CN116701371A - Method and device for interpolating missing values of atmospheric temperature data under covariance analysis - Google Patents
Method and device for interpolating missing values of atmospheric temperature data under covariance analysis Download PDFInfo
- Publication number
- CN116701371A CN116701371A CN202310682985.XA CN202310682985A CN116701371A CN 116701371 A CN116701371 A CN 116701371A CN 202310682985 A CN202310682985 A CN 202310682985A CN 116701371 A CN116701371 A CN 116701371A
- Authority
- CN
- China
- Prior art keywords
- temperature data
- atmospheric temperature
- data
- missing
- station
- 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
- 238000000034 method Methods 0.000 title claims abstract description 39
- 238000004458 analytical method Methods 0.000 title claims abstract description 34
- 239000011159 matrix material Substances 0.000 claims abstract description 42
- 238000012360 testing method Methods 0.000 claims abstract description 34
- 239000013598 vector Substances 0.000 claims abstract description 21
- 238000004519 manufacturing process Methods 0.000 claims abstract description 7
- 238000004364 calculation method Methods 0.000 claims description 11
- 238000012545 processing Methods 0.000 claims description 11
- 238000001556 precipitation Methods 0.000 claims description 7
- 238000006243 chemical reaction Methods 0.000 claims description 6
- 238000012217 deletion Methods 0.000 claims description 6
- 230000037430 deletion Effects 0.000 claims description 6
- 238000012544 monitoring process Methods 0.000 claims description 5
- 238000000342 Monte Carlo simulation Methods 0.000 claims description 3
- 230000008859 change Effects 0.000 claims description 3
- 230000008676 import Effects 0.000 claims description 2
- 238000012549 training Methods 0.000 abstract description 4
- 238000013179 statistical model Methods 0.000 abstract description 2
- 238000010276 construction Methods 0.000 description 2
- 238000010586 diagram Methods 0.000 description 2
- 238000010801 machine learning Methods 0.000 description 2
- 230000035945 sensitivity Effects 0.000 description 2
- 238000000528 statistical test Methods 0.000 description 2
- 238000013136 deep learning model Methods 0.000 description 1
- 230000007547 defect Effects 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 230000007246 mechanism Effects 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 230000035772 mutation Effects 0.000 description 1
- 238000004088 simulation Methods 0.000 description 1
- 230000009469 supplementation Effects 0.000 description 1
- XLYOFNOQVPJJNP-UHFFFAOYSA-N water Chemical compound O XLYOFNOQVPJJNP-UHFFFAOYSA-N 0.000 description 1
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/20—Information retrieval; Database structures therefor; File system structures therefor of structured data, e.g. relational data
- G06F16/21—Design, administration or maintenance of databases
- G06F16/215—Improving data quality; Data cleansing, e.g. de-duplication, removing invalid entries or correcting typographical errors
-
- 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
- G06F18/21—Design or setup of recognition systems or techniques; Extraction of features in feature space; Blind source separation
- G06F18/214—Generating training patterns; Bootstrap methods, e.g. bagging or boosting
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/27—Regression, e.g. linear or logistic regression
-
- 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
- Y02A—TECHNOLOGIES FOR ADAPTATION TO CLIMATE CHANGE
- Y02A90/00—Technologies having an indirect contribution to adaptation to climate change
- Y02A90/10—Information and communication technologies [ICT] supporting adaptation to climate change, e.g. for weather forecasting or climate simulation
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Data Mining & Analysis (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- General Engineering & Computer Science (AREA)
- Evolutionary Biology (AREA)
- Databases & Information Systems (AREA)
- Life Sciences & Earth Sciences (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Bioinformatics & Computational Biology (AREA)
- Mathematical Physics (AREA)
- Computational Mathematics (AREA)
- Mathematical Analysis (AREA)
- Mathematical Optimization (AREA)
- Evolutionary Computation (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Pure & Applied Mathematics (AREA)
- Artificial Intelligence (AREA)
- Operations Research (AREA)
- Software Systems (AREA)
- Algebra (AREA)
- Probability & Statistics with Applications (AREA)
- Quality & Reliability (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
- Complex Calculations (AREA)
Abstract
The application provides an atmospheric temperature data missing value interpolation method under covariance analysis, which comprises the following steps: constructing an influence factor of the atmospheric temperature data, manufacturing an atmospheric temperature data set, building a multiple regression model of a plurality of station positions, traversing and searching for an atmospheric temperature data missing data set, building an atmospheric temperature data sequence set, calculating a covariance matrix of temperature data reference sequence data, solving a variance weight vector, building a variance model of a plurality of station positions, building an atmospheric temperature data missing value interpolation model based on covariance analysis, and circularly interpolating the atmospheric temperature data missing value; and provides an interpolation device based on the same. The method synthesizes two large statistical models of multiple regression and covariance, introduces a plurality of influencing factors, and fully considers the correlation in time and space; and the atmospheric temperature data of different stations are divided into a missing sequence, a reference sequence and a test sequence for training, testing and missing value interpolation, so that the reliability and accuracy of data interpolation are improved.
Description
Technical Field
The application belongs to the technical field of interpolation of missing values of meteorological elements, in particular to an interpolation method and an interpolation device of missing values of atmospheric temperature data under covariance analysis.
Background
In order to monitor the atmospheric temperature data in real time, more and more atmospheric temperature monitoring stations are generally provided to obtain more accurate and complete atmospheric temperature data. The complete high-precision temperature observation data is an important input parameter for agricultural meteorological disaster monitoring and ecosystem simulation. Because of the limitation of field meteorological conditions, the phenomenon of meteorological data missing is normal, and the data interpolation method is a necessary processing step for meteorological data application.
The interpolation method of the atmospheric temperature data missing value commonly used at present includes a regression-based method, a geostatistical method and a machine learning method. However, in interpolating the atmospheric temperature data loss value, there is a significant disadvantage in accuracy of the above-described several methods: regression-based temperature data interpolation methods may generate over-fitting problems, and geostatistical methods cannot constrain the minimum estimation error, while machine learning methods, while being able to construct new temperature missing value interpolation deep learning models, perform high-precision interpolation on missing half-hour temperature observation data, require training through a large amount of atmospheric temperature data and lack clear mechanism meaning.
Patent CN113495913a discloses a method and device for interpolating air quality data missing values, which calculates covariance and weight based on a first air quality data sequence, a second air quality data sequence and a third air quality data sequence, calculates air quality data corresponding to a first time point in the first air quality data sequence according to a plurality of third air quality data sequences and corresponding weights, thereby completing the supplementation of the air quality data missing values, and calculates missing values based on a historical dataset by a covariance calculation weight method without considering the influence of other influencing parameters on interpolation results, so that the accuracy of interpolation cannot be ensured.
Therefore, in order to solve the problem, it is urgent and necessary to introduce a method for interpolating the atmospheric temperature data missing value under covariance analysis to improve the accuracy of the interpolation result of the atmospheric temperature missing value so as to ensure the accuracy and perfection of the application of the meteorological data.
Disclosure of Invention
The application provides an atmospheric temperature data missing value interpolation method and an interpolation device thereof under covariance analysis aiming at the defects in the prior art.
The method synthesizes a multiple regression model and a covariance model, introduces influencing factors such as time, atmospheric pressure, maximum wind speed, relative humidity, water vapor pressure, precipitation, reference temperature and the like, and fully considers the correlation in time and space; and the atmospheric temperature data of different stations are divided into a missing sequence, a reference sequence and a test sequence, so that the reliability and the statistical test sensitivity of the interpolated data are improved.
The application provides an atmospheric temperature data missing value interpolation method under covariance analysis, which comprises the following steps:
s1, constructing an influence factor of atmospheric temperature data;
the influence factors of the atmospheric temperature data are determined according to the climate change theory and the traditional temperature prediction scale, and are selected from the atmospheric temperature data set, and the influence factors comprise: time x 1 Atmospheric pressure x 2 Maximum wind speed x 3 Relative humidity x 4 Steam pressure x 5 Precipitation x 6 Reference temperature x 7 Seven influencing factors are added.
S2, making an atmospheric temperature data set and establishing a multiple regression model of a plurality of station positions;
the multiple regression model of the plurality of station locations is represented in matrix form as follows:
Y b ＝Xw (1)；
s3, traversing to find an atmospheric temperature data missing data set;
according to the Monte Carlo method, traversing all the atmospheric temperature data and station position information of the atmospheric temperature data set, finding out stations with all the atmospheric temperature data in the atmospheric temperature data set missing as stations to be interpolated, and extracting all the missing moments of the atmospheric temperature data of each station to be interpolated and the corresponding station position information;
s4, establishing an atmospheric temperature data sequence set for each station to be interpolated;
s41, combining the moments corresponding to the plurality of atmosphere temperature data which are continuously deleted to form a temperature data deletion sequence, wherein the initial moment of the specified temperature data deletion sequence is a first interpolation point;
s42, manufacturing temperature data missing sequence data, temperature data reference sequence data and temperature data test sequence data;
s5, calculating a covariance matrix of the temperature data reference sequence data;
the covariance of all the reference sequence data of any two temperature data is calculated to obtain a covariance matrix as follows
Wherein y is 1 ,y n A single column matrix representing the composition of all temperatures in the 1 st and nth station locations, respectively; c () represents covariance calculation;
s6, solving a variance weight vector, and establishing a variance model of a plurality of station positions;
s61, solving a variance weight vector r;
s62, establishing a variance model of a plurality of station positions through a covariance matrix and a variance weight vector of temperature data reference sequence data as follows:
Y a ＝Zr (3)；
s7, constructing an atmospheric temperature data missing value interpolation model based on covariance analysis;
s71, defining and solving an error remainder d;
s72, establishing an atmospheric temperature data missing value interpolation model of a plurality of station positions through a multiple regression model and a variance model of temperature data reference sequence data, wherein the interpolation model is as follows:
Y＝(Xw+Y 0 (Zr)+d)/2 (4)
wherein Y is 0 (Zr) represents a predicted value obtained based on the variance model;
s8, circularly interpolating the atmospheric temperature data missing value;
substituting the temperature data missing sequence data of all the stations to be interpolated into the interpolation model of the atmospheric temperature data missing values in the step S72, and repeatedly executing the steps S2 to S7 until the atmospheric temperature data in the atmospheric temperature data set is free of missing.
Preferably, the step S2 specifically includes the following steps:
s21, acquiring an atmospheric temperature data set based on a national geographic database and a temperature monitoring station real-time database, and manufacturing a data set containing time x 1 Atmospheric pressure x 2 Maximum wind speed x 3 Relative humidity x 4 Steam pressure x 5 Precipitation x 6 And a reference temperature x 7 Is a set of station location information;
s22, establishing a multiple regression model of the position of a single station in the following basic form:
Y b ＝w 1 x 1 +w 2 x 2 +…+w j x j (5)
wherein Y is b Representing a predicted atmospheric temperature; w (w) j Regression weight coefficients representing the jth influence factor;
s23, aiming at different station positions, performing multiple regression processing on the atmospheric temperature data set in the step S21 by utilizing a MATLAB software multiple regression module to obtain preliminary weight coefficients w of multiple regression models of different station positions j ；
S24, establishing a multiple regression model of a plurality of station positions in the following basic form:
Y ib ＝w 1 x i1 +w 2 x i2 +…+w j x ij (6)
wherein Y is ib A predicted atmospheric temperature representing an ith station location sample; x is x ij A j-th influencing factor representing the i-th station location sample;
preferably, the step S42 specifically includes the following steps:
s421, extracting the atmospheric temperature data of the station to be interpolated in a preset period as temperature data missing sequence data, wherein the missing value is replaced by an atmospheric temperature average value in the preset period;
s422, taking all stations except the station to be interpolated, which have no loss of atmospheric temperature data in the preset period, as reference stations;
s423, respectively extracting 80% of data from the atmospheric temperature data of each reference station in the preset period to obtain a plurality of temperature data reference sequence data, and correspondingly extracting 80% of temperature data missing sequence data to obtain temperature data reference prediction sequence data;
s424, respectively extracting the remaining 20% of data from the atmospheric temperature data of each reference station in the preset period to obtain a plurality of temperature data test sequence data, and correspondingly extracting 20% of temperature data missing sequence data to obtain temperature data test prediction sequence data;
preferably, the step S61 specifically includes the steps of:
s611, calculating covariance of all temperature data reference sequence data and temperature data missing sequence data to obtain a reference covariance matrix Z r The following are provided:
Z r ＝[C(y 1 y 0 )…C(y n y 0 )] T (7)
wherein y is 0 A single column matrix representing all temperatures in the station position to be interpolated; t is the transpose;
s612, defining a variance weight matrix r:
r＝[r 1 …r n ] T (8)
wherein r is 1 ,r n Variance weights of the 1 st and nth temperature data reference sequence data are respectively represented;
s613, based on the covariance matrix and the reference covariance matrix, the solving equation of the variance weight matrix r is as follows:
preferably, the step S71 specifically includes the steps of:
s711, defining error vectorWherein d 1 ,d n The 1 st and nth atmospheric temperature data loss value interpolation error values are shown, respectively.
S712, substituting the temperature data test sequence data into the variance model to obtain a temperature data test prediction sequence data predicted value, and comparing the temperature data test prediction sequence data predicted value with a true value behind the temperature data test prediction sequence data to obtain an error vector d;
s713, taking the average value of the error vector d by the error remainder d;
preferably, the step S21 specifically includes the following steps:
s211, collecting the atmospheric temperature data of all stations hour by hour, and constructing an atmospheric temperature data set containing station position information;
s212, converting the position information of all stations in the atmospheric temperature data set by utilizing coordinate conversion, and constructing an atmospheric temperature data layer on a map.
Preferably, there are 1 or more stations to be interpolated in the step S4, 1 or more first interpolation points in the stations to be interpolated in the step S41, and not less than 2 reference stations in the step S42.
Preferably, the data missing value interpolation model based on covariance analysis in step S7 is also applicable to other meteorological element missing value interpolation technical fields.
In another aspect of the present application, an interpolation device using the foregoing method for interpolating missing values of atmospheric temperature data under covariance analysis is provided, which includes an acquisition and display module, a search module, a processing module, and a calculation module, where the acquisition and display module acquires atmospheric temperature data of all stations from hour to hour and introduces the position information of all stations as a layer into ArcGIS software through coordinate conversion, and displays an atmospheric temperature data layer established on a map; the searching module searches all reference stations through ArcGIS software to establish an atmospheric temperature data sequence set, and traverses to search an atmospheric temperature data missing data set; the calculation module builds an atmospheric temperature data missing value interpolation model based on covariance analysis by building a multiple regression model of a plurality of station positions and a covariance matrix of temperature data reference sequence data; the processing module circularly interpolates the missing value of the atmospheric temperature data until the atmospheric temperature data in the atmospheric temperature data set is free of missing.
Compared with the prior art, the application has the technical effects that:
1. according to the atmospheric temperature data missing value interpolation method under covariance analysis, a multiple regression model and a covariance model in mathematical statistics are combined, an atmospheric temperature data missing value interpolation model based on covariance analysis is constructed, atmospheric temperature data of different stations are divided into missing sequences, reference sequences and test sequences, so that training, testing and missing value interpolation of the atmospheric temperature data are completed circularly, reliability and statistical test sensitivity of the interpolated data are improved, and error of atmospheric temperature data missing value interpolation is reduced.
2. According to the atmospheric temperature data missing value interpolation method under covariance analysis, influence factors such as seasons, weather, time and reference temperature are introduced, an atmospheric temperature data set is manufactured, multiple regression models of a plurality of station positions are built, and MATLAB software multiple regression modules are utilized to conduct multiple regression processing on the atmospheric temperature data set, so that preliminary weight coefficients of the multiple regression models of different station positions are obtained, the mutation influence of discontinuous variables on the atmospheric temperature data is weakened, and the stability, accuracy and efficiency of data interpolation are greatly improved.
3. According to the atmospheric temperature data missing value interpolation method under covariance analysis, an atmospheric temperature data sequence set is established for each station to be interpolated, a covariance matrix of temperature data reference sequence data is calculated, variance weight vectors are solved, variance models of a plurality of station positions are established, correlation of atmospheric temperature data in time and space is fully considered, and reliability and accuracy of data interpolation are improved.
Drawings
Other features, objects and advantages of the present application will become more apparent upon reading of the detailed description of non-limiting embodiments made with reference to the following drawings.
FIG. 1 is a flow chart of a method for interpolating missing values of atmospheric temperature data under covariance analysis according to the application;
FIG. 2a is a graph comparing predicted atmospheric temperature with true atmospheric temperature curves obtained by a multiple regression model of station number 1 in an embodiment of the application;
FIG. 2b is a graph comparing predicted atmospheric temperature with real atmospheric temperature curves obtained by a multiple regression model of station number 2 in an embodiment of the application;
FIG. 2c is a graph comparing predicted atmospheric temperature with true atmospheric temperature using a multiple regression model for station number 3 in accordance with an embodiment of the present application;
FIG. 2d is a graph comparing predicted atmospheric temperature with real atmospheric temperature curves obtained by a multiple regression model of station number 4 in an embodiment of the present application;
FIG. 2e is a graph comparing predicted atmospheric temperature with true atmospheric temperature using a multiple regression model for station number 5 in an embodiment of the present application;
FIG. 2f is a graph comparing predicted atmospheric temperature with true atmospheric temperature curves obtained by a multiple regression model of station number 6 in an embodiment of the application;
FIG. 3 is a schematic diagram of an atmospheric temperature data sequence set construction strategy of the present application;
FIG. 4 is a schematic diagram of an interpolation device according to the present application;
fig. 5 is a map of the atmospheric temperature data set of the present application.
Detailed Description
The application is described in further detail below with reference to the drawings and examples. It is to be understood that the specific embodiments described herein are merely illustrative of the application and are not limiting of the application. It should be noted that, for convenience of description, only the portions related to the present application are shown in the drawings.
It should be noted that, without conflict, the embodiments of the present application and features of the embodiments may be combined with each other. The application will be described in detail below with reference to the drawings in connection with embodiments.
Fig. 1 shows an atmospheric temperature data missing value interpolation method under covariance analysis according to the present application, which comprises the steps of:
s1, constructing an influence factor of the atmospheric temperature data.
The influence factors of the atmospheric temperature data are determined according to the climate change theory and the traditional temperature prediction scale, and are selected from the atmospheric temperature data set, and the influence factors comprise: time x 1 Atmospheric pressure x 2 Maximum wind speed x 3 Relative humidity x 4 Steam pressure x 5 Precipitation x 6 Reference temperature x 7 Seven influencing factors are added.
S2, making an atmospheric temperature data set and establishing a multiple regression model of a plurality of station positions.
S21, acquiring an atmospheric temperature data set based on a national geographic database and a temperature monitoring station real-time database, and manufacturing a data set containing time x 1 Atmospheric pressure x 2 Maximum wind speed x 3 Relative humidity x 4 Steam pressure x 5 Precipitation x 6 And a reference temperature x 7 Is provided, and a station location information set.
S211, collecting the atmospheric temperature data of all stations hour by hour, and constructing an atmospheric temperature data set containing station position information.
S212, converting the position information of all stations in the atmospheric temperature data set by using coordinate conversion, and constructing an atmospheric temperature data layer on a map.
S22, establishing a multiple regression model of the position of a single station in the following basic form:
Y b ＝w 1 x 1 +w 2 x 2 +…+w j x j (5)
wherein Y is b Representing a predicted atmospheric temperature; w (w) j Regression weight coefficients representing the jth influence factor.
S23, aiming at different station positions, performing multiple regression processing on the atmospheric temperature data set in the step S21 by utilizing a MATLAB software multiple regression module to obtain preliminary weight coefficients w of multiple regression models of different station positions j 。
S24, establishing a multiple regression model of a plurality of station positions in the following basic form:
Y ib ＝w 1 x i1 +w 2 x i2 +…+w j x ij (6)
wherein Y is ib A predicted atmospheric temperature representing an ith station location sample; x is x ij The jth influencing factor representing the ith station location sample.
The multiple regression model of the plurality of station locations is represented in matrix form as follows:
Y b ＝Xw (1)。
in a specific embodiment, the station with the latitude of 44.02 and the longitude of 114.95 is selected as the station 1, the station with the longitude of 119.93 and the latitude of 47.17 is selected as the station 2, the station with the longitude of 120.05 and the latitude of 43.88 of the interior Mongolian A Lu Keer is selected as the station 3, the station with the longitude of 119.92 and the latitude of 42.28 of the interior Mongolian A is selected as the station 4, the station with the longitude of 118.75 and the latitude of 41.52 of the interior Mongolian A is selected as the station 5, and the station with the longitude of 118.65 and the latitude of 43.53 of the interior Mongolian B is selected as the station 6.
Aiming at each item of data from the station 1 to the station 6 from the time of 6 months, 10 days, 0 to the time of 6 months, 10 days, 24, respectively establishing multiple regression models, and solving by Matlab to obtain regression weight coefficients of each influence factor, wherein the multiple regression models are respectively obtained as follows:
the graphs of the predicted atmospheric temperature and the true atmospheric temperature obtained by the multiple regression models of the stations 1 to 6 are shown in fig. 2a to 2f, respectively, so that the prediction results of the multiple regression models are effective.
S3, traversing and searching for an atmospheric temperature data missing data set.
According to the Monte Carlo method, traversing all the atmospheric temperature data and station position information of the atmospheric temperature data set, finding out stations with all the atmospheric temperature data in the atmospheric temperature data set missing as stations to be interpolated, and extracting all missing moments of the atmospheric temperature data of each station to be interpolated and corresponding station position information.
S4, establishing an atmospheric temperature data sequence set for each station to be interpolated, wherein FIG. 3 shows a construction strategy of the atmospheric temperature data sequence set. There are 1 or more stations to be interpolated.
S41, combining the moments corresponding to the plurality of atmosphere temperature data which are continuously deleted to form a temperature data deletion sequence, wherein the initial moment of the specified temperature data deletion sequence is a first interpolation point. There are 1 or more first interpolation points in the station to be interpolated.
S42, manufacturing temperature data missing sequence data, temperature data reference sequence data and temperature data test sequence data.
S421, extracting the atmospheric temperature data of the station to be interpolated in a preset period as temperature data missing sequence data, wherein the missing value is replaced by an atmospheric temperature average value in the preset period.
S422, taking all stations except the station to be interpolated, which have no loss of atmospheric temperature data in a preset period, as reference stations. The reference stations are not less than 2.
S423, extracting 80% of data from the atmospheric temperature data of each reference station in a preset period of time respectively to obtain a plurality of temperature data reference sequence data, and correspondingly extracting 80% of temperature data missing sequence data to obtain temperature data reference prediction sequence data.
S424, extracting the remaining 20% of data from the atmospheric temperature data of each reference station in a preset period of time respectively to obtain a plurality of temperature data test sequence data, and correspondingly extracting 20% of temperature data missing sequence data to obtain temperature data test prediction sequence data.
In one embodiment, assuming that station 2 starts to have air temperature data missing for 3 hours at 6 months and 10 days 0, and the missing data is a temperature data missing sequence, then the first interpolation point is determined at 6 months and 10 days 0.
Extracting atmospheric temperature time sequence data in a preset period (from 6 months, 10 days, 0 time to 6 months, 10 days, 24 time) from all other stations without missing atmospheric temperature data except station No. 2, randomly extracting 80% of the data to obtain 5 temperature data reference sequence data, and correspondingly extracting 80% of the temperature data missing sequence data to obtain 1 temperature data reference predicted sequence data; and extracting the remaining 20% of data to obtain 5 temperature data test sequence data, and correspondingly extracting 20% of temperature data missing sequence data to obtain 1 temperature data test prediction sequence data.
S5, calculating a covariance matrix of the temperature data reference sequence data.
The covariance of the reference sequence data of all any two temperature data is calculated, and a covariance matrix is obtained as follows:
wherein y is 1 ,y n A single column matrix representing the composition of all temperatures in the 1 st and nth station locations, respectively; c () represents covariance calculation.
In one embodiment, the covariance matrix calculated based on the temperature data reference sequence data for stations 1, 3, 4, 5, 6 is:
s6, solving the variance weight vector, and establishing a variance model of the positions of the plurality of stations.
S61, solving a variance weight vector r.
S611, calculating covariance of all temperature data reference sequence data and temperature data missing sequence data to obtain a reference covariance matrix Z r The following are provided:
Z r ＝[C(y 1 y 0 )…C(y n y 0 )] T (7)
wherein y is 0 A single column matrix representing all temperatures in the station position to be interpolated; t is the transpose.
S612, defining a variance weight matrix r:
r＝[r 1 …r n ] T (8)
wherein r is 1 ,r n The variance weights of the 1 st and nth temperature data reference sequence data are represented, respectively.
S613, based on the covariance matrix and the reference covariance matrix, the solving equation of the variance weight matrix r is as follows:
in one embodiment, the calculated reference covariance matrix Z is based on the temperature data reference sequence data of the stations 1, 3, 4, 5, 6 and the temperature data missing sequence data of the station 2 r The method comprises the following steps:
Z r ＝[5.9648 6.9089 6.8233 7.7380 7.7973] T (12)。
and (4) calculating a variance weight matrix r to obtain:
r＝[0.8600 -0.0313 -0.0556 -0.0487 0.4553] T (13)。
s62, establishing a variance model of a plurality of station positions through a covariance matrix and a variance weight vector of temperature data reference sequence data as follows:
Y a ＝Zr (3)。
s7, constructing an atmospheric temperature data missing value interpolation model based on covariance analysis.
S71, defining and solving an error remainder d.
S711, defining error vectorWherein d 1 ,d n The 1 st and nth atmospheric temperature data loss value interpolation error values are shown, respectively.
S712, substituting the temperature data test sequence data into the variance model to obtain a temperature data test prediction sequence data predicted value, and comparing the temperature data test prediction sequence data predicted value with a true value after the temperature data test prediction sequence data to obtain an error vector d.
S713, taking the average value of the error vector d by the error remainder d.
In one embodiment, the covariance matrix calculated based on the temperature data test sequence data for stations 1, 3, 4, 5, 6 is:
at this time, the variance model is:
Y a ＝[1.5945 1.2049 -0.6922 1.9456 0.7322] T (15)。
further, based on the temperature data test sequence data of the stations 1, 3, 4, 5, 6, the temperature data test predicted sequence data predicted value y is obtained by Matlab numerical calculation 20 ：
y 20 ＝[2.8798 2.4898 2.7146 4.2351 6.2836] T (16)。
And comparing the difference with the true value after the temperature data test prediction sequence data to obtain an error vector d:
d＝[0.4202 1.0102 1.4854 0.8649 0.4164] T (17)。
then the error margin d is 0.8394.
S72, establishing an atmospheric temperature data missing value interpolation model of a plurality of station positions through a multiple regression model and a variance model of temperature data reference sequence data, wherein the interpolation model is as follows:
Y＝(Xw+Y 0 (Zr)+d)/2 (4)
wherein Y is 0 (Zr) represents a predicted value obtained based on the variance model.
Furthermore, the data missing value interpolation model based on covariance analysis is also applicable to the technical field of other meteorological element missing value interpolation.
S8, circularly interpolating the atmospheric temperature data missing value.
Substituting the temperature data missing sequence data of all the stations to be interpolated into the interpolation model of the atmospheric temperature data missing values in the step S72, and repeatedly executing the steps S2 to S7 until the atmospheric temperature data in the atmospheric temperature data set is free of missing.
In one embodiment, the predictive value y is derived based on a multiple regression model 21 The method comprises the following steps:
y 21 ＝Xw＝[12.228412.167712.2077] T (18)。
predictive value y obtained based on variance model 22 The method comprises the following steps:
y 22 ＝Y 0 (Zr)＝[9.547012.863614.7894] T (19)。
taking the error remainder into consideration, station No. 2 predicts the air temperature data of 3 hours from day 0 of 6 months to 10 hours as [11.307412.935413.9183 ]] T By means of real air temperature data [1211.312.1 ] starting 3 hours at 0 on day 6 with station No. 2] T By contrast, the error is [ -0.69261.63541.8183] T This error is within (-1, 2), proving the effectiveness of the proposed method.
In another aspect of the present application, an interpolation device using the foregoing method for interpolating missing values of atmospheric temperature data under covariance analysis is provided, as shown in fig. 4, which includes a collection and display module, a search module, a processing module and a calculation module, where the collection and display module collects the atmospheric temperature data of all stations from hour to hour and imports the position information of all stations as one layer into ArcGIS software through coordinate conversion, and displays an atmospheric temperature data layer established on a map; the searching module searches all reference stations through ArcGIS software to establish an atmospheric temperature data sequence set, and traverses to search an atmospheric temperature data missing data set, and the distribution of each atmospheric temperature data sequence set is shown in figure 5; the calculation module builds an atmospheric temperature data missing value interpolation model based on covariance analysis by building a multiple regression model of a plurality of station positions and a covariance matrix of temperature data reference sequence data; the processing module circularly interpolates the missing value of the atmospheric temperature data until the atmospheric temperature data in the atmospheric temperature data set is free of missing.
The application provides an atmospheric temperature data missing value interpolation method under covariance analysis, which integrates two large statistical models of multiple regression and covariance, introduces influence factors such as seasons, weather, time and reference temperature, and fully considers the correlation in time and space; and the atmospheric temperature data of different stations are divided into a missing sequence, a reference sequence and a test sequence for training, testing and missing value interpolation, so that the reliability and accuracy of data interpolation are improved.
Finally, what should be said is: the above embodiments are merely for illustrating the technical aspects of the present application, and it should be understood by those skilled in the art that although the present application has been described in detail with reference to the above embodiments: modifications and equivalents may be made thereto without departing from the spirit and scope of the application, which is intended to be encompassed by the claims.
Claims (9)
1. The method for interpolating the atmospheric temperature data missing value under the covariance analysis is characterized by comprising the following steps of:
s1, constructing an influence factor of atmospheric temperature data;
the influence factors of the atmospheric temperature data are determined according to the climate change theory and the traditional temperature prediction scale, and are selected from the atmospheric temperature data set, and the influence factors comprise: time x 1 Atmospheric pressure x 2 Maximum wind speed x 3 Relative humidity x 4 Steam pressure x 5 Precipitation x 6 Reference temperature x 7 Seven influencing factors are added.
S2, making an atmospheric temperature data set and establishing a multiple regression model of a plurality of station positions;
the multiple regression model of the plurality of station locations is represented in matrix form as follows:
Y b ＝Xw (1)；
s3, traversing to find an atmospheric temperature data missing data set;
according to the Monte Carlo method, traversing all the atmospheric temperature data and station position information of the atmospheric temperature data set, finding out stations with all the atmospheric temperature data in the atmospheric temperature data set missing as stations to be interpolated, and extracting all the missing moments of the atmospheric temperature data of each station to be interpolated and the corresponding station position information;
s4, establishing an atmospheric temperature data sequence set for each station to be interpolated;
s41, combining the moments corresponding to the plurality of atmosphere temperature data which are continuously deleted to form a temperature data deletion sequence, wherein the initial moment of the specified temperature data deletion sequence is a first interpolation point;
s42, manufacturing temperature data missing sequence data, temperature data reference sequence data and temperature data test sequence data;
s5, calculating a covariance matrix of the temperature data reference sequence data;
the covariance of all the reference sequence data of any two temperature data is calculated to obtain a covariance matrix as follows
Wherein y is 1 ,y n A single column matrix representing the composition of all temperatures in the 1 st and nth station locations, respectively; c () represents covariance calculation;
s6, solving a variance weight vector, and establishing a variance model of a plurality of station positions;
s61, solving a variance weight vector r;
s62, establishing a variance model of a plurality of station positions through a covariance matrix and a variance weight vector of temperature data reference sequence data as follows:
Y a ＝Zr (3)；
s7, constructing an atmospheric temperature data missing value interpolation model based on covariance analysis;
s71, defining and solving an error remainder d;
s72, establishing an atmospheric temperature data missing value interpolation model of a plurality of station positions through a multiple regression model and a variance model of temperature data reference sequence data, wherein the interpolation model is as follows:
Y＝(Xw+Y 0 (Zr)+d)/2 (4)
wherein Y is 0 (Zr)Representing a predicted value obtained based on the variance model;
s8, circularly interpolating the atmospheric temperature data missing value;
substituting the temperature data missing sequence data of all the stations to be interpolated into the interpolation model of the atmospheric temperature data missing values in the step S72, and repeatedly executing the steps S2 to S7 until the atmospheric temperature data in the atmospheric temperature data set is free of missing.
2. The method for interpolating an atmospheric temperature data loss value under covariance analysis according to claim 1, wherein said step S2 comprises the steps of:
s21, acquiring an atmospheric temperature data set based on a national geographic database and a temperature monitoring station real-time database, and manufacturing a data set containing time x 1 Atmospheric pressure x 2 Maximum wind speed x 3 Relative humidity x 4 Steam pressure x 5 Precipitation x 6 And a reference temperature x 7 Is a set of station location information;
s22, establishing a multiple regression model of the position of a single station in the following basic form:
Y b ＝w 1 x 1 +w 2 x 2 +…+w j x j (5)
wherein Y is b Representing a predicted atmospheric temperature; w (w) j Regression weight coefficients representing the jth influence factor;
s23, aiming at different station positions, performing multiple regression processing on the atmospheric temperature data set in the step S21 by utilizing a MATLAB software multiple regression module to obtain preliminary weight coefficients w of multiple regression models of different station positions j ；
S24, establishing a multiple regression model of a plurality of station positions in the following basic form:
Y ib ＝w 1 x i1 +w 2 x i2 +…+w j x ij (6)
wherein Y is ib A predicted atmospheric temperature representing an ith station location sample; x is x ij Representing the i-th stationThe j-th influencing factor of the station position sample.
3. The method for interpolating an atmospheric temperature data loss value under covariance analysis according to claim 1, wherein said step S42 comprises the steps of:
s421, extracting the atmospheric temperature data of the station to be interpolated in a preset period as temperature data missing sequence data, wherein the missing value is replaced by an atmospheric temperature average value in the preset period;
s422, taking all stations except the station to be interpolated, which have no loss of atmospheric temperature data in the preset period, as reference stations;
s423, respectively extracting 80% of data from the atmospheric temperature data of each reference station in the preset period to obtain a plurality of temperature data reference sequence data, and correspondingly extracting 80% of temperature data missing sequence data to obtain temperature data reference prediction sequence data;
s424, respectively extracting the remaining 20% of data from the atmospheric temperature data of each reference station in the preset period to obtain a plurality of temperature data test sequence data, and correspondingly extracting 20% of temperature data missing sequence data to obtain temperature data test prediction sequence data.
4. The method for interpolating an atmospheric temperature data loss value under covariance analysis according to claim 1, wherein said step S61 comprises the steps of:
s611, calculating covariance of all temperature data reference sequence data and temperature data missing sequence data to obtain a reference covariance matrix Z r The following are provided:
Z r ＝[C(y 1 y 0 )…C(y n y 0 )] T (7)
wherein y is 0 A single column matrix representing all temperatures in the station position to be interpolated; t is the transpose;
s612, defining a variance weight matrix r:
r＝[r 1 …r n ] T (8)
wherein r is 1 ,r n Variance weights of the 1 st and nth temperature data reference sequence data are respectively represented;
s613, based on the covariance matrix and the reference covariance matrix, the solving equation of the variance weight matrix r is as follows:
5. the method for interpolating an atmospheric temperature data loss value under covariance analysis according to claim 1, wherein said step S71 comprises the steps of:
s711, defining error vectorWherein d 1 ,d n The 1 st and nth atmospheric temperature data loss value interpolation error values are shown, respectively.
S712, substituting the temperature data test sequence data into the variance model to obtain a temperature data test prediction sequence data predicted value, and comparing the temperature data test prediction sequence data predicted value with a true value behind the temperature data test prediction sequence data to obtain an error vector d;
s713, taking the average value of the error vector d by the error remainder d.
6. The method for interpolating an atmospheric temperature data loss value under covariance analysis according to claim 2, wherein said step S21 comprises the steps of:
s211, collecting the atmospheric temperature data of all stations hour by hour, and constructing an atmospheric temperature data set containing station position information;
s212, converting the position information of all stations in the atmospheric temperature data set by utilizing coordinate conversion, and constructing an atmospheric temperature data layer on a map.
7. The method according to claim 1, wherein the number of stations to be interpolated in step S4 is 1 or more, the number of first interpolation points in the stations to be interpolated in step S41 is 1 or more, and the number of reference stations in step S42 is not less than 2.
8. The method for interpolating missing values of atmospheric temperature data under covariance analysis according to claim 1, wherein the data missing value interpolation model based on covariance analysis in step S7 is also applicable to the technical field of interpolation of missing values of other meteorological elements.
9. An interpolation device using the interpolation method of atmospheric temperature data missing values under covariance analysis according to any one of claims 1 to 8, characterized in that it comprises an acquisition and display module, a search module, a processing module and a calculation module, wherein the acquisition and display module acquires the atmospheric temperature data of all stations hour by hour and imports the position information of all stations as one layer into ArcGIS software through coordinate conversion, and displays the atmospheric temperature data layer established on a map; the searching module searches all reference stations through ArcGIS software to establish an atmospheric temperature data sequence set, and traverses to search an atmospheric temperature data missing data set; the calculation module builds an atmospheric temperature data missing value interpolation model based on covariance analysis by building a multiple regression model of a plurality of station positions and a covariance matrix of temperature data reference sequence data; the processing module circularly interpolates the missing value of the atmospheric temperature data until the atmospheric temperature data in the atmospheric temperature data set is free of missing.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202310682985.XA CN116701371B (en) | 2023-06-09 | 2023-06-09 | Method and device for interpolating missing values of atmospheric temperature data under covariance analysis |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202310682985.XA CN116701371B (en) | 2023-06-09 | 2023-06-09 | Method and device for interpolating missing values of atmospheric temperature data under covariance analysis |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN116701371A true CN116701371A (en) | 2023-09-05 |
| CN116701371B CN116701371B (en) | 2024-03-22 |
Family
ID=87827182
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202310682985.XA Active CN116701371B (en) | 2023-06-09 | 2023-06-09 | Method and device for interpolating missing values of atmospheric temperature data under covariance analysis |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN116701371B (en) |
Citations (17)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103473408A (en) * | 2013-08-28 | 2013-12-25 | 河南大学 | Method for restoring missing air temperature records on basis of spatial-temporal information fusion |
| EP2704065A1 (en) * | 2012-08-06 | 2014-03-05 | ATS Group (IP Holdings) Limited | System and method for updating a data structure with sensor measurement data |
| US20150178633A1 (en) * | 2013-12-20 | 2015-06-25 | Johnson Controls Technology Company | Systems and methods for determining the uncertainty in parameters of an energy use model |
| US20160034615A1 (en) * | 2014-08-01 | 2016-02-04 | Tata Consultancy Services Limited | System and method for forecasting a time series data |
| US20160117423A1 (en) * | 2013-03-13 | 2016-04-28 | The Government Of The United States Of America, As Represented By The Secretary Of The Navy | System and method for correcting a model-derived vertical structure of ocean temperature and ocean salinity based on sea surface height observations |
| CN105678046A (en) * | 2014-11-18 | 2016-06-15 | 日本电气株式会社 | Missing data repairing method and device in time-space sequence data |
| US20160291203A1 (en) * | 2015-04-06 | 2016-10-06 | The Government Of The United States Of America, As Represented By The Secretary Of The Navy | Adaptive ecosystem climatology |
| US20170338802A1 (en) * | 2014-12-01 | 2017-11-23 | Harbin Engineering University | Actually-measured marine environment data assimilation method based on sequence recursive filtering three-dimensional variation |
| CN107577649A (en) * | 2017-09-26 | 2018-01-12 | 广州供电局有限公司 | The interpolation processing method and device of missing data |
| US20200073915A1 (en) * | 2018-08-29 | 2020-03-05 | Kabushiki Kaisha Toshiba | Information processing apparatus, information processing system, and information processing method |
| US20200193220A1 (en) * | 2018-12-18 | 2020-06-18 | National Sun Yat-Sen University | Method for data imputation and classification and system for data imputation and classification |
| AU2020104000A4 (en) * | 2020-12-10 | 2021-02-18 | Guangxi University | Short-term Load Forecasting Method Based on TCN and IPSO-LSSVM Combined Model |
| CN113312587A (en) * | 2021-06-16 | 2021-08-27 | 福建中锐网络股份有限公司 | Sensor acquisition data missing value processing method based on ARIMA prediction and regression prediction |
| CN113495913A (en) * | 2021-09-07 | 2021-10-12 | 中国科学院地理科学与资源研究所 | Air quality data missing value interpolation method and device |
| CN114840616A (en) * | 2021-12-28 | 2022-08-02 | 北京航空航天大学 | Dynamic atmospheric natural environment modeling method based on space-time interpolation |
| CN115203625A (en) * | 2022-07-29 | 2022-10-18 | 应急管理部国家减灾中心 | Drought and waterlogging index data missing value interpolation method and device |
| CN115391746A (en) * | 2022-10-28 | 2022-11-25 | 航天宏图信息技术股份有限公司 | Interpolation method, device, electronic device and medium for meteorological element data |
-
2023
- 2023-06-09 CN CN202310682985.XA patent/CN116701371B/en active Active
Patent Citations (17)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| EP2704065A1 (en) * | 2012-08-06 | 2014-03-05 | ATS Group (IP Holdings) Limited | System and method for updating a data structure with sensor measurement data |
| US20160117423A1 (en) * | 2013-03-13 | 2016-04-28 | The Government Of The United States Of America, As Represented By The Secretary Of The Navy | System and method for correcting a model-derived vertical structure of ocean temperature and ocean salinity based on sea surface height observations |
| CN103473408A (en) * | 2013-08-28 | 2013-12-25 | 河南大学 | Method for restoring missing air temperature records on basis of spatial-temporal information fusion |
| US20150178633A1 (en) * | 2013-12-20 | 2015-06-25 | Johnson Controls Technology Company | Systems and methods for determining the uncertainty in parameters of an energy use model |
| US20160034615A1 (en) * | 2014-08-01 | 2016-02-04 | Tata Consultancy Services Limited | System and method for forecasting a time series data |
| CN105678046A (en) * | 2014-11-18 | 2016-06-15 | 日本电气株式会社 | Missing data repairing method and device in time-space sequence data |
| US20170338802A1 (en) * | 2014-12-01 | 2017-11-23 | Harbin Engineering University | Actually-measured marine environment data assimilation method based on sequence recursive filtering three-dimensional variation |
| US20160291203A1 (en) * | 2015-04-06 | 2016-10-06 | The Government Of The United States Of America, As Represented By The Secretary Of The Navy | Adaptive ecosystem climatology |
| CN107577649A (en) * | 2017-09-26 | 2018-01-12 | 广州供电局有限公司 | The interpolation processing method and device of missing data |
| US20200073915A1 (en) * | 2018-08-29 | 2020-03-05 | Kabushiki Kaisha Toshiba | Information processing apparatus, information processing system, and information processing method |
| US20200193220A1 (en) * | 2018-12-18 | 2020-06-18 | National Sun Yat-Sen University | Method for data imputation and classification and system for data imputation and classification |
| AU2020104000A4 (en) * | 2020-12-10 | 2021-02-18 | Guangxi University | Short-term Load Forecasting Method Based on TCN and IPSO-LSSVM Combined Model |
| CN113312587A (en) * | 2021-06-16 | 2021-08-27 | 福建中锐网络股份有限公司 | Sensor acquisition data missing value processing method based on ARIMA prediction and regression prediction |
| CN113495913A (en) * | 2021-09-07 | 2021-10-12 | 中国科学院地理科学与资源研究所 | Air quality data missing value interpolation method and device |
| CN114840616A (en) * | 2021-12-28 | 2022-08-02 | 北京航空航天大学 | Dynamic atmospheric natural environment modeling method based on space-time interpolation |
| CN115203625A (en) * | 2022-07-29 | 2022-10-18 | 应急管理部国家减灾中心 | Drought and waterlogging index data missing value interpolation method and device |
| CN115391746A (en) * | 2022-10-28 | 2022-11-25 | 航天宏图信息技术股份有限公司 | Interpolation method, device, electronic device and medium for meteorological element data |
Non-Patent Citations (1)
| Title |
|---|
| 金志凤;李波;袁德辉;: "基于GIS的浙江省年平均气温空间模拟研究", 气象科学, no. 06, 15 December 2010 (2010-12-15) * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN116701371B (en) | 2024-03-22 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN108701274B (en) | Urban small-scale air quality index prediction method and system | |
| US20160203245A1 (en) | Method for simulating wind field of extreme arid region based on wrf | |
| CN110442911B (en) | High-dimensional complex system uncertainty analysis method based on statistical machine learning | |
| CN114399073A (en) | Ocean surface temperature field prediction method based on deep learning | |
| CN108764527B (en) | Screening method for soil organic carbon library time-space dynamic prediction optimal environment variables | |
| CN110909447A (en) | High-precision short-term prediction method for ionization layer region | |
| CN112070272A (en) | Method and device for predicting icing thickness of power transmission line | |
| CN114595876A (en) | Regional wind field prediction model generation method and device and electronic equipment | |
| CN117909888B (en) | Intelligent artificial intelligence climate prediction method | |
| CN110852415B (en) | Vegetation index prediction method, system and equipment based on neural network algorithm | |
| CN117933095B (en) | Earth surface emissivity real-time inversion and assimilation method based on machine learning | |
| CN115032720A (en) | Application of multi-mode integrated forecast based on random forest in ground air temperature forecast | |
| CN115472233A (en) | Semi-supervised integrated industrial process soft measurement modeling method and system based on thermal diffusion label propagation | |
| CN110909492B (en) | Sewage treatment process soft measurement method based on extreme gradient lifting algorithm | |
| CN116701371B (en) | Method and device for interpolating missing values of atmospheric temperature data under covariance analysis | |
| CN106157163A (en) | A kind of grain yield short term prediction method and device | |
| CN112580899A (en) | Medium-and-long-term wind power generation prediction method and system fused with machine learning model | |
| CN116629415A (en) | Employment post prediction method, equipment and medium based on geographic weighted regression | |
| CN116245018A (en) | Sea wave missing measurement data forecasting method based on bivariate long-short-term memory algorithm | |
| CN210895535U (en) | Air temperature prediction system based on convolution cyclic neural network | |
| CN113962432A (en) | Wind power prediction method and system integrating three-dimensional convolution and light-weight convolution threshold unit | |
| CN115809947A (en) | Water pollution time sequence data continuous missing value filling method | |
| CN111199283A (en) | Air temperature prediction system and method based on convolution cyclic neural network | |
| CN115186941B (en) | Variable optimization climate mode method based on multiple space-time indexes and comprehensive sequencing | |
| CN116449460B (en) | Regional month precipitation prediction method and system based on convolution UNet and transfer learning |
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 |