CN116341993B - State monitoring method and system for polystyrene production process - Google Patents

State monitoring method and system for polystyrene production process Download PDF

Info

Publication number
CN116341993B
CN116341993B CN202310611845.3A CN202310611845A CN116341993B CN 116341993 B CN116341993 B CN 116341993B CN 202310611845 A CN202310611845 A CN 202310611845A CN 116341993 B CN116341993 B CN 116341993B
Authority
CN
China
Prior art keywords
data
abnormal
data segment
parameter
segment
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Active
Application number
CN202310611845.3A
Other languages
Chinese (zh)
Other versions
CN116341993A (en
Inventor
华啸威
吴钧
龚官浩
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.)
Wuxi Xingda Foam Plastic New Materials Co ltd
Original Assignee
Wuxi Xingda Foam Plastic New Materials 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 Wuxi Xingda Foam Plastic New Materials Co ltd filed Critical Wuxi Xingda Foam Plastic New Materials Co ltd
Priority to CN202310611845.3A priority Critical patent/CN116341993B/en
Publication of CN116341993A publication Critical patent/CN116341993A/en
Application granted granted Critical
Publication of CN116341993B publication Critical patent/CN116341993B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
    • G06Q10/00—Administration; Management
    • G06Q10/06—Resources, workflows, human or project management; Enterprise or organisation planning; Enterprise or organisation modelling
    • G06Q10/063—Operations research, analysis or management
    • G06Q10/0639—Performance analysis of employees; Performance analysis of enterprise or organisation operations
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F18/00—Pattern recognition
    • G06F18/20—Analysing
    • G06F18/22—Matching criteria, e.g. proximity measures
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F18/00—Pattern recognition
    • G06F18/20—Analysing
    • G06F18/24—Classification techniques
    • G06F18/243—Classification techniques relating to the number of classes
    • G06F18/2433—Single-class perspective, e.g. one-against-all classification; Novelty detection; Outlier detection
    • 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—Information and communication technology [ICT] specially adapted for implementation of business processes of specific business sectors, e.g. utilities or tourism
    • G06Q50/04—Manufacturing

Landscapes

  • Engineering & Computer Science (AREA)
  • Business, Economics & Management (AREA)
  • Theoretical Computer Science (AREA)
  • Human Resources & Organizations (AREA)
  • Physics & Mathematics (AREA)
  • Data Mining & Analysis (AREA)
  • General Physics & Mathematics (AREA)
  • Economics (AREA)
  • Strategic Management (AREA)
  • General Business, Economics & Management (AREA)
  • Bioinformatics & Computational Biology (AREA)
  • General Engineering & Computer Science (AREA)
  • Tourism & Hospitality (AREA)
  • Marketing (AREA)
  • Evolutionary Computation (AREA)
  • Entrepreneurship & Innovation (AREA)
  • Educational Administration (AREA)
  • Evolutionary Biology (AREA)
  • Computer Vision & Pattern Recognition (AREA)
  • Bioinformatics & Cheminformatics (AREA)
  • Artificial Intelligence (AREA)
  • Development Economics (AREA)
  • Life Sciences & Earth Sciences (AREA)
  • Primary Health Care (AREA)
  • General Health & Medical Sciences (AREA)
  • Operations Research (AREA)
  • Health & Medical Sciences (AREA)
  • Manufacturing & Machinery (AREA)
  • Game Theory and Decision Science (AREA)
  • Quality & Reliability (AREA)
  • Testing And Monitoring For Control Systems (AREA)

Abstract

The invention relates to the technical field of data processing, and discloses a state monitoring method and a state monitoring system for a polystyrene production process.

Description

State monitoring method and system for polystyrene production process
Technical Field
The invention relates to the technical field of data analysis, in particular to a state monitoring method and system used in a polystyrene production process.
Background
In order to improve the production efficiency, ensure the product quality and the operation safety in the polystyrene production process, the process parameters of the production equipment are required to be monitored, analyzed and controlled in real time, so that workers are helped to quickly find and solve the problems of insufficient yield, quality problems, abnormal conditions in the production process and the like, and the fine control of key links in the production process is realized, so that the consistency, stability and quality of the product are ensured.
The method is characterized in that the method comprises the steps of setting a standard range corresponding to each parameter data, carrying out early warning and judgment based on the deviation degree of the parameter data in the standard range as the abnormality degree, wherein each parameter in the polymerization reaction is mutually influenced, when one parameter changes or is abnormal, equipment or parameters need to be regulated according to the abnormal parameter data, and the abnormal parameter is possibly interfered by abnormality of other parameters at the moment, so that abnormality is shown, and therefore, larger interference can occur to actual reason judgment at the moment, and the abnormal data segment in one parameter is directly extracted at present, and whether the abnormal data segment is influenced by other parameters is not judged, equipment corresponding to the parameter with the abnormal data segment is checked, and the problem of the abnormal data segment cannot be accurately detected.
Disclosure of Invention
The invention is used for solving the problem that whether the abnormal data segment of the parameter in the polystyrene production process is influenced by other parameters at present to cause data to be abnormal, and provides a state monitoring method for the polystyrene production process, which can accurately judge whether the abnormal data in the parameter is influenced by other parameters, and comprises the following steps: s is(s)
Acquiring a plurality of pieces of parameter data in the polystyrene production process; selecting any piece of parameter data as target parameter data;
performing EMD (empirical mode decomposition) on the target parameter data to obtain a plurality of first IMF components;
acquiring a first abnormal data segment in each first IMF component, and obtaining a curve fluctuation characteristic value of the first abnormal data segment by utilizing the slope of each data in each first abnormal data segment and the average slope of all data in the first abnormal data segment;
the method comprises the steps that the adjacent data segment of the next segment of a first abnormal data segment in each first IMF component, the length of the adjacent data segment is consistent with that of the first abnormal data segment, and the curve fluctuation characteristic value of the adjacent data segment of the first abnormal data segment is obtained by utilizing a method for obtaining the curve fluctuation characteristic value of the first abnormal data segment;
Acquiring a target data segment in each piece of other parameter data, wherein the target data segment and the first abnormal data segment belong to the data segment in the same time period, and acquiring an abnormal influence coefficient of each first IMF component in the target parameter data and each other parameter data by using a curve fluctuation characteristic value of each first abnormal data segment in each first IMF component, a curve fluctuation characteristic value of an adjacent data segment of the first abnormal data segment and a standard deviation of the target data segment in each piece of other parameter data corresponding to the first abnormal data segment;
acquiring a matching data segment matched with each first abnormal data segment in each first IMF component in each other parameter data;
acquiring the relevance of each first abnormal data segment and the matched data segment matched in each other parameter data by using the curve of each first abnormal data segment in each first IMF component in the target data and the curve of the matched data segment matched in each other parameter data;
obtaining the degree of embodying the influence of each piece of other parameter data on each piece of first IMF component in the target parameter data according to the abnormal influence coefficient of each piece of first IMF component and each piece of other parameter data and the relevance of each piece of first abnormal data segment in each piece of first IMF component and the matched parameter data segment in each piece of other parameter data;
And acquiring a data segment corresponding to the first IMF component with the largest embodying degree of all the first IMF components in the target parameter data as a required data segment, and performing abnormality judgment.
The expression for obtaining the curve fluctuation characteristic value of the first abnormal data segment is as follows:
wherein:representing the first of the target parameter dataFirst IMF component of the first IMF componentCurve fluctuation characteristic values of the first abnormal data segments;represent the firstThe first abnormal data segmentSlope of the individual data;represent the firstAverage slope of all data in the first abnormal data segment;represent the firstThe total number of all data in the first abnormal data segment;represent the firstThe first abnormal data segmentWeight difference of the individual data;representing normalization.
The method for obtaining the abnormal influence coefficient of each first IMF component and each other parameter data in the target parameter data comprises the following steps:
the expression of the anomaly impact coefficient is:
wherein:representing the first of the target parameter dataAbnormal influence coefficients of the first IMF component and the ith other parameter data;represent the firstFirst IMF component of the first IMF componentStandard deviation of the target data segment in the ith other parameter data corresponding to the first abnormal data segment; Representing the first of the target parameter dataThe total number of first abnormal data segments in the first IMF component;；representing the first of the target parameter dataFirst IMF component of the first IMF componentCurve fluctuation characteristic values of the first abnormal data segments;representing the first of the target parameter dataFirst IMF component of the first IMF componentCurve fluctuation characteristic values of adjacent data segments of the first abnormal data segment;representing normalization.
The method for acquiring the matched data segment matched with each first abnormal data segment in each first IMF component in the target parameter data in each other parameter data comprises the following steps:
performing EMD (empirical mode decomposition) on each piece of other parameter data to obtain a plurality of second IMF components, and obtaining a second abnormal data segment in each second IMF component;
respectively acquiring the mean time of the first abnormal data segment and the mean time of the second abnormal data segment by utilizing the moment of each data segment in each first abnormal data segment in each first IMF component and the moment of each data segment in each second abnormal data segment in the second IMF component;
and comparing the mean time of each first abnormal data segment with the mean time of each second abnormal data segment to obtain a second abnormal data segment matched with each first abnormal data segment, and taking the matched second abnormal data segment as a matched data segment.
The method for acquiring the relevance of each first abnormal data segment and the matched data segment matched in each other parameter data comprises the following steps:
acquiring a first curve of each first abnormal constant section and a second curve of a matched data section matched in each piece of other parameter data;
the centers of gravity of the first curve and the second curve are obtained, and the first curve or the second curve is moved to enable the centers of gravity of the first curve and the second curve to coincide;
acquiring the interval area of the first curve and the second curve after the centers of gravity are overlapped;
and acquiring the relevance of each first abnormal data segment and the matched data segment matched in each other parameter data by using the interval area of the first curve and the second curve.
The expression of the relevance is:
wherein:representing the first of the target parameter dataFirst IMF component of the first IMF componentThe first abnormal data segment is matched with the ith other parameter dataThe relevance of the individual matching data segments;representing the first of the target parameter dataFirst IMF component of the first IMF componentFirst data in the first abnormal data segment is matched with the ith other parameter dataA time span of the first data in the matching data segment;representing the first of the target parameter dataFirst IMF component of the first IMF component First abnormal data segment and ith barMatched first of other parameter dataThe time spans of the matching data segments;representing the first of the target parameter dataFirst IMF component of the first IMF componentFirst curve corresponding to first abnormal data segment and the first curve matched with the ith other parameter dataThe interval area of the second curve corresponding to each matching data segment.
The method for obtaining the degree of the influence of each piece of other parameter data on each piece of first IMF component in the target parameter data comprises the following steps:
wherein:representing the third of all other parameter data versus the target parameter dataDegree of embodiment of the influence of the first IMF component;representing the first of the target parameter dataAbnormal influence coefficients of the first IMF component and the ith other parameter data;representing the first of the target parameter dataFirst IMF component of the first IMF componentThe first abnormal data segment is matched with the ith other parameter dataThe relevance of the individual matching data segments;representing the amount of other parameter data;representing the first of the target parameter dataThe total number of first abnormal data segments in the first IMF component;characterization of the first embodimentThe other parameter data is the first one in the target parameter dataA first IMF componentThe associations in the individual anomaly segments are summed.
A state monitoring system used in the polystyrene production process; comprising the following steps:
and a data acquisition module: a plurality of pieces of parameter data for use in a polystyrene production process; selecting any piece of parameter data as target parameter data;
EMD decomposition module: EMD is carried out on the target parameter data and other parameter data to obtain a plurality of first IMF components and second IMF components;
abnormal influence coefficient acquisition module: acquiring a first abnormal data segment in each first IMF component, and obtaining a curve fluctuation characteristic value of the first abnormal data segment by utilizing the slope of each data in each first abnormal data segment and the average slope of all data in the first abnormal data segment;
the method comprises the steps that the adjacent data segment of the next segment of a first abnormal data segment in each first IMF component, the length of the adjacent data segment is consistent with that of the first abnormal data segment, and the curve fluctuation characteristic value of the adjacent data segment of the first abnormal data segment is obtained by utilizing a method for obtaining the curve fluctuation characteristic value of the first abnormal data segment;
acquiring a target data segment in each piece of other parameter data, wherein the target data segment and the first abnormal data segment belong to the data segment in the same time period, and acquiring an abnormal influence coefficient of each first IMF component in the target parameter data and each other parameter data by using a curve fluctuation characteristic value of each first abnormal data segment in each first IMF component, a curve fluctuation characteristic value of an adjacent data segment of the first abnormal data segment and a standard deviation of the target data segment in each piece of other parameter data corresponding to the first abnormal data segment;
And the relevance calculating module is used for: acquiring a matching data segment matched with each first abnormal data segment in each first IMF component in each other parameter data;
acquiring the relevance of each first abnormal data segment and the matched data segment matched in each other parameter data by using the curve of each first abnormal data segment in each first IMF component in the target data and the curve of the matched data segment matched in each other parameter data;
the embodiment calculating module is used for: obtaining the degree of embodying the influence of each piece of other parameter data on each piece of first IMF component in the target parameter data according to the abnormal influence coefficient of each piece of first IMF component and each piece of other parameter data and the relevance of each piece of first abnormal data segment in each piece of first IMF component and the matched parameter data segment in each piece of other parameter data;
the abnormality judgment module: and acquiring data segments corresponding to the first IMF components with the largest degree of embodiment of all the first IMF components in the target parameter data as required data segments, and judging whether the abnormality of the required data segments is influenced by other parameter data.
The beneficial effects of the invention are as follows: the method comprises the steps of carrying out EMD (empirical mode decomposition) on any one of a plurality of pieces of acquired parameter data to obtain an IMF component, obtaining an abnormal data segment in the IMF component, obtaining an abnormal parameter coefficient by utilizing the curve fluctuation characteristic values of the abnormal data segment and the next data segment of the abnormal data segment, obtaining the relevance of the abnormal data segment and the matched data segment in other parameter data, obtaining the degree of embodiment of IMF according to the abnormal parameter coefficient and the relevance of the abnormal data in the IMF component, determining whether the abnormal data in the IMF component is affected by other parameters or not by utilizing the degree of embodiment of IMF, carrying out EMD decomposition on the historical monitoring data of each parameter, and analyzing the degree of embodiment of the abnormal condition of one parameter in each IMF component, so that when the parameter is abnormal, according to the analysis of the degree of embodiment of different parameters in each component, whether the abnormal data in the parameters are affected by other parameters or not can be accurately judged, providing an accurate conclusion for the equipment abnormality by the staff, and further carrying out quick adjustment on the equipment or maintenance on the equipment in terms of quality due to the fact that the equipment is greatly influenced by the overlong parameter adjustment is carried out, thereby avoiding the quality adjustment on the equipment.
Drawings
In order to more clearly illustrate the embodiments of the invention or the technical solutions of the prior art, the drawings which are used in the description of the embodiments or the prior art will be briefly described, it being obvious that the drawings in the description below are only some embodiments of the invention, and that other drawings can be obtained according to these drawings without inventive faculty for a person skilled in the art.
FIG. 1 is a schematic flow chart of the present invention;
FIG. 2 is a block diagram of a system architecture of the present invention;
fig. 3 is a schematic diagram of the center of gravity coincidence of the first curve and the second curve in the embodiment of the present invention.
Detailed Description
The following description of the embodiments of the present invention will be made clearly and completely with reference to the accompanying drawings, in which it is apparent that the embodiments described are only some embodiments of the present invention, but not all embodiments. All other embodiments, which can be made by those skilled in the art based on the embodiments of the invention without making any inventive effort, are intended to be within the scope of the invention.
Examples
The embodiment as shown in fig. 1 provides a method for monitoring a state in a polystyrene production process, which comprises the following steps:
Acquiring a plurality of pieces of parameter data in the polystyrene production process; selecting any piece of parameter data as target parameter data; in this embodiment, the parameter data is obtained by collecting different parameter data in real time by using sensors of different parameters provided on the equipment for producing polystyrene, and in this embodiment, the pressure parameter and the temperature parameter are taken as examples to describe the embodiment, and when the parameters are collected, the data collection frequency is usually 10-20 times per second; after one parameter is analyzed, the other parameters are analyzed to analyze all the parameters, so as to judge whether the abnormal data in the parameters are caused by the other parameters or not, and further, the equipment can be accurately detected and maintained according to the abnormal data of the parameters; specifically, any one piece of parameter data in a plurality of pieces of parameter data is used as target parameter data, and after the target parameter data is detected, other parameters can be detected according to a target parameter detection method;
in the polystyrene production process, monitoring data in a polymerization reaction system and a vacuum system can interfere with each other along with the time and between parameters, so that short-period oscillation and long-period change trend appear in a data curve, and therefore the conditions of EMD are met;
In view of this, the present embodiment performs EMD decomposition on the target parameter data to obtain a plurality of first IMF components; the method specifically includes performing EMD decomposition on a curve formed by target parameter data, wherein the EMD decomposition adopted by the embodiment belongs to a conventional means of a person skilled in the art, and the EMD decomposition is performed on target data parameters (time sequence parameter data) without detailed description, so as to obtain a first IMF component after the EMD decomposition;
the polystyrene reactor is provided with a polymerization reaction monitoring system and a vacuum system, wherein parameters in the same polymerization reactor are monitored and adjusted, and the interference influence degree of each parameter on other parameters is different, namely the fluctuation condition shown on the data change is different; for example, for parameter a, when parameter a changes, it may be subjected to a change in parameters B, C, which in turn causes a change, but when parameters B, C change by the same magnitude, the effect on parameter a is different, it may be that B causes a strong fluctuation, and C causes a slow gradual change trend; thus, by performing EMD decomposition on a data curve in which certain parameter data, the influence of different parameters on a certain parameter will mainly occur in a certain component with different frequencies.
When abnormal fluctuation or deviation occurs to the target parameter data, in order to prevent the influence on the production quality or the danger caused by overlong time, the rest parameters are usually regulated in a short time to restore the normal level, so that the continuous duration corresponding to each fluctuation in the parameters is not longer, and the duration corresponding to the window size is set to be three hours according to the experience value.
The historical data of a certain parameter is subjected to the first step of the set window sizeThe IMF components are processed by sliding window, and the first is calculatedStandard deviation of data within individual windows, i.e.WhereinCharacterization of target parameter dataThe IMF component weightsWithin the window (th)Data amplitude. ThenThe meaning of the characterization is the first of the target parameter dataThe IMF component weightsThe data amplitude averages within the windows. The larger the standard deviation, the stronger the fluctuation of the data in this window, i.e. the higher the probability of the data having the above-mentioned abnormal situation such as deviation,
since it is not uncommon for deviations to occur during the production of polystyrene, but at the same time less frequent, the standard deviation calculated for each window is used here And (5) sorting in a descending order, and taking a window with the size of 30 percent before, wherein the data in the window is considered to have abnormal conditions. Meanwhile, since the recovery time of the abnormal condition of the data possibly occurs once exceeds the window size, judging whether each window meets the conditions or not by judging the adjacent window of each window, if so, considering that the two windows correspond to the same abnormal condition, judging the combination of the two windows, repeating until the window which also meets the abnormal condition judging conditions does not occur in the adjacent moment of each segment, and naming the combination of the segments as the same segmentAn abnormal section; and finally obtaining the first abnormal data segment in each first IMF component.
In a long-time historical time sequence range, although the target parameter data are changed and influence is caused on a plurality of other parameters, the corresponding change amplitude of the corresponding parameter which is mainly changed is larger, so that the relevance of the two parameters is analyzed by counting the condition that each parameter is fluctuated in a section corresponding to abnormal data fluctuation in a certain IMF component, and the condition that the corresponding other parameters are fluctuated in the section corresponding time according to the amplitude, the data fluctuation condition and the like;
For the first abnormal data segment in each IMF component decomposed by the target parameter data, obtaining a curve fluctuation characteristic value of the first abnormal data segment by utilizing the slope of each data in each first abnormal data segment and the average slope of all data in the first abnormal data segment;
the expression of the curve fluctuation characteristic value of the first abnormal data segment is as follows:
wherein:representing the first of the target parameter dataFirst IMF component of the first IMF componentCurve fluctuation characteristic values of the first abnormal data segments;represent the firstThe first abnormal data segmentSlope of the individual data;represent the firstAverage slope of all data in the first abnormal data segment;represent the firstThe total number of all data in the first abnormal data segment;represent the firstThe first abnormal data segmentWeight difference of the individual data;representation normalization; the saidThe standard value range is set directly by the monitoring personnel according to the experience value, and no specific value is given here.
After the curve fluctuation characteristic value of a first abnormal data segment in the first IMF component is obtained, taking the time sequence length corresponding to the first abnormal data segment as the window size, and obtaining the data segment of the next segment next to the first abnormal data segment as an adjacent data segment, wherein the length of the adjacent data segment is consistent with that of the first abnormal data segment;
Acquiring the curve fluctuation characteristic value of the adjacent data segment of the first abnormal data segment by utilizing the method for acquiring the curve fluctuation characteristic value of the first abnormal data segment；
In view of this, obtain the goal data field in every other parameter data, and this goal data field and first unusual data field belong to the data field of the same time quantum, specifically obtain the time sequence that this goal data field is located, utilize the same time sequence to obtain the data field of other parameters, regard the data field of this other parameters and first unusual data field of time sequence as the goal data field, after obtaining the goal data field in every other parameter, utilize the curve fluctuation characteristic value of every first unusual data field in every first IMF component, curve fluctuation characteristic value of the adjacent data field of this first unusual data field, and standard deviation of the goal data field in every other parameter data that the first unusual data field corresponds to obtain the unusual influence coefficient of every first IMF component and every other parameter data in the goal parameter data;
the expression of the abnormal influence coefficient is as follows:
wherein:representing the first of the target parameter dataAbnormal influence coefficients of the first IMF component and the ith other parameter data; Represent the firstFirst IMF component of the first IMF componentStandard deviation of the target data segment in the ith other parameter data corresponding to the first abnormal data segment;representing the first of the target parameter dataThe total number of first abnormal data segments in the first IMF component;；representing the first of the target parameter dataFirst IMF component of the first IMF componentCurve fluctuation characteristic values of the first abnormal data segments;representing the first of the target parameter dataFirst IMF component of the first IMF componentCurve fluctuation characteristic values of adjacent data segments of the first abnormal data segment;representing normalization. And scaling the value to be within the value range of 0-1 according to the relation of the size proportion by a maximum and minimum value normalization algorithm. When the value is larger, the condition that the target parameter data is abnormal and the data of the target parameter data is fluctuated is characterized, and when the current parameter is influenced and the data is abnormal, the reflected influence is represented under the current component, and compared with other parameters, the influence is mainly caused by the abnormal influence of the target parameter data.
When the target parameter data is changed and the corresponding parameter is affected, the target parameter data is not changed at the same time, but has certain hysteresis, so that the abnormal data segment of the parameter which affects the parameter is advanced compared with the parameter, and meanwhile, the influence of the changed parameter on the current parameter has certain similarity from the change of the data curve, so that the influence ratio of a certain parameter in a plurality of components can not be completely used as a reference value only according to the influence ratio of the certain parameter, and the most important component is found according to the corresponding influence change in different components.
In view of this, it is necessary to acquire a matching data segment in the other parameter data that matches the first abnormal data segment in the target parameter data;
the method for acquiring the matched data segment comprises the following steps:
performing EMD (empirical mode decomposition) on each piece of other parameter data to obtain a plurality of second IMF components, and obtaining a second abnormal data segment in each second IMF component; the method for acquiring the second IMF component of each other parameter data and the second abnormal data segment in each second IMF component is the same as the method for acquiring the first IMF component of the target parameter data and the first abnormal data segment in each first IMF component, and a detailed description is not given to how to acquire the second IMF component and the second abnormal data segment in the second IMF component in this embodiment;
then, respectively acquiring the mean time of the first abnormal data segment and the mean time of the second abnormal data segment by utilizing the time of each data segment in each first abnormal data segment in each first IMF component and the time of each data segment in each second abnormal data segment in the second IMF component; the method for acquiring the mean time of the first abnormal data segment in this embodiment is illustrated, the mean time of the second abnormal data segment is the same as the method for acquiring the mean time of the first abnormal data segment, and will not be described in detail in this embodiment
The expression for obtaining the mean time of the first abnormal data segment is:
wherein:characterizing the first abnormal data segmentThe time of day of the individual data,characterization of the firstThe deviation value of the data, namely the difference value from the intermediate value in the standard range, belongs to an empirical value and is set by a worker;indicating the number of all data in the first abnormal data segment at this time; when calculating the moment corresponding to the central position of the first abnormal data segment, the local abnormality at the moment corresponding to the final mean time is maximum by giving the weight of the data abnormal value corresponding to the moment to each moment.
Repeating the above operation to obtain a second abnormal data segment in a second IMF component in other parameter dataMean time of (1)。
Comparing the mean time of each first abnormal data segment with the mean time of each second abnormal data segment to obtain a second abnormal data segment matched with each first abnormal data segment, and taking the matched second abnormal data segment as a matched data segment;
specifically, the time span of the first abnormal data segment in the first IMF component and the second abnormal data segment in the second IMF component in each piece of other parameter data is calculated SelectingThe first abnormal data segment and the second abnormal data segment corresponding to the value of 0 are the data matched with each other, and the second abnormal data segment is taken as the first abnormal data segmentMatching the abnormal data segment when the abnormal data segment is matchedWhen the first data abnormal section in the target parameter data section is considered later than the second abnormal data section in the second IMF component in the other parameter data. That is, at this time, it is considered that the first abnormal data segment in the first IMF component in the target parameter data is abnormal due to the influence of the second abnormal data segment in the corresponding other parameter data after matching, while satisfyingWhen the first data anomaly segment in the target parameter data segment is considered earlier than the second anomaly data segment in the second IMF component in the other parameter data. That is, at this time, it is considered that the first abnormal data segment in the first IMF component in the target parameter data is abnormal without being affected by the second abnormal data segment in the corresponding other parameter data after matching.
After the matched abnormal data segments in each piece of other parameter data after the first abnormal data segments in the target data segments are matched are obtained, namely if the other parameter data is three, the matched abnormal data segments matched by the first abnormal data segments are three segments;
Dynamically planning two data curves of each obtained first abnormal data segment and a data curve formed by the matched data segment by using a DTW algorithm, aligning the lengths, and mapping the lengths into the same coordinate system;
specifically, a first curve of each first abnormal constant section and a second curve of matched data sections matched in each other parameter data are obtained
Aligning the first curve and the second curve as shown in fig. 3, obtaining the centers of gravity of the first curve and the second curve, and then moving the first curve or the second curve to coincide with the centers of gravity of the first curve and the second curve;
acquiring the interval area of the first curve and the second curve after the barycenter is overlapped by using the images with the barycenter overlapped; the specific algorithm is that a fixed integral algorithm is adopted to obtain the interval area, and the calculated area is a shadow part shown in figure 3; the fixed integral algorithm belongs to a conventional means in the art, and the calculation of the interval area by the algorithm in this embodiment belongs to common general knowledge, and will not be described in detail again.
Calculating the relevance of the first abnormal data segment and the matched data segment matched with each other parameter data by the time difference between the two curves and the area difference between the corresponding curves:
The expression of the relevance is:
wherein:representing the first of the target parameter dataFirst IMF component of the first IMF componentThe first abnormal data segment is matched with the ith other parameter dataThe relevance of the individual matching data segments;representing the first of the target parameter dataFirst IMF component of the first IMF componentFirst data in the first abnormal data segment is matched with the ith other parameter dataA time span of the first data in the matching data segment;representing the first of the target parameter dataFirst IMF component of the first IMF componentThe first abnormal data segment is matched with the ith other parameter dataThe time spans of the matching data segments;representing the first of the target parameter dataFirst IMF component of the first IMF componentFirst curve corresponding to first abnormal data segment and the first curve matched with the ith other parameter dataThe interval area of the second curve corresponding to each matching data segment.
Due toRepresenting the first of the target parameter dataFirst IMF component of the first IMF componentThe first abnormal data segment is matched with the ith other parameter dataThe time span of each matching data segment does not immediately cause other parameter data to change when the target parameter data is abnormal, but gradually characterizes the affected condition, so that the difference between the first moments is equal to true The closer the time span between the times at which the positive main abnormality occurs, i.e., the more the first abnormal data segment of the corresponding target parameter data and the second abnormal data segment in the other parameter data are in time series to conform to the affected feature, and thus are used herein as weight values.The first of the target parameter dataFirst IMF component of the first IMF componentFirst curve corresponding to first abnormal data segment and the first curve matched with the ith other parameter dataThe more similar the data curves of the first abnormal data segment of the target parameter data and the second abnormal data segment in the other parameter data are, namely, the more similar the change characteristics of the data amplitude reflected from the trend are. I.e. the more the second anomalous data segment in the other parametric data is caused by the first anomalous data segment of the target parametric data. Thus by combining the two features described above with a weight value multiplied by the magnitude of the trend difference.
Obtaining the degree of embodying the influence of each piece of other parameter data on each piece of first IMF component in the target parameter data according to the abnormal influence coefficient of each piece of first IMF component and each piece of other parameter data and the relevance of each piece of first abnormal data segment in each piece of first IMF component and the matched parameter data segment in each piece of other parameter data;
Expression of degree of embodiment:
wherein:representing all other parameter dataFor the first of the target parameter dataDegree of embodiment of the influence of the first IMF component;representing the first of the target parameter dataAbnormal influence coefficients of the first IMF component and the ith other parameter data;representing the first of the target parameter dataFirst IMF component of the first IMF componentThe first abnormal data segment is matched with the ith other parameter dataThe relevance of the individual matching data segments;representing the amount of other parameter data;representing the first of the target parameter dataThe total number of first abnormal data segments in the first IMF component;characterization of the first embodimentThe other parameter data is the first one in the target parameter dataA first IMF componentThe associations in the individual anomaly segments are summed.
And multiplying the two characteristic values to obtain the degree of the embodiment of the influence of the IMF component in the target parameter data in each IMF component in other parameter data. Since the target parameter data, although likely to have its effect represented mainly in the 1 st IMF component, may be represented more on its effect in the current IMF component than in the 2 nd IMF component, is represented almost entirely by the first parameter, although less highly than in the first IMF component, because it is more prominent than the rest of the parameters, it is more suitable to observe in the 2 nd IMF component when observing the reaction representation of the effect of the first parameter on the current one. Thus by calculating the first The above-mentioned characteristic of one parameter and the rest of the parameters are at the current point of this parameterThe degree of embodiment in the IMF components is convenient for subsequent real abnormal condition analysis.
Acquiring data segments corresponding to first IMF components with the largest degree of embodiment of all first IMF components in target parameter data as required data segments, performing abnormality judgment, wherein the explanation of the data segment with the largest abnormality is most likely to be influenced by other parameter data, so that the largest degree of embodiment is compared with a set threshold value, if the degree of embodiment is larger than the set threshold value, the abnormality is not caused by equipment corresponding to the parameter itself but caused by the abnormality of other parameters, and further, when subsequent equipment is maintained or parameter is adjusted, the equipment corresponding to the parameter is not required to be adjusted and maintained, the detection workload is reduced, and the detection efficiency is improved; the threshold value of this embodiment is set to 0.7.
The method comprises the steps of carrying out EMD (empirical mode decomposition) on any one of a plurality of pieces of acquired parameter data to obtain an IMF component, obtaining an abnormal data segment in the IMF component, obtaining an abnormal parameter coefficient by utilizing the curve fluctuation characteristic values of the abnormal data segment and the next data segment of the abnormal data segment, obtaining the relevance of the abnormal data segment and the matched data segment in other parameter data, obtaining the degree of reflection of the IMF according to the abnormal parameter coefficient and the relevance of the abnormal data in the IMF component, determining whether the abnormal data in the IMF component is affected by other parameters or not by utilizing the degree of reflection of the IMF, carrying out EMD decomposition on the historical monitoring data of each parameter, and analyzing the degree of reflection of the abnormal condition of one parameter in each IMF component, so that when the parameter is abnormal, according to the analysis of the degree of reflection of different parameters in each component, whether the abnormal data in the parameters is affected by other parameters or not can be accurately judged, providing an accurate conclusion for the worker to the abnormality of equipment, and further carrying out quick adjustment on the equipment or maintenance on the equipment when the equipment is adjusted, thereby greatly avoiding the influence on the quality of polystyrene on the equipment due to the quality.
The embodiment as shown in fig. 2 also provides a state monitoring system used in the polystyrene production process; comprising the following steps:
and a data acquisition module: a plurality of pieces of parameter data for use in a polystyrene production process; selecting any piece of parameter data as target parameter data;
EMD decomposition module: EMD is carried out on the target parameter data and other parameter data to obtain a plurality of first IMF components and second IMF components;
abnormal influence coefficient acquisition module: acquiring a first abnormal data segment in each first IMF component, and obtaining a curve fluctuation characteristic value of the first abnormal data segment by utilizing the slope of each data in each first abnormal data segment and the average slope of all data in the first abnormal data segment;
the method comprises the steps that the adjacent data segment of the next segment of a first abnormal data segment in each first IMF component, the length of the adjacent data segment is consistent with that of the first abnormal data segment, and the curve fluctuation characteristic value of the adjacent data segment of the first abnormal data segment is obtained by utilizing a method for obtaining the curve fluctuation characteristic value of the first abnormal data segment;
acquiring a target data segment in each piece of other parameter data, wherein the target data segment and the first abnormal data segment belong to the data segment in the same time period, and acquiring an abnormal influence coefficient of each first IMF component in the target parameter data and each other parameter data by using a curve fluctuation characteristic value of each first abnormal data segment in each first IMF component, a curve fluctuation characteristic value of an adjacent data segment of the first abnormal data segment and a standard deviation of the target data segment in each piece of other parameter data corresponding to the first abnormal data segment;
And the relevance calculating module is used for: acquiring a matching data segment matched with each first abnormal data segment in each first IMF component in each other parameter data;
acquiring the relevance of each first abnormal data segment and the matched data segment matched in each other parameter data by using the curve of each first abnormal data segment in each first IMF component in the target data and the curve of the matched data segment matched in each other parameter data;
the embodiment calculating module is used for: obtaining the degree of embodying the influence of each piece of other parameter data on each piece of first IMF component in the target parameter data according to the abnormal influence coefficient of each piece of first IMF component and each piece of other parameter data and the relevance of each piece of first abnormal data segment in each piece of first IMF component and the matched parameter data segment in each piece of other parameter data;
the abnormality judgment module: and acquiring data segments corresponding to the first IMF components with the largest degree of embodiment of all the first IMF components in the target parameter data as required data segments, and judging whether the abnormality of the required data segments is influenced by other parameter data.
The EMD analysis module carries out EMD component on any parameter data in a plurality of parameter data acquired by the data acquisition module to obtain an IMF component, an abnormal data section in the IMF component is obtained, the abnormal influence coefficient acquisition module utilizes the abnormal data section and the curve fluctuation characteristic value of the next section of the abnormal data section to obtain an abnormal parameter coefficient, the relevance calculation module acquires the relevance of the matched data section in other parameter data of the abnormal data section, the relevance calculation module obtains the degree of the appearance of the IMF according to the abnormal parameter coefficient and the relevance of the abnormal data in the IMF component, the abnormal judgment module utilizes the degree of the appearance of the IMF to determine whether the abnormal data in the IMF component is influenced by other parameters, the EMD analysis is carried out through the history monitoring data of each parameter, and the abnormal condition degree of a certain parameter in each IMF component is analyzed, so that when the parameters are abnormal, the degree of the different parameters in each component is used as the analysis of related factors, whether the abnormal data in the parameters is influenced by other parameters can be accurately judged, the abnormal degree of the abnormal data in the parameters is provided, the equipment can be accurately regulated, the quality of the equipment can be greatly regulated when the equipment is greatly influenced by the abnormal equipment is guaranteed, and the quality of the equipment is greatly regulated, and the abnormal equipment can be greatly regulated, and the quality of the equipment can be greatly regulated when the equipment is greatly is adjusted.
The foregoing description of the preferred embodiments of the invention is not intended to be limiting, but rather is intended to cover all modifications, equivalents, alternatives, and improvements that fall within the spirit and scope of the invention.

Claims (7)

1. A method for monitoring conditions during the production of polystyrene, comprising:
acquiring a plurality of pieces of parameter data in the polystyrene production process; selecting any piece of parameter data as target parameter data;
performing EMD (empirical mode decomposition) on the target parameter data to obtain a plurality of first IMF components;
acquiring a first abnormal data segment in each first IMF component, and obtaining a curve fluctuation characteristic value of the first abnormal data segment by utilizing the slope of each data in each first abnormal data segment and the average slope of all data in the first abnormal data segment;
the method comprises the steps that the adjacent data segment of the next segment of a first abnormal data segment in each first IMF component, the length of the adjacent data segment is consistent with that of the first abnormal data segment, and the curve fluctuation characteristic value of the adjacent data segment of the first abnormal data segment is obtained by utilizing a method for obtaining the curve fluctuation characteristic value of the first abnormal data segment;
acquiring a target data segment in each piece of other parameter data, wherein the target data segment and the first abnormal data segment belong to the data segment in the same time period, and acquiring an abnormal influence coefficient of each first IMF component in the target parameter data and each other parameter data by using a curve fluctuation characteristic value of each first abnormal data segment in each first IMF component, a curve fluctuation characteristic value of an adjacent data segment of the first abnormal data segment and a standard deviation of the target data segment in each piece of other parameter data corresponding to the first abnormal data segment;
Acquiring a matching data segment matched with each first abnormal data segment in each first IMF component in each other parameter data;
acquiring the relevance of each first abnormal data segment and the matched data segment matched in each other parameter data by using the curve of each first abnormal data segment in each first IMF component in the target data and the curve of the matched data segment matched in each other parameter data;
obtaining the degree of embodying the influence of each piece of other parameter data on each piece of first IMF component in the target parameter data according to the abnormal influence coefficient of each piece of first IMF component and each piece of other parameter data and the relevance of each piece of first abnormal data segment in each piece of first IMF component and the matched parameter data segment in each piece of other parameter data;
the method for obtaining the degree of the influence of each piece of other parameter data on each piece of first IMF component in the target parameter data comprises the following steps:
wherein:indicating the +.f. of all other parameter data to the target parameter data>Degree of embodiment of the influence of the first IMF component;representing the +.>Abnormal influence coefficients of the first IMF component and the ith other parameter data; />Representing the +. >The first IMF component is +.>The first abnormal data segment is matched with the ith part of the other parameter data in the ith part of the other parameter data>The relevance of the individual matching data segments; />Representing the amount of other parameter data; />Representing the +.>The total number of first abnormal data segments in the first IMF component; />Characterization of->The other parameter data are +.>First IMF component +.>The associative sums in the abnormal segments;
and acquiring data segments corresponding to the first IMF components with the largest degree of embodiment of all the first IMF components in the target parameter data as required data segments, and judging whether the abnormality of the required data segments is influenced by other parameter data.
2. The method for monitoring conditions in a polystyrene production process according to claim 1, wherein the expression for obtaining the characteristic value of curve fluctuation of the first abnormal data segment is:
wherein:representing the +.>The first IMF component is +.>Curve fluctuation characteristic values of the first abnormal data segments; />Indicate->The first abnormal data segment is +.>Slope of the individual data; />Indicate->Average slope of all data in the first abnormal data segment; / >Indicate->The total number of all data in the first abnormal data segment; />Indicate->The first abnormal data segment is +.>Weight difference of the individual data; />Representing normalization.
3. The method for monitoring the state in the polystyrene production process according to claim 1, wherein the method for obtaining the abnormal influence coefficient of each first IMF component in the target parameter data and each other parameter data comprises:
the expression of the anomaly impact coefficient is:
wherein:representing the +.>Abnormal influence coefficients of the first IMF component and the ith other parameter data; />Indicate->The first IMF component is +.>Standard deviation of the target data segment in the ith other parameter data corresponding to the first abnormal data segment; />Representing the +.>The total number of first abnormal data segments in the first IMF component;；/>representing the +.>The first IMF component is +.>Curve fluctuation characteristic values of the first abnormal data segments; />Representing the +.>The first IMF component is +.>Curve fluctuation characteristic values of adjacent data segments of the first abnormal data segment; />Representing normalization.
4. The method of claim 1, wherein the step of obtaining a matching data segment of each other parameter data that matches each first anomalous data segment of each first IMF component of the target parameter data comprises:
Performing EMD (empirical mode decomposition) on each piece of other parameter data to obtain a plurality of second IMF components, and obtaining a second abnormal data segment in each second IMF component;
respectively acquiring the mean time of the first abnormal data segment and the mean time of the second abnormal data segment by utilizing the moment of each data segment in each first abnormal data segment in each first IMF component and the moment of each data segment in each second abnormal data segment in the second IMF component;
and comparing the mean time of each first abnormal data segment with the mean time of each second abnormal data segment to obtain a second abnormal data segment matched with each first abnormal data segment, and taking the matched second abnormal data segment as a matched data segment.
5. The method for monitoring the state in the polystyrene production process according to claim 1 or 4, wherein the method for acquiring the correlation of each first abnormal data segment with the matched data segment matched with each other parameter data comprises:
acquiring a first curve of each first abnormal constant section and a second curve of a matched data section matched in each piece of other parameter data;
the centers of gravity of the first curve and the second curve are obtained, and the first curve or the second curve is moved to enable the centers of gravity of the first curve and the second curve to coincide;
Acquiring the interval area of the first curve and the second curve after the centers of gravity are overlapped;
and acquiring the relevance of each first abnormal data segment and the matched data segment matched in each other parameter data by using the interval area of the first curve and the second curve.
6. The method for monitoring conditions in a polystyrene production process according to claim 5, wherein said correlation is expressed as:
wherein:representing the +.>The first IMF component is +.>The first abnormal data segment is matched with the ith part of the other parameter data in the ith part of the other parameter data>The relevance of the individual matching data segments; />Representing the +.>The first IMF component is +.>First difference ofFirst data in constant data section and the (i) th ∈th matching in the other parameter data of the (i) th item>A time span of the first data in the matching data segment; />Representing the +.>The first IMF component is +.>The first abnormal data segment is matched with the ith part of the other parameter data in the ith part of the other parameter data>The time spans of the matching data segments; />Representing the +.>The first IMF component is +.>First curve corresponding to first abnormal data segment and the (i) th matched parameter data >The interval area of the second curve corresponding to each matching data segment.
7. A state monitoring system used in the polystyrene production process; characterized by comprising the following steps:
and a data acquisition module: a plurality of pieces of parameter data for use in a polystyrene production process; selecting any piece of parameter data as target parameter data;
EMD decomposition module: EMD is carried out on the target parameter data and other parameter data to obtain a plurality of first IMF components and second IMF components;
abnormal influence coefficient acquisition module: acquiring a first abnormal data segment in each first IMF component, and obtaining a curve fluctuation characteristic value of the first abnormal data segment by utilizing the slope of each data in each first abnormal data segment and the average slope of all data in the first abnormal data segment;
the method comprises the steps that the adjacent data segment of the next segment of a first abnormal data segment in each first IMF component, the length of the adjacent data segment is consistent with that of the first abnormal data segment, and the curve fluctuation characteristic value of the adjacent data segment of the first abnormal data segment is obtained by utilizing a method for obtaining the curve fluctuation characteristic value of the first abnormal data segment;
acquiring a target data segment in each piece of other parameter data, wherein the target data segment and the first abnormal data segment belong to the data segment in the same time period, and acquiring an abnormal influence coefficient of each first IMF component in the target parameter data and each other parameter data by using a curve fluctuation characteristic value of each first abnormal data segment in each first IMF component, a curve fluctuation characteristic value of an adjacent data segment of the first abnormal data segment and a standard deviation of the target data segment in each piece of other parameter data corresponding to the first abnormal data segment;
And the relevance calculating module is used for: acquiring a matching data segment matched with each first abnormal data segment in each first IMF component in each other parameter data;
acquiring the relevance of each first abnormal data segment and the matched data segment matched in each other parameter data by using the curve of each first abnormal data segment in each first IMF component in the target data and the curve of the matched data segment matched in each other parameter data;
the embodiment calculating module is used for: obtaining the degree of embodying the influence of each piece of other parameter data on each piece of first IMF component in the target parameter data according to the abnormal influence coefficient of each piece of first IMF component and each piece of other parameter data and the relevance of each piece of first abnormal data segment in each piece of first IMF component and the matched parameter data segment in each piece of other parameter data;
the method for obtaining the degree of the influence of each piece of other parameter data on each piece of first IMF component in the target parameter data comprises the following steps:
wherein:indicating the +.f. of all other parameter data to the target parameter data>Degree of embodiment of the influence of the first IMF component;representing the +.>Abnormal influence coefficients of the first IMF component and the ith other parameter data; / >Representing the +.>The first IMF component is +.>The first abnormal data segment is matched with the ith part of the other parameter data in the ith part of the other parameter data>The relevance of the individual matching data segments; />Representing the amount of other parameter data; />Representing the +.>The total number of first abnormal data segments in the first IMF component; />Characterization of->The other parameter data are +.>First IMF component +.>The associative sums in the abnormal segments;
the abnormality judgment module: and acquiring data segments corresponding to the first IMF components with the largest degree of embodiment of all the first IMF components in the target parameter data as required data segments, and judging whether the abnormality of the required data segments is influenced by other parameter data.
CN202310611845.3A 2023-05-29 2023-05-29 State monitoring method and system for polystyrene production process Active CN116341993B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202310611845.3A CN116341993B (en) 2023-05-29 2023-05-29 State monitoring method and system for polystyrene production process

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202310611845.3A CN116341993B (en) 2023-05-29 2023-05-29 State monitoring method and system for polystyrene production process

Publications (2)

Publication Number Publication Date
CN116341993A CN116341993A (en) 2023-06-27
CN116341993B true CN116341993B (en) 2023-07-25

Family

ID=86893330

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202310611845.3A Active CN116341993B (en) 2023-05-29 2023-05-29 State monitoring method and system for polystyrene production process

Country Status (1)

Country Link
CN (1) CN116341993B (en)

Citations (8)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN105928701A (en) * 2016-04-29 2016-09-07 石家庄铁道大学 Valid IMF determining method in EMD process on the basis of correlation analysis
JP2018120407A (en) * 2017-01-25 2018-08-02 Ntn株式会社 State monitoring method and state monitoring apparatus
CN110226140A (en) * 2017-01-25 2019-09-10 Ntn株式会社 State monitoring method and state monitoring apparatus
CN111324639A (en) * 2020-02-11 2020-06-23 京东数字科技控股有限公司 Data monitoring method and device and computer readable storage medium
CN213563746U (en) * 2020-11-04 2021-06-29 福建天原化工有限公司 A quick cooling device for producing transparent polystyrene
CN113065498A (en) * 2021-04-13 2021-07-02 杭州哲达科技股份有限公司 Abnormal data detection method based on improved EMD and neural network model
WO2022252505A1 (en) * 2021-06-02 2022-12-08 杭州安脉盛智能技术有限公司 Device state monitoring method based on multi-index cluster analysis
CN115659160A (en) * 2022-12-28 2023-01-31 北京中航路通科技有限公司 Data quality measurement method for optimizing digital twin model

Family Cites Families (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
EP2624091B1 (en) * 2012-02-06 2014-07-30 C.R.F. Società Consortile per Azioni A method for monitoring the quality of industrial processes and system therefrom
EP3418831B1 (en) * 2017-06-19 2023-08-16 C.R.F. Società Consortile per Azioni A method for performing a noise removal operation on a signal acquired by a sensor and system therefrom

Patent Citations (8)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN105928701A (en) * 2016-04-29 2016-09-07 石家庄铁道大学 Valid IMF determining method in EMD process on the basis of correlation analysis
JP2018120407A (en) * 2017-01-25 2018-08-02 Ntn株式会社 State monitoring method and state monitoring apparatus
CN110226140A (en) * 2017-01-25 2019-09-10 Ntn株式会社 State monitoring method and state monitoring apparatus
CN111324639A (en) * 2020-02-11 2020-06-23 京东数字科技控股有限公司 Data monitoring method and device and computer readable storage medium
CN213563746U (en) * 2020-11-04 2021-06-29 福建天原化工有限公司 A quick cooling device for producing transparent polystyrene
CN113065498A (en) * 2021-04-13 2021-07-02 杭州哲达科技股份有限公司 Abnormal data detection method based on improved EMD and neural network model
WO2022252505A1 (en) * 2021-06-02 2022-12-08 杭州安脉盛智能技术有限公司 Device state monitoring method based on multi-index cluster analysis
CN115659160A (en) * 2022-12-28 2023-01-31 北京中航路通科技有限公司 Data quality measurement method for optimizing digital twin model

Non-Patent Citations (2)

* Cited by examiner, † Cited by third party
Title
一种改进的多模态过程故障检测方法;杨青 等;沈阳理工大学学报;第36卷(第3期);48-53 *
基于高斯模型的工业过程数据的故障预测;杨为惠 等;太原理工大学学报;第1卷(第49期);86-93 *

Also Published As

Publication number Publication date
CN116341993A (en) 2023-06-27

Similar Documents

Publication Publication Date Title
CN105975443B (en) Network anomalous behaviors detection method and system based on Lasso
CN110895526A (en) Method for correcting data abnormity in atmosphere monitoring system
CN117290802B (en) Host power supply operation monitoring method based on data processing
CN116400126B (en) Low-voltage power box with data processing system
CN114167838B (en) Multi-scale health assessment and fault prediction method for servo system
CN113762329A (en) Method and system for constructing state prediction model of large rolling mill
CN117708748B (en) Operation monitoring system and method for centrifugal fan
CN116702093B (en) Marine target positioning method based on big data fusion
CN115860579B (en) Production quality monitoring system for flour processing
CN111913443A (en) Industrial equipment fault early warning method based on similarity
CN117972614B (en) Real-time temperature monitoring method in charging and discharging process of mobile power supply
CN117439827B (en) Network flow big data analysis method
CN117034178B (en) Online monitoring system for detergent production equipment
CN118168652B (en) Cable vibration signal time-frequency analysis system
CN106600076B (en) Monitoring data analysis and early warning method for turret type RTO waste gas treatment equipment
CN117289778B (en) Real-time monitoring method for health state of industrial control host power supply
CN117971625B (en) Performance data intelligent monitoring system based on computer cloud platform
CN114429238A (en) Wind turbine generator fault early warning method based on space-time feature extraction
CN116341993B (en) State monitoring method and system for polystyrene production process
CN117556364A (en) Mining ore pressure safety intelligent monitoring system
CN117872900A (en) Non-woven fabric production equipment remote monitoring system based on internet of things technology
CN116523321A (en) Intelligent helmet security control system based on cloud platform
CN113408383A (en) Audible noise invalid data judgment method based on bounded beta (g, h) distribution and MWKPCA
CN118442125B (en) Coal mine safety monitoring and early warning system based on Internet of things
CN117556274B (en) Temperature data anomaly analysis method for heat pipe backboard

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