CN116756529B - White piece production facility running state detecting system hangs - Google Patents
White piece production facility running state detecting system hangs Download PDFInfo
- Publication number
- CN116756529B CN116756529B CN202311047135.9A CN202311047135A CN116756529B CN 116756529 B CN116756529 B CN 116756529B CN 202311047135 A CN202311047135 A CN 202311047135A CN 116756529 B CN116756529 B CN 116756529B
- Authority
- CN
- China
- Prior art keywords
- fitting
- preset
- discrete
- data
- data points
- 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
Links
- 238000004519 manufacturing process Methods 0.000 title claims abstract description 49
- 230000002159 abnormal effect Effects 0.000 claims abstract description 80
- 230000035772 mutation Effects 0.000 claims abstract description 74
- 238000011156 evaluation Methods 0.000 claims abstract description 52
- 238000000034 method Methods 0.000 claims abstract description 47
- 238000000354 decomposition reaction Methods 0.000 claims abstract description 45
- 230000001932 seasonal effect Effects 0.000 claims abstract description 37
- 238000009826 distribution Methods 0.000 claims abstract description 29
- 238000001514 detection method Methods 0.000 claims abstract description 15
- 238000010606 normalization Methods 0.000 claims description 16
- 238000012545 processing Methods 0.000 claims description 9
- 238000000605 extraction Methods 0.000 claims description 6
- 238000013507 mapping Methods 0.000 claims description 4
- 230000008569 process Effects 0.000 abstract description 17
- 238000012937 correction Methods 0.000 abstract description 16
- 230000000694 effects Effects 0.000 abstract description 8
- 230000008859 change Effects 0.000 description 20
- 238000002360 preparation method Methods 0.000 description 13
- WSFSSNUMVMOOMR-UHFFFAOYSA-N Formaldehyde Chemical compound O=C WSFSSNUMVMOOMR-UHFFFAOYSA-N 0.000 description 11
- 238000006243 chemical reaction Methods 0.000 description 9
- 238000002955 isolation Methods 0.000 description 7
- XWGJFPHUCFXLBL-UHFFFAOYSA-M rongalite Chemical compound [Na+].OCS([O-])=O XWGJFPHUCFXLBL-UHFFFAOYSA-M 0.000 description 7
- 238000003786 synthesis reaction Methods 0.000 description 7
- 238000004458 analytical method Methods 0.000 description 6
- 230000015572 biosynthetic process Effects 0.000 description 6
- 238000010586 diagram Methods 0.000 description 6
- 239000004973 liquid crystal related substance Substances 0.000 description 6
- 230000035945 sensitivity Effects 0.000 description 6
- HCHKCACWOHOZIP-UHFFFAOYSA-N Zinc Chemical compound [Zn] HCHKCACWOHOZIP-UHFFFAOYSA-N 0.000 description 5
- 230000000737 periodic effect Effects 0.000 description 5
- GRWZHXKQBITJKP-UHFFFAOYSA-L dithionite(2-) Chemical compound [O-]S(=O)S([O-])=O GRWZHXKQBITJKP-UHFFFAOYSA-L 0.000 description 4
- XUIVKWAWICCWIQ-UHFFFAOYSA-M sodium;formaldehyde;hydrogen sulfite Chemical compound [Na+].O=C.OS([O-])=O XUIVKWAWICCWIQ-UHFFFAOYSA-M 0.000 description 4
- DWAQJAXMDSEUJJ-UHFFFAOYSA-M Sodium bisulfite Chemical compound [Na+].OS([O-])=O DWAQJAXMDSEUJJ-UHFFFAOYSA-M 0.000 description 3
- HEMHJVSKTPXQMS-UHFFFAOYSA-M Sodium hydroxide Chemical compound [OH-].[Na+] HEMHJVSKTPXQMS-UHFFFAOYSA-M 0.000 description 3
- 235000010267 sodium hydrogen sulphite Nutrition 0.000 description 3
- 229910052725 zinc Inorganic materials 0.000 description 3
- 239000011701 zinc Substances 0.000 description 3
- RAHZWNYVWXNFOC-UHFFFAOYSA-N Sulphur dioxide Chemical compound O=S=O RAHZWNYVWXNFOC-UHFFFAOYSA-N 0.000 description 2
- 230000004913 activation Effects 0.000 description 2
- 238000004364 calculation method Methods 0.000 description 2
- 230000007774 longterm Effects 0.000 description 2
- 239000002994 raw material Substances 0.000 description 2
- 230000035484 reaction time Effects 0.000 description 2
- LFQSCWFLJHTTHZ-UHFFFAOYSA-N Ethanol Chemical compound CCO LFQSCWFLJHTTHZ-UHFFFAOYSA-N 0.000 description 1
- 235000000177 Indigofera tinctoria Nutrition 0.000 description 1
- 230000005856 abnormality Effects 0.000 description 1
- 230000009471 action Effects 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 230000005540 biological transmission Effects 0.000 description 1
- 238000004061 bleaching Methods 0.000 description 1
- 239000007844 bleaching agent Substances 0.000 description 1
- 239000003638 chemical reducing agent Substances 0.000 description 1
- 239000003795 chemical substances by application Substances 0.000 description 1
- 238000013480 data collection Methods 0.000 description 1
- 239000000975 dye Substances 0.000 description 1
- 238000004043 dyeing Methods 0.000 description 1
- 238000009499 grossing Methods 0.000 description 1
- 238000010438 heat treatment Methods 0.000 description 1
- 229940097275 indigo Drugs 0.000 description 1
- COHYTHOBJLSHDF-UHFFFAOYSA-N indigo powder Natural products N1C2=CC=CC=C2C(=O)C1=C1C(=O)C2=CC=CC=C2N1 COHYTHOBJLSHDF-UHFFFAOYSA-N 0.000 description 1
- 230000001788 irregular Effects 0.000 description 1
- 238000005259 measurement Methods 0.000 description 1
- 239000000203 mixture Substances 0.000 description 1
- 238000012544 monitoring process Methods 0.000 description 1
- 239000000843 powder Substances 0.000 description 1
- 238000004321 preservation Methods 0.000 description 1
- 230000000750 progressive effect Effects 0.000 description 1
- 230000000630 rising effect Effects 0.000 description 1
- 238000012216 screening Methods 0.000 description 1
- 239000004289 sodium hydrogen sulphite Substances 0.000 description 1
- 239000000126 substance Substances 0.000 description 1
- 230000002194 synthesizing effect Effects 0.000 description 1
- XLYOFNOQVPJJNP-UHFFFAOYSA-N water Substances O XLYOFNOQVPJJNP-UHFFFAOYSA-N 0.000 description 1
Classifications
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01K—MEASURING TEMPERATURE; MEASURING QUANTITY OF HEAT; THERMALLY-SENSITIVE ELEMENTS NOT OTHERWISE PROVIDED FOR
- G01K13/00—Thermometers specially adapted for specific purposes
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Data Mining & Analysis (AREA)
- Theoretical Computer Science (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Artificial Intelligence (AREA)
- Bioinformatics & Computational Biology (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Evolutionary Biology (AREA)
- Evolutionary Computation (AREA)
- General Engineering & Computer Science (AREA)
- Life Sciences & Earth Sciences (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The invention relates to the technical field of temperature detection, in particular to a system for detecting the running state of hanging white block production equipment. STL decomposition is carried out on temperature time sequence data to obtain seasonal items in the decomposition process; determining discrete data points according to the fluctuation deviation degree of the data points; obtaining mutation evaluation factors through the data value deviation degree and the distribution density condition of the discrete data points in the preset local range of each discrete data point; updating the fitting weight of the local weighted regression algorithm according to the mutation evaluation factor, and fitting the seasonal term to obtain a fitting trend term; and adjusting STL decomposition according to the fitting trend term to obtain a correction season term, and detecting the running state of the hanging white block production equipment according to the fitting trend term and the correction season term. According to the invention, by adjusting the fitting weight of the abnormal data in STL decomposition, the obtained trend item and the correction season item are more accurate, and the effect of detecting the running state of the hanging white block production equipment is better.
Description
Technical Field
The invention relates to the technical field of temperature detection, in particular to a system for detecting the running state of hanging white block production equipment.
Background
The white hanging block is also called as rongalite, is prepared by combining formalin with sodium bisulphite and then reducing the mixture, and has the chemical name of formaldehyde sodium bisulphite, presents white block or crystalline powder, is easy to dissolve in water and is slightly soluble in alcohol. Because of its extremely strong reducibility at high temperature, it has bleaching action, and is therefore mainly used as an industrial bleaching agent, as a discharge agent and a reducing agent in the printing and dyeing industry, for producing indigo dyes, etc. The traditional process for preparing the sodium formaldehyde sulfoxylate takes sulfur dioxide, zinc powder, formaldehyde and sodium hydroxide as raw materials, and synthesizes the sodium formaldehyde sulfoxylate through three steps of reaction, and finally synthesizes the sodium formaldehyde sulfoxylate in a closed reaction kettle. In the production process, besides the great influence on the synthesis reaction caused by the content and the addition amount of the raw materials, the synthesis temperature in the production equipment can also have great influence, and strict control and monitoring are required.
However, STL decomposition methods are often used for detecting the change of the synthesis temperature in the equipment in the product preparation scene, and the STL decomposition methods can help identify and understand different components in the temperature data, and know the temperature change mode and periodicity of the equipment in the preparation process during operation. When the production equipment is used for preparing formaldehyde sodium bisulphite, a sensor for collecting temperature data is interfered by surrounding magnetic fields, power sources or chemical reactions and other environments, the situation that the temperature data is changed temporarily and abnormally occurs at a certain time node in the preparation process can exist, the phenomenon can affect the STL decomposition method, because a local weighted regression algorithm used when the STL decomposition method is used for fitting data points is very sensitive to the abnormal situation, fitting results are interfered to generate offset, fitting results are inaccurate, errors of data states represented by trend items and seasonal items obtained after time sequence decomposition are larger, and finally, the detection effect on the temperature change state of the production equipment is poor.
Disclosure of Invention
In order to solve the technical problems that in the prior art, the errors of the state of the trend item and the season item representation data obtained after the temperature time sequence data in the operation of the hanging white block production equipment are decomposed are large, and finally the detection effect on the temperature change state in the operation of the production equipment is poor, the invention aims to provide a hanging white block production equipment operation state detection system, which adopts the following technical scheme:
the invention provides a system for detecting the running state of hanging white block production equipment, which comprises the following components:
the data acquisition module is used for acquiring temperature time sequence data in the operation of the hanging white block production equipment and acquiring seasonal items of the temperature time sequence data through STL decomposition;
the evaluation factor extraction module is used for determining discrete data points in all data points according to the fluctuation deviation degree of each data point on the out-of-season term in a preset fitting window; obtaining mutation evaluation factors of each discrete data point according to the data value offset degree between the discrete data point and other data points and the distribution density condition of the discrete data points in a preset local range of each discrete data point;
the fitting module is used for updating the fitting weights of all the data points in a preset fitting window according to the mutation evaluation factors of the discrete data points; fitting the seasonal items through a local weighted regression algorithm to obtain fitting trend items;
the state detection module is used for adjusting the season term in STL decomposition according to the fitting trend term to obtain a corrected season term corresponding to the temperature time sequence data; and detecting the running state of the hanging white block production equipment through fitting the trend item and correcting the season item.
Further, the determining discrete data points in all data points according to the fluctuation deviation degree of each data point on the out-of-season item comprises:
in a preset fitting window, acquiring a connecting line between each data point and the previous data point on the out-of-season item, and taking an included angle between the connecting line and a horizontal line as the fluctuation degree of each data point; taking the fluctuation degree of the second data point in the preset fitting window as the fluctuation degree of the first data point;
calculating the difference between the fluctuation degree of each data point and the average value of all fluctuation degrees in a preset fitting window, and taking the difference after normalization processing as a discrete evaluation factor of each data point; discrete ones of the data points are determined based on the discrete evaluation factors of the data points.
Further, the method for obtaining the mutation evaluation factor comprises the following steps:
calculating the data value difference between any one discrete data point and other data points in the preset local range of the discrete data point; calculating the average value of all data value differences corresponding to the discrete data points and carrying out normalization processing to obtain the data mutation index of the discrete data points;
carrying out negative correlation mapping and normalization processing on the quantity proportion of all discrete data points in the preset local range corresponding to each discrete data point to obtain a distribution mutation index of each discrete data point;
and carrying out weighted summation on the data mutation index and the distribution mutation index of each discrete data point to obtain a mutation evaluation factor of each discrete data point.
Further, updating the fitting weights of all data points in a preset fitting window according to the mutation evaluation factors of the discrete data points by the local weighted regression algorithm comprises the following steps:
when the mutation evaluation factor of the discrete data points is larger than or equal to a preset mutation threshold value, setting the weight adjustment value of the corresponding discrete data points as a preset first adjustment value; when the mutation evaluation factor of the discrete data points is smaller than a preset mutation threshold value, setting the weight adjustment value of the corresponding discrete data points as a preset second adjustment value;
setting the weight adjustment value of the non-discrete data point as a preset third adjustment value; the preset first adjustment value is smaller than the preset third adjustment value, and the preset third adjustment value is smaller than the preset second adjustment value;
and updating the fitting weight of each data point in the local weighted regression algorithm according to the weight adjustment value of each data point in the preset fitting window.
Further, updating the fitting weight of each data point in the local weighted regression algorithm according to the weight adjustment value of each data point in the preset fitting window, including:
and adding the fitting weight of each data point in the local weighted regression algorithm to the corresponding weight adjustment value to obtain the updated fitting weight of each data point.
Further, the detecting the operation state of the hanging white block production device through fitting the trend item and the modified season item comprises the following steps:
obtaining abnormal values in the fitting trend item according to the Laida criterion, and counting the number of the abnormal values to obtain the number of the trend abnormal values; obtaining abnormal values in the corrected season items according to the Laida criterion, and counting the number of the abnormal values to obtain the number of the season abnormal values;
when the number of the trend abnormal values and the number of the season abnormal values are smaller than a preset abnormal threshold value, the running state of the hanging white block production equipment is recorded as normal; when the number of the trend abnormal values or the number of the season abnormal values is larger than or equal to a preset abnormal threshold value, the running state of the hanging white block production equipment is marked as abnormal.
Further, the seasonal out-of-season term for obtaining temperature time series data through STL decomposition includes:
in the time sequence data of the decomposition temperature through STL, fitting the time sequence data of the temperature to obtain a season term; subtracting the seasonal term from the temperature time sequence data to obtain a seasonal term.
Further, the determining discrete ones of the data points based on the discrete evaluation factors of the data points includes:
and taking the data points with the discrete evaluation factors being greater than or equal to a preset discrete threshold value as discrete data points.
Further, the preset local range is smaller than or equal to half of the size of a preset fitting window.
Further, the fitting method of the seasonal term is a moving average method.
The invention has the following beneficial effects:
according to the method, STL decomposition is carried out on temperature time sequence data, seasonal items in the decomposition process are obtained firstly, in the fitting process of the seasonal items, the sensitivity problem of the fitting weight of a local weighted regression algorithm to abnormal data is considered, the fitting weight of the abnormal data is adjusted, and firstly, discrete data points with abnormal conditions are screened out according to the fluctuation deviation degree of the data points in a preset fitting window. In the production and preparation process of the hanging white block, the abnormal condition of temperature data consists of two conditions of abnormal points and error points, and fitting sensitivity required by the error points and the abnormal points is different, so that the mutation evaluation factor of each discrete data point is obtained according to the characteristic that the error points have mutation isolation and in the preset local range of each discrete data point through the two aspects of the data value deviation degree and the distribution density condition of the discrete data point. According to the mutation evaluation factors, the discrete data points can be represented as error points or abnormal points, and then the fitting weight of the local weighted regression algorithm is updated, so that the fitting of the seasonal items is completed to obtain fitting trend items, and the fitting trend items are more accurate. Finally, the season term in STL decomposition is adjusted according to the fitting trend term, a correction season term is obtained, the abnormal condition in the temperature data is fully considered, the fitting weight of the abnormal data in the STL decomposition is adjusted, the obtained fitting trend term and the correction season term are more accurate and reliable, and the effect of detecting the running state of the hanging white block production equipment through the fitting trend term and the correction season term is better and more reliable.
Drawings
In order to more clearly illustrate the embodiments of the invention 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 invention, and other drawings can be obtained according to the drawings without inventive effort for a person skilled in the art.
FIG. 1 is a diagram of a system for detecting the operation state of a white hanging block production device according to an embodiment of the present invention;
FIG. 2 is a schematic view of obtaining the fluctuation degree according to an embodiment of the present invention;
FIG. 3 is a schematic diagram of a fitting trend term of STL decomposition according to an embodiment of the present invention;
fig. 4 is a schematic diagram of a modified season term for STL decomposition according to an embodiment of the present invention.
Detailed Description
In order to further describe the technical means and effects adopted by the invention to achieve the preset aim, the following detailed description is given below of a system for detecting the running state of hanging white block production equipment according to the invention, which is specific to the implementation, structure, characteristics and effects thereof, with reference to the attached 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 invention belongs.
The following specifically describes a specific scheme of the running state detection system of the hanging white block production equipment provided by the invention with reference to the accompanying drawings.
Referring to fig. 1, a block diagram of an operation state detection system of a white hanging block production device according to an embodiment of the present invention is shown, where the operation state detection system of the white hanging block production device includes: the system comprises a data acquisition module 101, an evaluation factor extraction module 102, a fitting module 103 and a state detection module 104.
The data acquisition module 101 is used for acquiring temperature time sequence data in the operation of the hanging white block production equipment, and obtaining seasonal items of the temperature time sequence data through STL decomposition.
In the preparation process of sodium formaldehyde sulfoxylate, namely sodium formaldehyde sulfoxylate, there are two key stages: an activation stage and a synthesis stage. The activation stage is a reaction stage in which zinc powder is added to the sodium bisulfite solution but formaldehyde is not added, and is mainly a reaction stage in which zinc reacts with sodium bisulfite to generate dithionite. Dithionite is an important intermediate for synthesizing sodium formaldehyde sulfoxylate, and whether the reaction is sufficiently performed or not greatly affects the synthesis result. Since dithionite is particularly sensitive to temperature, when abnormality occurs in the temperature in the apparatus, dithionite is decomposed in large amounts, resulting in poor synthesis. And after formaldehyde is added in the synthesis stage, the activity of zinc is required to be improved by heating, and finally formaldehyde sodium bisulphite is generated, but the exothermic reaction occurs when the zinc is heated, so that the temperature in the equipment is increased, and the preparation reaction process is influenced.
Therefore, the invention needs to detect the temperature in the equipment in real time so as to find abnormal temperature change in time, and in order to comprehensively detect the change state in the preparation process, the STL (Seaseal-Trend decomposition procedures based on Loess) decomposition method is used for decomposing temperature time sequence data, wherein the STL decomposition method is a method capable of decomposing a time sequence into trend items, season items and residual items, comprises a series of local weighted regression smoother, has higher calculation speed and can cope with very large time sequence data. The long-term trend of the temperature in the equipment along with the change of the reaction time is reflected by the trend item obtained through decomposition, the rising or falling state of the temperature of the equipment is observed more clearly, and the periodic state of the fluctuation of the temperature in the equipment can be reflected by the season item obtained through decomposition.
Firstly, temperature change time sequence data of the hanging white block production equipment in the preparation process is acquired through a temperature sensor, all the acquired data are mapped into a time sequence space, the temperature time sequence data are obtained, and STL decomposition is further carried out on the temperature time sequence data. Wherein the time sequence space is a time sequence coordinate system, the horizontal axis is time, and the vertical axis is temperature value.
In the embodiment of the invention, the STL decomposition process mainly comprises the following steps: and obtaining a season term, fitting a trend term, and carrying out four processes of season term correction and residual term obtaining. The season term is obtained first: the seasonal items in the time sequence data are obtained through a moving average method, the seasonal modes of the time sequence data can be estimated through the moving average method, and then the seasonal items are fitted, and the seasonal items reflect the seasonality of the time sequence data. The fitting trend term is: subtracting the seasonal term from the time sequence data to obtain a seasonal term. The seasonal term includes a trend term and a residual term, a fitting trend term is obtained by a smoothing mode through a local weighted regression (Locally weighted regression, loess) algorithm, and the fitting trend term carries out preliminary judgment on trend conditions of temperature time sequence data. The season term correction is carried out as follows: and correcting by combining the fitting trend term and the season term to obtain a corrected season term, and adjusting the season term of each time point to enable the sum of the season terms to be matched with the original time sequence data. The residual term is obtained as: and deleting the fitting trend item and the correction season item from the original time sequence data to obtain a residual item. It should be noted that, the sensor data collection, the moving average method, the STL decomposition method, etc. are all known means known to those skilled in the art, and other known means such as seasonal index method may be used for fitting the seasonal term, which will not be described herein.
In consideration of the fact that when the production equipment is used for preparing formaldehyde sodium bisulphite, the sensor is likely to be interfered by a magnetic field, a power supply or a chemical reaction, abnormal data of abrupt change of temperature time sequence data transmission is extremely easy to collect, the abnormal data belong to error points, the interference of the error points to the Loess algorithm is extremely high, fitting weights of data points in the Loess algorithm are acquired based on adjacent data points, the error points are also used as abnormal points of the temperature for distributing the fitting weights, and therefore the invention analyzes the change condition of the data and carries out self-adaptive adjustment on the fitting weights in the Loess algorithm so as to obtain better fitting trend items for subsequent decomposition.
Therefore, in the STL decomposition process of the temperature time sequence data, the seasonal term is obtained by fitting the temperature time sequence data to obtain the seasonal term, and subtracting the seasonal term from the temperature time sequence data to obtain the seasonal term. And then analyzing and adjusting the fitting trend item process after the season item is obtained.
The evaluation factor extraction module 102 is configured to determine, in a preset fitting window, discrete data points in all data points according to the fluctuation deviation degree of each data point; and obtaining the mutation evaluation factor of each discrete data point according to the data value offset degree between the discrete data point and other data points and the distribution density condition of the discrete data points in the preset local range of each discrete data point.
The Loess algorithm mainly obtains a final fitting result by carrying out weighted fitting on each data point in a local window and adjusting fitting weights according to initial fitting conditions, and the calculation process of the Loess algorithm is a technical means well known to those skilled in the art. When the data change condition is analyzed, the method is also based on a preset fitting window in the Loess algorithm. In the embodiment of the invention, the size of the preset fitting window is set to be 20, namely, each 20 continuous time points are correspondingly one preset fitting window, 20 data points are arranged in one preset fitting window, and a specific numerical value implementation person can regulate and control the window by himself. The points of temperature anomaly are usually points requiring higher sensitivity, but since the points in which error caused by external factors exist are also regarded as points of high sensitivity, the abnormal points of temperature are distinguished from the error points by the analysis of the change of data, and the fitting weight is adjusted.
Firstly, determining discrete data points in a preset fitting window according to the fluctuation deviation degree of each data point, wherein the discrete data points are points with abnormal conditions of temperature change, namely abnormal points and error points. Firstly, obtaining a discrete evaluation factor of each data point according to the fluctuation deviation condition, and screening out the data points with larger fluctuation degree difference through the fluctuation deviation condition. Preferably, in the preset fitting window, a connection line between each data point and a previous data point is obtained, and an included angle between the connection line and a horizontal line is taken as the fluctuation degree of each data point. Referring to FIG. 2, a schematic view of obtaining fluctuation degree according to one embodiment of the present invention is shown, whereinAnd->For two adjacent data points +.>Denoted as->The degree of fluctuation of the data points.
Because the data points are time-series data points, each data point corresponds to a time node, analysis of each data point and the previous data point does not affect analysis of the next newly added data point, in the embodiment of the invention, since the first data point in the preset fitting window does not correspond to the previous data point, the fluctuation degree cannot be calculated, and therefore the fluctuation degree of the second data point is taken as the fluctuation degree of the first data point, and each data point corresponds to one fluctuation degree.
The discrete condition of each data point is reflected by the deviation of the fluctuation degree of each data point relative to the overall fluctuation degree. Preferably, the difference between the fluctuation degree of each data point and the average value of all fluctuation degrees in the preset fitting window is calculated, and the difference after normalization treatment is taken as a discrete evaluation factor of each data point. In the embodiment of the invention, the specific expression of the discrete evaluation factor is as follows:
in the method, in the process of the invention,denoted as +.>Discrete evaluation factors for data points +.>Denoted as +.>The degree of fluctuation of the data points,denoted as +.>Degree of fluctuation of data points +.>Expressed as total number of data points in a preset fit window, +.>Expressed as absolute value extraction function,/->Expressed as an S-shaped growth curve, i.e. normalization function, according to +.>The function normalizes the range of values of the discrete evaluation factors to +.>. It should be noted that->The method of function normalization is a technical means well known to those skilled in the art, and will not be described in detail herein.
Wherein, the liquid crystal display device comprises a liquid crystal display device,expressed as the average of all the fluctuation degrees in the preset fit window,denoted as +.>The difference between the degree of fluctuation of the data points and the average value of all the degrees of fluctuation in the preset fitting window. When the difference is larger, the inconsistent fluctuation condition of the corresponding data point and the whole data point is indicated, and the more abnormal data points in the temperature time sequence data are likely to be corresponding data points.
Further, discrete data points among the data points are determined according to the discrete evaluation factors of the data points, and in one embodiment of the present invention, the data points with the discrete evaluation factors being greater than or equal to a preset discrete threshold value are used as the discrete data points. The discrete data points are abnormal data points in the temperature time sequence data, wherein the preset discrete threshold value is 0.5, and a specific numerical value implementation person can adjust according to specific implementation conditions.
The white hanging block production equipment is extremely easy to be interfered by external factors in the preparation process, so that the temperature sensor can acquire some abrupt error points. Therefore, when the temperature is abnormal, the isolation of the abnormal point is low, no obvious abrupt change condition occurs, and the isolation of the data point is strong for the error point acquired by the sensor caused by external factors. Because the abnormal points and the error points have obvious mutation characteristics in numerical value and distribution density, the analysis of mutation conditions is carried out on the screened discrete data points in two aspects, so that the abnormal points and the error points are further distinguished.
When the variability of each discrete data point is analyzed, based on the distribution condition of each discrete data point in the local range, a corresponding preset local range is taken for each discrete data point, in one embodiment of the invention, the value of the preset local range is less than or equal to half of the size of a preset fitting window, so that the analysis of the discrete data point can be ensured to be in the corresponding preset fitting window, therefore, the size of the preset local range is set to be 10, namely, each 10 continuous time points are one preset local range, 10 data points are arranged in the preset local range, and a specific numerical value implementation person can regulate and control by himself.
In the embodiment of the invention, the preset local range of each discrete data point specifically comprises: and taking 9 time points backwards from the time point corresponding to each discrete data point as a preset local range of each discrete data point, and continuously selecting forwards from the time point corresponding to each discrete data point when the time point behind the discrete data point in a preset fitting window is less than 9 time points, so that the continuous time points meet the size of the preset local range, for example, the preset fitting window size is 20, the preset local range size is 5, and when the discrete data point is the 18 th data point, the preset local range corresponding to the discrete data point is finally formed by the 16 th, 17, 18, 19 and 20 th data points in the preset fitting window.
Firstly, analyzing the mutation degree of each discrete data point and other data points in a preset local range, according to the deviation degree of the data value between the discrete data point and the other data points, obtaining a data mutation index of the discrete data point, preferably, analyzing any one discrete data point, calculating the data value difference between the discrete data point and the other data points in the preset local range of any one discrete data point, wherein the larger the data difference is, the larger the deviation degree of the data value of the corresponding discrete data point is, the average value of all the data value differences corresponding to the discrete data point is obtained, and carrying out normalization processing to obtain the data mutation index of the discrete data point, and the larger the data mutation index is, the more obvious the mutation situation of the discrete data is indicated, and in the embodiment of the invention, the expression of the data mutation index is as follows:
in the method, in the process of the invention,denoted as +.>Data mutation index of discrete data points, +.>Denoted as +.>The number of discrete data points corresponds to the total number of other data points within the preset local range, +.>Denoted as +.>Data value of discrete data points, +.>Denoted as +.>The discrete data points correspond to the +.>Data values of the other data points, +.>Expressed as absolute value extraction function,/->It should be noted that, normalization is a technical means well known to those skilled in the art, and the normalization function may be selected by linear normalization or standard normalization, and the specific normalization method is not limited herein.
Wherein, the liquid crystal display device comprises a liquid crystal display device,denoted as +.>Discrete data points and->The greater the difference in data values between the discrete data point and all other data points, the greater the degree of deviation of the discrete data point, which is the greater the degree of mutation of the discrete data point.
Further, the discrete data points are analyzed in the distribution density degree, when the discrete data points are more likely to be abrupt error points, the distribution of the discrete data points is isolated, and for abnormal points of temperature time sequence data, as the temperature change is stable and changed in the preparation process, the data points with abnormal temperature have certain continuous stability, and the isolation possibility is lower. Therefore, under the condition of dense distribution of the discrete data points, the distribution mutation index of the discrete data points can be obtained, and preferably, the quantity proportion of all the discrete data points in the preset local range corresponding to each discrete data point is subjected to negative correlation mapping and normalization processing to obtain the distribution mutation index of each discrete data point. The distribution isolation of each discrete data point is reflected by the distribution mutation index, and in the embodiment of the invention, the specific expression of the distribution mutation index is as follows:
in the method, in the process of the invention,denoted as +.>Mutation index of the distribution of discrete data points, +.>Denoted as +.>The number of discrete data points corresponds to the total number of other data points within the preset local range, +.>Denoted as +.>The number of discrete data points corresponds to the total number of discrete data points within a preset local range, < >>Represented as an exponential function with a base of natural constant.
Wherein, the liquid crystal display device comprises a liquid crystal display device,denoted as +.>The number of discrete data points corresponds to the total number of data points within a predetermined local range, i.e. the size of the predetermined local range,/->Expressed as a negative correlation mapping by an exponential function based on a natural constant and normalized,/treatment>Denoted as +.>The number of the discrete data points corresponds to the number of the discrete data points in the preset local range, the distribution density condition of the discrete data points in the preset local range is reflected through the number of the discrete data points, and when the number of the discrete data points corresponding to the preset local range is larger, the weaker the isolation of the discrete data points corresponding to the preset local range is, so that the distribution mutation index is smaller. Conversely, when fewer discrete data points are provided, the smaller the data duty ratio is, which means that the more isolated the discrete data points are, the higher the mutation degree isThe greater the index of the distribution mutation.
And combining the deviation degree of the data values and the distribution density condition of the discrete data points to comprehensively obtain mutation evaluation factors of each discrete data point, and preferably, weighting and summing the data mutation indexes of the discrete data points and the distribution mutation indexes to obtain the mutation evaluation factors of the discrete data points. In the embodiment of the invention, the specific expression of the mutation evaluation factor is:
in the method, in the process of the invention,denoted as +.>Mutation assessment factor of discrete data points, +.>Denoted as +.>Data mutation index of discrete data points, +.>Denoted as +.>Mutation index of the distribution of discrete data points, +.>And->Expressed as adjusting weights, the confidence of the consideration for numerical and dense cases is different, in the present embodiment,/-in>Set to 0.6%>Set to 0.4, the practitioner can adjust to the specific implementation, and is not limited herein.
When the mutation evaluation factor is larger, the mutation degree of the corresponding discrete data point is larger, and the error point is more likely, otherwise, when the mutation evaluation factor is smaller, the mutation degree of the corresponding discrete data point is smaller, the data distribution in the local range is more stable, and the discrete data point is more likely to be an abnormal point.
Thus, the analysis of all data points is completed, and a discrete evaluation factor of each data point and a mutation evaluation factor of the discrete data point are obtained.
The fitting module 103 is configured to update the fitting weights of all the data points in the preset fitting window according to the discrete evaluation factors of the data points and the mutation evaluation factors of the discrete data points; fitting the seasonal term by a local weighted regression algorithm to obtain a fitting trend term.
Firstly, the discrete data points can be divided according to the mutation evaluation factors of the discrete data points, for abnormal points in the temperature time sequence data, the weight in the fitting needs to be increased, so that the abnormal information can be well represented, for error points in the temperature time sequence data, the weight in the fitting needs to be reduced, the influence of errors on the fitting is avoided, and the fitting accuracy is improved.
In one embodiment of the present invention, when the mutation evaluation factor of the discrete data points is greater than or equal to the preset mutation threshold, the corresponding discrete data points are indicated as error points, and the weight adjustment value of the corresponding discrete data points is set as the preset first adjustment value. When the mutation evaluation factor of the discrete data points is smaller than the preset mutation threshold value, the corresponding discrete data points are indicated to be abnormal points, and the weight adjustment value of the corresponding discrete data points is set to be a preset second adjustment value. And setting the weight adjustment value of the non-discrete data point as a preset third adjustment value, wherein the non-discrete data point is a normal point, namely a point at which no abnormal condition occurs. Wherein the preset mutation threshold value is 0.7, and the implementer can adjust the mutation threshold value according to specific conditions.
In the embodiment of the method, the preset first adjustment value is smaller than the preset third adjustment value, and the preset third adjustment value is smaller than the preset second adjustment value, wherein the fitting weight of the normal point is not adjusted, so that the preset third adjustment value is 0. For the adjustment of error points, the fitting weight is smaller, and the consideration is smaller in the fitting process, so that the confidence of the fitting weight is lower than that of normal fitting weight, and the preset first adjustment value is-0.3. For the adjustment of the abnormal points, the fitting weight is larger, more abnormal conditions need to be considered for preservation during fitting, so that the confidence of the fitting weight is higher than that of normal fitting, the preset second adjustment value is 0.3, and a specific numerical value implementation person can adjust according to specific implementation conditions.
Finally, according to the weight adjustment value of each data point in the preset fit window, the fit weight of each data point in the local weighted regression algorithm is updated, and preferably, in the embodiment of the invention, the fit weight of each data point in the Loess algorithm is added with the corresponding weight adjustment value, so as to obtain the updated fit weight of each data point.
Fitting the seasonal term through a Loess algorithm after self-adaptively updating the fitting weight to obtain a fitting trend term, and adjusting the Loess algorithm to ensure that the obtained fitting trend term is less influenced by error points, so that the obtained decomposition term is more accurate when STL decomposition is continued subsequently. It should be noted that, the method of obtaining the fitting weight of each data point by the Loess algorithm is a well-known technique known to those skilled in the art, for example, a method of using a trigonometric kernel function, etc., which will not be described herein.
The state detection module 104 is configured to adjust a season term in STL decomposition according to the fitting trend term to obtain a modified season term corresponding to the temperature time sequence data; and detecting the running state of the hanging white block production equipment through fitting the trend item and correcting the season item.
According to the fitting trend term obtained by the fitting module 103, the season term obtained by STL decomposition is adjusted, and it is to be noted that the main process is already described in the data obtaining module 101, which is not described in detail herein, the fitting trend term can be finally obtained, the season term and the residual term are corrected, the long-term trend of the temperature in the device along with the change of the reaction time length is reflected by the fitting trend term, the fluctuation periodic state of the temperature in the device can be reflected by the corrected season term, and the residual term is represented as the residual term of irregular fluctuation. Referring to fig. 3, a schematic diagram of a trend term of STL decomposition according to an embodiment of the present invention is shown, where in fig. 3, the abscissa indicates time and the ordinate indicates a trend value of temperature. Referring to fig. 4, a schematic diagram of an STL decomposition modified season term according to an embodiment of the present invention is shown, wherein in fig. 4, the abscissa indicates time and the ordinate indicates a period change degree value. It should be noted that, the temperature trend value and the periodic variation degree value are both a measurement standard value of the temperature data, and may be represented by a differential value or a percentage value, which are well known to those skilled in the art, and are not described herein. Because the temperature change of the hanging white block production equipment in the preparation process is relatively regular, the fitting trend term and the correction season term obtained by decomposition have regular periodic changes.
In one embodiment of the invention, the outliers in the fitting trend term are obtained according to the Laida criterion, the number of outliers is counted to obtain the number of trend outliers, the outliers in the correction season term are obtained according to the Laida criterion, and the number of season outliers is counted to obtain the number of season outliers. Because the fitting trend term and the correction season term are curves with regular periodic changes and are similar to normal distribution, abnormal value conditions of the fitting trend term and the correction season term are detected through the Laida criterion, abnormal degree of the regular changes is reflected through statistics of abnormal value quantity, and it is required to be noted that the Laida criterion is a technical means well known to a person skilled in the art, and details are omitted here.
Preferably, the preset abnormal threshold is set to 3, a specific numerical value implementation can regulate and control by oneself, when the number of the trend abnormal values and the number of the season abnormal values are smaller than the preset abnormal threshold, the temperature data are stable, the temperature change is normal, and the running state of the suspended white block production equipment is recorded as normal. When the number of the trend abnormal values or the number of the season abnormal values is larger than or equal to a preset abnormal threshold, the change of the temperature data is changed, and the abnormal condition possibly occurs in the production and preparation process, the running state of the hanging white block production equipment is marked as abnormal, and relevant staff is warned.
In summary, STL decomposition is performed on temperature time sequence data, a seasonal term is obtained in the decomposition process, in the fitting process of the seasonal term, the sensitivity problem of the fitting weight of a local weighted regression algorithm to abnormal data is considered, the fitting weight of the abnormal data is adjusted, and firstly, discrete data points with abnormal conditions are screened out according to the fluctuation deviation degree of the data points in a preset fitting window. In the production and preparation process of the hanging white block, the abnormal condition of temperature data consists of two conditions of abnormal points and error points, and fitting sensitivity required by the error points and the abnormal points is different, so that the mutation evaluation factor of each discrete data point is obtained according to the characteristic that the error points have mutation isolation and in the preset local range of each discrete data point through the two aspects of the data value deviation degree and the distribution density condition of the discrete data point. According to the mutation evaluation factors, the discrete data points can be represented as error points or abnormal points, and then the fitting weight of the local weighted regression algorithm is updated, so that the fitting of the seasonal items is completed to obtain fitting trend items, and the fitting trend items are more accurate. Finally, the season term of STL decomposition is adjusted according to the fitting trend term, a correction season term is obtained, the abnormal condition in the temperature data is fully considered, the fitting weight of the abnormal data in the STL decomposition is adjusted, the obtained fitting trend term and correction season term are more accurate and reliable, and the effect of detecting the running state of the hanging white block production equipment through the trend term and the correction season term is better and more reliable.
It should be noted that: the sequence of the embodiments of the present invention is only for description, and does not represent the advantages and disadvantages of the embodiments. The processes depicted in the accompanying drawings 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.
Claims (10)
1. A white hanging block production facility running state detection system, the system comprising:
the data acquisition module is used for acquiring temperature time sequence data in the operation of the hanging white block production equipment and acquiring seasonal items of the temperature time sequence data through STL decomposition;
the evaluation factor extraction module is used for determining discrete data points in all data points according to the fluctuation deviation degree of each data point on the out-of-season term in a preset fitting window; obtaining mutation evaluation factors of each discrete data point according to the data value offset degree between the discrete data point and other data points and the distribution density condition of the discrete data points in a preset local range of each discrete data point;
the fitting module is used for updating the fitting weights of all the data points in a preset fitting window according to the mutation evaluation factors of the discrete data points; fitting the seasonal items through a local weighted regression algorithm to obtain fitting trend items;
the state detection module is used for adjusting the season term in STL decomposition according to the fitting trend term to obtain a corrected season term corresponding to the temperature time sequence data; and detecting the running state of the hanging white block production equipment through fitting the trend item and correcting the season item.
2. A hanging white block production facility operational status detection system in accordance with claim 1 wherein said determining discrete ones of all data points based on the degree of fluctuating deviations of each data point on the out-of-season term comprises:
in a preset fitting window, acquiring a connecting line between each data point and the previous data point on the out-of-season item, and taking an included angle between the connecting line and a horizontal line as the fluctuation degree of each data point; taking the fluctuation degree of the second data point in the preset fitting window as the fluctuation degree of the first data point;
calculating the difference between the fluctuation degree of each data point and the average value of all fluctuation degrees in a preset fitting window, and taking the difference after normalization processing as a discrete evaluation factor of each data point; discrete ones of the data points are determined based on the discrete evaluation factors of the data points.
3. The system for detecting the operation state of hanging white block production equipment according to claim 1, wherein the method for acquiring the mutation evaluation factor comprises the following steps:
calculating the data value difference between any one discrete data point and other data points in the preset local range of the discrete data point; calculating the average value of all data value differences corresponding to the discrete data points and carrying out normalization processing to obtain the data mutation index of the discrete data points;
carrying out negative correlation mapping and normalization processing on the quantity proportion of all discrete data points in the preset local range corresponding to each discrete data point to obtain a distribution mutation index of each discrete data point;
and carrying out weighted summation on the data mutation index and the distribution mutation index of each discrete data point to obtain a mutation evaluation factor of each discrete data point.
4. The system for detecting the running state of a hanging white block production device according to claim 1, wherein updating the fitting weights of all data points in a preset fitting window by a local weighted regression algorithm according to the mutation evaluation factor of the discrete data points comprises:
when the mutation evaluation factor of the discrete data points is larger than or equal to a preset mutation threshold value, setting the weight adjustment value of the corresponding discrete data points as a preset first adjustment value; when the mutation evaluation factor of the discrete data points is smaller than a preset mutation threshold value, setting the weight adjustment value of the corresponding discrete data points as a preset second adjustment value;
setting the weight adjustment value of the non-discrete data point as a preset third adjustment value; the preset first adjustment value is smaller than the preset third adjustment value, and the preset third adjustment value is smaller than the preset second adjustment value;
and updating the fitting weight of each data point in the local weighted regression algorithm according to the weight adjustment value of each data point in the preset fitting window.
5. The system for detecting the operation state of a hanging white block production device according to claim 4, wherein updating the fitting weight of each data point in the local weighted regression algorithm according to the weight adjustment value of each data point in the preset fitting window comprises:
and adding the fitting weight of each data point in the local weighted regression algorithm to the corresponding weight adjustment value to obtain the updated fitting weight of each data point.
6. The system for detecting the operation state of the hanging white block producing apparatus according to claim 1, wherein the detecting the operation state of the hanging white block producing apparatus by fitting a trend term and a modified season term comprises:
obtaining abnormal values in the fitting trend item according to the Laida criterion, and counting the number of the abnormal values to obtain the number of the trend abnormal values; obtaining abnormal values in the corrected season items according to the Laida criterion, and counting the number of the abnormal values to obtain the number of the season abnormal values;
when the number of the trend abnormal values and the number of the season abnormal values are smaller than a preset abnormal threshold value, the running state of the hanging white block production equipment is recorded as normal; when the number of the trend abnormal values or the number of the season abnormal values is larger than or equal to a preset abnormal threshold value, the running state of the hanging white block production equipment is marked as abnormal.
7. The system according to claim 1, wherein the seasonal term for obtaining temperature time series data by STL decomposition comprises:
in the time sequence data of the decomposition temperature through STL, fitting the time sequence data of the temperature to obtain a season term; subtracting the seasonal term from the temperature time sequence data to obtain a seasonal term.
8. A hanging white block production facility operational status detection system in accordance with claim 2 wherein said determining discrete ones of the data points based on discrete evaluation factors for the data points comprises:
and taking the data points with the discrete evaluation factors being greater than or equal to a preset discrete threshold value as discrete data points.
9. The system for detecting the running state of hanging white block production equipment according to claim 1, wherein the preset local range is less than or equal to half of the size of a preset fitting window.
10. The system for detecting the operation state of hanging white block producing equipment according to claim 7, wherein the fitting method of the seasonal term is a moving average method.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202311047135.9A CN116756529B (en) | 2023-08-21 | 2023-08-21 | White piece production facility running state detecting system hangs |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202311047135.9A CN116756529B (en) | 2023-08-21 | 2023-08-21 | White piece production facility running state detecting system hangs |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN116756529A CN116756529A (en) | 2023-09-15 |
| CN116756529B true CN116756529B (en) | 2023-10-20 |
Family
ID=87948218
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202311047135.9A Active CN116756529B (en) | 2023-08-21 | 2023-08-21 | White piece production facility running state detecting system hangs |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN116756529B (en) |
Families Citing this family (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN117057517B (en) * | 2023-10-12 | 2024-01-12 | 国网吉林省电力有限公司长春供电公司 | Efficient processing method and system for electric power data based on digital twin |
| CN117235678B (en) * | 2023-11-14 | 2024-01-26 | 山东交通学院 | Method for detecting blade abrasion degree in wood cutting process |
| CN117556386B (en) * | 2024-01-12 | 2024-04-05 | 济南海德热工有限公司 | Temperature data monitoring method in furnace refining process |
| CN117591542B (en) * | 2024-01-18 | 2024-03-22 | 准检河北检测技术服务有限公司 | Intelligent detection method for database software data security |
| CN117931094A (en) * | 2024-03-21 | 2024-04-26 | 山东奥斯瑞特检验检测有限公司 | Block chain-based reliable storage method for ambient air monitoring data |
Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US6745150B1 (en) * | 2000-09-25 | 2004-06-01 | Group 1 Software, Inc. | Time series analysis and forecasting program |
| CN104090280A (en) * | 2014-06-04 | 2014-10-08 | 郑州轻工业学院 | Ionosphere delay correction forecasting method based on regional CORS |
| CN111428908A (en) * | 2020-02-21 | 2020-07-17 | 上海海洋大学 | Sea surface temperature prediction algorithm based on ST L-NN model |
| CN113792366A (en) * | 2021-09-07 | 2021-12-14 | 上海交通大学 | Tunnel convergence deformation prediction method based on STL decomposition |
| CN114091768A (en) * | 2021-11-25 | 2022-02-25 | 安徽大学 | STL (Standard template library) and LSTM (local Scale TM) with attention mechanism based tourism demand prediction method |
| CN115587666A (en) * | 2022-11-04 | 2023-01-10 | 元始智能科技(南通)有限公司 | Load prediction method and system based on seasonal trend decomposition and hybrid neural network |
Family Cites Families (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US11567914B2 (en) * | 2018-09-14 | 2023-01-31 | Verint Americas Inc. | Framework and method for the automated determination of classes and anomaly detection methods for time series |
| US11579951B2 (en) * | 2018-09-27 | 2023-02-14 | Oracle International Corporation | Disk drive failure prediction with neural networks |
| US20210326235A1 (en) * | 2020-04-01 | 2021-10-21 | Zoho Corporation Private Limited | Time series forecasting |
-
2023
- 2023-08-21 CN CN202311047135.9A patent/CN116756529B/en active Active
Patent Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US6745150B1 (en) * | 2000-09-25 | 2004-06-01 | Group 1 Software, Inc. | Time series analysis and forecasting program |
| CN104090280A (en) * | 2014-06-04 | 2014-10-08 | 郑州轻工业学院 | Ionosphere delay correction forecasting method based on regional CORS |
| CN111428908A (en) * | 2020-02-21 | 2020-07-17 | 上海海洋大学 | Sea surface temperature prediction algorithm based on ST L-NN model |
| CN113792366A (en) * | 2021-09-07 | 2021-12-14 | 上海交通大学 | Tunnel convergence deformation prediction method based on STL decomposition |
| CN114091768A (en) * | 2021-11-25 | 2022-02-25 | 安徽大学 | STL (Standard template library) and LSTM (local Scale TM) with attention mechanism based tourism demand prediction method |
| CN115587666A (en) * | 2022-11-04 | 2023-01-10 | 元始智能科技(南通)有限公司 | Load prediction method and system based on seasonal trend decomposition and hybrid neural network |
Non-Patent Citations (3)
| Title |
|---|
| P. A. Pickers and A. C. Manning.Investigating bias in the application of curve fitting programs to atmospheric time series.《Atmos. Meas. Tech.》.2015,全文. * |
| 贺琪 等.基于STL的海表面温度预测算法.《海洋环境科学》.2020,全文. * |
| 郭锦桥 等.基于时间序列分解的信号挖掘与预测.《太赫兹科学与电子信息学报》.2023,全文. * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN116756529A (en) | 2023-09-15 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN116756529B (en) | White piece production facility running state detecting system hangs | |
| WO2021179572A1 (en) | Operation and maintenance system anomaly index detection model optimization method and apparatus, and storage medium | |
| CN109459993A (en) | A kind of process flow industry process online adaptive Fault monitoring and diagnosis method | |
| CN107137093B (en) | A kind of continuous blood sugar monitoring device comprising abnormal plasma glucose probability alarm | |
| JP6442398B2 (en) | Method and system for detecting microbial growth in a sample container | |
| CN116659589A (en) | Donkey-hide gelatin cake preservation environment monitoring method based on data analysis | |
| CN117196353B (en) | Environmental pollution assessment and monitoring method and system based on big data | |
| CN104200113A (en) | Internet of Things data uncertainty measurement, prediction and outlier-removing method based on Gaussian process | |
| CN116881745B (en) | Pressure transmitter abnormality monitoring method based on big data | |
| CN115931055B (en) | Rural water supply operation diagnosis method and system based on big data analysis | |
| CN104921736A (en) | Continuous blood glucose monitoring device comprising parameter estimation function filtering module | |
| CN117094611B (en) | Quality safety traceability management method and system for food processing | |
| CN114970688A (en) | Landslide monitoring data preprocessing method based on LSTMAD algorithm and Hermite interpolation method | |
| CN115167323A (en) | Industrial control equipment feedback information instruction transmission system based on digital factory | |
| CN117278150B (en) | Indoor wireless network signal measurement and calculation method, equipment and medium | |
| CN111542010A (en) | WSN data fusion method based on classification adaptive estimation weighting fusion algorithm | |
| CN116881646B (en) | Stevioside production monitoring data management system | |
| CN117349664B (en) | On-line monitoring method and system for sprouting vegetable growth environment | |
| CN109325065A (en) | Multi-sampling rate flexible measurement method based on dynamic latent variable model | |
| CN116204805B (en) | Micro-pressure oxygen cabin and data management system | |
| Alvarez et al. | Batch process monitoring in the original measurement's space | |
| CN113449789A (en) | Quality control method for monitoring water quality by full-spectrum water quality monitoring equipment based on big data | |
| CN114557690A (en) | Calibration system and method for continuous glucose monitoring | |
| CN110377007A (en) | Fault Diagnosis for Chemical Process method based on pivot analysis | |
| CN117224088B (en) | User sleep quality management system and method based on detection data analysis |
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 |