CN113468751A - Recursion Lasso-based flowmeter anomaly online monitoring method and system and storage medium - Google Patents

Recursion Lasso-based flowmeter anomaly online monitoring method and system and storage medium Download PDF

Info

Publication number
CN113468751A
CN113468751A CN202110765956.0A CN202110765956A CN113468751A CN 113468751 A CN113468751 A CN 113468751A CN 202110765956 A CN202110765956 A CN 202110765956A CN 113468751 A CN113468751 A CN 113468751A
Authority
CN
China
Prior art keywords
flow
flowmeter
autoregressive model
lasso
model
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
Application number
CN202110765956.0A
Other languages
Chinese (zh)
Other versions
CN113468751B (en
Inventor
刘颖
刘穗君
任淑军
卢成
曾九孙
蔡晋辉
崔廷
崔岩
刘磊
杨林超
陈建中
李春松
纪晓楠
李金洺
文金昉
李新会
李显红
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
China Jiliang University
China Tobacco Henan Industrial Co Ltd
Original Assignee
China Jiliang University
China Tobacco Henan Industrial Co Ltd
Priority date (The priority date 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 date listed.)
Filing date
Publication date
Application filed by China Jiliang University, China Tobacco Henan Industrial Co Ltd filed Critical China Jiliang University
Priority to CN202110765956.0A priority Critical patent/CN113468751B/en
Publication of CN113468751A publication Critical patent/CN113468751A/en
Application granted granted Critical
Publication of CN113468751B publication Critical patent/CN113468751B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Images

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F30/00—Computer-aided design [CAD]
    • G06F30/20—Design optimisation, verification or simulation
    • 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
    • G06Q10/20—Administration of product repair or maintenance
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
    • G06Q50/00—Systems or methods specially adapted for specific business sectors, e.g. utilities or tourism
    • G06Q50/04—Manufacturing
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F2111/00—Details relating to CAD techniques
    • G06F2111/10—Numerical modelling
    • 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
    • Y02P90/30—Computing systems specially adapted for manufacturing

Abstract

The invention discloses a recursion Lasso-based flowmeter abnormity on-line monitoring method, a system and a storage medium, wherein the method comprises the following steps: establishing an autoregressive model according to historical flow data of the flow meter, and determining a model coefficient of the autoregressive model through an offline Lasso algorithm; updating the autoregressive model by a recursion Lasso algorithm by utilizing the real-time flow of the flowmeter; and determining whether the flowmeter is abnormal according to the predicted flow obtained based on the updated autoregressive model and the actual flow of the flowmeter at the next moment. The recursion Lasso-based flowmeter abnormity on-line monitoring method updates the autoregressive model through the recursion Lasso algorithm, monitors the abnormity of the flowmeter according to the predicted flow and the actual flow obtained based on the updated autoregressive model, has the characteristics of high accuracy, convenient operation, real-time tracking and the like, provides scientific, objective and reliable technical support for on-line monitoring of the humidifying water flow in the tobacco shred manufacturing process, and ensures the metering performance stability of the flowmeter.

Description

Recursion Lasso-based flowmeter anomaly online monitoring method and system and storage medium
Technical Field
The invention relates to the technical field of tobacco processing, in particular to a recursion Lasso-based flow meter abnormity online monitoring method, a recursion Lasso-based flow meter abnormity online monitoring system and a storage medium.
Background
The measuring instrument product is abnormal in the using process, and if the abnormal measuring instrument product is not checked and maintained in time, the accuracy of the measuring instrument product is affected, so that huge hidden dangers are brought to the subsequent industrial production process, and the management expenditure cost of enterprises is increased.
In order to ensure the accuracy of the data measured by the measuring instrument, the data measured by the measuring instrument needs to be monitored in real time during the working process of the measuring instrument. Once abnormity occurs, such as data jumping and the like, whether the measuring instrument has a fault or the adjustment of production conditions needs to be judged; if the measuring instrument is in failure, the measuring instrument needs to be subjected to abnormal investigation and maintenance.
The on-line flowmeter is one of the most important basic data acquisition sources in the traceability chain of the cigarette manufacturing process, and the detection performance of the on-line flowmeter needs to be maintained within an allowable range all the time in the production process. At present, the abnormal monitoring mode of the flow meter by the production line of the cigarette enterprise lacks the specialty, on one hand, the conventional point inspection regulation is simply referred to, and the online monitoring characteristic in the production process is ignored; on the other hand, the subjective prejudgment is carried out only by the maintenance experience of technicians, the randomness is certain, and the error risk caused by human intervention exists, so that whether the flowmeter really fails or not cannot be ensured. Therefore, due to the lack of scientific basis and data support, the conventional flowmeter abnormity monitoring method cannot well achieve the function of fault early warning, and cannot find out the flowmeter fault in time, so that great potential production hazards exist.
Therefore, a method, a system and a storage medium for online monitoring of flow meter anomalies based on recursive Lasso are needed.
Disclosure of Invention
The invention aims to provide a recursion Lasso-based flowmeter anomaly online monitoring method, a recursion Lasso-based flowmeter anomaly online monitoring system and a storage medium, so that the problems in the prior art are solved, and the anomaly condition of a flowmeter can be monitored in time.
The invention provides a recursion Lasso-based flowmeter abnormity on-line monitoring method, which comprises the following steps:
establishing an autoregressive model according to historical flow data of the flow meter, and determining a model coefficient of the autoregressive model through an offline Lasso algorithm;
updating the autoregressive model by a recursion Lasso algorithm by utilizing the real-time flow of the flowmeter;
and determining whether the flowmeter is abnormal according to the predicted flow obtained based on the updated autoregressive model and the actual flow of the flowmeter at the next moment.
The method for monitoring abnormality of a flow meter on line based on recursive Lasso as described above, wherein preferably, the establishing an autoregressive model according to historical flow data of the flow meter, and determining a model coefficient of the autoregressive model by an offline Lasso algorithm specifically includes:
collecting a data sample of the flow of the flowmeter to obtain a historical flow data set;
constructing an autoregressive model according to the historical flow data set;
solving a regression coefficient of the autoregressive model based on an offline Lasso algorithm;
and determining the regularization parameters of the autoregressive model by using a cross validation method.
The method for monitoring flow meter anomaly online based on recursive Lasso as described above, wherein preferably, the optimal condition of the offline Lasso algorithm is determined by the following formula, and the regression coefficient and the regularization parameter of the autoregressive model are determined according to the optimal condition of the offline Lasso algorithm:
Figure BDA0003148342660000021
where p is the hysteresis coefficient of the autoregressive model, ynIndicating the stream to be predicted at the current timeMagnitude, yn-iRepresents the flow rate value at the historical time, alpha represents the regression coefficient of the autoregressive model, mun-1The regularization parameters of the autoregressive model are represented.
The method for monitoring abnormality of a flow meter on line based on recursive Lasso as described above, wherein preferably, the updating the autoregressive model by using the real-time flow rate of the flow meter through a recursive Lasso algorithm specifically includes:
taking the model coefficient of the autoregressive model determined by an offline Lasso algorithm as an initial model coefficient of the autoregressive model in an updating process;
updating the model parameters of the autoregressive model according to the new real-time flow of the flowmeter every time a new real-time flow of the flowmeter is read in, and updating the autoregressive model by the following formula:
Figure BDA0003148342660000031
wherein, Yn-1A set of historical flow data, Z, representing the measured flow of the flowmetern-1Autoregressive term, y, representing an autoregressive modelnRepresenting the flow value to be predicted at the current moment, ZnAnd the value of t is 0-1, the value of t is changed from 0 to 1 after new data is read, when t is 1, the new data is completely read, parameters alpha and mu in the autoregressive model are updated, the new data at the next moment is ready to be read, and after the new data at the next moment is read, the value of t is changed from 0 to 1 again.
The method for monitoring abnormality of a flow meter on line based on recursive Lasso as described above, wherein preferably, the determining whether the flow meter is abnormal according to the predicted flow rate obtained based on the updated autoregressive model and the actual flow rate of the flow meter at the next time includes:
predicting the flow of the flowmeter at the next moment in real time according to the updated autoregressive model to obtain predicted flow;
obtaining an estimation curve graph of the predicted flow according to the obtained predicted flow corresponding to different moments;
obtaining a residual error map according to the deviation between the predicted flow and the actual flow of the flowmeter at the next moment;
and determining whether the flow meter is abnormal according to the residual error map.
The method for online monitoring of flow meter abnormality based on recursive Lasso as described above, wherein preferably, the determining whether the flow meter is abnormal according to the residual error map includes:
when the ordinate of the residual error map exceeds a preset threshold value, determining that the flow data of the flowmeter is abnormal;
and determining whether the reason causing the flow data abnormity of the flow meter is the abnormity of the flow meter or the batch replacement of the cut stem product according to the production record.
The method for monitoring the flow meter abnormality on line based on the recursive Lasso as described above, wherein, preferably, the method for monitoring the flow meter abnormality on line based on the recursive Lasso is used for monitoring whether the flow rate of the humidifying water in the tobacco shred making process is abnormal,
the flow meter abnormity online monitoring method based on recursion Lasso further comprises the following steps:
and determining the cut stem product specification at the current moment based on a preset flow threshold of the cut stem product batch and the predicted flow.
The invention also provides a recursion Lasso-based flowmeter abnormity on-line monitoring system adopting the method, which comprises the following steps:
the autoregressive model establishing module is used for establishing an autoregressive model according to historical flow data of the flow meter and determining a model coefficient of the autoregressive model through an offline Lasso algorithm;
the autoregressive model updating module is used for updating the autoregressive model through a recursion Lasso algorithm by utilizing the real-time flow of the flowmeter;
and the flow monitoring module is used for determining whether the flowmeter is abnormal according to the predicted flow obtained based on the updated autoregressive model and the actual flow of the flowmeter at the next moment.
The present invention also provides a computer-readable storage medium, in which a computer program is stored, which, when running on a computer, causes the computer to execute the above-mentioned method for online monitoring of flow meter anomalies based on recursive Lasso.
The invention also provides a computer program product, which is characterized in that when the computer program product runs on a terminal device, the terminal device executes the flow meter abnormity online monitoring method based on recursion Lasso.
The invention provides a recursion-based flowmeter anomaly online monitoring method, which updates an autoregressive model through a recursion-based Lasso algorithm, monitors the anomaly condition of a flowmeter according to the predicted flow and the actual flow obtained based on the updated autoregressive model, has the characteristics of high accuracy, convenient operation, real-time tracking and the like, provides scientific, objective and reliable technical support for online monitoring of the humidifying water flow in the tobacco shred manufacturing process, and further can ensure the metering performance stability of the flowmeter.
Drawings
In order to make the objects, technical solutions and advantages of the present invention more apparent, the present invention will be further described with reference to the accompanying drawings, in which:
FIG. 1 is a flow chart of an embodiment of a recursive Lasso-based flow meter anomaly online monitoring method provided by the present invention;
FIG. 2 is a graph illustrating an estimated flow rate provided by an embodiment of the present invention;
FIG. 3 is a residual map provided by an embodiment of the present invention;
fig. 4 is a structural block diagram of an embodiment of an online flow meter anomaly monitoring system based on recursive Lasso according to the present invention.
Detailed Description
Various exemplary embodiments of the present disclosure will now be described in detail with reference to the accompanying drawings. The description of the exemplary embodiments is merely illustrative and is in no way intended to limit the disclosure, its application, or uses. The present disclosure may be embodied in many different forms and is not limited to the embodiments described herein. These embodiments are provided so that this disclosure will be thorough and complete, and will fully convey the scope of the disclosure to those skilled in the art. It should be noted that: the relative arrangement of parts and steps, the composition of materials, numerical expressions and numerical values set forth in these embodiments are to be construed as merely illustrative, and not as limitative, unless specifically stated otherwise.
As used in this disclosure, "first", "second": and the like, do not denote any order, quantity, or importance, but rather are used to distinguish one element from another. The word "comprising" or "comprises", and the like, means that the element preceding the word covers the element listed after the word, and does not exclude the possibility that other elements are also covered. "upper", "lower", and the like are used merely to indicate relative positional relationships, and when the absolute position of the object being described is changed, the relative positional relationships may also be changed accordingly.
In the present disclosure, when a specific component is described as being located between a first component and a second component, there may or may not be intervening components between the specific component and the first component or the second component. When it is described that a specific component is connected to other components, the specific component may be directly connected to the other components without having an intervening component, or may be directly connected to the other components without having an intervening component.
All terms (including technical or scientific terms) used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this disclosure belongs unless specifically defined otherwise. It will be further understood that terms, such as those defined in commonly used dictionaries, should be interpreted as having a meaning that is consistent with their meaning in the context of the relevant art and will not be interpreted in an idealized or overly formal sense unless expressly so defined herein.
Techniques, methods, and apparatus known to those of ordinary skill in the relevant art may not be discussed in detail, but are intended to be part of the specification where appropriate.
The invention provides a recursion Lasso-based flowmeter anomaly online monitoring method for online monitoring a flowmeter, aiming at a flowmeter which is a sensing device frequently used in the cigarette production process, based on an actual application scene and an objective scientific thought, so that faults are found in time and loss is reduced.
The method for monitoring the abnormity of the flowmeter based on the recursion Lasso is used for monitoring whether the flow rate of the humidifying water in the tobacco shred making process is abnormal or not, and as shown in fig. 1, the method for monitoring the abnormity of the flowmeter based on the recursion Lasso provided by the embodiment specifically comprises the following steps in the actual implementation process:
step S1, establishing an autoregressive model according to the historical flow data of the flow meter, and determining the model coefficient of the autoregressive model through an offline Lasso algorithm.
In an embodiment of the online anomaly monitoring method for a flowmeter based on recursive Lasso, the step S1 may specifically include:
and step S11, collecting data samples of the flow meter to obtain a historical flow data set.
Specifically, data collected by a humidified water flow meter in the cut stem charging unit is collected as a data source for online monitoring according to a preset time length, so that a historical flow data set is obtained.
And step S12, constructing an autoregressive model according to the historical flow data set.
And step S13, solving the regression coefficient of the autoregressive model based on an offline Lasso algorithm.
And step S14, determining the regularization parameters of the autoregressive model by using a cross validation method.
Specifically, the optimal condition of the offline Lasso algorithm is determined through the following formula, and the regression coefficient and the regularization parameter of the autoregressive model are determined according to the optimal condition of the offline Lasso algorithm:
Figure BDA0003148342660000061
where p is the hysteresis coefficient of the autoregressive model, ynRepresenting the flow value to be predicted at the current moment, yn-iRepresents the flow rate value at the historical time, alpha represents the regression coefficient of the autoregressive model, mun-1The regularization parameters of the autoregressive model are represented.
Will l1Introducing the norm into an autoregressive model, and assuming that flow data measured by the flowmeter is y ∈ Rn-1The regression coefficient of the model is p,
Figure BDA0003148342660000071
wherein epsiloniRepresenting noisy data measured by the flow meter. Then the optimal solution for Lasso becomes:
Figure BDA0003148342660000072
wherein mun-1Is a regularization parameter, the form of the solution of the formula (1) is sparse, only a few elements are non-zero, and the purpose of the Lasso algorithm is to highlight some non-zero important variables by setting some non-important variables to zero under the condition of a large number of variables, and the non-zero variables are variables useful for prediction, so that it can be determined which historical time flows are helpful for predicting the current time flows according to the non-zero variables.
The indices of the non-zero elements of the regression coefficient vector α define an "active set" which is preceded for simplicity of notation, e.g., αT＝(α1 T,0T)、vT＝(v1 T,v2 T). Wherein the step function v is satisfied for all i1i＝sgn(α1i) (ii) a For all j, -1 ≦ v2j1, where i and j are indices that are separately represented for different sets, so confusion can be avoided. While Y is divided into Y ═ Y (Y) according to the "active set1 Y2) Wherein Y represents Y in the formula (1)n-1To yn-pSet of (2)The elements in each set correspond to a regression coefficient alpha, Y1Representing elements in the corresponding Y for which the regression coefficient alpha is not zero, Y2Refers to the element in the corresponding Y that represents a regression coefficient α of zero. If the determined solution is unique, Y1 TY1Is reversible, then the optimal conditions are
Figure BDA0003148342660000073
It should be noted that α can be calculated in a closed form by knowing the sign of the coefficients in the "active set" and the feature vector α.
Step S2, updating the autoregressive model by using the real-time flow rate of the flow meter and a recursive Lasso (Least absolute convergence and selection operator) algorithm.
The model parameters of the autoregressive model are finely adjusted according to the read new data, so that the prediction effect of the autoregressive model can be improved. In an embodiment of the online anomaly monitoring method for a flowmeter based on recursive Lasso, the step S2 may specifically include:
and step S21, taking the model coefficient of the autoregressive model determined by the offline Lasso algorithm as the initial model coefficient of the autoregressive model in the updating process.
Step S22, when a new real-time flow rate of the flowmeter is read in, updating the model parameters of the autoregressive model according to the new real-time flow rate of the flowmeter, and updating the autoregressive model according to the following formula:
Figure BDA0003148342660000081
wherein, Yn-1A set of historical flow data, Z, representing the measured flow of the flowmetern-1Autoregressive term, y, representing an autoregressive modelnRepresenting the flow value to be predicted at the current moment, ZnAnd the value of t is 0-1, the value of t is changed from 0 to 1 after new data is read, when t is 1, the new data is completely read, parameters alpha and mu in the autoregressive model are updated, the new data at the next moment is ready to be read, and after the new data at the next moment is read, the value of t is changed from 0 to 1 again.
Specifically, assume that the solution α of equation (1) at time n-1 has been calculated(n-1)Then, new observation data y is obtainednThen y can be utilizednAnd historical data to predict data y at time n +1n+1Then the target is changed to calculate alpha(n)The value of (c). Thereby leading to a recursive Lasso solution. Note zn＝(yn,yn-1,...,yn-p)，Zn＝(z1,z2,...,zn)T，Yn-1＝(y1,y2,...yn-1)TWherein z isnRepresenting historical time flow, Z, needed to predict current flownRepresenting the set of traffic at all previous historical moments, Yn-1Representing all historical time flows, the optimized objective function becomes:
Figure BDA0003148342660000082
the entire update path becomes alpha(n-1)＝α(0,μn-1) To alpha(n)＝α(1,μn) According to the update path, the update method can be divided into two steps:
the first step is as follows: when t is 0, from mun-1To munAnd updating the regularization parameters, which is equivalent to calculating the regularization path between the two by adopting a minimum angle regression method.
The second step is that: when mu is munThen, t is calculated from 0 to 1.
The solving process of the second step is as follows: it should be first demonstrated that α (t, μ) is a piecewise smooth function for t. To make the notation simpler, let α (t) be α (t, μ), ifKnowing the "active set" and the sign of the intra α coefficients, the Lasso solution can be calculated. In a first step, the "active set" and the sign of the intra- α coefficients have been computed, and when t ∈ [0, t*) When (wherein, t)*Is expressed in t ∈ [0, t ∈ [ ]*) When α (t) is a smooth curve), the sign of the "active set" and the α -inner coefficient remain unchanged, and the solution α (t) of Lasso is smooth. One point at which the "active set" changes is called the turning point, and then it is analyzed how to calculate this point:
when t is equal to t*The sign of the "active set" and the coefficients in α are updated and kept constant until the next turning point is reached, and the process is iterated until t is 1, so that the required solution α (t) can be calculated.
Thus, an online update algorithm for Lasso with added observations is obtained:
STEP 1: calculating alpha(n-1)＝α(0,μn-1) To alpha(n)＝α(1,μn)；
STEP 2: initializing alpha (0, mu)n) To the "active set" by making v ═ sgn (α (0, μm)n) Let v) make v1And zn,1Is v and znThe sub-vectors divided according to the "active set",
Figure BDA0003148342660000091
is that
Figure BDA0003148342660000092
Wherein the columns are "active sets", initializing
Figure BDA0003148342660000093
Initializing turning point t' ═ 0;
STEP 3: the next turning point t' is calculated. If the inflection point is less than the previous inflection point, or if the inflection point is greater than 1, STEP5 is skipped,
the first case is: alpha first1(t') the ith element becomes 0; then i is removed from the "active set"; then v is converted intoi Setting 0;
second kindThe situation is that: firstly, omega is measured2(t') the absolute value of the jth element reaches 1; j is then added to the "active set"; next, if the element reaches 1 (or-1), v will bejPut 1 (or-1).
STEP 4: updating v according to updated' active set1，
Figure BDA0003148342660000094
And zn,1Update
Figure BDA0003148342660000095
STEP 5: calculating the final result when t is 1, where α(n)Is given a value of
Figure BDA0003148342660000096
Is given.
When the observed values in the data set are too many, the calculation time of STEP1 is prolonged, and in the subsequent calculation process, the observed point data which is more distant from the current time has little influence on the value of the predicted current time, so the initial historical data is removed after a period of time, and the calculation time of the model is stable within a certain range.
Suppose that the solution α at time n has been calculated(n)Later, the first data needs to be removed before the new observation data is read in, thereby leading to a solution for recursive Lasso removal of historical data. Note z1＝(y0,y-1,...,y-p)，Z＝(z2,...,zn)T，Y＝(y2,y3,...,yn)TThen the optimized objective function becomes:
Figure BDA0003148342660000101
the entire update path becomes alpha(n)＝α(1,μn) To alpha(n')＝α(0,μn') According to the update path, the update method can be divided into two steps: STEP 1: when t isWhen 1, from munTo mun' updating the regularization parameters, which is equivalent to computing the regularization path between the two using a least-angle regression method.
STEP 2: when mu is mun'Then, t is calculated from 1 to 0.
The calculation procedure of STEP2 is the same as the procedure of adding the observation data.
And step S3, determining whether the flowmeter is abnormal according to the predicted flow obtained based on the updated autoregressive model and the actual flow of the flowmeter at the next moment.
In an embodiment of the online anomaly monitoring method for a flowmeter based on recursive Lasso, the step S3 may specifically include:
and step S31, predicting the flow of the flowmeter at the next moment in real time according to the updated autoregressive model to obtain the predicted flow.
And step S32, obtaining an estimation curve chart of the predicted flow according to the obtained predicted flow corresponding to different moments.
The dashed line in fig. 2 is an estimated graph of the predicted flow rate.
And step S33, obtaining a residual error map according to the deviation between the predicted flow and the actual flow of the flowmeter at the next moment.
The deviation is the absolute value of the difference between the predicted value and the actual flow rate. The residual map is used for representing whether the flow rate is abnormal under the preset use environment. In some embodiments of the present invention, the residual error map is shown in FIG. 3, in which the straight lines are 50kg/h, 25kg/h, and 10kg/h from top to bottom.
And step S34, determining whether the flow meter is abnormal according to the residual error map.
And estimating the change of the online flow in the working process by using the deviation value, and determining abnormal operation of the flow when the deviation value is overlarge. In an embodiment of the online anomaly monitoring method for a flowmeter based on recursive Lasso, the step S34 may specifically include:
and step S341, when the ordinate of the residual error map exceeds a preset threshold value, determining that the flow data of the flowmeter is abnormal.
In fig. 3, when the flow meter failed at the 7000 th time, the deviation between the predicted value and the actual value was large, and the residual (ordinate of the residual map) was larger than 50kg/h, and it was assumed that the flow rate data was abnormal (the abnormal point in the map was around 100 kg/h).
And step S342, determining whether the reason causing the flow data abnormity of the flow meter is the abnormity of the flow meter or the replacement of the cut stem product batch according to the production record.
If the cut stem product batch is replaced, determining that the reason for causing the flow data abnormity of the flowmeter is production adjustment; if the cut stem product batch is not replaced, determining that the reason for causing the flow data abnormity of the flowmeter is the abnormity of the flowmeter, and at the moment, timely maintaining the flowmeter is needed. In the invention, the state performance of the flowmeter is tracked in real time by using the residual error map, and when the flow is abnormal, the flow can be monitored in time and an alarm is given out.
Further, in some embodiments of the present invention, the method for online monitoring of an abnormality of a flowmeter based on recursive Lasso further includes:
and step S4, determining the cut stem product specification at the current moment based on the preset flow threshold of the cut stem product batch and the predicted flow.
And determining the product batch entering the cut stem charging unit at the current moment based on a preset flow threshold value of the cut stem product batch and an estimation curve chart drawn based on the predicted flow at a plurality of moments. Because each cut stem product has different batches and different humidifying water flow rates, the cut stem product specification can be judged according to the flow rate estimated by the autoregressive model by setting the range of the humidifying water flow rates.
The cut stems in the data sample are divided into two types, one is common stems, and the other is special stems for golden leaves. When the material supply is stable, the humidifying water flow of the common stalks is within the range of 110-130 kg/h, and the humidifying water flow of the special golden leaf stalks is within the range of 140-170 kg/h. Therefore, the product specification of the cut stems can be judged according to the predicted flow value. In fig. 2, the dotted line is a predicted value, and the solid curve is an actual value; the solid straight line is a dividing line of 140kg/h, if the material supply is stable, the predicted flow is less than 140kg/h, the stem is judged as a common stem, and the measured flow is more than 140kg/h, the stem is judged as a special stem for golden leaf.
The recursion Lasso-based flowmeter anomaly online monitoring method provided by the embodiment of the invention updates the autoregressive model through the recursion Lasso algorithm, monitors the anomaly condition of the flowmeter according to the predicted flow and the actual flow obtained based on the updated autoregressive model, has the characteristics of high accuracy, convenient operation, real-time tracking and the like, provides scientific, objective and reliable technical support for online monitoring of the humidifying water flow in the tobacco shred manufacturing process, and further can ensure the metering performance stability of the flowmeter.
Correspondingly, as shown in fig. 2, the present invention further provides an online monitoring system for flow meter anomaly based on recursive Lasso, including:
the system comprises an autoregressive model establishing module 1, a flow meter and a flow control module, wherein the autoregressive model establishing module is used for establishing an autoregressive model according to historical flow data of the flow meter and determining a model coefficient of the autoregressive model through an offline Lasso algorithm;
the autoregressive model updating module 2 is used for updating the autoregressive model through a recursion Lasso algorithm by utilizing the real-time flow of the flowmeter;
and the flow monitoring module 3 is used for determining whether the flowmeter is abnormal according to the predicted flow obtained based on the updated autoregressive model and the actual flow of the flowmeter at the next moment.
According to the recursion-based flowmeter anomaly online monitoring system provided by the embodiment of the invention, the autoregressive model is updated by the autoregressive model updating module through the recursion-based Lasso algorithm, and the flow monitoring module monitors the anomaly condition of the flowmeter according to the predicted flow and the actual flow obtained based on the updated autoregressive model, so that the system has the characteristics of high accuracy, convenience in operation, real-time tracking and the like, provides scientific, objective and reliable technical support for online monitoring of the humidifying water flow in the tobacco shred making process, further can ensure the metering performance stability of the flowmeter, is also suitable for detection scenes of other metering devices, and has a wide application prospect.
It should be understood that the division of the components of the recursive Lasso-based flow meter anomaly online monitoring system shown in fig. 4 is merely a logical functional division, and the actual implementation may be wholly or partially integrated into one physical entity or physically separated. And these components may all be implemented in software invoked by a processing element; or may be implemented entirely in hardware; and part of the components can be realized in the form of calling by the processing element in software, and part of the components can be realized in the form of hardware. For example, a certain module may be a separate processing element, or may be integrated into a certain chip of the electronic device. Other components are implemented similarly. In addition, all or part of the components can be integrated together or can be independently realized. In implementation, each step of the above method or each component above may be implemented by an integrated logic circuit of hardware in a processor element or an instruction in the form of software.
The present invention also provides a computer-readable storage medium, in which a computer program is stored, which, when running on a computer, causes the computer to execute the above-mentioned method for online monitoring of flow meter anomalies based on recursive Lasso.
The invention also provides a computer program product, which is characterized in that when the computer program product runs on a terminal device, the terminal device executes the flow meter abnormity online monitoring method based on recursion Lasso.
In the above embodiments, the implementation may be wholly or partly realized by software, hardware, firmware, or any combination thereof. When implemented in software, may be implemented in whole or in part in the form of a computer program product. The computer program product includes one or more computer instructions. When loaded and executed on a computer, cause the processes or functions described in accordance with the embodiments of the application to occur, in whole or in part. The computer may be a general purpose computer, a special purpose computer, a network of computers, or other programmable device. The computer instructions may be stored on a computer readable storage medium or transmitted from one computer readable storage medium to another, for example, from one website, computer, server, or data center to another website, computer, server, or data center via wire (e.g., coaxial cable, fiber optic, Digital Subscriber Line (DSL)) or wireless (e.g., infrared, wireless, microwave, etc.). The computer-readable storage medium can be any available medium that can be accessed by a computer or a data storage device, such as a server, a data center, etc., that incorporates one or more of the available media. The usable medium may be a magnetic medium (e.g., floppy Disk, hard Disk, magnetic tape), an optical medium (e.g., Digital Versatile Disk (DVD)), or a semiconductor medium (e.g., Solid State Disk (SSD)), among others.
Thus, various embodiments of the present disclosure have been described in detail. Some details that are well known in the art have not been described in order to avoid obscuring the concepts of the present disclosure. It will be fully apparent to those skilled in the art from the foregoing description how to practice the presently disclosed embodiments.
Although some specific embodiments of the present disclosure have been described in detail by way of example, it should be understood by those skilled in the art that the foregoing examples are for purposes of illustration only and are not intended to limit the scope of the present disclosure. It will be understood by those skilled in the art that various changes may be made in the above embodiments or equivalents may be substituted for elements thereof without departing from the scope and spirit of the present disclosure. The scope of the present disclosure is defined by the appended claims.

Claims (10)

1. A method for monitoring flow meter abnormity on line based on recursion Lasso is characterized by comprising the following steps:
establishing an autoregressive model according to historical flow data of the flow meter, and determining a model coefficient of the autoregressive model through an offline Lasso algorithm;
updating the autoregressive model by a recursion Lasso algorithm by utilizing the real-time flow of the flowmeter;
and determining whether the flowmeter is abnormal according to the predicted flow obtained based on the updated autoregressive model and the actual flow of the flowmeter at the next moment.
2. The method for on-line monitoring of flow meter anomaly based on recursive Lasso according to claim 1, wherein the establishing of an autoregressive model according to historical flow data of the flow meter and the determining of model coefficients of the autoregressive model by an offline Lasso algorithm specifically comprises:
collecting a data sample of the flow of the flowmeter to obtain a historical flow data set;
constructing an autoregressive model according to the historical flow data set;
solving a regression coefficient of the autoregressive model based on an offline Lasso algorithm;
and determining the regularization parameters of the autoregressive model by using a cross validation method.
3. The recursive Lasso-based flow meter anomaly online monitoring method according to claim 2, wherein the optimal conditions of the offline Lasso algorithm are determined by the following formula, and the regression coefficients and the regularization parameters of the autoregressive model are determined according to the optimal conditions of the offline Lasso algorithm:
Figure FDA0003148342650000011
where p is the hysteresis coefficient of the autoregressive model, ynRepresenting the flow value to be predicted at the current moment, yn-iRepresents the flow rate value at the historical time, alpha represents the regression coefficient of the autoregressive model, mun-1The regularization parameters of the autoregressive model are represented.
4. The method for on-line monitoring of flow meter anomaly based on recursive Lasso according to claim 1, wherein the updating of the autoregressive model by the recursive Lasso algorithm using the real-time flow of the flow meter specifically comprises:
taking the model coefficient of the autoregressive model determined by an offline Lasso algorithm as an initial model coefficient of the autoregressive model in an updating process;
updating the model parameters of the autoregressive model according to the new real-time flow of the flowmeter every time a new real-time flow of the flowmeter is read in, and updating the autoregressive model by the following formula:
Figure FDA0003148342650000021
wherein, Yn-1A set of historical flow data, Z, representing the measured flow of the flowmetern-1Autoregressive term, y, representing an autoregressive modelnRepresenting the flow value to be predicted at the current moment, ZnAnd the value of t is 0-1, the value of t is changed from 0 to 1 after new data is read, when t is 1, the new data is completely read, parameters alpha and mu in the autoregressive model are updated, the new data at the next moment is ready to be read, and after the new data at the next moment is read, the value of t is changed from 0 to 1 again.
5. The method for online monitoring of flowmeter anomaly based on recursive Lasso as claimed in claim 1, wherein the determining whether the flowmeter is anomalous according to the predicted flow rate obtained based on the updated autoregressive model and the actual flow rate of the flowmeter at the next time specifically comprises:
predicting the flow of the flowmeter at the next moment in real time according to the updated autoregressive model to obtain predicted flow;
obtaining an estimation curve graph of the predicted flow according to the obtained predicted flow corresponding to different moments;
obtaining a residual error map according to the deviation between the predicted flow and the actual flow of the flowmeter at the next moment;
and determining whether the flow meter is abnormal according to the residual error map.
6. The method for online monitoring of flow meter anomaly based on recursive Lasso according to claim 5, wherein the determining whether the flow meter is anomalous according to the residual error map specifically comprises:
when the ordinate of the residual error map exceeds a preset threshold value, determining that the flow data of the flowmeter is abnormal;
and determining whether the reason causing the flow data abnormity of the flow meter is the abnormity of the flow meter or the batch replacement of the cut stem product according to the production record.
7. The recursion Lasso-based flowmeter anomaly online monitoring method as claimed in claim 1, wherein the recursion Lasso-based flowmeter anomaly online monitoring method is used for monitoring whether the flow rate of the humidifying water in the tobacco shredding process is abnormal,
the flow meter abnormity online monitoring method based on recursion Lasso further comprises the following steps:
and determining the cut stem product specification at the current moment based on a preset flow threshold of the cut stem product batch and the predicted flow.
8. A recursive Lasso based flowmeter anomaly on-line monitoring system employing the method of any of claims 1-7, comprising:
the autoregressive model establishing module is used for establishing an autoregressive model according to historical flow data of the flow meter and determining a model coefficient of the autoregressive model through an offline Lasso algorithm;
the autoregressive model updating module is used for updating the autoregressive model through a recursion Lasso algorithm by utilizing the real-time flow of the flowmeter;
and the flow monitoring module is used for determining whether the flowmeter is abnormal according to the predicted flow obtained based on the updated autoregressive model and the actual flow of the flowmeter at the next moment.
9. A computer-readable storage medium, having stored thereon a computer program which, when run on a computer, causes the computer to execute the recursive Lasso-based flowmeter anomaly online monitoring method according to any of claims 1 to 7.
10. A computer program product, which, when run on a terminal device, causes the terminal device to perform the method for online monitoring of flow meter anomalies based on recursive Lasso according to any of claims 1 to 7.
CN202110765956.0A 2021-07-05 2021-07-05 Recursion Lasso-based flowmeter anomaly online monitoring method and system and storage medium Active CN113468751B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202110765956.0A CN113468751B (en) 2021-07-05 2021-07-05 Recursion Lasso-based flowmeter anomaly online monitoring method and system and storage medium

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202110765956.0A CN113468751B (en) 2021-07-05 2021-07-05 Recursion Lasso-based flowmeter anomaly online monitoring method and system and storage medium

Publications (2)

Publication Number Publication Date
CN113468751A true CN113468751A (en) 2021-10-01
CN113468751B CN113468751B (en) 2022-12-27

Family

ID=77878744

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202110765956.0A Active CN113468751B (en) 2021-07-05 2021-07-05 Recursion Lasso-based flowmeter anomaly online monitoring method and system and storage medium

Country Status (1)

Country Link
CN (1) CN113468751B (en)

Citations (8)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20160026738A1 (en) * 2014-07-23 2016-01-28 International Business Machines Corporation Modeling and visualizing a dynamic interpersonal relationship from social media
CN105975443A (en) * 2016-05-04 2016-09-28 西南大学 Lasso-based anomaly detection method and system
CN107610464A (en) * 2017-08-11 2018-01-19 河海大学 A kind of trajectory predictions method based on Gaussian Mixture time series models
CN109443419A (en) * 2018-08-31 2019-03-08 广州市世科高新技术有限公司 A kind of rectifier on-line monitoring method based on machine learning
US20190286506A1 (en) * 2018-03-13 2019-09-19 Nec Laboratories America, Inc. Topology-inspired neural network autoencoding for electronic system fault detection
CN111628961A (en) * 2020-03-30 2020-09-04 西安交大捷普网络科技有限公司 DNS (Domain name Server) anomaly detection method
CN111737249A (en) * 2020-08-24 2020-10-02 国网浙江省电力有限公司 Abnormal data detection method and device based on Lasso algorithm
CN112631250A (en) * 2020-12-15 2021-04-09 中国计量大学 Fault isolation and identification method in nonlinear process based on denoising autoencoder

Patent Citations (8)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20160026738A1 (en) * 2014-07-23 2016-01-28 International Business Machines Corporation Modeling and visualizing a dynamic interpersonal relationship from social media
CN105975443A (en) * 2016-05-04 2016-09-28 西南大学 Lasso-based anomaly detection method and system
CN107610464A (en) * 2017-08-11 2018-01-19 河海大学 A kind of trajectory predictions method based on Gaussian Mixture time series models
US20190286506A1 (en) * 2018-03-13 2019-09-19 Nec Laboratories America, Inc. Topology-inspired neural network autoencoding for electronic system fault detection
CN109443419A (en) * 2018-08-31 2019-03-08 广州市世科高新技术有限公司 A kind of rectifier on-line monitoring method based on machine learning
CN111628961A (en) * 2020-03-30 2020-09-04 西安交大捷普网络科技有限公司 DNS (Domain name Server) anomaly detection method
CN111737249A (en) * 2020-08-24 2020-10-02 国网浙江省电力有限公司 Abnormal data detection method and device based on Lasso algorithm
CN112631250A (en) * 2020-12-15 2021-04-09 中国计量大学 Fault isolation and identification method in nonlinear process based on denoising autoencoder

Non-Patent Citations (2)

* Cited by examiner, † Cited by third party
Title
PEMOGEN AUTOMATIC ADAPTIVE PERFORMANCE MODELING DURING PROGRAM: "PEMOGEN:automatic adaptive performance modeling during program runtime", 《PACT "14: PROCEEDINGS OF THE 23RD INTERNATIONAL CONFERENCE ON PARALLEL ARCHITECTURES AND COMPILATION》 *
何勇 等: "基于向量自回归模型的移动通信基站流量预测", 《工业工程与管理》 *

Also Published As

Publication number Publication date
CN113468751B (en) 2022-12-27

Similar Documents

Publication Publication Date Title
CN110442936B (en) Equipment fault diagnosis method, device and system based on digital twin model
CN107765347B (en) Short-term wind speed prediction method based on Gaussian process regression and particle filtering
CN111108362B (en) Abnormal sound detection device, abnormal model learning device, abnormal sound detection method, and recording medium
Li et al. Wiener-based remaining useful life prediction of rolling bearings using improved Kalman filtering and adaptive modification
EP3182346A1 (en) A system for maintenance recommendation based on performance degradation modeling and monitoring
KR101335896B1 (en) Advanced process control system and method utilizing virtual metrology with reliance index
Zhao et al. Online fault prognosis with relative deviation analysis and vector autoregressive modeling
CN116757534B (en) Intelligent refrigerator reliability analysis method based on neural training network
KR20090121177A (en) System and method for automatic virtual metrology
US11257001B2 (en) Prediction model enhancement
JP4706608B2 (en) Manufacturing process analysis method
Chen et al. Soft sensor model maintenance: A case study in industrial processes
CN114403487B (en) Water adding control method for loosening and dampening
CN116388150A (en) Matching short-term load prediction method based on digital twin and neural network technology
CN113468751B (en) Recursion Lasso-based flowmeter anomaly online monitoring method and system and storage medium
Li et al. An adaptive prognostics method based on a new health index via data fusion and diffusion process
Iskandar et al. Maintenance of virtual metrology models
CN114254904B (en) Method and device for evaluating operation health degree of engine room of wind turbine generator
Kakati et al. Remaining useful life predictions for turbofan engine degradation using online long short-term memory network
CN115829089A (en) Load composition analysis method, device and equipment
CN115274004A (en) Knowledge reuse-based fermentation process thallus concentration prediction method and system
WO2021140942A1 (en) Diagnosing device, diagnosing method, and program
Sharma et al. Robust online algorithm for adaptive linear regression parameter estimation and prediction
Qi et al. Remaining useful life prediction of rolling element bearings based on unscented Kalman filter
CN112748660A (en) Memory, heating furnace outlet temperature control method, device and equipment

Legal Events

Date Code Title Description
PB01 Publication
PB01 Publication
SE01 Entry into force of request for substantive examination
SE01 Entry into force of request for substantive examination
GR01 Patent grant
GR01 Patent grant