CN111209537A - Space fine-grained pollution inference method based on Gaussian regression - Google Patents

Space fine-grained pollution inference method based on Gaussian regression Download PDF

Info

Publication number
CN111209537A
CN111209537A CN201911397406.7A CN201911397406A CN111209537A CN 111209537 A CN111209537 A CN 111209537A CN 201911397406 A CN201911397406 A CN 201911397406A CN 111209537 A CN111209537 A CN 111209537A
Authority
CN
China
Prior art keywords
data
pollution
gaussian
fine
gaussian regression
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
Application number
CN201911397406.7A
Other languages
Chinese (zh)
Other versions
CN111209537B (en
Inventor
李倩
刘锐
谢涛
席春秀
张丽伟
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
China Science Mapuniverse Tchndogy Co ltd
Original Assignee
China Science Mapuniverse Tchndogy Co ltd
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by China Science Mapuniverse Tchndogy Co ltd filed Critical China Science Mapuniverse Tchndogy Co ltd
Priority to CN201911397406.7A priority Critical patent/CN111209537B/en
Publication of CN111209537A publication Critical patent/CN111209537A/en
Application granted granted Critical
Publication of CN111209537B publication Critical patent/CN111209537B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Images

Classifications

    • 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
    • G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
    • G06N5/00—Computing arrangements using knowledge-based models
    • G06N5/04—Inference or reasoning models
    • G06N5/041—Abduction

Abstract

The invention discloses a space fine-grained pollution inference method based on Gaussian regression, which relates to the technical field of atmospheric pollution models and comprises the following steps: s1, defining data of all monitoring point locations in a given monitoring area, and deducing PM2.5 values of all unknown point locations; s2, determining a selected Gaussian regression model, and performing data training by using the Gaussian regression model; and S3, selecting training data and testing data, and obtaining a predicted value of the to-be-inferred spatial fine-grained pollution by using the training data and the testing data. Compared with other pollution estimation methods, the method has higher accuracy and stability, and is more suitable for carrying out space inference on fine-grained PM 2.5; the fine pollution heat map makes subsequent fine pollution management and health risk assessment more likely.

Description

Space fine-grained pollution inference method based on Gaussian regression
Technical Field
The invention relates to the field of environmental monitoring, in particular to a method for deducing spatial fine-grained pollution based on Gaussian regression.
Background
In order to study the generation and diffusion rules of PM2.5 in a fine manner, a monitoring system which is more densely deployed is needed, accurate national control stations are deployed in each main city in China for monitoring at present, but the monitoring density is still very sparse, for example, only 35 national control stations are deployed in about ten thousand square kilometers of beijing for monitoring, which poses great challenges for accurate spatial inference, fine control and health risk assessment. Relevant studies have shown that even in two places that are close together, there may be a large gap in their PM 2.5.
In order to infer pollution monitoring data for a space, two main types of methods have been proposed in recent years. The first category is the traditional diffusion models such as the model of the gaussian plume, the three-dimensional street valley model and the computational fluid dynamics model. These models generally integrate numerous data such as meteorological information, street geographic feature information, traffic information, and the like, and perform complex data modeling, but such models generally require a relatively strong assumption about the physical environment and various latitude-oriented monitoring data, which are relatively difficult to obtain in the air pollution monitoring field. The second type of model is based on space inference, and the second type of model is based on data of monitored and fallen sparse national control sites in cities, and is combined with data of weather, geographic positions, traffic information and the like to establish a space statistical inference model, so that the pollution value of an unknown place is inferred. However, data of an area where a domestic control station is not deployed cannot be accurately inferred, so how to effectively utilize limited data to speculate an undeployed site with the accuracy of 50 meters is a problem to be solved at present.
Disclosure of Invention
The invention aims to provide a method for deducing the spatial fine-grained pollution based on Gaussian regression, so as to solve the problems in the prior art.
In order to achieve the purpose, the technical scheme adopted by the invention is as follows:
1. a method for deducing spatial fine-grained pollution based on Gaussian regression comprises the following steps:
s1, defining data of all monitoring point locations in a given monitoring area, and deducing PM2.5 values of all unknown point locations;
s2, determining a selected Gaussian regression model, and performing data training by using the Gaussian regression model;
and S3, selecting training data and testing data, and obtaining a predicted value of the to-be-inferred spatial fine-grained pollution by using the training data and the testing data.
Preferably, the definition of the given data in step S1 is specifically:
Figure BDA0002346673540000021
xirepresenting the latitude and longitude of the ith monitoring station in the monitoring area, using yiA value representing PM2.5 for the monitored site;
the formula for inferring the values of PM2.5 for all unknown points is:
Figure BDA0002346673540000022
wherein eiRepresenting noise.
The method is aimed at
Figure BDA0002346673540000023
And (4) learning the data to obtain a correct function f, so that the corresponding y can be predicted for any given x.
The gaussian process is a statistical distribution, which is a combination of a series of random variables in a continuous domain (time or space), and follows a gaussian distribution for each random variable at a time point or a space point. In the gaussian regression problem, the function f distribution follows a gaussian distribution (normal distribution),
preferably, the specific process in step S2 includes:
s21, definition fi＝f(xi)；
S22, when x satisfies the following condition:
Figure BDA0002346673540000024
where K is a covariance matrix, where Kij＝k(xi,xj)，k(x1,x2) Can be any kernel that satisfies a semi-positive definite characteristic, where K is a covariance matrix.
Preferably, the kernel function is selected from the following square exponential covariance functions:
Figure BDA0002346673540000025
where l represents the scale of the change in the level of the function.
Preferably, step S3 specifically includes:
s31, making f ═ f1,f2,...,fn],f*＝[f*1,f*2,...,f*n]Respectively representing training data and test data;
s32, using bayesian theory, yielding:
Figure BDA0002346673540000031
s33, obtaining posterior probability distribution according to the formula:
Figure BDA0002346673540000032
s34, the prior probability and the likelihood function are independent distribution and obey gaussian distribution:
Figure BDA0002346673540000033
Figure BDA0002346673540000034
wherein delta2Is the noise variance, I is the identity matrix;
the integral in equation (5) can thus result in a complete solution, whose solution also obeys a Gaussian distribution
Figure BDA0002346673540000035
μ*＝K*,f(Kf,f+δ2I)-1y (7)
Σ*＝K*,*-K*,f(Kf,f+δ2I)-1Kf,*(8)
μ*Is a predicted mean, sigma*Is its corresponding prediction variance, i.e. confidence corresponding to the predicted value, in our usage scenario we use μ*iAs we are for yiThe predicted value of (2).
The invention has the beneficial effects that:
the invention discloses a method for deducing spatial fine-grained pollution based on Gaussian regression, which has higher accuracy and stability compared with other pollution conjecture methods and is more suitable for carrying out spatial deduction on fine-grained PM 2.5; the fine pollution heat map makes subsequent fine pollution management and health risk assessment more likely.
Drawings
FIG. 1 is a data deployment diagram of a PM2.5 monitoring site deployed in Beijing;
FIG. 2 is a graph showing the difference between two point data in example 2;
FIG. 3 is a relationship between a horizontal scale parameter and a spatial extrapolation absolute error in example 2;
FIG. 4 is the absolute error distribution statistics of spatial extrapolation using three different methods in example 2;
fig. 5 is a histogram of spatial extrapolation errors for 8 monitored sites in example 2.
Detailed Description
In order to make the objects, technical solutions and advantages of the present invention more apparent, the present invention is further described in detail below with reference to the accompanying drawings. It should be understood that the detailed description and specific examples, while indicating the invention, are intended for purposes of illustration only and are not intended to limit the scope of the invention.
Example 1
The embodiment discloses a method for deducing spatial fine-grained pollution based on Gaussian regression, which comprises the following steps:
s1, defining data of all monitoring point locations in a given monitoring area, and deducing PM2.5 values of all unknown point locations;
using xiRepresenting the latitude and longitude of the ith monitoring station in the monitoring area, using yiA value representing PM2.5 for the monitored site. The problem may be defined as data for all monitoring points within a given monitoring area
Figure BDA0002346673540000041
And deducing the PM2.5 value of all unknown points. This is a typical numerical regression problem, which can be defined as follows:
Figure BDA0002346673540000042
wherein eiRepresenting noise. The method is aimed at
Figure BDA0002346673540000043
And (4) learning the data to obtain a correct function f, so that the corresponding y can be predicted for any given x.
Step two, determining a Gaussian regression model (determining a covariance function)
The gaussian process is a statistical distribution, which is a combination of a series of random variables in a continuous domain (time or space), and follows a gaussian distribution for each random variable at a time point or a space point. In the gaussian regression problem, the function f distribution follows a gaussian distribution (normal distribution) when x satisfies the following condition:
Figure BDA0002346673540000044
wherein f isi＝f(xi) Represents the function f above, where K is a covariance matrix, where K isij＝k(xi,xj). k(x1,x2) Can be any kernel function that satisfies the semi-positive characteristic.
Thirdly, obtaining a predicted value by using the training data and the test data
In the inference process of the algorithm, let f ═ f1,f2,...,fn],f*＝[f*1,f*2,...,f*n]Training data and test data are represented separately. Further, using bayes theory, one can obtain:
Figure BDA0002346673540000045
by the above formula, the posterior probability distribution can be further obtained:
Figure RE-GDA0002457740300000052
because the prior probability and the likelihood function are both independently distributed and both obey a gaussian distribution:
Figure RE-GDA0002457740300000053
wherein delta2Is the noise variance and I is the identity matrix. The integral in equation (4) can thus result in a complete solution, whose solution also obeys a Gaussian distribution
Figure BDA0002346673540000053
μ*＝K*,f(Kf,f+δ2I)-1y(6)
Σ*＝K*,*-K*,f(Kf,f+δ2I)-1Kf,*(7)
Wherein mu*Is a predicted mean, sigma*Is the prediction variance corresponding to it, i.e. the confidence corresponding to the predicted value. In our usage scenario, we use μ*iAs we are for yiThe predicted value of (2).
Example 2
In the embodiment, data of a PM2.5 monitoring site deployed in beijing city is used, the data is updated once per hour, and the inference performance of a gaussian regression model is analyzed as shown in fig. 1 below.
Fig. 2 shows a distribution diagram of the difference between two point location data, and it can be seen from fig. 2 that although two point locations are not far apart, there may still be a large difference between PM2.5 data of the two point locations, and the two point locations are 6 km apart, but the absolute error is greater than 100 in more than 21% cases.
In actual prediction, one point location is randomly removed, and the data of the rest point locations is used for predicting the point location.
Step one, giving specific definition of problems
Using xiRepresenting the latitude and longitude of the ith monitoring station in the monitoring area, using yiA value representing PM2.5 for the monitored site. The problem may be defined as data for all monitoring points within a given monitoring area
Figure BDA0002346673540000054
And deducing the PM2.5 value of all unknown points. This is a typical numerical regression problem, which can be defined as follows:
Figure BDA0002346673540000055
wherein eiRepresenting noise. The aim of the model is to provide for
Figure BDA0002346673540000056
And (4) learning the data to obtain a correct function f, so that the corresponding y can be predicted for any given x.
Step two, determining a Gaussian regression model (determining a covariance function)
The gaussian process is a statistical distribution, which is a combination of a series of random variables in a continuous domain (time or space), and follows a gaussian distribution for each random variable at a time point or a space point. In the gaussian regression problem, the function f distribution follows a gaussian distribution (normal distribution) when x satisfies the following condition:
Figure BDA0002346673540000061
wherein f isi＝f(xi) Represents the function f above, where K is a covariance matrix, where K isij＝k(xi,xj). k(x1,x2) Can be any kernel function that satisfies the semi-positive characteristic.
The following squared exponential covariance function was chosen:
Figure BDA0002346673540000062
where l represents the scale of the change in the level of the function. As l becomes larger, the corresponding characteristic latitude becomes relatively unimportant, and vice versa. When a larger value l is selected, but the spatial inference effect is still better, the distribution of the PM2.5 in the area range is reflected to be relatively gentle, otherwise, the distribution nonlinearity of the PM2.5 in the area is reflected, and therefore, the selection of the value l reflects the characteristic of the PM2.5 distribution.
Fig. 3 shows the relationship between the horizontal scale parameter and the spatial extrapolation absolute error. As l increases, the error of the spatial inference decreases from 27.6 to 21.9, which indicates that the distribution of PM2.5 is non-linear in most cases, and there may be a large gap in the data when the two locations are far apart.
Thirdly, obtaining a predicted value by using the training data and the test data
In the inference process of the algorithm, let f ═ f1,f2,...,fn],f*＝[f*1,f*2,...,f*n]Training data and test data are represented separately. Further, using bayes theory, one can obtain:
Figure BDA0002346673540000063
by the above formula, the posterior probability distribution can be further obtained:
Figure BDA0002346673540000071
because the prior probability and the likelihood function are both independently distributed and both obey a gaussian distribution:
Figure BDA0002346673540000072
Figure BDA0002346673540000073
wherein delta2Is the noise variance and I is the identity matrix. The integral in equation (5) can thus result in a complete solution, whose solution also obeys a Gaussian distribution
Figure BDA0002346673540000074
μ*＝K*,f(Kf,f+δ2I)-1y (7)
Σ*＝K*,*-K*,f(Kf,f+δ2I)-1Kf,*(8)
Wherein mu*Is a predicted mean, sigma*Is the prediction variance corresponding to it, i.e. the confidence corresponding to the predicted value. In the present invention*iAs we are for yiThe predicted value of (2).
Fig. 4 shows the absolute error distribution statistics for spatial extrapolation using three different methods. It can be seen that the inference method using the gaussian regression model in this embodiment is significantly better than the linear model and the polynomial inference model. The 65% result of the gaussian extrapolation method is within 20% error, corresponding to 52% for the linear model and 46% for the polynomial model. Meanwhile, it can be seen that the linear model has better effect than the polynomial model, but does not represent that PM2.5 is linearly distributed in space. A histogram of the spatial extrapolation error for 8 monitored sites is shown in fig. 5. It can be seen that the linear model at stations 24, 25, 38 is better than the polynomial model, and the polynomial model at stations 7, 12, 47 performs better than the linear model, indicating that at some locations the distribution of PM2.5 is shifted from a linear distribution, and correspondingly, at other locations the distribution may be non-linear. In substantially all cases, the gaussian regression model was better than the linear and polynomial models, which also demonstrates the effectiveness of the gaussian regression model in PM2.5 space inference.
Table 1 shows the error statistics of spatial inference for three different methods.
Figure BDA0002346673540000075
Figure BDA0002346673540000081
As can be seen from the table, the Gaussian inference model is significantly better than the linear and polynomial models. In the chebyshev norm error statistics, both the linear and polynomial models are as high as 266.92, while the gaussian model is reduced to 154.14, which indicates that the gaussian inference model has better stability in the spatial inference of PM 2.5.
By adopting the technical scheme disclosed by the invention, the following beneficial effects are obtained:
the invention discloses a method for deducing fine-grained pollution in space based on a Gaussian regression file, which is characterized by firstly scientifically defining problems and deducing PM2.5 values of all unknown point positions, then analyzing PM2.5 data obtained by 300 network nodes by using a Gaussian regression model so as to deduce the PM2.5 values of all unknown point positions, and finally measuring distance similarity by using a covariance function. The model provided by the invention is a Gaussian process model, which provides prediction variance to indicate the effectiveness of prediction in a local random area, but a traditional linear model polynomial model is not.
Compared with the prior art, the invention has the following advantages: 1. compared with the inference method adopting other models, the method adopting the Gaussian regression model has higher accuracy and model stability, and is more suitable for carrying out space inference on fine-grained PM 2.5; 2. the fine pollution heat map makes subsequent fine pollution management and health risk assessment more likely.
The foregoing is only a preferred embodiment of the present invention, and it should be noted that, for those skilled in the art, various modifications and improvements can be made without departing from the principle of the present invention, and such modifications and improvements should also be considered within the scope of the present invention.

Claims (5)

1. A method for deducing spatial fine-grained pollution based on Gaussian regression is characterized by comprising the following steps:
s1, defining data of all monitoring point locations in a given monitoring area, and deducing PM2.5 values of all unknown point locations;
s2, determining a selected Gaussian regression model, and performing data training by using the Gaussian regression model;
and S3, selecting training data and testing data, and obtaining a predicted value of the to-be-inferred spatial fine-grained pollution by using the training data and the testing data.
2. The method for inferring pollution with spatial fine granularity based on gaussian regression as claimed in claim 1, wherein the given data defined in step S1 are specifically:
Figure FDA0002346673530000011
xirepresenting the latitude and longitude of the ith monitoring station in the monitoring area, using yiA value representing PM2.5 for the monitored site;
the formula for inferring the values of PM2.5 for all unknown points is:
Figure FDA0002346673530000012
wherein eiRepresenting noise.
3. The method for inferring the spatial fine-grained pollution based on the gaussian regression as claimed in claim 1, wherein the specific process in the step S2 includes:
s21, definition fi＝f(xi)；
S22, when x satisfies the following condition:
Figure FDA0002346673530000013
where K is a covariance matrix, where Kij＝k(xi,xj)，k(x1,x2) Can be any kernel that satisfies a semi-positive definite characteristic, where K is a covariance matrix.
4. The method of claim 3, wherein the kernel function is selected from the following square exponential covariance functions:
Figure FDA0002346673530000014
where l represents the scale of the change in the level of the function.
5. The method according to claim 1, wherein the step S3 specifically includes:
s31, making f ═ f1,f2,…,fn],f*＝[f*1,f*2,…,f*n]Respectively representing training data and test data;
s32, using bayesian theory, yielding:
Figure FDA0002346673530000021
s33, obtaining posterior probability distribution according to the formula:
Figure FDA0002346673530000022
s34, the prior probability and the likelihood function are independent distribution and obey gaussian distribution:
Figure FDA0002346673530000023
Figure FDA0002346673530000024
wherein delta2Is the noise variance, I is the identity matrix;
the integral in equation (5) can thus result in a complete solution, whose solution also obeys a Gaussian distribution
Figure FDA0002346673530000025
μ*＝K*,f(Kf,f+δ2I)-1y (7)
Σ*＝K*,*-K*,f(Kf,f+δ2I)-1Kf,*(8)
μ*Is a predicted mean, sigma*Is its corresponding prediction variance, i.e. confidence corresponding to the predicted value, in the use scenario, mu*iAs for yiThe predicted value of (2).
CN201911397406.7A 2019-12-30 2019-12-30 Space fine-grained pollution inference method based on Gaussian regression Active CN111209537B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN201911397406.7A CN111209537B (en) 2019-12-30 2019-12-30 Space fine-grained pollution inference method based on Gaussian regression

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN201911397406.7A CN111209537B (en) 2019-12-30 2019-12-30 Space fine-grained pollution inference method based on Gaussian regression

Publications (2)

Publication Number Publication Date
CN111209537A true CN111209537A (en) 2020-05-29
CN111209537B CN111209537B (en) 2023-04-07

Family

ID=70786579

Family Applications (1)

Application Number Title Priority Date Filing Date
CN201911397406.7A Active CN111209537B (en) 2019-12-30 2019-12-30 Space fine-grained pollution inference method based on Gaussian regression

Country Status (1)

Country Link
CN (1) CN111209537B (en)

Citations (7)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN104798090A (en) * 2012-08-06 2015-07-22 Agt国际有限公司 System and method for updating a data structure with sensor measurement data
CN105181898A (en) * 2015-09-07 2015-12-23 李岩 Atmospheric pollution monitoring and management method as well as system based on high-density deployment of sensors
US20160125307A1 (en) * 2013-06-05 2016-05-05 Yu Zheng Air quality inference using multiple data sources
US20170261645A1 (en) * 2016-03-10 2017-09-14 The Climate Corporation Long-range temperature forecasting
WO2018225030A1 (en) * 2017-06-09 2018-12-13 Sense Square S.R.L.S. Atmospheric pollution source mapping and tracking of pollutants by using air quality monitoring networks having high space-time resolution
CN110210681A (en) * 2019-06-11 2019-09-06 西安电子科技大学 A kind of prediction technique of the monitoring station PM2.5 value based on distance
CN110378520A (en) * 2019-06-26 2019-10-25 浙江传媒学院 A kind of PM2.5 concentration prediction and method for early warning

Patent Citations (7)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN104798090A (en) * 2012-08-06 2015-07-22 Agt国际有限公司 System and method for updating a data structure with sensor measurement data
US20160125307A1 (en) * 2013-06-05 2016-05-05 Yu Zheng Air quality inference using multiple data sources
CN105181898A (en) * 2015-09-07 2015-12-23 李岩 Atmospheric pollution monitoring and management method as well as system based on high-density deployment of sensors
US20170261645A1 (en) * 2016-03-10 2017-09-14 The Climate Corporation Long-range temperature forecasting
WO2018225030A1 (en) * 2017-06-09 2018-12-13 Sense Square S.R.L.S. Atmospheric pollution source mapping and tracking of pollutants by using air quality monitoring networks having high space-time resolution
CN110210681A (en) * 2019-06-11 2019-09-06 西安电子科技大学 A kind of prediction technique of the monitoring station PM2.5 value based on distance
CN110378520A (en) * 2019-06-26 2019-10-25 浙江传媒学院 A kind of PM2.5 concentration prediction and method for early warning

Non-Patent Citations (2)

* Cited by examiner, † Cited by third party
Title
危浩: "空气质量模拟与监测数据融合方法优化及应用", 《中国优秀硕士学位论文全文数据库》 *
齐甜方等: "面向多源数据沈阳市PM2.5浓度预测研究及实证分析", 《系统工程》 *

Also Published As

Publication number Publication date
CN111209537B (en) 2023-04-07

Similar Documents

Publication Publication Date Title
Wen et al. Spatial heterogeneity in implicit housing prices: Evidence from Hangzhou, China
CN104809333A (en) Capacity predicating method and system based on Kalman filter
CN110346654B (en) Electromagnetic spectrum map construction method based on common kriging interpolation
Bargaoui et al. Seasonal precipitation variability in regional climate simulations over Northern basins of Tunisia
Zhang et al. Groundwater depth predictions by GSM, RBF, and ANFIS models: a comparative assessment
Ni et al. Evolutionary modeling for streamflow forecasting with minimal datasets: a case study in the West Malian River, China
Zounemat-Kermani Investigating chaos and nonlinear forecasting in short term and mid-term river discharge
Núñez et al. Resolving regional frequency analysis of precipitation at large and complex scales using a bottom-up approach: The Latin America and the Caribbean Drought Atlas
Huang et al. Improvement of two-dimensional flow-depth prediction based on neural network models by preprocessing hydrological and geomorphological data
CN114236480A (en) Airborne platform sensor system error registration algorithm
Sane et al. Internal vs forced variability metrics for geophysical flows using information theory
JP2017033198A (en) Space-time variable prediction device and program
Masanta et al. Proposal and evaluation of nonstationary versions of SPEI and SDDI based on climate covariates for regional drought analysis
Roux et al. Use of parameter optimization to estimate a flood wave: Potential applications to remote sensing of rivers
CN102122004A (en) Railway wind disaster prevention prediction method
CN111209537B (en) Space fine-grained pollution inference method based on Gaussian regression
Smith et al. Testing probabilistic adaptive real‐time flood forecasting models
Liu et al. Gridded statistical downscaling based on interpolation of parameters and predictor locations for summer daily precipitation in North China
Mazzoleni et al. Data assimilation in hydrologic routing: Impact of model error and sensor placement on flood forecasting
Petrov et al. Maximum entropy estimates of extreme significant wave heights from satellite altimeter data
CN108256261B (en) Stability evaluation method and device for offshore floating structure
Long et al. Dynamic data driven simulation with soft data
Assis et al. A model-based site selection approach associated with regional frequency analysis for modeling extreme rainfall depths in Minas Gerais state, Southeast Brazil
White et al. A model for Antarctic surface mass balance and ice core site selection
McGuire et al. Characterizing sensor datasets with multi-granular spatio-temporal intervals

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
PE01 Entry into force of the registration of the contract for pledge of patent right

Denomination of invention: A spatial fine-grained pollution inference method based on Gaussian regression

Effective date of registration: 20230627

Granted publication date: 20230407

Pledgee: Zhongguancun Beijing technology financing Company limited by guarantee

Pledgor: CHINA SCIENCE MAPUNIVERSE TCHNDOGY Co.,Ltd.

Registration number: Y2023990000326

PE01 Entry into force of the registration of the contract for pledge of patent right