CN116823125A - White spirit storage abnormality early warning system based on multisource data - Google Patents
White spirit storage abnormality early warning system based on multisource data Download PDFInfo
- Publication number
- CN116823125A CN116823125A CN202311112658.7A CN202311112658A CN116823125A CN 116823125 A CN116823125 A CN 116823125A CN 202311112658 A CN202311112658 A CN 202311112658A CN 116823125 A CN116823125 A CN 116823125A
- Authority
- CN
- China
- Prior art keywords
- data
- characteristic
- suspected
- line
- characteristic 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.)
- Granted
Links
- 230000005856 abnormality Effects 0.000 title claims abstract description 43
- 230000007613 environmental effect Effects 0.000 claims abstract description 39
- 239000011159 matrix material Substances 0.000 claims abstract description 34
- 230000002159 abnormal effect Effects 0.000 claims abstract description 24
- 238000012544 monitoring process Methods 0.000 claims abstract description 14
- 238000001514 detection method Methods 0.000 claims abstract description 10
- 238000012545 processing Methods 0.000 claims abstract description 9
- 238000000034 method Methods 0.000 claims description 47
- 230000008569 process Effects 0.000 claims description 28
- 235000020097 white wine Nutrition 0.000 claims description 7
- 238000010606 normalization Methods 0.000 claims description 3
- 238000012937 correction Methods 0.000 claims description 2
- 230000000694 effects Effects 0.000 description 3
- 230000009286 beneficial effect Effects 0.000 description 2
- 230000008859 change Effects 0.000 description 2
- 230000002547 anomalous effect Effects 0.000 description 1
- 238000012512 characterization method Methods 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
- 238000005530 etching Methods 0.000 description 1
- 238000011156 evaluation Methods 0.000 description 1
- 238000005286 illumination Methods 0.000 description 1
- 244000005700 microbiome Species 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 230000000750 progressive effect Effects 0.000 description 1
- 230000001105 regulatory effect Effects 0.000 description 1
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
- G06Q10/08—Logistics, e.g. warehousing, loading or distribution; Inventory or stock management
- G06Q10/087—Inventory or stock management, e.g. order filling, procurement or balancing against orders
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F17/00—Digital computing or data processing equipment or methods, specially adapted for specific functions
- G06F17/10—Complex mathematical operations
- G06F17/18—Complex mathematical operations for evaluating statistical data, e.g. average values, frequency distributions, probability functions, regression analysis
-
- 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
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/25—Fusion techniques
- G06F18/251—Fusion techniques of input or preprocessed data
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02P—CLIMATE CHANGE MITIGATION TECHNOLOGIES IN THE PRODUCTION OR PROCESSING OF GOODS
- Y02P90/00—Enabling technologies with a potential contribution to greenhouse gas [GHG] emissions mitigation
- Y02P90/30—Computing systems specially adapted for manufacturing
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- Data Mining & Analysis (AREA)
- Theoretical Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Business, Economics & Management (AREA)
- General Engineering & Computer Science (AREA)
- Evolutionary Biology (AREA)
- Bioinformatics & Computational Biology (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Life Sciences & Earth Sciences (AREA)
- Operations Research (AREA)
- Mathematical Analysis (AREA)
- Evolutionary Computation (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Artificial Intelligence (AREA)
- Mathematical Physics (AREA)
- Pure & Applied Mathematics (AREA)
- Mathematical Optimization (AREA)
- Economics (AREA)
- Computational Mathematics (AREA)
- Entrepreneurship & Innovation (AREA)
- Marketing (AREA)
- Development Economics (AREA)
- Tourism & Hospitality (AREA)
- Probability & Statistics with Applications (AREA)
- Human Resources & Organizations (AREA)
- Algebra (AREA)
- Finance (AREA)
- Databases & Information Systems (AREA)
- Software Systems (AREA)
- Accounting & Taxation (AREA)
- Quality & Reliability (AREA)
- General Business, Economics & Management (AREA)
- Strategic Management (AREA)
- Alarm Systems (AREA)
- Complex Calculations (AREA)
Abstract
The application relates to the field of data processing, in particular to a white spirit storage abnormity early warning system based on multi-source data, which comprises the following steps: collecting an environmental characteristic data sequence of white spirit storage; obtaining the close constraint degree of each suspected point according to the occurrence frequency and the position distribution of the suspected point in each line of characteristic data; obtaining sets of characteristic data of each row according to the close constraint degree among the suspected points, and calculating the close combination coefficient of each set; obtaining characteristic weight coefficients according to the close combination coefficients of each set of each row of characteristic data; using an SOS anomaly detection algorithm to take the weighted characteristic parameters as an input matrix, and outputting to obtain outlier probability of each data point; judging whether each data point is an abnormal data point according to the outlier probability and the outlier threshold value of each data point, and finishing the early warning of the white spirit storage abnormality. The data monitoring precision is improved, dimension disasters are avoided, and errors of the white spirit storage abnormity early warning result are reduced.
Description
Technical Field
The application relates to the field of data processing, in particular to a white spirit storage abnormity early warning system based on multi-source data.
Background
The environment such as temperature, humidity and the like needs to be controlled in the optimal range in the process of storing the white spirit, but the storage environment is continuously changed, if the white spirit is not found in time, various environmental factors are caused to exceed the optimal range, the quality of the white spirit is damaged, and the quality of the white spirit is influenced. Therefore, it is necessary to provide a sensor system in an unsupervised warehouse, observe data information generated by the sensor system, identify an abnormal state in the data information, and perform early warning on the abnormal state.
Based on an anomaly detection algorithm SOS based on statistics, the traditional early warning system takes each data of environmental monitoring in the white spirit storage process as an input matrix to finish evaluation of abnormal data points. However, the method has higher algorithm complexity when the high-dimensional data is to be monitored, and causes dimension disasters; meanwhile, when the multi-line data in the environment are calculated, the influence degree of each line of data on the white spirit storage abnormality is not combined, so that the abnormality monitoring and early warning result error is larger.
In summary, the application provides a white spirit storage abnormality pre-warning system based on multi-source data, which adopts a sensor system to collect the environmental characteristic data of white spirit storage, combines the distribution characteristics of suspected points in each environmental characteristic data sequence and the characteristics of characteristic data, and performs SOS abnormality detection to complete white spirit storage abnormality pre-warning.
Disclosure of Invention
In order to solve the technical problems, the application provides a white spirit storage abnormality early warning system based on multi-source data, which comprises:
and a data acquisition module: collecting an environmental characteristic data sequence in the white spirit storage process according to a sensor system;
and a data processing module: forming an environmental feature matrix by each row of feature data sequence; respectively deriving a function for each line of characteristic data of the environment characteristic matrix to obtain an environment characteristic slope matrix, and counting suspected points in the environment characteristic slope matrix; obtaining the close constraint degree of each suspected point according to the occurrence frequency and the position distribution of each suspected point of the characteristic data; obtaining each set of each line of characteristic data according to the close constraint degree among each suspected point of each line of characteristic data; obtaining span coefficients of each set according to the position information of the data points in each set of each row of characteristic data; obtaining the close combination coefficient of each set according to each close constraint degree in each set of each line of characteristic data and the span coefficient of each set;
obtaining characteristic influence coefficients of each row of characteristic data according to the close combination coefficients of each set of each row of characteristic data; obtaining a characteristic weight coefficient according to the characteristic influence coefficient of each row of characteristic data;
an anomaly monitoring module: obtaining a weighted characteristic parameter of each data point according to the characteristic weight coefficient of each row of characteristic data; using an SOS anomaly detection algorithm to take the weighted characteristic parameters as an input matrix, and outputting to obtain outlier probability of each data point; and obtaining different data points according to the outlier probability and the outlier threshold of each data point, and combining the different data points to finish early warning of the abnormal storage of the white wine.
Preferably, the specific steps of counting suspected points in the environmental characteristic slope matrix are as follows:
the data point corresponding to the environmental characteristic matrix with the value of 0 in the environmental characteristic slope matrix is marked as a first suspected point;
the data point corresponding to the environmental characteristic matrix with the highest value in the environmental characteristic slope matrix is marked as a second suspected point;
the first and second suspected points are noted as suspected points.
Preferably, the expression for obtaining the close constraint degree of each suspected point according to the occurrence frequency and the position distribution of each suspected point of the feature data is as follows:
in the method, in the process of the application,、/>respectively +.>The suspected point is at->Horizontal coordinate, vertical coordinate, and/or +.>、Respectively +.>The suspected point is at->Horizontal coordinate, vertical coordinate, and/or +.>Is->Frequency of occurrence of each suspected point in each line of characteristic data, +.>Is->Frequency of occurrence of each suspected point in each line of characteristic data, +.>Is->Line characteristic data>The first part of the suspected points and the rest of the suspected points>Close constraints of the individual suspected points.
Preferably, the specific steps of obtaining each set of each line of feature data according to the degree of close constraint among each suspected point of each line of feature data are as follows:
in each line of feature data, for any suspected point, the suspected point with the maximum close constraint degree obtained by calculating the rest suspected points is formed into a small set;
and summing all small sets with intersections to obtain each set of the characteristic data of each row.
Preferably, the specific step of obtaining the span coefficient of each set according to the position information of the data points in each set of each line of characteristic data is as follows:
in the method, in the process of the application,is->Line characteristic data>The left-most suspected point in the set is on the abscissa in the spatial distribution of the feature data, +.>Is->Line characteristic data>The right-most suspected point in the set is on the abscissa in the spatial distribution of the feature data, +.>Is->Line characteristic data>Maximum longitudinal span of the sets in the longitudinal direction, < > about->Is->Line characteristic data>Maximum longitudinal span distance in longitudinal axis direction between the nearest two peaks and valleys on the left and right sides of each set, +.>Is->Line characteristic data>Span coefficients of the sets.
Preferably, the expression for obtaining the close combination coefficient of each set according to each close constraint degree in each set of each line of characteristic data and the span coefficient of each set is:
in the method, in the process of the application,is->Line characteristic data>Span coefficient of individual sets, +.>Is->Line characteristic data>Number of suspected points of the collection, +.>Is->Line characteristic data>First->Degree of close constraint->Is->Line characteristic data>Closely combining coefficients of the sets.
Preferably, the expression for obtaining the characteristic influence coefficient of each line of characteristic data according to the close combination coefficient of each set of each line of characteristic data is as follows:
in the method, in the process of the application,for correction factor +.>For normalization function->Is->The number of sets in the line profile, +.>Is->Line characteristic data>Closely-combined coefficients of the individual sets,/->Is->Line characteristic data>Number of suspected points of the collection, +.>Is->The characteristic influence coefficient of the line characteristic data.
Preferably, the specific steps of obtaining the characteristic weight coefficient according to the characteristic influence coefficient of each line of characteristic data are as follows:
summing the characteristic influence coefficients of each row of characteristic data to obtain denominators;
and (3) comparing the characteristic influence coefficient of each line of characteristic data with a denominator to obtain a characteristic weight coefficient of the corresponding line of characteristic data.
Preferably, the specific step of obtaining the weighted characteristic parameter of each data point according to the characteristic weight coefficient of each row of characteristic data is as follows:
and carrying out weighted summation on each data point corresponding to each column in the environment feature matrix by using the feature weight coefficient of each row of feature data to obtain weighted feature parameters corresponding to each column of time points.
Preferably, the specific steps of obtaining each different data point according to the outlier probability and the outlier threshold of each data point are as follows:
and acquiring the outlier probability of each data point, setting an outlier threshold, and taking the data point with the outlier probability larger than the outlier threshold as an abnormal data point.
The application has at least the following beneficial effects:
compared with the traditional SOS anomaly detection algorithm, the method provided by the application has the advantages that the suspected points are calculated according to the environmental parameters obtained by white wine storage, and the different distribution of the suspected points in each row of environmental parameter data sequences is combined, so that the monitoring of the anomaly data interval in the environmental characteristic data sequence in the white wine storage process is facilitated, and the data monitoring precision is improved;
meanwhile, the characteristic influence coefficient of each row of data is obtained by combining the influence degree of each row of characteristic data sequence on the white spirit storage abnormality, and the high-dimensional data can be effectively fused into a group of data under different weights, so that dimension disasters are avoided, the time complexity of an algorithm is reduced, and the error of a white spirit storage abnormality early warning result is reduced.
Drawings
In order to more clearly illustrate the embodiments of the application or the technical solutions and advantages of the prior art, the following description will briefly explain the drawings used in the embodiments or the description of the prior art, and it is obvious that the drawings in the following description are only some embodiments of the application, and other drawings can be obtained according to the drawings without inventive effort for a person skilled in the art.
Fig. 1 is a flow chart of a white spirit storage abnormality early warning system based on multi-source data.
Detailed Description
In order to further explain the technical means and effects adopted by the application to achieve the preset aim, the following detailed description is given below of the multi-source data-based white spirit storage abnormality early warning system according to the application, which is based on the specific implementation, structure, characteristics and effects thereof, with reference to the accompanying drawings and the preferred embodiment. In the following description, different "one embodiment" or "another embodiment" means that the embodiments are not necessarily the same. Furthermore, the particular features, structures, or characteristics of one or more embodiments may be combined in any suitable manner.
Unless defined otherwise, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this application belongs.
The specific scheme of the white spirit storage abnormality early warning system based on the multi-source data provided by the application is specifically described below with reference to the accompanying drawings.
The application provides a white spirit storage abnormality early warning system based on multi-source data, which comprises a data acquisition module, a data processing module and an abnormality monitoring module.
Specifically, the internet of things engineering information acquisition processing system of the present embodiment provides a multi-source data-based white spirit storage abnormality early warning system, referring to fig. 1, which includes:
and the data acquisition module is used for carrying out abnormal monitoring and early warning on the white spirit warehouse mainly through a data processing technology. In general, in the process of storing white spirit, environmental factors which have a great influence on the quality of white spirit include temperature, humidity, illumination, air quality and the like. The quality of the white spirit can be influenced by the excessively high temperature and the excessively low temperature, and if the humidity exceeds a specified range, conditions are provided for the propagation of microorganisms, so that the quality of the white spirit is deteriorated, and if the air quality is poor, the quality of the white spirit can be indirectly influenced.
The embodiment of the application adopts the temperature, the humidity and the likeThe sensor system composed of the group sensor is used for carrying out abnormal monitoring on the environmental factors in the white spirit storage process to obtain the +.f. for possibly influencing the white spirit quality in the white spirit storage process>A row context feature data sequence. Wherein, the characteristic data of each group are collected in the same time>Data, the time interval between each data is +.>。
The sensor system can be used for collecting an environment characteristic data sequence of the white spirit warehouse and analyzing characteristic data abnormality in the white spirit warehouse process.
Data processing modules, in commonThe environmental characteristic data sequence of the white spirit warehouse is shown in the row, and each row is provided with +.>Data points, thereby constructing and obtaining an environment characteristic matrix +.>。
In the method, in the process of the application,is->The +.>Data points.
Matrix of environmental characteristicsSolving a first order derivative function of each line of data in the array to obtain an environment characteristic slope matrix +.>。
Because the regulating equipment of the environmental characteristic data such as the white spirit storage temperature, the humidity and the like is abnormal, the white spirit storage environment is abnormal, the environmental characteristic data obtained by the sensor system at the moment can have larger difference compared with the surrounding moment, and the slope matrix of the environmental characteristic data is counted at the momentCorresponding timeThe value of the etching should be 0; meanwhile, the difference between the numerical values before and after the data points of abnormal change of the sensor is obvious, and at the moment, the slope matrix of the characteristic of the statistical environment is +.>The value at the corresponding time should be the largest.
Thus, the slope matrix of the environmental characteristic is countedThe maximum value and zero point appearing in each data parameter of the data and recording these points as the suspected points of the environmental characteristic data, thereby obtaining +.>Suspected points (I)>. This->The suspected points are->The data points in the row environment feature data where anomalies are most likely to occur.
Counting the occurrence frequency of each suspected point，/>. First->The frequency of occurrence of each suspected dot is +.>I.e. as a weight coefficient for the suspected point.
Since these suspected points are scattered data points, the degree of abnormality of the suspected points cannot be fully characterized. If an anomaly occurs at a certain point in time, there is a greater likelihood that the data points near that point in time will also be anomalous.
For this case, according to the environment feature matrixThe%>Line environmental data parameters, calculate +.>+.>The suspected points and the more closely related suspected points form a set according to the +.>+.>Spatial coordinates of the respective suspected points and the suspected points->At->The frequency of occurrence in the line characteristic data calculates the degree of affinity constraint +.>Calculate +.>Suspected dot and suspected dot->Degree of close constraint between->And taking the suspected point where the maximum value is located to form a set. Wherein->First->The first part of the suspected points and the rest of the suspected points>The expression of the close constraint degree of each suspected point is as follows:
in the method, in the process of the application,、/>respectively +.>The suspected point is at->Horizontal coordinate, vertical coordinate, and/or +.>、Respectively +.>The suspected point is at->Horizontal coordinate, vertical coordinate, and/or +.>Is->Frequency of occurrence of each suspected point in each line of characteristic data, +.>Is->Frequency of occurrence of each suspected point in each line of characteristic data, +.>Is->Line characteristic data>The first part of the suspected points and the rest of the suspected points>Close constraints of the individual suspected points.
In the first placeIn the line feature data, if the closer the spatial distribution distance between two suspected points is, and the higher the frequency of the suspected points at the current moment appearing in different feature data is, the calculated +.>Line characteristic data>The first part of the suspected points and the rest of the suspected points>The greater the value of the degree of closeness constraint for each suspected point. The two suspected points are assembled into a set. />The larger the set, the more abnormal the set of the two suspected points.
And repeating the steps to obtain the close constraint degree of each suspected point in each line of characteristic data.
For a pair ofThe suspected points are used for obtaining the sum of the suspected points and the residual->The suspected points with the greatest close constraint degree among the suspected points form a set to obtain +.>A set of sets. For this->And (3) collecting intersection sets of each set and the rest sets, merging the two sets by solving the sum if the two sets are not empty, and not calculating if the two sets are empty.
In the method, in the process of the application,、/>is->First->、/>A set of suspected points.
If the first isCollections and->If there are intersections in the sets, the two sets are combined to obtain a new set. Thereby obtaining a relatively close-spaced and provided withA set of a plurality of suspected points associated with a degree of abnormality.
Repeating the above steps until the condition that the intersection of any two sets is not empty can not be found, stopping merging the sets, so that the characteristic data of each line can be obtainedA set of sets.
Because of the different distribution of the suspected points in each set, the set may be more abnormal if the spatial span of the suspected points in the set is larger on the abscissa and the ordinate.
In this case, the span coefficients of the respective sets are obtained from the position information of the data points in the respective sets of the feature data of each line. And calculating the difference value of the vertical coordinates of the two suspected points with the largest difference in the longitudinal axis direction in the set to obtain the longitudinal span of the set by calculating the transverse coordinate difference value of the two suspected points at the left and right sides in the set. However, the longitudinal span is limited in comparison with the transverse span, and the different change ranges of the original data are limited, and the maximum span distance of the set in the longitudinal axis direction is obtained by taking the peaks and the troughs adjacent to the left side and the right side of the set and obtaining the maximum difference value of all data points in the range in the longitudinal axis direction, so that the span coefficient of the set can be calculated.
In the method, in the process of the application,is->Line characteristic data>The left-most suspected point in the set is on the abscissa in the spatial distribution of the feature data, +.>Is->Line characteristic data>The right-most suspected point in the set is on the abscissa in the spatial distribution of the feature data, +.>Is->Line characteristic data>Maximum longitudinal span of the sets in the longitudinal direction, < > about->Is->Line characteristic data>Maximum longitudinal span distance in longitudinal axis direction between the nearest two peaks and valleys on the left and right sides of each set, +.>Is->Line characteristic data>Span coefficients of the sets.
It should be noted that if the transverse span of the setThe larger the cross distribution range of the suspected points in the set is, the larger the cross distribution range is, namely the more abnormal the set is; by calculating the longitudinal span of the set and the maximum longitudinal span around the setRatio between spans->And obtaining the longitudinal difference degree of the longitudinal distribution range of the suspected points in the set in comparison with the normal value range of the line of characteristic data. Combining the lateral span with the degree of longitudinal variation to obtain a span coefficient characterizing the set, < >>The larger the set, the more abnormal the set is with respect to the line of feature data.
To characterize the anomalies of the suspected points contained in each set in each row of feature data, the degree of anomalies of each set needs to be calculated in combination with the suspected points in that set.
For the first+.>The suspected points are->Sets, each set having +>The suspected points are based on +.>The close constraint degree of each suspected point is combined with the span coefficient of the set to obtain the +.>Closely combined coefficients of individual sets。
In the method, in the process of the application,is->Line characteristic data>Span coefficient of individual sets, +.>Is->Line characteristic data>Number of suspected points of the collection, +.>Is->Line characteristic data>First->Degree of close constraint->Is->Line characteristic data>Closely combining coefficients of the sets.
Note that, the firstLine characteristic data>The greater the sum of the close constraints of the respective suspicious points of the collection, the +.>The probability of abnormal data in the sets is high; at the same time if%>The larger the span coefficient in the set, the larger the abnormal data interval in the set, and the abnormal point to be considered in the interval is also abnormal.
Repeating the above steps to obtain the close combination coefficient of each set in each line of characteristic data。
Because the close combination coefficients of the sets of the characteristic data of each row are different, when the influence of the characteristic data of each row on the white wine storage abnormality is to be represented, the influence degree of the characteristic data of each row needs to be evaluated by integrating the information of each set.
For this case, calculate the firstThe influence degree of the close combination coefficient of each set in the line characteristic data on the number of suspected points of each set is obtained, and the average value of the number of the suspected points corrected by each set of each line characteristic data is obtained to obtain the characterization +.>Characteristic influence coefficient of line characteristic data +.>。
In the method, in the process of the application,to correct coefficient0.5，/>For normalization function->Is->The number of sets in the line profile, +.>Is->Line characteristic data>Closely-combined coefficients of the individual sets,/->Is->Line characteristic data>Number of suspected points of the collection, +.>Is->The characteristic influence coefficient of the line characteristic data.
Note that, the firstThe greater the close combination coefficient of the set in the line characteristic data, the amplifying effect is presented to the number of suspected points in the set when the normalized value is greater than 0.5, and the +.>The greater the degree of abnormality of the line characteristic data, byEach set in the line of characteristic data is averaged to obtain +.>And the characteristic influence coefficient of the row characteristic data on the white spirit storage abnormality.
Repeating the steps to obtain the characteristic influence coefficient of each line of characteristic data。
For this purposeNormalizing the characteristic influence coefficients of the line characteristic data to obtain characteristic weight coefficients of each line characteristic data respectively>。
In the method, in the process of the application,is->Characteristic influence coefficient of line characteristic data, +.>For the number of lines of the characteristic data>Is->Characteristic weight coefficients of the line characteristic data.
Wherein,,the larger, the description of->Line characteristic data in->The weight in the line characteristic data is larger, i.e.>The influence of the line characteristic data on the detection of the white spirit storage abnormality is large.
So far, the characteristic weight coefficient of each row of environment characteristic data in the white spirit storage process is obtained and is used for realizing the early warning of the abnormal white spirit storage.
The abnormality monitoring module is used for monitoring the abnormality according to the environmental characteristic matrixObtaining the influence weight of each row of characteristic data on the storage abnormality, thereby obtaining +.>+.>Characteristic parameters weighted by individual characteristic parameters +.>Wherein->Data points are +.>。
In the method, in the process of the application,、/>、/>1 st, 2 nd,/-th, respectively>Characteristic weight coefficient of line characteristic data, +.>、/>、/>1 st, 2 nd,/-th, respectively>Line characteristic data>Values of individual data points>For the +.>Values for the individual data points.
Thereby obtaining the dimension-reduced productInput matrix of size->I.e. the input matrix->There is->Data points, each data point having a characteristic parameter +.>Is a two-dimensional feature space distribution of (a).
Storing Chinese liquorInput matrix obtained from environmental feature data in a computerInput to SOS abnormality detection algorithm, output +.>Outlier probability for each of the data points>Setting an outlier threshold, and setting the outlier probability +.>And taking the data points larger than the outlier threshold as abnormal data points, and carrying out early warning on the abnormal data points. Wherein the outlier threshold is set to 0.5.
So far, can realize the unusual early warning of white spirit storage according to this embodiment.
In summary, the embodiment of the application provides a white spirit storage abnormality pre-warning system based on multi-source data, which adopts a sensor system to collect the environmental characteristic data of white spirit storage, combines the distribution characteristics of suspected points in each environmental characteristic data sequence and the characteristics of characteristic data, and performs SOS abnormality detection to complete white spirit storage abnormality pre-warning.
Compared with the traditional SOS anomaly detection algorithm, the method provided by the embodiment of the application obtains each environmental parameter according to the white wine storage to calculate the suspected points, and is beneficial to monitoring the anomaly data interval in the environmental characteristic data sequence in the white wine storage process by combining different distribution of each suspected point in each row of the environmental parameter data sequence, so that the data monitoring precision is improved;
meanwhile, the characteristic influence coefficient of each row of data is obtained by combining the influence degree of each row of characteristic data sequence on the white spirit storage abnormality, and the high-dimensional data can be effectively fused into a group of data under different weights, so that dimension disasters are avoided, the time complexity of an algorithm is reduced, and the error of a white spirit storage abnormality early warning result is reduced.
It should be noted that: the sequence of the embodiments of the present application is only for description, and does not represent the advantages and disadvantages of the embodiments. And the foregoing description has been directed to specific embodiments of this specification. In addition, the processes depicted in the accompanying figures do not necessarily require the particular order shown, or sequential order, to achieve desirable results. In some embodiments, multitasking and parallel processing are also possible or may be advantageous.
In this specification, each embodiment is described in a progressive manner, and identical and similar parts of each embodiment are all referred to each other, and each embodiment mainly describes differences from other embodiments.
The foregoing description of the preferred embodiments of the present application is not intended to be limiting, but rather, any modifications, equivalents, improvements, etc. that fall within the principles of the present application are intended to be included within the scope of the present application.
Claims (10)
1. White spirit storage abnormality early warning system based on multisource data, which is characterized in that the system includes:
and a data acquisition module: collecting an environmental characteristic data sequence in the white spirit storage process according to a sensor system;
and a data processing module: forming an environmental feature matrix by each row of feature data sequence; respectively deriving a function for each line of characteristic data of the environment characteristic matrix to obtain an environment characteristic slope matrix, and counting suspected points in the environment characteristic slope matrix; obtaining the close constraint degree of each suspected point according to the occurrence frequency and the position distribution of each suspected point of the characteristic data; obtaining each set of each line of characteristic data according to the close constraint degree among each suspected point of each line of characteristic data; obtaining span coefficients of each set according to the position information of the data points in each set of each row of characteristic data; obtaining the close combination coefficient of each set according to each close constraint degree in each set of each line of characteristic data and the span coefficient of each set;
obtaining characteristic influence coefficients of each row of characteristic data according to the close combination coefficients of each set of each row of characteristic data; obtaining a characteristic weight coefficient according to the characteristic influence coefficient of each row of characteristic data;
an anomaly monitoring module: obtaining a weighted characteristic parameter of each data point according to the characteristic weight coefficient of each row of characteristic data; using an SOS anomaly detection algorithm to take the weighted characteristic parameters as an input matrix, and outputting to obtain outlier probability of each data point; and obtaining different data points according to the outlier probability and the outlier threshold of each data point, and combining the different data points to finish early warning of the abnormal storage of the white wine.
2. The multi-source data-based white spirit storage abnormality early warning system according to claim 1, wherein the specific steps of counting suspected points in the environmental characteristic slope matrix are as follows:
the data point corresponding to the environmental characteristic matrix with the value of 0 in the environmental characteristic slope matrix is marked as a first suspected point;
the data point corresponding to the environmental characteristic matrix with the highest value in the environmental characteristic slope matrix is marked as a second suspected point;
the first and second suspected points are noted as suspected points.
3. The multi-source data-based white spirit storage abnormality early warning system according to claim 1, wherein the expression for obtaining the degree of close constraint of each suspected point according to the frequency and the position distribution of the suspected point of each line of characteristic data is:
in the method, in the process of the application,、/>respectively +.>The suspected point is at->Horizontal coordinate, vertical coordinate, and/or +.>、/>Respectively +.>The suspected point is at->Horizontal coordinate, vertical coordinate, and/or +.>Is->Frequency of occurrence of each suspected point in each line of characteristic data, +.>Is->Frequency of occurrence of each suspected point in each line of characteristic data, +.>Is->Line characteristic data>The first part of the suspected points and the rest of the suspected points>Close constraint of individual suspected pointsDegree.
4. The white spirit storage abnormality pre-warning system based on multi-source data according to claim 1, wherein the specific steps of obtaining each set of each line of characteristic data according to the degree of close constraint among each suspected point of each line of characteristic data are as follows:
in each line of feature data, for any suspected point, the suspected point with the maximum close constraint degree obtained by calculating the rest suspected points is formed into a small set;
and summing all small sets with intersections to obtain each set of the characteristic data of each row.
5. The white spirit storage abnormality pre-warning system based on multi-source data according to claim 1, wherein the specific step of obtaining the span coefficient of each set according to the position information of the data points in each set of each row of characteristic data is as follows:
in the method, in the process of the application,is->Line characteristic data>The left-most suspected point in the set is on the abscissa in the spatial distribution of the feature data, +.>Is->Line characteristic data>The right-most suspected point in the set is on the abscissa in the spatial distribution of the feature data, +.>Is->Line characteristic data>Maximum longitudinal span of the sets in the longitudinal direction, < > about->Is->Line characteristic data>Maximum longitudinal span distance in longitudinal axis direction between the nearest two peaks and valleys on the left and right sides of each set, +.>Is->Line characteristic data>Span coefficients of the sets.
6. The multi-source data-based white spirit storage abnormality pre-warning system according to claim 1, wherein the expression for obtaining the close combination coefficients of the sets according to each close constraint degree in each set of each row of characteristic data and the span coefficient of each set is:
in the method, in the process of the application,is->Line characteristic data>Span coefficient of individual sets, +.>Is->Line characteristic data>Number of suspected points of the collection, +.>Is->Line characteristic data>First->Degree of close constraint->Is->Line characteristic data>Personal setAnd (5) closely combining coefficients.
7. The multi-source data-based white spirit storage abnormality pre-warning system according to claim 1, wherein the expression for obtaining the characteristic influence coefficient of each line of characteristic data according to the close combination coefficient of each set of each line of characteristic data is:
in the method, in the process of the application,for correction factor +.>For normalization function->Is->The number of sets in the line profile, +.>Is->Line characteristic data>Closely-combined coefficients of the individual sets,/->Is->Line characteristic data>The number of suspected points in the set,is->The characteristic influence coefficient of the line characteristic data.
8. The white spirit storage abnormality pre-warning system based on multi-source data according to claim 1, wherein the specific steps of obtaining the characteristic weight coefficient according to the characteristic influence coefficient of each row of characteristic data are as follows:
summing the characteristic influence coefficients of each row of characteristic data to obtain denominators;
and (3) comparing the characteristic influence coefficient of each line of characteristic data with a denominator to obtain a characteristic weight coefficient of the corresponding line of characteristic data.
9. The white spirit storage abnormality pre-warning system based on multi-source data according to claim 1, wherein the specific step of obtaining the weighted characteristic parameter of each data point according to the characteristic weight coefficient of each row of characteristic data is as follows:
and carrying out weighted summation on each data point corresponding to each column in the environment feature matrix by using the feature weight coefficient of each row of feature data to obtain weighted feature parameters corresponding to each column of time points.
10. The white spirit storage abnormality pre-warning system based on multi-source data according to claim 1, wherein the specific step of obtaining each different data point according to the outlier probability and the outlier threshold of each data point is as follows:
and acquiring the outlier probability of each data point, setting an outlier threshold, and taking the data point with the outlier probability larger than the outlier threshold as an abnormal data point.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202311112658.7A CN116823125B (en) | 2023-08-31 | 2023-08-31 | White spirit storage abnormality early warning system based on multisource data |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202311112658.7A CN116823125B (en) | 2023-08-31 | 2023-08-31 | White spirit storage abnormality early warning system based on multisource data |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN116823125A true CN116823125A (en) | 2023-09-29 |
| CN116823125B CN116823125B (en) | 2023-10-31 |
Family
ID=88114952
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202311112658.7A Active CN116823125B (en) | 2023-08-31 | 2023-08-31 | White spirit storage abnormality early warning system based on multisource data |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN116823125B (en) |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN117282261A (en) * | 2023-11-23 | 2023-12-26 | 天津恩纳社环保有限公司 | Microorganism waste gas treatment system |
| CN118195105A (en) * | 2024-05-17 | 2024-06-14 | 江苏云机汇软件科技有限公司 | Supply chain flow optimizing system based on artificial intelligence |
Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20160156652A1 (en) * | 2012-04-20 | 2016-06-02 | Numerica Corporaition | Pattern detection in sensor networks |
| CN206787602U (en) * | 2017-01-10 | 2017-12-22 | 浙江思平智能科技有限公司 | A kind of storage environment monitor and alarm system |
| CN112465948A (en) * | 2020-11-24 | 2021-03-09 | 山东科技大学 | Vehicle-mounted laser pavement point cloud rarefying method capable of retaining spatial features |
| CN112990706A (en) * | 2021-03-15 | 2021-06-18 | 上海应用技术大学 | Dangerous chemical storage environment risk early warning method |
| CN114254705A (en) * | 2021-12-21 | 2022-03-29 | 上海观安信息技术股份有限公司 | Abnormal data detection method and device, storage medium and computer equipment |
| CN114912852A (en) * | 2022-07-18 | 2022-08-16 | 山东鑫亚装备制造有限公司 | Abnormity monitoring and early warning system for storage tank of refrigeration bin |
-
2023
- 2023-08-31 CN CN202311112658.7A patent/CN116823125B/en active Active
Patent Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20160156652A1 (en) * | 2012-04-20 | 2016-06-02 | Numerica Corporaition | Pattern detection in sensor networks |
| CN206787602U (en) * | 2017-01-10 | 2017-12-22 | 浙江思平智能科技有限公司 | A kind of storage environment monitor and alarm system |
| CN112465948A (en) * | 2020-11-24 | 2021-03-09 | 山东科技大学 | Vehicle-mounted laser pavement point cloud rarefying method capable of retaining spatial features |
| CN112990706A (en) * | 2021-03-15 | 2021-06-18 | 上海应用技术大学 | Dangerous chemical storage environment risk early warning method |
| CN114254705A (en) * | 2021-12-21 | 2022-03-29 | 上海观安信息技术股份有限公司 | Abnormal data detection method and device, storage medium and computer equipment |
| CN114912852A (en) * | 2022-07-18 | 2022-08-16 | 山东鑫亚装备制造有限公司 | Abnormity monitoring and early warning system for storage tank of refrigeration bin |
Non-Patent Citations (1)
| Title |
|---|
| 杜航原等: "融合知识的多视图属性网络异常检测模型", 自动化学报, vol. 49, no. 8 * |
Cited By (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN117282261A (en) * | 2023-11-23 | 2023-12-26 | 天津恩纳社环保有限公司 | Microorganism waste gas treatment system |
| CN117282261B (en) * | 2023-11-23 | 2024-02-23 | 天津恩纳社环保有限公司 | Microorganism waste gas treatment system |
| CN118195105A (en) * | 2024-05-17 | 2024-06-14 | 江苏云机汇软件科技有限公司 | Supply chain flow optimizing system based on artificial intelligence |
| CN118195105B (en) * | 2024-05-17 | 2024-08-02 | 江苏云机汇软件科技有限公司 | Supply chain flow optimizing system based on artificial intelligence |
Also Published As
| Publication number | Publication date |
|---|---|
| CN116823125B (en) | 2023-10-31 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN116823125B (en) | White spirit storage abnormality early warning system based on multisource data | |
| US11740619B2 (en) | Malfunction early-warning method for production logistics delivery equipment | |
| CN105467975B (en) | A kind of equipment fault diagnosis method | |
| CN111444241A (en) | Data mining-based accurate positioning method for line loss abnormity associated users of distribution room | |
| CN106548270B (en) | Photovoltaic power station power abnormity data identification method and device | |
| CN112039903B (en) | Network security situation assessment method based on deep self-coding neural network model | |
| CN113255848A (en) | Water turbine cavitation sound signal identification method based on big data learning | |
| CN113923104A (en) | Network fault diagnosis method, equipment and storage medium based on wavelet neural network | |
| CN117609929B (en) | Industrial production line fault online diagnosis method and system based on big data | |
| CN111680107A (en) | Financial prediction system based on artificial intelligence and block chain | |
| CN116842410B (en) | Intelligent helmet antitheft management method and system based on dynamic perception | |
| CN116956197B (en) | Deep learning-based energy facility fault prediction method and device and electronic equipment | |
| Ghanaatiyan et al. | Multi-objective economic-statistical design of VSSI-MEWMA-DWL control chart with multiple assignable causes | |
| CN115310565B (en) | Network security monitoring method based on artificial intelligence | |
| CN116701846A (en) | Hydropower station dispatching operation data cleaning method based on unsupervised learning | |
| CN117877587A (en) | Deep learning algorithm of whole genome prediction model | |
| CN116313111A (en) | Breast cancer risk prediction method, system, medium and equipment based on combined model | |
| CN115034443A (en) | Boiler performance prediction method based on working condition division and neural network | |
| CN115184250A (en) | Optical fiber aging state evaluation system and evaluation method thereof | |
| CN112906751A (en) | Method for identifying abnormal value through unsupervised learning | |
| CN117648657B (en) | Urban planning multi-source data optimization processing method | |
| CN118171135B (en) | Aquaculture environment monitoring method based on multi-source data fusion | |
| CN118277935A (en) | Distributed photovoltaic abnormal data score decision method based on time sequence clustering | |
| Szatkownik et al. | Towards creating longer genetic sequences with GANs: Generation in principal component space | |
| CN118052540B (en) | Intelligent management method and system for equipment maintenance scheme |
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 |