CN112578740A - Fault diagnosis and processing method and system in industrial production process - Google Patents
Fault diagnosis and processing method and system in industrial production process Download PDFInfo
- Publication number
- CN112578740A CN112578740A CN201910940882.2A CN201910940882A CN112578740A CN 112578740 A CN112578740 A CN 112578740A CN 201910940882 A CN201910940882 A CN 201910940882A CN 112578740 A CN112578740 A CN 112578740A
- Authority
- CN
- China
- Prior art keywords
- production data
- real
- fault diagnosis
- historical
- data
- 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.)
- Pending
Links
Images
Classifications
-
- 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]
- G05B19/41875—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 quality surveillance of production
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02P—CLIMATE CHANGE MITIGATION TECHNOLOGIES IN THE PRODUCTION OR PROCESSING OF GOODS
- Y02P90/00—Enabling technologies with a potential contribution to greenhouse gas [GHG] emissions mitigation
- Y02P90/02—Total factory control, e.g. smart factories, flexible manufacturing systems [FMS] or integrated manufacturing systems [IMS]
Abstract
The invention discloses a fault diagnosis and processing method and system in an industrial production process. The method comprises the following steps: acquiring historical production data in an industrial production process, performing pretreatment and dimension reduction treatment, and visualizing by using a scatter diagram; determining a fault diagnosis closed area of the industrial production process; acquiring real-time production data and performing dimensionality reduction treatment; performing fault diagnosis on the real-time production data subjected to the dimensionality reduction according to the fault diagnosis closed area to obtain a diagnosis result; if the diagnosis result indicates that a fault occurs, calculating to obtain the offset of the real-time production data according to the real-time production data after dimensionality reduction and the fault diagnosis closed area; and adjusting the production data of the industrial production process to eliminate faults. The real-time fault diagnosis and processing in the whole industrial production process are realized by the fault diagnosis of the real-time production data in the whole industrial production process and the calculation of the offset of the real-time production data, and the industrial production efficiency is improved.
Description
Technical Field
The invention relates to the technical field of process industrial production, in particular to a fault diagnosis and processing method and system in an industrial production process.
Background
At present, in the field of process industrial production, a Distributed Control System (DCS) is mainly used to Control the processes and devices of industrial production. DCS is a multi-stage computer system in the tie of a communication network, consisting of a process control stage and a process monitoring stage. The DCS integrates the technologies of computer, communication, display, control and the like, the whole technological process is monitored, operated and managed in a centralized mode through the operation station, and all parts of the technological process are controlled in a decentralized mode through the control station.
The DCS is distributed, so that the state of a single device or a single production process can be controlled, systematic control of the industrial production process is lacked, when the industrial production process breaks down, a control station needs to be checked to determine a fault point, the fault point needs to be eliminated manually, real-time fault diagnosis and treatment cannot be carried out on the whole industrial production process, and the problem of low efficiency of the industrial production process exists.
Disclosure of Invention
The invention aims to provide a fault diagnosis and processing method and system in an industrial production process, and industrial production efficiency is improved.
In order to achieve the purpose, the invention provides the following scheme:
a method of fault diagnosis and handling for an industrial process, the method comprising:
acquiring historical production data in the industrial production process;
preprocessing the historical production data to obtain preprocessed historical production data;
performing dimensionality reduction on the preprocessed historical production data to obtain dimensionality-reduced historical production data;
visualizing the historical production data subjected to dimensionality reduction by using a scatter diagram to obtain a historical production data scatter diagram subjected to dimensionality reduction;
determining a fault diagnosis closed area of the industrial production process on the dimensionality reduced historical production data scatter diagram according to the historical production data and the dimensionality reduced historical production data;
acquiring real-time production data of an industrial production process;
performing dimensionality reduction on the real-time production data to obtain dimensionality-reduced real-time production data;
performing fault diagnosis on the real-time production data subjected to the dimensionality reduction according to the fault diagnosis closed area to obtain a diagnosis result;
if the diagnosis result indicates that a fault occurs, calculating to obtain the offset of the real-time production data according to the real-time production data after the dimensionality reduction and the fault diagnosis closed area;
and adjusting the production data of the industrial production process according to the offset to eliminate faults.
Optionally, the preprocessing the historical production data to obtain the preprocessed historical production data specifically includes:
deleting abnormal data in the historical production data to obtain historical production data after the abnormal data are deleted; the exception data includes undisplayed data and delayed data;
and performing low-pass filtering on the historical production data from which the abnormal data are deleted to obtain the preprocessed historical production data.
Optionally, the performing the dimensionality reduction on the preprocessed historical production data to obtain the dimensionality reduced historical production data specifically includes: and performing dimensionality reduction on the preprocessed historical production data by adopting a principal component analysis method.
Optionally, the determining the fault diagnosis closed area in the industrial production process on the dimensionality reduced historical production data scatter diagram according to the historical production data and the dimensionality reduced historical production data specifically includes:
acquiring and calculating parameters of the fault diagnosis closed area; the parameters include: the number of the historical production data points after dimensionality reduction, the dimensionality of the historical production data after dimensionality reduction and different significant horizontal values;
according to the parameters, respectively calculating prediction fault diagnosis closed areas corresponding to different significant level values on the dimensionality-reduced historical production data scatter diagram by adopting a Hotelling T square threshold calculation method or a prediction error square threshold calculation method;
respectively calculating the corresponding predicted fault accuracy rate of each predicted fault diagnosis closed region according to the historical production data and different predicted fault diagnosis closed regions;
and selecting the prediction fault diagnosis closed area corresponding to the maximum value of the prediction fault accuracy as the fault diagnosis closed area in the industrial production process.
Optionally, the performing the dimension reduction processing on the real-time production data to obtain the dimension-reduced real-time production data specifically includes: and performing dimensionality reduction on the real-time production data by adopting a principal component analysis method.
Optionally, the performing fault diagnosis on the real-time production data after the dimensionality reduction according to the fault diagnosis closed area to obtain a diagnosis result specifically includes:
visualizing the real-time production data subjected to dimensionality reduction on the historical production data scatter diagram subjected to dimensionality reduction;
judging whether the real-time production data point after dimensionality reduction falls within the range of the fault diagnosis closed area or not to obtain a first judgment result;
if the first judgment result is yes, the diagnosis result is that no fault occurs;
and if the first judgment result is negative, the diagnosis result is that the fault occurs.
Optionally, the calculating the offset of the real-time production data according to the reduced real-time production data and the fault diagnosis closed area specifically includes:
calculating the distance between the real-time production data after the dimensionality reduction and the central point between the fault diagnosis closed areas to obtain the offset of the real-time production data after the dimensionality reduction;
and carrying out inverse dimensionality reduction on the real-time production data offset subjected to dimensionality reduction to obtain the offset of the real-time production data.
A fault diagnosis and treatment system for an industrial process, the system comprising:
the historical data acquisition module is used for acquiring historical production data in the industrial production process;
the preprocessing module is used for preprocessing the historical production data to obtain preprocessed historical production data;
the first dimension reduction processing module is used for carrying out dimension reduction processing on the preprocessed historical production data to obtain the dimension-reduced historical production data;
the scatter diagram generating module is used for visualizing the scatter diagram for the historical production data after dimensionality reduction to obtain a scatter diagram for the historical production data after dimensionality reduction;
the fault diagnosis closed area calculation module is used for determining a fault diagnosis closed area in the industrial production process on the dimensionality reduced historical production data scatter diagram according to the historical production data and the dimensionality reduced historical production data;
the real-time data acquisition module is used for acquiring real-time production data in the industrial production process;
the second dimension reduction processing module is used for carrying out dimension reduction processing on the real-time production data to obtain dimension-reduced real-time production data;
the diagnosis module is used for carrying out fault diagnosis on the real-time production data subjected to the dimensionality reduction according to the fault diagnosis closed area to obtain a diagnosis result;
the offset calculation module is used for calculating the offset of the real-time production data according to the real-time production data after the dimensionality reduction and the fault diagnosis closed area if the diagnosis result shows that the fault occurs;
and the fault elimination module is used for eliminating faults according to the production data of the offset adjustment industrial production process.
According to the specific embodiment provided by the invention, the invention discloses the following technical effects:
according to the fault diagnosis and processing method of the industrial production process, firstly, dimension reduction processing is carried out on historical production data in the industrial production process, the dimension reduced historical production data is visualized by using a scatter diagram, fault diagnosis closed areas of the industrial production process are determined according to the historical production data and the dimension reduced historical production data, dimension reduction processing is carried out on real-time production data in the industrial production process, fault diagnosis is carried out on the dimension reduced real-time production data according to the fault diagnosis closed areas, and if a diagnosis result shows that a fault occurs, an offset of the real-time production data is obtained through calculation according to the dimension reduced real-time production data and the fault diagnosis closed areas; and then adjusting the production data corresponding to the industrial production process according to the offset to eliminate the fault.
The real-time production data of the whole industrial production process are diagnosed and processed in real time through the fault diagnosis of the real-time production data and the calculation of the offset of the real-time production data, and the industrial production efficiency is improved.
Drawings
In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings needed in the embodiments will be briefly described below, and it is obvious that the drawings in the following description are only some embodiments of the present invention, and it is obvious for those skilled in the art to obtain other drawings without creative efforts.
FIG. 1 is a flow chart of a method for diagnosing and handling a fault in an industrial process according to an embodiment of the present invention;
FIG. 2 is a diagram showing the main expressions of a principal component analysis method according to an embodiment of the present invention;
FIG. 3 is a diagram illustrating a distribution of historical production data after dimensionality reduction according to an embodiment of the present invention;
FIG. 4 is a dimension reduction plane identification diagram provided by an embodiment of the present invention;
FIG. 5 is a state diagram of real-time production data on a dimension reduction plane according to an embodiment of the present invention;
fig. 6 is a block diagram of a fault diagnosis and processing system of an industrial process according to an embodiment of the present invention.
Detailed Description
The technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the drawings in the embodiments of the present invention, and it is obvious that the described embodiments are only a part of the embodiments of the present invention, and not all of the embodiments. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present invention.
The invention aims to provide a fault diagnosis and processing method and system in an industrial production process, and industrial production efficiency is improved.
In order to make the aforementioned objects, features and advantages of the present invention comprehensible, embodiments accompanied with figures are described in further detail below.
Fig. 1 is a flowchart of a method for diagnosing and processing a fault in an industrial process according to an embodiment of the present invention, and as shown in fig. 1, the method for diagnosing and processing a fault in an industrial process includes:
s101: historical production data including temperature, pressure or flow rate in an industrial process is obtained.
S102: and preprocessing the historical production data to obtain preprocessed historical production data.
S103: and performing dimensionality reduction on the preprocessed historical production data to obtain dimensionality-reduced historical production data.
S104: and visualizing the historical production data subjected to dimensionality reduction by using a scatter diagram to obtain the historical production data scatter diagram subjected to dimensionality reduction.
S105: and determining a fault diagnosis closed area of the industrial production process on the dimensionality reduced historical production data scatter diagram according to the historical production data and the dimensionality reduced historical production data.
S106: real-time production data of an industrial production process is obtained, the real-time production data including real-time temperature, real-time pressure, or real-time flow.
S107: and performing dimensionality reduction on the real-time production data to obtain the dimensionality-reduced real-time production data.
S108: and carrying out fault diagnosis on the real-time production data subjected to the dimensionality reduction according to the fault diagnosis closed area to obtain a diagnosis result.
S109: and if the diagnosis result indicates that a fault occurs, calculating to obtain the offset of the real-time production data according to the real-time production data after the dimensionality reduction and the fault diagnosis closed area.
S110: and adjusting the production data of the industrial production process according to the offset to eliminate faults.
The step S101: the specific process for acquiring historical production data in the industrial production process comprises the following steps: and (4) carrying out data acquisition on the historical data of the monitored equipment or process according to the sampling frequency which is 10 times of the shortest effective information frequency. Wherein the historical production data can be the basic data such as temperature, pressure or flow rate.
The step S102: preprocessing the historical production data to obtain preprocessed historical production data specifically comprises the following steps:
deleting abnormal data in the historical production data to obtain historical production data after the abnormal data are deleted; the exception data includes undisplayed data and delayed data;
and performing low-pass filtering on the historical production data from which the abnormal data are deleted to obtain the preprocessed historical production data.
Specifically, before low-pass filtering, the method further comprises the steps of carrying out mode classification on the collected historical production data according to the process running state, and carrying out multi-point interpolation by using a mode consistent interpolation method to obtain a historical production data set. And then, filtering the historical production data set by adopting first-order low-pass filtering, removing noise by using a sliding outlier average method, and selecting the size of a filtering sliding window according to the process requirements. For example, half an hour is taken as a parameter of a sliding window in the general flow chemical industry. The filter formula is:
wherein M is the size of the sliding window, x is a variable sample, y is a value obtained by sliding filter calculation, i and j both represent data dimensions, and i is 1,2,3, …; j is 1,2,3, ….
And then, normalizing the filtered historical production data according to a formula X ═ X-mu)/sigma, wherein mu is a sample mean value, sigma is a sample standard deviation, and X is a data value obtained after normalization.
The above step S103: performing dimensionality reduction on the preprocessed historical production data to obtain dimensionality-reduced historical production data specifically comprises the following steps: and performing dimensionality reduction on the preprocessed historical production data by adopting a principal component analysis method.
In particular, principal component analysis is performed on the pre-processed historical production data set using Partial Least Squares (NIPALS), Singular Value Decomposition (SVD), or covariance algorithms. Taking the SVD method as an example, each row of the preprocessed historical production data matrix X is a sampling point, and each column is a variable. The data matrix X is divided into a form of multiplication of a matrix of U, sigma and V through singular value decomposition, and the formula is as follows:
X＝U·∑·VT with UTU＝I,VTV＝I (1)
u, sigma and V are matrixes obtained by SVD, U and V are matrixes formed by eigenvectors, and sigma is a real diagonal matrix.
Rewriting equation 1 yields:
V＝XT·U·∑-1 (2)
let matrix P consist of the column vectors of matrix V:
P＝[v1 v2 … vk] (3)
let T be U · ∑ Σ-1Equation 2 is collated as: and T is XP, wherein T is historical production data after dimensionality reduction.
As shown in fig. 2, the main expression of the principal component analysis method is constructed as a graph, where E is a reconstructed residual matrix. And establishing a principal component analysis model for the data after the dimensionality reduction by taking the corresponding eigenvector and the corresponding eigenvalue as a projection coordinate axis of the principal component analysis model, and obtaining a distribution diagram of historical production data after the dimensionality reduction through principal component analysis as shown in fig. 3.
The above step S105: determining a fault diagnosis closed area of the industrial production process on the dimensionality reduced historical production data scatter diagram according to the historical production data and the dimensionality reduced historical production data specifically comprises the following steps:
acquiring and calculating parameters of the fault diagnosis closed area; the parameters include: the number of the historical production data points after dimensionality reduction, the dimensionality of the historical production data after dimensionality reduction and different significant horizontal values; in this example, the significance level value is set to 0.1%, 1%, 5%, 10%, or 25%;
according to the parameters, respectively calculating prediction fault diagnosis closed areas corresponding to different significant level values on the dimensionality-reduced historical production data scatter diagram by adopting a Hotelling T square threshold calculation method or a prediction error square threshold calculation method;
respectively calculating the corresponding predicted fault accuracy rate of each predicted fault diagnosis closed region according to the historical production data and different predicted fault diagnosis closed regions;
and selecting the prediction fault diagnosis closed area corresponding to the maximum value of the prediction fault accuracy as the fault diagnosis closed area in the industrial production process.
Specifically, the step of respectively calculating the predictive fault diagnosis closed areas corresponding to different significant level values on the dimensionality-reduced historical production data scatter diagram by adopting a Hotelling T-square threshold calculation method comprises the following steps: according to the formula
Performing a calculation, wherein Λ is a diagonal matrix, and Λ ═ diag (λ)1,…,λk)，tiIs a vector of T.
Respectively calculating the prediction fault diagnosis closed areas corresponding to different significant level values on the dimensionality-reduced historical production data scatter diagram by adopting a prediction error square threshold calculation method, wherein the prediction fault diagnosis closed areas comprise the following steps: according to the formula
Carry out the calculation of eiTo reconstruct the vector in the residual E.
And then, automatically running the software developed in the steps from S101 to S105, calculating the real-time data and performing dimension reduction processing to obtain the result of the real-time data.
In particular, the real-time device and process data sets are sampled and pre-processed for a specified period of time. FIG. 4 is a dimension reduction plane identification diagram provided by an embodiment of the present invention; fig. 5 is a state diagram of the real-time production data on the dimensionality reduction plane according to the embodiment of the present invention, and as shown in fig. 4 and 5, the real-time production data is subjected to dimensionality reduction calculation on the dimensionality reduced historical production data scatter diagram, and the result is output to the dimensionality reduced historical production data scatter diagram.
And then, the color of the dimension reduction plane is changed according to the change of the index variable by using the identification technology of the flow index in the dimension compression plane. Wherein, the process index is a pre-defined industrial operation variable with comparative significance, such as cost, yield, health state, efficiency and the like, and the defined index variable is divided into 3-5 grades according to the numerical value;
and performing dimension reduction calculation on the defined process indexes in a dimension reduction space, and marking the dimension reduction plane with different colors according to areas where the indexes are positioned in different grades. And (4) carrying out different transformations on the index variables to achieve the most obvious identification discrimination on the dimension reduction plane.
The step S108: performing fault diagnosis on the real-time production data subjected to dimensionality reduction according to the fault diagnosis closed area, wherein the obtained diagnosis result specifically comprises the following steps:
visualizing the real-time production data subjected to dimensionality reduction on the historical production data scatter diagram subjected to dimensionality reduction;
judging whether the real-time production data point after dimensionality reduction falls within the range of the fault diagnosis closed area or not to obtain a first judgment result;
if the first judgment result is yes, the diagnosis result is that no fault occurs;
and if the first judgment result is negative, the diagnosis result is that the fault occurs.
The above step S109: the step of calculating the offset of the real-time production data according to the reduced real-time production data and the fault diagnosis closed area specifically comprises the following steps:
calculating the distance between the real-time production data after the dimensionality reduction and the central point between the fault diagnosis closed areas to obtain the offset of the real-time production data after the dimensionality reduction;
and carrying out inverse dimensionality reduction on the real-time production data offset subjected to dimensionality reduction to obtain the offset of the real-time production data.
Specifically, if the real-time production data point after dimensionality reduction fails, the real-time production data point is firstly subjected to dimensionality reduction on a dimensionality reduction plane (a historical production data scatter diagram after dimensionality reduction) according to a formula
And calculating the center coordinate c of the optimal operation area, wherein t is the data point falling in the optimal operation area, and n is the number of the data points falling in the optimal operation area.
On the dimension reduction plane, calculating the coordinate t of the real-time data point after dimension reductioniCalculating the coordinates c and tiT is equal to the distance betweeni-c. Then, the delta t is projected back to the X variable space (the data set before dimension reduction processing) to obtain delta X-delta tPTAnd taking the first 5 variables in the delta x as suggested change values for steady state optimization according to the principal component contribution degree as a sequence, namely correspondingly adjusting the first 5 variable values of the real-time production data to the first 5 variable values in the delta x, adjusting fault points and eliminating faults.
In addition, in this embodiment, the coordinate of the real-time production data point after the dimensionality reduction on the historical production data plane after the dimensionality reduction is marked as tiThe coordinate of the past data points in the dimension reduction plane is marked as tjJ is 1,2,3, …, m is the number of data points in the time window. Calculating the coordinate distance of every two adjacent points and averaging
Will taveObtaining x in the variable space before projection dimension reduction processingave＝tavepTAnd taking x in order of principal component contributionaveThe first 5 middle variables are the main variables of the dynamic influence of the current process.
In the industrial production process, areas expressing industrial indexes by different colors and the position of real-time data on a dimensionality reduction plane are used, and the numerical values of the first 5 variables in the delta x are sent to a basic control system of an industrial device.
Identifying the estimated time and probability of the real-time data point reaching the boundary of the fault diagnosis closed region, and alarming or pushing character information according to the physical variable state represented by the data; the calculation process and the message pushing of the fault diagnosis and processing method in the industrial production process can be automatically carried out on the basis of a developed real-time software platform.
The present embodiment further provides a system for diagnosing and processing a fault in an industrial process, as shown in fig. 6, the system includes:
the historical data acquisition module 1 is used for acquiring historical production data in an industrial production process, wherein the historical production data comprises temperature, pressure or flow;
the preprocessing module 2 is used for preprocessing the historical production data to obtain preprocessed historical production data;
the first dimension reduction processing module 3 is used for performing dimension reduction processing on the preprocessed historical production data to obtain dimension-reduced historical production data;
the scatter diagram generating module 4 is used for visualizing the scatter diagram for the historical production data after the dimensionality reduction to obtain a scatter diagram for the historical production data after the dimensionality reduction;
a fault diagnosis closed area calculation module 5, configured to determine a fault diagnosis closed area in the industrial production process on the dimensionality reduced historical production data scatter diagram according to the historical production data and the dimensionality reduced historical production data;
the real-time data acquisition module 6 is used for acquiring real-time production data of an industrial production process, wherein the real-time production data comprises real-time temperature, real-time pressure or real-time flow;
the second dimension reduction processing module 7 is used for performing dimension reduction processing on the real-time production data to obtain dimension-reduced real-time production data;
the diagnosis module 8 is used for carrying out fault diagnosis on the real-time production data subjected to the dimensionality reduction according to the fault diagnosis closed area to obtain a diagnosis result;
an offset calculation module 9, configured to calculate, if the diagnosis result indicates that a fault occurs, an offset of the real-time production data according to the real-time production data after the dimensionality reduction and the fault diagnosis closed area;
and the fault elimination module 10 is used for eliminating faults according to the production data of the offset adjustment industrial production process.
The fault diagnosis and processing method and system in the industrial production process also have the following effects:
1) the complex mechanism model building work is avoided, and the data set is quickly fused with the physical significance of the industrial index;
2) the process optimization, the fault prediction, the operation monitoring and the stability are fused in a plane diagram, so that industrial operators can easily understand and use the plane diagram;
3) solidifying the historical operation experience by using a data dimension reduction plane, and keeping the whole industrial process running in an optimal state;
4) the visualization method can be used in the fields of devices, equipment, management, safety and the like, and is a universal structural framework platform;
5) a simple and clear quick decision support system is provided for operators and managers, quick processing optimization can be guaranteed, control safety is improved, and industrial production efficiency is improved.
For the system disclosed by the embodiment, the description is relatively simple because the system corresponds to the method disclosed by the embodiment, and the relevant points can be referred to the method part for description.
The principles and embodiments of the present invention have been described herein using specific examples, which are provided only to help understand the method and the core concept of the present invention; meanwhile, for a person skilled in the art, according to the idea of the present invention, the specific embodiments and the application range may be changed. In view of the above, the present disclosure should not be construed as limiting the invention.
Claims (8)
1. A method for diagnosing and handling a fault in an industrial process, the method comprising:
acquiring historical production data in the industrial production process;
preprocessing the historical production data to obtain preprocessed historical production data;
performing dimensionality reduction on the preprocessed historical production data to obtain dimensionality-reduced historical production data;
visualizing the historical production data subjected to dimensionality reduction by using a scatter diagram to obtain a historical production data scatter diagram subjected to dimensionality reduction;
determining a fault diagnosis closed area of the industrial production process on the dimensionality reduced historical production data scatter diagram according to the historical production data and the dimensionality reduced historical production data;
acquiring real-time production data of an industrial production process;
performing dimensionality reduction on the real-time production data to obtain dimensionality-reduced real-time production data;
performing fault diagnosis on the real-time production data subjected to the dimensionality reduction according to the fault diagnosis closed area to obtain a diagnosis result;
if the diagnosis result indicates that a fault occurs, calculating to obtain the offset of the real-time production data according to the real-time production data after the dimensionality reduction and the fault diagnosis closed area;
and adjusting the production data of the industrial production process according to the offset to eliminate faults.
2. The method for diagnosing and processing the fault in the industrial production process according to claim 1, wherein the step of preprocessing the historical production data to obtain the preprocessed historical production data specifically comprises:
deleting abnormal data in the historical production data to obtain historical production data after the abnormal data are deleted; the exception data includes undisplayed data and delayed data;
and performing low-pass filtering on the historical production data from which the abnormal data are deleted to obtain the preprocessed historical production data.
3. The method of claim 1, wherein the fault diagnosis and treatment method comprises,
the performing dimension reduction processing on the preprocessed historical production data to obtain the dimension-reduced historical production data specifically comprises: and performing dimensionality reduction on the preprocessed historical production data by adopting a principal component analysis method.
4. The method of claim 1, wherein the fault diagnosis and treatment method comprises,
the step of determining the fault diagnosis closed area of the industrial production process on the dimensionality reduced historical production data scatter diagram according to the historical production data and the dimensionality reduced historical production data specifically comprises the following steps:
acquiring and calculating parameters of the fault diagnosis closed area; the parameters include: the number of the historical production data points after dimensionality reduction, the dimensionality of the historical production data after dimensionality reduction and different significant horizontal values;
according to the parameters, respectively calculating prediction fault diagnosis closed areas corresponding to different significant level values on the dimensionality-reduced historical production data scatter diagram by adopting a Hotelling T square threshold calculation method or a prediction error square threshold calculation method;
respectively calculating the corresponding predicted fault accuracy rate of each predicted fault diagnosis closed region according to the historical production data and different predicted fault diagnosis closed regions;
and selecting the prediction fault diagnosis closed area corresponding to the maximum value of the prediction fault accuracy as the fault diagnosis closed area in the industrial production process.
5. The method according to claim 1, wherein the performing dimension reduction on the real-time production data to obtain the dimension-reduced real-time production data specifically comprises: and performing dimensionality reduction on the real-time production data by adopting a principal component analysis method.
6. The method of claim 1, wherein the fault diagnosis and treatment method comprises,
the fault diagnosis of the real-time production data subjected to the dimensionality reduction according to the fault diagnosis closed area to obtain a diagnosis result specifically comprises the following steps:
visualizing the real-time production data subjected to dimensionality reduction on the historical production data scatter diagram subjected to dimensionality reduction;
judging whether the real-time production data point after dimensionality reduction falls within the range of the fault diagnosis closed area or not to obtain a first judgment result;
if the first judgment result is yes, the diagnosis result is that no fault occurs;
and if the first judgment result is negative, the diagnosis result is that the fault occurs.
7. The method of claim 1, wherein the fault diagnosis and treatment method comprises,
the step of calculating the offset of the real-time production data according to the reduced real-time production data and the fault diagnosis closed area specifically comprises the following steps:
calculating the distance between the real-time production data after the dimensionality reduction and the central point between the fault diagnosis closed areas to obtain the offset of the real-time production data after the dimensionality reduction;
and carrying out inverse dimensionality reduction on the real-time production data offset subjected to dimensionality reduction to obtain the offset of the real-time production data.
8. A fault diagnosis and treatment system for an industrial process, the system comprising:
the historical data acquisition module is used for acquiring historical production data in the industrial production process;
the preprocessing module is used for preprocessing the historical production data to obtain preprocessed historical production data;
the first dimension reduction processing module is used for carrying out dimension reduction processing on the preprocessed historical production data to obtain the dimension-reduced historical production data;
the scatter diagram generating module is used for visualizing the scatter diagram for the historical production data after dimensionality reduction to obtain a scatter diagram for the historical production data after dimensionality reduction;
the fault diagnosis closed area calculation module is used for determining a fault diagnosis closed area in the industrial production process on the dimensionality reduced historical production data scatter diagram according to the historical production data and the dimensionality reduced historical production data;
the real-time data acquisition module is used for acquiring real-time production data in the industrial production process;
the second dimension reduction processing module is used for carrying out dimension reduction processing on the real-time production data to obtain dimension-reduced real-time production data;
the diagnosis module is used for carrying out fault diagnosis on the real-time production data subjected to the dimensionality reduction according to the fault diagnosis closed area to obtain a diagnosis result;
the offset calculation module is used for calculating the offset of the real-time production data according to the real-time production data after the dimensionality reduction and the fault diagnosis closed area if the diagnosis result shows that the fault occurs;
and the fault elimination module is used for eliminating faults according to the production data of the offset adjustment industrial production process.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910940882.2A CN112578740A (en) | 2019-09-30 | 2019-09-30 | Fault diagnosis and processing method and system in industrial production process |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910940882.2A CN112578740A (en) | 2019-09-30 | 2019-09-30 | Fault diagnosis and processing method and system in industrial production process |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN112578740A true CN112578740A (en) | 2021-03-30 |
Family
ID=75116312
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201910940882.2A Pending CN112578740A (en) | 2019-09-30 | 2019-09-30 | Fault diagnosis and processing method and system in industrial production process |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN112578740A (en) |
Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101158873A (en) * | 2007-09-26 | 2008-04-09 | 东北大学 | Non-linearity process failure diagnosis method |
| CN103902706A (en) * | 2014-03-31 | 2014-07-02 | 东华大学 | Method for classifying and predicting big data on basis of SVM (support vector machine) |
| CN104914847A (en) * | 2015-04-09 | 2015-09-16 | 东北大学 | Industrial process fault diagnosis method based on direction kernel partial least square |
| CN105893700A (en) * | 2016-04-26 | 2016-08-24 | 陆新建 | Chemical production on-line fault detection and diagnosis technique based on physical-large data hybrid model |
| CN108266219A (en) * | 2018-01-23 | 2018-07-10 | 辽宁工程技术大学 | Mine ventilation system resistive-switching single fault source diagnostic method based on air quantity feature |
| CN109675935A (en) * | 2019-03-06 | 2019-04-26 | 北京科技大学 | A kind of IPCA operation of rolling on-line fault diagnosis method becoming control limit |
| US20190188584A1 (en) * | 2017-12-19 | 2019-06-20 | Aspen Technology, Inc. | Computer System And Method For Building And Deploying Models Predicting Plant Asset Failure |
-
2019
- 2019-09-30 CN CN201910940882.2A patent/CN112578740A/en active Pending
Patent Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101158873A (en) * | 2007-09-26 | 2008-04-09 | 东北大学 | Non-linearity process failure diagnosis method |
| CN103902706A (en) * | 2014-03-31 | 2014-07-02 | 东华大学 | Method for classifying and predicting big data on basis of SVM (support vector machine) |
| CN104914847A (en) * | 2015-04-09 | 2015-09-16 | 东北大学 | Industrial process fault diagnosis method based on direction kernel partial least square |
| CN105893700A (en) * | 2016-04-26 | 2016-08-24 | 陆新建 | Chemical production on-line fault detection and diagnosis technique based on physical-large data hybrid model |
| US20190188584A1 (en) * | 2017-12-19 | 2019-06-20 | Aspen Technology, Inc. | Computer System And Method For Building And Deploying Models Predicting Plant Asset Failure |
| CN108266219A (en) * | 2018-01-23 | 2018-07-10 | 辽宁工程技术大学 | Mine ventilation system resistive-switching single fault source diagnostic method based on air quantity feature |
| CN109675935A (en) * | 2019-03-06 | 2019-04-26 | 北京科技大学 | A kind of IPCA operation of rolling on-line fault diagnosis method becoming control limit |
Non-Patent Citations (5)
| Title |
|---|
| FENG EN-BO等: "PPP over ATM based on ADSL and G.Lite", 《TIANJIN COMMUNICATIONS TECHNOLOGY》 * |
| QI ZHAO等: "An improved fault diagnosis approach based on support vector machine", 《2016 IEEE INTERNATIONAL CONFERENCE ON PROGNOSTICS AND HEALTH MANAGEMENT(ICPHM)》 * |
| 旷天亮: "基于深度神经网络的复杂化工过程故障检测研究", 《中国优秀博硕士学位论文全文数据库（硕士） 工程科技I辑》 * |
| 白世星等: "某油田陆地井控试压程序及注意事项", 《西部探矿工程》 * |
| 鞠浩: "基于偏最小二乘法的非线性工业过程监测方法研究", 《中国优秀硕士学位论文全文数据库 信息科技辑》 * |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN109459993B (en) | Online adaptive fault monitoring and diagnosing method for process industrial process | |
| WO2021004154A1 (en) | Method for predicting remaining life of numerical control machine tool | |
| CN107632592B (en) | Nonlinear time-varying process fault monitoring method based on efficient recursion kernel principal component analysis | |
| CN105425779B (en) | ICA-PCA multi-state method for diagnosing faults based on local neighborhood standardization and Bayesian inference | |
| CN108763729B (en) | Process industry electromechanical system coupling state evaluation method based on network structure entropy | |
| CN109675935B (en) | Online fault diagnosis method for IPCA rolling process with variable control limit | |
| CN110466381B (en) | Method for evaluating state of charging pile | |
| CN109033513B (en) | Power transformer fault diagnosis method and power transformer fault diagnosis device | |
| CN113344133A (en) | Method and system for detecting abnormal fluctuation of time sequence behavior | |
| CN110807245B (en) | Automatic modeling method and system for equipment fault early warning | |
| CN113239132A (en) | Online out-of-tolerance identification method for voltage transformer | |
| CN114978956A (en) | Method and device for detecting abnormal performance mutation points of network equipment in smart city | |
| CN112858804A (en) | Power supply screen fault diagnosis method and device for railway vehicle | |
| CN116086537A (en) | Equipment state monitoring method, device, equipment and storage medium | |
| CN115664038A (en) | Intelligent power distribution operation and maintenance monitoring system for electrical safety management | |
| CN111639304A (en) | CSTR fault positioning method based on Xgboost regression model | |
| CN108759745B (en) | Body-in-white fault detection method and device based on multivariate statistical analysis | |
| CN112947649B (en) | Multivariate process monitoring method based on mutual information matrix projection | |
| CN112418577A (en) | Visual monitoring method and system for industrial product production process | |
| CN111158338A (en) | Chemical risk monitoring method based on principal component analysis | |
| CN112578740A (en) | Fault diagnosis and processing method and system in industrial production process | |
| CN116542956B (en) | Automatic detection method and system for fabric components and readable storage medium | |
| CN117216481A (en) | Remote monitoring method and system for electric appliance | |
| CN109887253B (en) | Correlation analysis method for petrochemical device alarm | |
| CN114936614B (en) | Operation risk identification method and system based on neural network |
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 | ||
| RJ01 | Rejection of invention patent application after publication | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20210330 |