pseudo-random production of the abovementioned transmission coefficient d Z,1,j for each of the productive subsystems S1,j at the step E1, using the abovementioned probability distribution for the elementary failures, Pr1,j (x z,T,r,W z,1,j,a1,j) for the productive subsystems s1,j, that is, we electronically generate a pseudo-random number uz,1,j in the interval [0,1), and we electronically apply it to the inverse distribution function ~1,j (x z,T,r,W z,1,j,a1,j) (u z,1,j) giving the probability of elementary failure Pr1,j (x z,T,r,WZ,1,j,a1,j) for the productive subsystems S1,j, c) we measure the level of the abovementioned self stress W2,k (x z) for each of the subsystems S2,k at step E2 for the industrial production, d) we electronically evaluate the abovementioned aggregated stress for each subsystem S2,k at step E2 using the formula:
W z,2,k = W2,k (S z) + .SIGMA. j W' z,2,j,k=
W2,k (x2) + .SIGMA.w2,1,j(X z) ~ d2,1,j ~ q2,j,k, - we iterate these operations (b through d) for each step until we get the aggregated stress W z,m,j for the productive subsystems S m,j at step E m, - and we deduce the multivariable industrial impact I (x Z, T,) (r) linked to the sample with index z;
in such a manner that for each Monte-Carlo sample which constitutes a multivariable environmental factor x z and the emitted transmission coefficients d z,j,j:
- we measure the industrial impact variable I
while keeping track of the chained stress cascade W z,i,j at each production step and of the uncertain character of the transmission coefficients d z,i,j, - from this we obtain a more precise measurement of the probability of violating a level given by the supplied value V by the impact I(r) and thus, of the maximal impact VaR (p, T) (r), in such a manner that - the effective violation frequencies for the authorized limits fox the elementary components of the impact, I n, are closer to the target values, (1-ph) /T, - this allows a reduction of the industrial safety margins to be applied at the resistance level M, and, as a consequence, increases the production flaw F(r) while still respecting the given regulations, - we can build, thanks to this control and automation process, a more efficient production control system.
2) Process according to the claim 1, regulating a multistep and multilinked industrial production flow F, aiming to optimize this production flow F, by acting on an industrial action parameter (r), while mastering an industrial impact variable I resulting in a uncertain stress chain at different production steps, the abovementioned procedure being characterised additionally by that we electronically impose that the transmission coefficients d z,i,j for the productive subsystems Si,j be higher when the aggregated stress W z,i,j is larger, and we use the facts that:
- for any threshold value < 1, we electronically fix the probability law Pri,j (x z, T, r, W z,i,j, ai,j) (d z,I,J>d] such that it increases with the aggregated stress W z, I, j, and/or we electronically fix the parameters for the description of the inverse distribution function .PHI. i,j (x z,T,r,W z,i,j,ai,j) (u z,i,j) in such a manner that it becomes an increasing function of the parameter W z,i,j, all other parameters and variables, including u z,i, j, remaining fixed, such that:
- at each production step, we induce a larger transmission probability for a given proportion of stresses, towards the next step, - we take into account the fact that, in the majority of cases, the increased transmission coefficients will appear precisely when the stress is large, pulling with it a sensitive increase in the average level of stress transmitted to the next higher step, and, as a consequence, the value of the globally induced effect, such that:
- we correct one of the shortcomings of classical control systems, which by not using this complementary process, bring a much too low evaluation of the number of cases where the industrial impact variable violates the authorised limit M, - and we avoid one of the shortcomings of classical control systems which, in order to respect industrial norms, tend to use much larger safety margins, and thus to reduce the production flow.
- for any threshold value < 1, we electronically fix the probability law Pri,j (x z, T, r, W z,i,j, ai,j) (d z,I,J>d] such that it increases with the aggregated stress W z, I, j, and/or we electronically fix the parameters for the description of the inverse distribution function .PHI. i,j (x z,T,r,W z,i,j,ai,j) (u z,i,j) in such a manner that it becomes an increasing function of the parameter W z,i,j, all other parameters and variables, including u z,i, j, remaining fixed, such that:
- at each production step, we induce a larger transmission probability for a given proportion of stresses, towards the next step, - we take into account the fact that, in the majority of cases, the increased transmission coefficients will appear precisely when the stress is large, pulling with it a sensitive increase in the average level of stress transmitted to the next higher step, and, as a consequence, the value of the globally induced effect, such that:
- we correct one of the shortcomings of classical control systems, which by not using this complementary process, bring a much too low evaluation of the number of cases where the industrial impact variable violates the authorised limit M, - and we avoid one of the shortcomings of classical control systems which, in order to respect industrial norms, tend to use much larger safety margins, and thus to reduce the production flow.
3) Procedure according to the patent claim 1, regulation of a multi-step multi-linked industrial production flow F, aiming to optimise the production flow F, by acting on an industrial action parameter (r), while mastering an industrial impact I resulting from the uncertain chained stress at different production stages, this regulatory process being specifically applicable to a production of chained diversified stress, that is a production where the productive subsystems Si,j have independent size and reliability characteristics, - that is to say that certain productive subsystems Si,j can be of small size, in terms of subflow as well as in terms of their "self" stress, being - reliable at their scale, that is to say their transmission coefficients for the aggregated stress of the same order of magnitude as their "self" stress will be small on average, - but being such that a large aggregated stress can bring with it almost certain failure - while other productive subsystems Si,j have the reverse property, that is to say - their productive subflows are just as high as their "self" stress, - their transmission coefficient has a high average, but it is relatively stable, even when the aggregated stress is large, - the abovementioned procedure being characterised additionally by that we electronically fix the confidence coefficient a i,j un the form which possesses at least two independent components for the transmission coefficient:
- we electronically link the first component of the confidence coefficient a ij to the size of the productive subsystem S i,j, - and we electronically link the second component of the confidence coefficient to the reliability of the subsystem S i,j relative to its size, in such a way that - we obtain a more reliable and precise estimate of the reality of the chained production failures, which allows a reduction of the safety margins, and, as a consequence, an increase in the production flow F(r).
- we electronically link the first component of the confidence coefficient a ij to the size of the productive subsystem S i,j, - and we electronically link the second component of the confidence coefficient to the reliability of the subsystem S i,j relative to its size, in such a way that - we obtain a more reliable and precise estimate of the reality of the chained production failures, which allows a reduction of the safety margins, and, as a consequence, an increase in the production flow F(r).
4) Process according to patent claim 1, regulation of a multi-step multi-linked industrial production flow F, aiming to optimise the production flow F, by acting on an industrial action parameter (r), while mastering an industrial impact I resulting from an uncertain stress chain in different production steps, this control process is applicable specifically to production where the multivariable environmental production factor X is subject, with a small probability, to large unpredictable movements, the abovementioned procedure being additionally characterised by that (to electronically perform the abovementioned sampling of the state of production by a pseudo-random method of "Monte-Carlo" type, where the number of random draws is Z) we proceed electronically using a combination of history and catastrophes characterised as follows:
- we make a choice (validated by adequate classical tests) of instantaneous behaviour models for the different uncertain production subfactors (Xi), (they can be for example a normal distribution, log-normal, or more generally, a distribution of levels of uncertain production subfactors (Xi) justified by observations of historical data for these uncertain values), to which a standard probability P S of occurrence is attributed and for which we generate Z S
standard pseudo-random samples of the state vector x Z =
(x z,1, x z,2,..., x z,N) ( for z - 1, ..., Z s possible values of the abovementioned common uncertain subfactors indexed to production (X1)], and we apply to each of these Z S -called standard samples - a weight called standard, m s = P s/Z s.
- we make a choice of one or several so called "catastrophic scenarios" with different uncertain production factors (X1); indeed the subfamilies of catastrophic situations from which the characteristic averages and deviations are defined - either in the absolute - or by referring to the characteristics of distribution obtained from analysis of the historical record, - on which we apply an occurrence probability P c (or a set of probabilities (p c1,...,P cn) if there are several catastrophic scenarios) and for which we electronically generate Z c pseudo-random samples for the index vector X Z - (X Z,1,X Z,2,...,X Z, N), [Z = 1,...,Z the possible values for the abovementioned common factors for the uncertain production subfactors (X1)] (or the sub-series composed of Zc1,...Zcn samples) and we apply to each of the Z c, called "standard samples", a weight, called catastrophic, m c = P c/Z c (or weights m c1,...,m cn):
- we electronically determine for each of the Z
= Z p + Z c1 + ... Z cn) pseudo-random samples of the abovementioned specific state vector x z, and as a function of the abovementioned action parameter (r), the level corresponding to the multivariable industrial impact vector I (x z, T) (r), - we electronically reorganise and bring together the Z results, and, for each elementary impact component I h we calculate for each value V which can reach the component I h, the weight Z Ph (V) of the abovementioned samples [sum of the weights of electronic samples for which the abovementioned impact component I h (x2, T) (r) violates (too low or too high, depending on whether production of type 1 or 2 is controlled) the value V multiplied by the associated weights M S or M c)) and then we electronically calculate the pseudo-probability p'h (V) /Z ph of violating the value V by this impact component Ih, then, in a classical fashion - from this, we deduce electronically the variations of VaR h(ph,T)(r) for the imposed value of ph, this a function of the abovementioned action variable (r), [defined by Pr[I h (X, T) (r) > (or <) VaR h ] = (1-ph)], - we determine the extremal multi-value (r max) or (r min) depending on the type of system, for which the multivalued estimator VaR (p, T) (r) takes exactly the value allowed by regulation, or of breakage M, and - we adjust by a servo activator or servo motor the value of the action parameter (r) to this level, possibly multivariable;
such that - we correct the bias observed between the actual probability distribution of the uncertain values (X1,...,X N) and those of the values taken during the historical recording of the data, - if certain events (with grave consequences and whose probabilities cannot be considered negligible) do not in fact happen during the period when the historical record was made, we nevertheless impose on the control automation to take them into account, - if, by a compensatory phenomenon, the simulation of a definite event did not any important industrial impact, thanks to the simulation of the sub-family of events, we will nevertheless avoid this fortuitous compensation, and keep the real risk linked to this catastrophe.
- we make a choice (validated by adequate classical tests) of instantaneous behaviour models for the different uncertain production subfactors (Xi), (they can be for example a normal distribution, log-normal, or more generally, a distribution of levels of uncertain production subfactors (Xi) justified by observations of historical data for these uncertain values), to which a standard probability P S of occurrence is attributed and for which we generate Z S
standard pseudo-random samples of the state vector x Z =
(x z,1, x z,2,..., x z,N) ( for z - 1, ..., Z s possible values of the abovementioned common uncertain subfactors indexed to production (X1)], and we apply to each of these Z S -called standard samples - a weight called standard, m s = P s/Z s.
- we make a choice of one or several so called "catastrophic scenarios" with different uncertain production factors (X1); indeed the subfamilies of catastrophic situations from which the characteristic averages and deviations are defined - either in the absolute - or by referring to the characteristics of distribution obtained from analysis of the historical record, - on which we apply an occurrence probability P c (or a set of probabilities (p c1,...,P cn) if there are several catastrophic scenarios) and for which we electronically generate Z c pseudo-random samples for the index vector X Z - (X Z,1,X Z,2,...,X Z, N), [Z = 1,...,Z the possible values for the abovementioned common factors for the uncertain production subfactors (X1)] (or the sub-series composed of Zc1,...Zcn samples) and we apply to each of the Z c, called "standard samples", a weight, called catastrophic, m c = P c/Z c (or weights m c1,...,m cn):
- we electronically determine for each of the Z
= Z p + Z c1 + ... Z cn) pseudo-random samples of the abovementioned specific state vector x z, and as a function of the abovementioned action parameter (r), the level corresponding to the multivariable industrial impact vector I (x z, T) (r), - we electronically reorganise and bring together the Z results, and, for each elementary impact component I h we calculate for each value V which can reach the component I h, the weight Z Ph (V) of the abovementioned samples [sum of the weights of electronic samples for which the abovementioned impact component I h (x2, T) (r) violates (too low or too high, depending on whether production of type 1 or 2 is controlled) the value V multiplied by the associated weights M S or M c)) and then we electronically calculate the pseudo-probability p'h (V) /Z ph of violating the value V by this impact component Ih, then, in a classical fashion - from this, we deduce electronically the variations of VaR h(ph,T)(r) for the imposed value of ph, this a function of the abovementioned action variable (r), [defined by Pr[I h (X, T) (r) > (or <) VaR h ] = (1-ph)], - we determine the extremal multi-value (r max) or (r min) depending on the type of system, for which the multivalued estimator VaR (p, T) (r) takes exactly the value allowed by regulation, or of breakage M, and - we adjust by a servo activator or servo motor the value of the action parameter (r) to this level, possibly multivariable;
such that - we correct the bias observed between the actual probability distribution of the uncertain values (X1,...,X N) and those of the values taken during the historical recording of the data, - if certain events (with grave consequences and whose probabilities cannot be considered negligible) do not in fact happen during the period when the historical record was made, we nevertheless impose on the control automation to take them into account, - if, by a compensatory phenomenon, the simulation of a definite event did not any important industrial impact, thanks to the simulation of the sub-family of events, we will nevertheless avoid this fortuitous compensation, and keep the real risk linked to this catastrophe.
Applications Claiming Priority (3)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| FR99/05074 | 1999-04-21 | ||
| FR9905074A FR2792746B1 (en) | 1999-04-21 | 1999-04-21 | METHOD AND AUTOMATION OF REGULATION OF A STAGE INDUSTRIAL PRODUCTION WITH CONTROL OF A RANDOM STRESS STRESS, APPLICATION TO THE CONTROL OF THE NOISE AND THE RISK OF A COMPENSATION CHAMBER |
| PCT/FR2000/001059 WO2000065418A2 (en) | 1999-04-21 | 2000-04-21 | Method and automatic control for regulating a multiple-stage industrial production controlling random chained stress, application to noise and value at risk control of a clearing house |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CA2406899A1 true CA2406899A1 (en) | 2000-11-02 |
| CA2406899C CA2406899C (en) | 2009-11-24 |
Family
ID=9544723
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CA002406899A Expired - Fee Related CA2406899C (en) | 1999-04-21 | 2000-04-21 | Process and automation of an industrial process in steps, with mastery of an uncertain stress chain, and its application for control of noise and of the risk (value-at-risk, var) of a clearing house |
Country Status (8)
| Country | Link |
|---|---|
| US (1) | US7644005B1 (en) |
| EP (1) | EP2062111B1 (en) |
| AT (1) | ATE463780T1 (en) |
| AU (1) | AU7193600A (en) |
| CA (1) | CA2406899C (en) |
| DE (1) | DE60044163D1 (en) |
| FR (1) | FR2792746B1 (en) |
| WO (1) | WO2000065418A2 (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2005033987A1 (en) * | 2003-10-06 | 2005-04-14 | Corporate Integration Systems And Method | Risk assessment system and method |
Families Citing this family (13)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US7937313B2 (en) * | 2001-06-29 | 2011-05-03 | Goldman Sachs & Co. | Method and system for stress testing simulations of the behavior of financial instruments |
| US9729639B2 (en) | 2001-08-10 | 2017-08-08 | Rockwell Automation Technologies, Inc. | System and method for dynamic multi-objective optimization of machine selection, integration and utilization |
| US20090210081A1 (en) * | 2001-08-10 | 2009-08-20 | Rockwell Automation Technologies, Inc. | System and method for dynamic multi-objective optimization of machine selection, integration and utilization |
| US8914300B2 (en) | 2001-08-10 | 2014-12-16 | Rockwell Automation Technologies, Inc. | System and method for dynamic multi-objective optimization of machine selection, integration and utilization |
| US8788247B2 (en) * | 2008-08-20 | 2014-07-22 | International Business Machines Corporation | System and method for analyzing effectiveness of distributing emergency supplies in the event of disasters |
| US8972067B2 (en) | 2011-05-11 | 2015-03-03 | General Electric Company | System and method for optimizing plant operations |
| US9031892B2 (en) | 2012-04-19 | 2015-05-12 | Invensys Systems, Inc. | Real time safety management system and method |
| WO2014145705A2 (en) | 2013-03-15 | 2014-09-18 | Battelle Memorial Institute | Progression analytics system |
| US10637240B2 (en) * | 2014-01-24 | 2020-04-28 | Fujitsu Limited | Energy curtailment event implementation based on uncertainty of demand flexibility |
| EP3324254A1 (en) * | 2016-11-17 | 2018-05-23 | Siemens Aktiengesellschaft | Device and method for determining the parameters of a control device |
| SG11202104066UA (en) * | 2018-10-26 | 2021-05-28 | Dow Global Technologies Llc | Deep reinforcement learning for production scheduling |
| EP3908807B1 (en) | 2020-03-17 | 2023-11-29 | Freeport-McMoRan Inc. | Methods and systems for deploying equipment required to meet defined production targets |
| CN111723093A (en) * | 2020-06-17 | 2020-09-29 | 江苏海平面数据科技有限公司 | Uncertain interval data query method based on data division |
Family Cites Families (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US5148365A (en) * | 1989-08-15 | 1992-09-15 | Dembo Ron S | Scenario optimization |
| AU6398994A (en) * | 1993-03-09 | 1994-09-26 | C*Ats Software Inc. | An object-oriented system for creating, structuring, manipulating and evaluating a financial instrument |
| EP0845123A4 (en) * | 1995-08-15 | 2001-04-11 | Univ Columbia | Estimation method and system for financial securities trading |
| US6278981B1 (en) * | 1997-05-29 | 2001-08-21 | Algorithmics International Corporation | Computer-implemented method and apparatus for portfolio compression |
| US7188075B1 (en) * | 2000-06-29 | 2007-03-06 | Oracle International Corporation | Extended product configuration techniques |
| US7155399B2 (en) * | 2001-04-03 | 2006-12-26 | Witness Systems, Inc. | System and method for complex schedule generation |
| US7043444B2 (en) * | 2001-04-13 | 2006-05-09 | I2 Technologies Us, Inc. | Synchronization of planning information in a high availability planning and scheduling architecture |
-
1999
- 1999-04-21 FR FR9905074A patent/FR2792746B1/en not_active Expired - Fee Related
-
2000
- 2000-04-21 EP EP00922712A patent/EP2062111B1/en not_active Expired - Lifetime
- 2000-04-21 AU AU71936/00A patent/AU7193600A/en not_active Abandoned
- 2000-04-21 CA CA002406899A patent/CA2406899C/en not_active Expired - Fee Related
- 2000-04-21 US US10/149,960 patent/US7644005B1/en not_active Expired - Fee Related
- 2000-04-21 DE DE60044163T patent/DE60044163D1/en not_active Expired - Lifetime
- 2000-04-21 WO PCT/FR2000/001059 patent/WO2000065418A2/en active Application Filing
- 2000-04-21 AT AT00922712T patent/ATE463780T1/en not_active IP Right Cessation
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2005033987A1 (en) * | 2003-10-06 | 2005-04-14 | Corporate Integration Systems And Method | Risk assessment system and method |
Also Published As
| Publication number | Publication date |
|---|---|
| US7644005B1 (en) | 2010-01-05 |
| EP2062111B1 (en) | 2010-04-07 |
| FR2792746A1 (en) | 2000-10-27 |
| DE60044163D1 (en) | 2010-05-20 |
| WO2000065418A3 (en) | 2001-04-12 |
| ATE463780T1 (en) | 2010-04-15 |
| EP2062111A2 (en) | 2009-05-27 |
| FR2792746B1 (en) | 2003-10-17 |
| WO2000065418A2 (en) | 2000-11-02 |
| CA2406899C (en) | 2009-11-24 |
| AU7193600A (en) | 2000-11-10 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CA2406899A1 (en) | Process and automation of an industrial process in steps, with mastery of an uncertain stress chain, and its application for control of noise and of the risk (value-at-risk, var) of a clearing house | |
| Demetriou et al. | Incipient fault diagnosis of dynamical systems using online approximators | |
| Li et al. | Adaptive fuzzy output-constrained fault-tolerant control of nonlinear stochastic large-scale systems with actuator faults | |
| Capilla et al. | Integration of statistical and engineering process control in a continuous polymerization process | |
| Ibadov | Fuzzy estimation of activities duration in construction projects | |
| US8261132B2 (en) | Method for error tree analysis | |
| Lajmi et al. | Fault diagnosis of uncertain systems based on interval fuzzy PETRI net | |
| Haesaert et al. | Correct-by-design output feedback of LTI systems | |
| Elsayed et al. | Design of optimum simple step-stress accelerated life testing plans | |
| CN111598328A (en) | Power load prediction method considering epidemic situation events | |
| Navone et al. | Forecasting chaos from small data sets: a comparison of different nonlinear algorithms | |
| CN115427767A (en) | Improved pattern recognition techniques for data-driven fault detection within a process plant | |
| Sendur et al. | A model accuracy and validation algorithm | |
| US20040024559A1 (en) | Method, system and computer product for building calibration lookup tables from sparse data | |
| Aguila-Camacho et al. | Switched fractional order model reference adaptive control for unknown linear time invariant systems | |
| Zhao et al. | A new safety assessment method based on evidential reasoning rule with a prewarning function | |
| Cooper et al. | Pattern-based adaptive process control | |
| CN113404742B (en) | Electro-hydraulic servo mechanism health assessment method and system based on test data | |
| CN110942259B (en) | Community gas equipment risk assessment method and device | |
| CN114046456B (en) | Corrosion evaluation method and system for fusing fuzzy reasoning and neural network | |
| Witczak et al. | Observers and genetic programming in the identification and fault diagnosis of non-linear dynamic systems | |
| Quelhas et al. | Soft sensor models: Bias updating revisited | |
| Horak | Experimental estimation of modeling errors in dynamic systems | |
| Vanli et al. | Closed-loop system identification for small samples with constraints | |
| Bastani et al. | A software reliability model for artificial intelligence programs |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| EEER | Examination request | ||
| MKLA | Lapsed |
Effective date: 20160421 |