CN116702586A - Weather forecast method and system based on multi-spatial scale values - Google Patents
Weather forecast method and system based on multi-spatial scale values Download PDFInfo
- Publication number
- CN116702586A CN116702586A CN202310452599.1A CN202310452599A CN116702586A CN 116702586 A CN116702586 A CN 116702586A CN 202310452599 A CN202310452599 A CN 202310452599A CN 116702586 A CN116702586 A CN 116702586A
- Authority
- CN
- China
- Prior art keywords
- weather
- data acquisition
- results
- space scale
- data
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
- 238000000034 method Methods 0.000 title claims abstract description 40
- 238000007621 cluster analysis Methods 0.000 claims abstract description 46
- 238000012545 processing Methods 0.000 claims description 26
- 230000009467 reduction Effects 0.000 claims description 26
- 239000011159 matrix material Substances 0.000 claims description 21
- 238000010606 normalization Methods 0.000 claims description 15
- 238000004422 calculation algorithm Methods 0.000 claims description 13
- 239000013598 vector Substances 0.000 claims description 12
- 238000004364 calculation method Methods 0.000 claims description 11
- 238000013528 artificial neural network Methods 0.000 claims description 7
- 238000000513 principal component analysis Methods 0.000 claims description 6
- 238000004891 communication Methods 0.000 claims description 5
- 238000012549 training Methods 0.000 claims description 5
- 238000010276 construction Methods 0.000 claims description 4
- 230000000694 effects Effects 0.000 abstract description 4
- 238000004458 analytical method Methods 0.000 description 18
- 230000008859 change Effects 0.000 description 9
- 238000010586 diagram Methods 0.000 description 5
- 238000005070 sampling Methods 0.000 description 5
- 230000008569 process Effects 0.000 description 4
- 230000002159 abnormal effect Effects 0.000 description 2
- 230000009286 beneficial effect Effects 0.000 description 2
- 238000012986 modification Methods 0.000 description 2
- 230000004048 modification Effects 0.000 description 2
- 238000007781 pre-processing Methods 0.000 description 2
- 238000003672 processing method Methods 0.000 description 2
- 238000010845 search algorithm Methods 0.000 description 2
- 230000004075 alteration Effects 0.000 description 1
- QVFWZNCVPCJQOP-UHFFFAOYSA-N chloralodol Chemical compound CC(O)(C)CC(C)OC(O)C(Cl)(Cl)Cl QVFWZNCVPCJQOP-UHFFFAOYSA-N 0.000 description 1
- 239000003086 colorant Substances 0.000 description 1
- 238000013480 data collection Methods 0.000 description 1
- 238000000354 decomposition reaction Methods 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 238000011156 evaluation Methods 0.000 description 1
- 230000010354 integration Effects 0.000 description 1
- 238000003064 k means clustering Methods 0.000 description 1
- 238000012821 model calculation Methods 0.000 description 1
- 238000003062 neural network model Methods 0.000 description 1
- 238000005457 optimization Methods 0.000 description 1
- 239000002245 particle Substances 0.000 description 1
- 238000004088 simulation Methods 0.000 description 1
- 230000000007 visual effect Effects 0.000 description 1
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F30/00—Computer-aided design [CAD]
- G06F30/20—Design optimisation, verification or simulation
- G06F30/27—Design optimisation, verification or simulation using machine learning, e.g. artificial intelligence, neural networks, support vector machines [SVM] or training a model
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01W—METEOROLOGY
- G01W1/00—Meteorology
- G01W1/02—Instruments for indicating weather conditions by measuring two or more variables, e.g. humidity, pressure, temperature, cloud cover or wind speed
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01W—METEOROLOGY
- G01W1/00—Meteorology
- G01W1/10—Devices for predicting weather conditions
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/23—Clustering techniques
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q10/00—Administration; Management
- G06Q10/04—Forecasting or optimisation specially adapted for administrative or management purposes, e.g. linear programming or "cutting stock problem"
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q50/00—Information and communication technology [ICT] specially adapted for implementation of business processes of specific business sectors, e.g. utilities or tourism
- G06Q50/10—Services
- G06Q50/26—Government or public services
-
- 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)
- Business, Economics & Management (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Environmental & Geological Engineering (AREA)
- Strategic Management (AREA)
- Human Resources & Organizations (AREA)
- Life Sciences & Earth Sciences (AREA)
- Economics (AREA)
- Evolutionary Computation (AREA)
- Tourism & Hospitality (AREA)
- Data Mining & Analysis (AREA)
- General Engineering & Computer Science (AREA)
- Marketing (AREA)
- Environmental Sciences (AREA)
- Ecology (AREA)
- Biodiversity & Conservation Biology (AREA)
- Development Economics (AREA)
- Atmospheric Sciences (AREA)
- Artificial Intelligence (AREA)
- General Business, Economics & Management (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Quality & Reliability (AREA)
- Health & Medical Sciences (AREA)
- Operations Research (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Medical Informatics (AREA)
- Software Systems (AREA)
- Educational Administration (AREA)
- Entrepreneurship & Innovation (AREA)
- General Health & Medical Sciences (AREA)
- Primary Health Care (AREA)
- Game Theory and Decision Science (AREA)
- Computer Hardware Design (AREA)
- Geometry (AREA)
- Evolutionary Biology (AREA)
- Bioinformatics & Computational Biology (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
A weather forecast method and system based on a multi-space scale value belongs to the meteorological field, and comprises the following steps: determining N space scale sets by dividing the space scale according to the distance, and carrying out grade identification on the N space scale sets by a longitude and latitude acquisition unit; respectively carrying out data acquisition on the N space scale sets to obtain N data acquisition results; performing cluster analysis on the N data acquisition results according to the grade identification to obtain M cluster analysis results; performing coverage traversal on the M cluster analysis results to obtain coverage traversal results; inputting the coverage traversal result into a weather numerical model to obtain a weather numerical value; and weather forecast is conducted on the N space scale sets based on the weather values. The method solves the technical problems of low accuracy and incomplete weather forecast in the prior art, and achieves the technical effects of improving the accuracy and comprehensiveness of weather forecast.
Description
Technical Field
The application relates to the field of weather, in particular to a weather forecast method and system based on a multi-space scale value.
Background
Along with the rapid development of socioeconomic performance, the requirements of people on weather forecast are also higher and higher requirements are also put forward on weather forecast due to climate change and frequent disaster weather. At present, a weather forecast mode adopts numerical forecast, but because of the complexity and uncertainty of a weather system, the dynamic change of weather is difficult to comprehensively reflect through single numerical simulation, so that how to improve the accuracy and the comprehensiveness of the weather forecast becomes an important problem to be solved at present.
Disclosure of Invention
The application provides a weather forecast method and a weather forecast system based on a multi-space scale value, and aims to solve the technical problems of low accuracy and incomplete weather forecast in the prior art.
In view of the above problems, the embodiment of the application provides a weather forecast method and a weather forecast system based on a multi-spatial scale value.
In a first aspect of the disclosure, a weather forecast method based on a multi-spatial scale value is provided, the method comprising: determining N space scale sets by carrying out region division on the space scales according to the distance, wherein N is a positive integer greater than 1; carrying out grade identification on the N space scale sets through the longitude and latitude acquisition unit, wherein the N space scale sets and the grade identification are in one-to-one correspondence; respectively carrying out data acquisition on the N space scale sets to obtain N data acquisition results; performing cluster analysis on the N data acquisition results according to the grade identification to obtain M cluster analysis results; performing coverage traversal on the M cluster analysis results to obtain coverage traversal results; inputting the coverage traversal result into a weather numerical model to obtain a weather numerical value; and weather forecast is conducted on the N space scale sets based on the weather values.
In another aspect of the disclosure, a weather forecast system based on a multi-spatial scale value is provided, the system comprising: the space region dividing module is used for determining N space dimension sets by dividing the space dimension according to the distance, wherein N is a positive integer greater than 1; the identification set grade module is used for carrying out grade identification on the N space scale sets through the longitude and latitude acquisition unit, wherein the N space scale sets and the grade identification are in one-to-one correspondence; the data acquisition result module is used for respectively carrying out data acquisition on the N space scale sets to obtain N data acquisition results; the data cluster analysis module is used for carrying out cluster analysis on N data acquisition results according to the grade identification to obtain M cluster analysis results; the coverage traversal result module is used for performing coverage traversal on the M cluster analysis results to obtain coverage traversal results; the weather value acquisition module is used for inputting the coverage traversal result into the weather value model to obtain a weather value; and the aggregate weather forecast module is used for carrying out weather forecast on the N space scale aggregates based on the weather values.
One or more technical schemes provided by the application have at least the following technical effects or advantages:
the technical scheme of weather forecast based on a plurality of weather values solves the technical problems of weather forecast accuracy and incompleteness in the prior art, and achieves the technical effects of improving the weather forecast accuracy and comprehensiveness.
The foregoing description is only an overview of the present application, and is intended to be implemented in accordance with the teachings of the present application in order that the same may be more clearly understood and to make the same and other objects, features and advantages of the present application more readily apparent.
Drawings
FIG. 1 is a schematic diagram of a possible flow chart of a weather forecast method based on a multi-spatial scale value according to an embodiment of the present application;
FIG. 2 is a schematic flow chart of a possible level identification of a set in a weather forecast method based on multiple spatial scale values according to an embodiment of the present application;
FIG. 3 is a schematic flow chart of a possible process for obtaining a cluster analysis result in a weather forecast method based on multiple spatial scale values according to an embodiment of the present application;
fig. 4 is a schematic diagram of a possible structure of a weather forecast system based on multiple spatial scale values according to an embodiment of the present application.
Reference numerals illustrate: the system comprises a space region dividing module 11, an identification collection level module 12, a data acquisition result module 13, a data cluster analysis module 14, an overlay traversal result module 15, a weather value acquisition module 16 and a collection weather forecast module 17.
Detailed Description
The technical scheme provided by the application has the following overall thought:
and dividing the forecast area into N different spatial scale sets through regional division and grade identification of the spatial scale, and carrying out data acquisition and grade division on each set. And then, classifying the space scale sets of different levels by adopting cluster analysis to obtain M cluster analysis results. And then, performing overlay traversal on the cluster analysis results to obtain a final overlay traversal result. And finally, inputting the coverage traversal result into a weather numerical model to obtain a plurality of weather numerical values, and carrying out weather forecast based on the plurality of weather numerical values.
Having described the basic principles of the present application, various non-limiting embodiments of the present application will now be described in detail with reference to the accompanying drawings.
Example 1
As shown in fig. 1, the embodiment of the application provides a weather forecast method based on a multi-spatial scale value, which is applied to a weather forecast system, wherein the weather forecast system is in communication connection with a longitude and latitude acquisition unit;
specifically, the weather forecast method based on the multi-space scale value is used in a weather forecast system, the system is communicated with a longitude and latitude acquisition unit in a wired mode and a wireless mode, and the longitude and latitude acquisition unit is composed of a group of sensors, data collectors and the like and is used for acquiring weather data in a designated area, such as temperature, humidity, longitude, latitude, wind direction, air pressure and the like. The wireless communication modes comprise Bluetooth, wi-Fi, zigBee, loRa and the like; the wired communication modes include RS232, RS485, ethernet and the like. The longitude and latitude acquisition unit acquires longitude and latitude information, data are transmitted into the weather forecast system, and the weather forecast system calculates according to the data, so that support is provided for weather forecast conditions.
The weather forecast method comprises the following steps:
step S100: determining N space scale sets by carrying out region division on the space scales according to the distance, wherein N is a positive integer greater than 1;
specifically, first, the latitude and longitude ranges of the prediction area are determined, the WGS84 coordinate system is selected, uniform sampling is performed in the entire prediction area, and the sampling points cover the entire prediction area. And calculating the distance between each two points by using a Manhattan distance method according to the longitude and latitude coordinates of each sampling point to obtain a distance matrix. Secondly, clustering analysis is carried out on the distance matrix by adopting a hierarchical clustering method, the points with the relatively close distance are clustered into one cluster, and the points are gradually combined to obtain N clusters. Then, dividing the spatial scale according to the number of clusters to obtain N spatial scale sets. Wherein, N is a positive integer greater than 1, which means that the value range of N is a positive integer, and only the case that N is greater than 1 is considered, because hierarchical division of spatial scale can be realized only when N is greater than 1, and a multi-scale model is constructed.
The space-time evolution rule in the region can be comprehensively known by dividing the space scale according to the distance, so that basis is provided for the follow-up weather forecast, and the forecast accuracy and comprehensiveness are improved.
Step S200: performing grade identification on the N space scale sets through the longitude and latitude acquisition unit, wherein the N space scale sets and the grade identification are in one-to-one correspondence;
specifically, cluster analysis is performed on sampling points in a forecast area through a hierarchical clustering method, N space scale sets are obtained, particle positioning is performed on the N space scale sets, longitude information and latitude information of the N space scale sets are obtained through a longitude and latitude acquisition unit, namely, position information of each space scale set on the earth surface, historical weather information of the position information is obtained according to the position information, and the information comprises data of different weather elements such as temperature, rainfall, air pressure, wind direction and wind force.
Determining weather element grade standards according to historical weather data and position information of the space scale sets, calculating the change amplitude of each space scale set on different weather elements according to the grade standards, and carrying out grade identification on N space scale sets by using the calculated weather element change amplitude. Different levels are set according to the magnitude of the variation, for example, a set of spatial scales with smaller variation is set to a low level, and a set of spatial scales with larger variation is set to a high level. Finally, each spatial scale set corresponds to one grade identifier, so that a one-to-one correspondence relationship with the spatial scale sets is established.
And grading the N space scale sets, classifying the variation amplitude of different space scale sets on the weather elements, and generating the space scale sets with different grades. The method is beneficial to better distinguishing space scale sets of different grades, so that the subsequent data acquisition and processing are more accurate, and the accuracy and precision of weather forecast are improved.
Step S300: respectively carrying out data acquisition on the N space scale sets to obtain N data acquisition results;
specifically, the positions of N space scale sets are determined, corresponding weather stations are connected according to the position information, and weather element data such as temperature, rainfall, air pressure, wind direction and wind force of the corresponding space scale sets are collected by using corresponding sensors or instruments. Preprocessing the acquired weather element data, processing missing values, abnormal values and the like in the data, and normalizing the weather element data in a linear normalization or nonlinear normalization mode to enable the weather element data to be comparable. And acquiring corresponding N data acquisition results by carrying out data acquisition, data preprocessing and data normalization on the N space scale sets one by one, and providing data support for the follow-up weather forecast.
Step S400: performing cluster analysis on the N data acquisition results according to the grade identification to obtain M cluster analysis results;
specifically, the N data acquisition results are preprocessed, including operations such as outlier removal, missing value filling, feature standardization and the like. The N data acquisition results are divided into a plurality of subsets according to the grade identifiers, and each subset contains data with the same grade identifier, so that the data are divided into different grades. The characteristics for cluster analysis are selected from N data acquisition results, and the characteristics with important significance, high correlation and strong distinguishing degree, such as humidity, temperature and the like, are selected so as to improve the accuracy and the effectiveness of the cluster analysis. And analyzing each subset by adopting a proper clustering algorithm to obtain M clustering analysis results, wherein the clustering algorithm comprises hierarchical clustering, K-means clustering, DBSCAN clustering and the like. The N data acquisition results are subjected to clustering analysis through the grade identification, so that differences and characteristics in the data can be better mined, more accurate and reliable data information is provided for weather forecast, and the decision referential of the data acquisition results is facilitated to be improved.
Step S500: performing coverage traversal on the M cluster analysis results to obtain coverage traversal results;
specifically, first, an overlay matrix is constructed. And constructing an M multiplied by M coverage matrix of the M clustering analysis results, wherein the element of the ith row and the jth column in the matrix represents whether the ith clustering analysis result can cover the jth clustering analysis result or not. The value of the element of row i, column j is 1 if it can be overridden, otherwise 0. Wherein the coverage matrix may be in the form of a sparse matrix or a dense matrix. Secondly, analyzing and traversing the coverage matrix by adopting a coverage traversal algorithm to find a group of subsets covering all clustering analysis results, wherein the coverage traversal algorithm comprises a greedy algorithm, a depth-first search algorithm, a breadth-first search algorithm and the like. And then, obtaining coverage traversal results according to the coverage subset found by the coverage traversal algorithm. And finally, carrying out visual analysis and display on the coverage traversing result by adopting a thermodynamic diagram, a scatter diagram, a network diagram and other methods.
And performing coverage traversal on the M clustering analysis results to obtain coverage traversal results, eliminating the same weather values, simplifying the calculation of a weather numerical model, and improving the accuracy of model result output.
Step S600: inputting the coverage traversal result into a weather numerical model to obtain a weather numerical value;
specifically, the coverage traversal result is preprocessed, required features are extracted and converted into digital form, and the digital form is conveniently input into a weather numerical model, for example, the coverage traversal result may contain the following features: temperature, humidity, visibility, wind direction, wind force, etc., these characteristics need to be digitized. And inputting the feature data after the digitization into a constructed weather numerical model, and running the model to obtain a weather numerical value. The output layer of the model will calculate the weather value from the input features, where the maximum, minimum, average, etc. values can be selected as desired.
And meanwhile, carrying out subsequent processing on the weather values. For example, weather variation trends are intuitively expressed by means of charts and the like, and compared with historical weather data to verify the accuracy of the results. Or when the weather data needs to be updated, new data is input into the weather numerical model again, and the accuracy and the practicability of the model are continuously optimized. And converting the data of the multiple spatial scales into corresponding weather values through a weather numerical model, and providing data support for weather forecast through specific weather values.
Step S700: and weather forecast is conducted on the N space scale sets based on the weather values.
Specifically, the model output data is compared with the actual data, and an error is calculated. The error evaluation index includes an average error (ME), a Root Mean Square Error (RMSE), an average absolute error (MAE), and the like. And (3) evaluating the accuracy of the weather forecast by calculating an error index, and classifying and authenticating the weather values. For weather values with high accuracy, the weather values can be directly released to users after integration, and for weather values with low accuracy, further optimization and adjustment are needed to be performed for release. The weather forecast based on a plurality of weather values can be linked among different spatial scales, so that the accuracy and the comprehensiveness of the weather forecast are improved.
Further, as shown in fig. 2, the embodiment of the present application further includes:
step S210: determining N pieces of longitude information and N pieces of latitude information of the N space scale sets through the longitude and latitude acquisition unit;
step S220: acquiring historical weather information corresponding to the N space scale sets by connecting an weather table;
step S230: acquiring weather variation amplitudes of the N space scale set areas based on the historical weather information, the N pieces of longitude information and the N pieces of latitude information;
step S240: and carrying out grade identification on the N space scale sets based on the weather variation amplitude.
Specifically, the sampling points are divided into N clusters through hierarchical clustering, the N clusters are N space scale sets, the centroids of the N clusters are calculated, the longitude and latitude acquisition units are used for positioning the centroids of the N clusters, the longitude and latitude of the centroids of the N clusters are obtained, and N pieces of longitude information and N pieces of latitude information of the N space scale sets are obtained. The weather forecast system is connected with an API provided by a weather bureau, and historical weather information of a forecast area is obtained according to longitude and latitude information of N space scale sets, wherein the historical weather information is weather information within two years.
Combining the historical weather information and longitude and latitude information to form a data set, sorting according to the time and space scale set, dividing the historical weather data according to the space scale set, dividing the data into a plurality of groups, and dividing different data types from the plurality of groups of data, wherein the data types comprise temperature, air pressure, humidity, wind speed, wind power, rainfall and the like. And calculating the average value, standard deviation, maximum value and minimum value of each data, visually displaying the calculated data in the forms of a map, a line graph and the like, and displaying the weather variation amplitude of different space scale sets.
After the weather variation amplitude of different space scale sets is obtained, the definition and meaning of the grade identification are defined. Taking the temperature change amplitude as an example, dividing the weather change amplitude into five grade marks, namely 'very big', 'medium', 'small', 'tiny', and the corresponding change amplitudes are respectively: judging the temperature of the space scale collection area with the average temperature of the space scale collection area being more than 20 ℃, 15-20 ℃, 10-15 ℃, 5-10 ℃ and less than 5 ℃ as a grade standard, and determining the space scale grade identification. Traversing all the historical weather data types and N space scale collection areas, and displaying the calculated grade marks in different colors or symbols to obtain weather variation amplitudes of the N space scale collection areas.
Historical weather information is divided according to the space scale sets and different data types, grade standards are formulated, the grade standards are divided, data support is provided for analyzing weather change conditions of the different space scale sets, analysis results are accurate and comprehensive, and accuracy of weather forecast is improved.
Further, the embodiment of the application further comprises:
step S310: acquiring temperature data of the N space scale sets to acquire N temperature data acquisition results;
step S320: the rainfall data acquisition is carried out on the N space scale sets, and N rainfall data acquisition results are obtained;
step S330: air pressure data acquisition is carried out on the N space scale sets, and N air pressure data acquisition results are obtained;
step S340: wind direction and wind power data acquisition is carried out on the N space scale sets, and N wind direction and wind power data acquisition results are obtained;
step S350: and carrying out normalization processing on the N temperature data acquisition results, the N rainfall data acquisition results, the N air pressure data acquisition results and the N wind direction wind power data acquisition results, and adding the processing results into the N data acquisition results.
Specifically, the positions of N space scale sets are determined to form position set data, the position set data comprises all space scale sets in a forecast interval, and the position set data of the set is transmitted to a weather forecast system. The weather forecast system is connected with the corresponding weather stations according to the given position set, sends out temperature data acquisition instructions, rainfall data acquisition instructions, air pressure data acquisition instructions and wind direction and wind power data acquisition instructions, uses the corresponding sensors or instruments to acquire data according to the instructions, traverses the position set until the data of all the space scale sets are acquired, and feeds back to the weather forecast system.
After the data are collected, firstly, data abnormal conditions such as missing values, outliers, data skewness and the like are processed, and the data reliability is enhanced. Secondly, determining the normalization processing method and range, wherein the normalization processing method such as linear normalization, logarithmic normalization and the like can be adopted to process, and determining the range value which can be obtained by each factor, such as the temperature range of-40 ℃ to 50 ℃, the rainfall range of 0 to 500mm and the like. And then, carrying out normalization processing on the acquisition result of each factor.
Taking the linear normalization as an example, for each factor x, assuming its range between [ min, max ], the calculation formula for normalizing it to a value x' between [0,1] is:
where x' is the normalized result and min and max are the minimum and maximum values that this factor may take. And after x 'is calculated, storing the calculated x' into a corresponding data acquisition result. Traversing all the space scale sets, carrying out normalization processing on the temperature, rainfall, air pressure, wind direction, wind force and other data of each space scale set, and adding the processing result into N data acquisition results for subsequent data processing and analysis.
By carrying out normalization processing on the data such as temperature, rainfall, air pressure, wind direction and wind force, the data such as temperature, rainfall, air pressure, wind direction and wind force are comparable, and subsequent data processing and analysis are convenient.
Further, as shown in fig. 3, the embodiment of the present application further includes:
step S410: performing data dimension reduction on the N data acquisition results through a principal component analysis algorithm based on the grade identification;
step S420: and carrying out cluster analysis on the N data acquisition results after dimension reduction to obtain M cluster analysis results.
Specifically, setting a grade standard, grading each factor to 1 to 5 grades by adopting a grade grading method, carrying out grade identification on N data acquisition results, carrying out dimension reduction on the N data acquisition results after grade identification by using a principal component analysis algorithm, reserving principal components with the accumulated contribution rate reaching 80% -90%, converting a plurality of related variables into a small number of irrelevant principal components by using a matrix decomposition method, and obtaining N data acquisition results after dimension reduction.
After N data acquisition results after dimension reduction are obtained, determining the clustering number M by adopting methods such as an elbow method and a contour coefficient, and gathering the N data acquisition results after dimension reduction into M clusters by utilizing a K-means cluster analysis algorithm, classifying similar data into one type, and classifying dissimilar data into different types. And classifying each group of data acquisition results into corresponding clusters according to the results of the cluster analysis, marking the cluster identification for each group of data acquisition results, and obtaining M cluster analysis results.
The data dimension reduction is carried out through a principal component analysis algorithm, the data set is converted into a low-dimension data set, the clustering analysis is carried out on the data set after the dimension reduction, the correlation and the collinearity between the features are eliminated, the noise and the interference in the processing process are avoided, and the accuracy of the follow-up weather forecast is improved.
Further, the embodiment of the application further comprises:
step S411: performing decentralization processing on the N data acquisition results to obtain V data acquisition results;
step S412: obtaining covariance matrixes of the V data acquisition results;
step S413: calculating the covariance matrix to obtain a feature vector corresponding to a feature value of the covariance matrix;
step S414: projecting the N data acquisition results onto the feature vector to obtain a dimension reduction data set, wherein the dimension reduction data set is a feature data set obtained after dimension reduction of the N data acquisition results.
Specifically, the average value of the temperature, the air pressure, the rainfall and the wind direction and the wind force of the forecasting area is determined, and in N data acquisition results, the average value of corresponding weather elements in the forecasting area is subtracted from the normalized data of the temperature, the air pressure, the rainfall and the wind direction and the wind force of the forecasting area, so that the N data acquisition results are subjected to decentralization processing, and V data acquisition results are obtained. And calculating covariance among all factors, and filling more covariance into a V multiplied by V matrix to form a covariance matrix of V data acquisition results. And calculating eigenvalues and eigenvectors of the covariance matrix by using a linear algebraic method. And sorting all the feature vectors according to the sizes of the corresponding feature values of the feature vectors, and selecting the feature vector with the largest first K corresponding feature values as a reserved feature vector, wherein K is the determined dimension reduction target dimension. And projecting N data acquisition results onto the K reserved characteristic vectors in a matrix multiplication or vector inner product mode, wherein the projected data points form a new dimension reduction data set. The new data set comprises K dimensions, wherein each dimension corresponds to a feature vector and is used for representing the structure and the features of the original data set, and the new data set is a feature data set after dimension reduction.
Through carrying out decentralization processing and data dimension reduction processing on N data acquisition results, the problem that the model is fitted excessively in the follow-up process is avoided, and the accuracy and generalization performance of the model are improved, so that the accuracy and comprehensiveness of weather forecast are improved.
Further, the embodiment of the application further comprises:
step S610: adopting space scale data with grade identification in the coverage traversal result as construction data;
step S620: constructing and training to obtain the weather numerical model based on a BP neural network;
step S630: the weather numerical model comprises a data input layer, a weather evolution calculation layer and a numerical grade output layer;
step S640: and inputting the spatial scale data with the grade identification in the coverage traversal result into the weather numerical model, and outputting the weather numerical value.
Specifically, the spatial scale region is divided into N spatial scale sets, class identifiers are allocated to the sets, M clustering analysis results are obtained after cluster analysis, the same sets are removed through coverage traversal, and spatial scale data with the class identifiers are used as construction data. Spatial scale data with a grade identification is selected from the historical weather data record, and the data is subjected to calibration processing. Setting a BP neural network architecture, wherein the BP neural network architecture comprises a data input layer, a weather evolution calculation layer and a numerical grade output layer, initializing the weight and the bias value of the weather evolution calculation layer, using historical weather data records comprising past time and space data, further training the neural network, continuously adjusting the node weight and the bias value to improve the accuracy of a model, and using an error back propagation algorithm to conduct supervision training so as to calculate weather changes at each time point in a time sequence.
The spatial scale data with the class identification is input into a trained BP neural network model, the input layer receives the data and passes it to a weather evolution calculation layer, which calculates weather results using historical data records in combination with the input data. The output layer converts the weather result of the calculation layer into a digital form and outputs the weather result as a weather value result, and the weather value at each time point can be calculated and output.
The method has the advantages that the weather change rule is learned from a large amount of historical weather data through the BP neural network, the weather numerical model is constructed, the accuracy and the practical feasibility of the model can be improved by de-duplication and identification of input data, and the accuracy of weather forecast can be improved.
In summary, the weather forecast method based on the multi-spatial scale values provided by the embodiment of the application has the following technical effects:
n space scale sets are determined by dividing the space scale according to the distance, wherein N is a positive integer greater than 1, and the large-range weather data are effectively divided into different small areas for targeted analysis and processing. The longitude and latitude acquisition units are used for carrying out grade identification on the N space scale sets, wherein the N space scale sets and the grade identification are in one-to-one correspondence, and a basis is provided for subsequent analysis and processing according to weather data of different grades; and respectively carrying out data acquisition on the N space scale sets to obtain N data acquisition results, classifying the data with higher similarity, and being beneficial to further improving the accuracy of weather forecast. And carrying out cluster analysis on N data acquisition results according to the grade identification, obtaining M cluster analysis results, and determining the relation between different data sets so as to carry out more accurate analysis and calculation. And performing coverage traversal on the M clustering analysis results to obtain coverage traversal results, removing the same weather data, simplifying model calculation, and improving prediction accuracy. The coverage traversing result is input into a weather numerical model to obtain weather numerical values, calculation is carried out for each area, so that follow-up weather forecast is more comprehensive and accurate, weather forecast is carried out on N space scale sets based on the weather numerical values, and weather forecast based on a plurality of weather numerical values can be linked among different space scales, so that accuracy and comprehensiveness of the weather forecast are improved.
Example two
Based on the same inventive concept as the weather forecast method based on the multi-spatial scale values in the foregoing embodiment, as shown in fig. 4, an embodiment of the present application provides a weather forecast system based on the multi-spatial scale values, where the weather forecast system is communicatively connected with a longitude and latitude collecting unit, and includes:
the spatial region dividing module 11 is configured to determine N sets of spatial dimensions by performing region division on the spatial dimensions according to distances, where N is a positive integer greater than 1;
an identification set grade module 12, configured to grade identify the N spatial scale sets through the longitude and latitude acquisition unit, where the N spatial scale sets and the grade identify are in a one-to-one correspondence;
the data acquisition result module 13 is used for respectively carrying out data acquisition on the N space scale sets to obtain N data acquisition results;
the data cluster analysis module 14 is used for carrying out cluster analysis on the N data acquisition results according to the grade identification to obtain M cluster analysis results;
the overlay traversal result module 15 is configured to perform overlay traversal on the M cluster analysis results, and obtain an overlay traversal result;
a weather value acquisition module 16, configured to input the coverage traversal result into a weather value model to obtain a weather value;
and the aggregate weather forecast module 17 is used for carrying out weather forecast on the N space scale aggregates based on the weather values.
Further, the embodiment of the application further comprises:
the collection information determining module is used for determining N pieces of longitude information and N pieces of latitude information of the N pieces of space scale collection through the longitude and latitude collecting unit;
the historical weather information module is connected with the weather table to acquire historical weather information corresponding to the N space scale sets;
the weather variation amplitude module is used for acquiring weather variation amplitudes of the N space scale collection areas based on the historical weather information, the N pieces of longitude information and the N pieces of latitude information;
and the grade identification module is used for carrying out grade identification on the N space scale sets based on the weather variation amplitude.
Further, the embodiment of the application further comprises:
the temperature data acquisition module is used for acquiring temperature data of the N space scale sets and acquiring N temperature data acquisition results;
the rainfall data acquisition module is used for acquiring rainfall data of the N space scale sets and acquiring N rainfall data acquisition results;
the air pressure data acquisition module is used for acquiring air pressure data of the N space scale sets and acquiring N air pressure data acquisition results;
the wind direction and wind power data module is used for collecting wind direction and wind power data of the N space scale sets and obtaining N wind direction and wind power data collection results;
and the data normalization processing module is used for carrying out normalization processing on the N temperature data acquisition results, the N rainfall data acquisition results, the N air pressure data acquisition results and the N wind direction wind power data acquisition results, and adding the processing results into the N data acquisition results.
Further, the embodiment of the application further comprises:
the principal component analysis module is used for carrying out data dimension reduction on the N data acquisition results through a principal component analysis algorithm based on the grade identification;
and the cluster analysis module is used for carrying out cluster analysis on the N data acquisition results after the dimension reduction to obtain M cluster analysis results.
Further, the embodiment of the application further comprises:
the decentralizing processing module is used for decentralizing the N data acquisition results to obtain V data acquisition results;
the covariance matrix module is used for obtaining covariance matrixes of the V data acquisition results;
the eigenvector module is used for operating the covariance matrix to obtain eigenvectors corresponding to eigenvalues of the covariance matrix;
and the dimension reduction data set module is used for projecting the N data acquisition results onto the feature vectors to obtain a dimension reduction data set, wherein the dimension reduction data set is a feature data set obtained after dimension reduction of the N data acquisition results.
Further, the embodiment of the application further comprises:
the spatial scale data module adopts spatial scale data with grade marks in the coverage traversal result as construction data;
the weather numerical model module is used for constructing and training to obtain the weather numerical model based on the BP neural network;
the weather numerical model comprises a data input layer, a weather evolution calculation layer and a numerical grade output layer;
and the weather numerical value output module is used for inputting the spatial scale data with the grade identification in the coverage traversal result into the weather numerical model and outputting the weather numerical value.
Any of the steps of the methods described above may be stored as computer instructions or programs in a non-limiting computer memory and may be called by a non-limiting computer processor to identify any method for implementing an embodiment of the present application, without unnecessary limitations.
It will be apparent to those skilled in the art that various modifications and variations can be made to the present application without departing from the scope of the application. Thus, the present application is intended to include such modifications and alterations insofar as they come within the scope of the application or the equivalents thereof.
Claims (7)
1. The weather forecast method based on the multi-space scale values is characterized by being applied to a weather forecast system, wherein the weather forecast system is in communication connection with a longitude and latitude acquisition unit and comprises the following steps of:
determining N space scale sets by carrying out region division on the space scales according to the distance, wherein N is a positive integer greater than 1;
performing grade identification on the N space scale sets through the longitude and latitude acquisition unit, wherein the N space scale sets and the grade identification are in one-to-one correspondence;
respectively carrying out data acquisition on the N space scale sets to obtain N data acquisition results;
performing cluster analysis on the N data acquisition results according to the grade identification to obtain M cluster analysis results;
performing coverage traversal on the M cluster analysis results to obtain coverage traversal results;
inputting the coverage traversal result into a weather numerical model to obtain a weather numerical value;
and weather forecast is conducted on the N space scale sets based on the weather values.
2. The method of claim 1, wherein performing the rank identification comprises:
determining N pieces of longitude information and N pieces of latitude information of the N space scale sets through the longitude and latitude acquisition unit;
acquiring historical weather information corresponding to the N space scale sets by connecting an weather table;
acquiring weather variation amplitudes of the N space scale set areas based on the historical weather information, the N pieces of longitude information and the N pieces of latitude information;
and carrying out grade identification on the N space scale sets based on the weather variation amplitude.
3. The method of claim 1, wherein obtaining the N data acquisition results comprises:
acquiring temperature data of the N space scale sets to acquire N temperature data acquisition results;
the rainfall data acquisition is carried out on the N space scale sets, and N rainfall data acquisition results are obtained;
air pressure data acquisition is carried out on the N space scale sets, and N air pressure data acquisition results are obtained;
wind direction and wind power data acquisition is carried out on the N space scale sets, and N wind direction and wind power data acquisition results are obtained;
and carrying out normalization processing on the N temperature data acquisition results, the N rainfall data acquisition results, the N air pressure data acquisition results and the N wind direction wind power data acquisition results, and adding the processing results into the N data acquisition results.
4. The method of claim 1, wherein obtaining the M cluster analysis results comprises:
performing data dimension reduction on the N data acquisition results through a principal component analysis algorithm based on the grade identification;
and carrying out cluster analysis on the N data acquisition results after dimension reduction to obtain M cluster analysis results.
5. The method of claim 4, wherein the steps of:
performing decentralization processing on the N data acquisition results to obtain V data acquisition results;
obtaining covariance matrixes of the V data acquisition results;
calculating the covariance matrix to obtain a feature vector corresponding to a feature value of the covariance matrix;
projecting the N data acquisition results onto the feature vector to obtain a dimension reduction data set, wherein the dimension reduction data set is a feature data set obtained after dimension reduction of the N data acquisition results.
6. The method of claim 1, wherein obtaining the weather value comprises:
adopting space scale data with grade identification in the coverage traversal result as construction data;
constructing and training to obtain the weather numerical model based on a BP neural network;
the weather numerical model comprises a data input layer, a weather evolution calculation layer and a numerical grade output layer;
and inputting the spatial scale data with the grade identification in the coverage traversal result into the weather numerical model, and outputting the weather numerical value.
7. A weather forecast system based on a multi-spatial scale value, wherein the weather forecast system is in communication connection with a longitude and latitude acquisition unit, comprising:
the space region dividing module is used for determining N space scale sets by dividing the space scale according to the distance, wherein N is a positive integer greater than 1;
the identification set grade module is used for carrying out grade identification on the N space scale sets through the longitude and latitude acquisition unit, wherein the N space scale sets and the grade identification are in one-to-one correspondence;
the data acquisition result module is used for respectively carrying out data acquisition on the N space scale sets to obtain N data acquisition results;
the data cluster analysis module is used for carrying out cluster analysis on the N data acquisition results according to the grade identification to obtain M cluster analysis results;
the coverage traversal result module is used for performing coverage traversal on the M cluster analysis results to obtain coverage traversal results;
the weather value acquisition module is used for inputting the coverage traversing result into a weather value model to obtain a weather value;
and the aggregate weather forecast module is used for carrying out weather forecast on the N space scale aggregates based on the weather values.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202310452599.1A CN116702586A (en) | 2023-04-25 | 2023-04-25 | Weather forecast method and system based on multi-spatial scale values |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202310452599.1A CN116702586A (en) | 2023-04-25 | 2023-04-25 | Weather forecast method and system based on multi-spatial scale values |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN116702586A true CN116702586A (en) | 2023-09-05 |
Family
ID=87832961
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202310452599.1A Pending CN116702586A (en) | 2023-04-25 | 2023-04-25 | Weather forecast method and system based on multi-spatial scale values |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN116702586A (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN117893722A (en) * | 2024-01-22 | 2024-04-16 | 中科三清科技有限公司 | Weather object identification method, system, medium and electronic equipment |
-
2023
- 2023-04-25 CN CN202310452599.1A patent/CN116702586A/en active Pending
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN117893722A (en) * | 2024-01-22 | 2024-04-16 | 中科三清科技有限公司 | Weather object identification method, system, medium and electronic equipment |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN110346517B (en) | Smart city industrial atmosphere pollution visual early warning method and system | |
| CN106257948B (en) | A kind of basin Rainfall Monitoring wireless sensor network node Optimal Deployment Method | |
| CN114254561A (en) | Waterlogging prediction method, waterlogging prediction system and storage medium | |
| CN110705115B (en) | Weather forecast method and system based on deep belief network | |
| CN110427993B (en) | High-speed train navigation blind area positioning method based on meteorological parameters | |
| CN114529097B (en) | Multi-scale crop phenological period remote sensing dimensionality reduction prediction method | |
| CN116702586A (en) | Weather forecast method and system based on multi-spatial scale values | |
| CN113836808A (en) | PM2.5 deep learning prediction method based on heavy pollution feature constraint | |
| CN111209968A (en) | Multi-meteorological factor mode forecast temperature correction method and system based on deep learning | |
| CN116822185B (en) | Daily precipitation data space simulation method and system based on HASM | |
| CN115775053A (en) | Distributed photovoltaic power short-term prediction method based on improved similar time method | |
| CN118013300B (en) | Short-term wind power prediction method and system for wind turbine generator | |
| CN114691661B (en) | Assimilation-based cloud air guide and temperature and humidity profile pretreatment analysis method and system | |
| Gündoğdu et al. | Application of feed forward and cascade forward neural network models for prediction of hourly ambient air temperature based on MERRA-2 reanalysis data in a coastal area of Turkey | |
| CN109116300B (en) | Extreme learning positioning method based on insufficient fingerprint information | |
| CN117370785A (en) | Environment monitoring method, terminal and environment monitoring system for smart city | |
| CN110704789B (en) | Population dynamic measurement and calculation method and system based on 'urban superconcephalon' computing platform | |
| CN117710802A (en) | Gravity field direction suitability analysis method based on image texture features | |
| CN117911883A (en) | GEE time sequence analysis-based large-area ecological vulnerability remote sensing dynamic assessment method | |
| CN114679683B (en) | Indoor intelligent positioning method based on derived fingerprint migration | |
| CN113645565B (en) | Indoor positioning method based on hexagonal closest packing structure | |
| CN115442887A (en) | Indoor positioning method based on cellular network RSSI | |
| CN115600498A (en) | Wind speed forecast correction method based on artificial neural network | |
| CN114880954A (en) | Landslide sensitivity evaluation method based on machine learning | |
| CN115080629A (en) | Meteorological element forecasting system and method based on remote sensing technology |
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 |