Visibility hierarchical prediction model based on correlation analysis and data equalization
A correlation analysis and prediction model technology, applied in character and pattern recognition, biological neural network models, instruments, etc., can solve problems such as sample imbalance and inaccurate low visibility forecast, so as to improve accuracy and generalization ability , the effect of reducing the error
Inactive Publication Date: 2020-10-27
南京信大气象科学技术研究院有限公司
View PDF6 Cites 7 Cited by
- Summary
- Abstract
- Description
- Claims
- Application Information
AI Technical Summary
Problems solved by technology
[0005] The method of predicting visibility in the above-mentioned patent fails to solve the problem of unbalanced samples and inaccurate forecasting of low visibility
Method used
the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
View moreImage
Smart Image Click on the blue labels to locate them in the text.
Smart ImageViewing Examples
Examples
Experimental program
Comparison scheme
Effect test
Embodiment
[0045] The present invention provides a hierarchical forecasting model of visibility based on correlation analysis and data balance. The steps of establishing the model are as follows:
[0046] S1. Collect the observation data of the weather station, and select the factors with greater correlation with visibility through correlation analysis to form data samples;
[0047] S2. Process the data samples in step 1 to obtain training samples and test samples;
[0048] S3. Statistically analyze the training samples obtained in step 2 according to the classification criteria of visibility categories, and obtain new training samples by balancing various samples by random down-sampling;
[0049] S4. Carry out sample classification to the new training samples obtained in step 3 by the long-short-term memory neural network (LSTM) classification model;
[0050] S5. Input the classification results in step 4 and the corresponding category training samples into the LSTM-based regression mode...
the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to View More PUM
Login to View More Abstract
A visibility hierarchical prediction model based on correlation analysis and data equalization comprises the following model building steps of collecting meteorological station observation data, selecting factors with large correlation with visibility out through correlation analysis, and forming a data sample; processing the data sample to obtain a training sample and a test sample; dividing standard training samples according to visibility categories to perform statistical analysis, and balancing various samples through a random down-sampling method to obtain new training samples; performingsample classification on the new training sample through a long short-term memory (LSTM) neural network classification model; and inputting the classification result and the corresponding class training sample into an LSTM-based regression model, selecting a subclass sample corresponding to each class, and finally regressing the visibility. According to the visibility hierarchical prediction algorithm model based on correlation analysis and data equalization, the generalization ability of the network is improved, so that the accuracy of visibility category prediction is improved, the error ofvisibility prediction is reduced, and the application value is relatively high.
Description
technical field [0001] The invention relates to the field of meteorological monitoring, in particular to a hierarchical forecasting model of visibility based on correlation analysis and data balance. Background technique [0002] Atmospheric visibility (Visibility) is an index reflecting the transparency of the atmosphere. It is generally defined as the maximum ground horizontal distance at which a person with normal vision can still see the outline of the target clearly under the prevailing weather conditions. Atmospheric visibility is an important indicator in meteorological monitoring, and it is widely used in road traffic, navigation, aviation and environmental protection monitoring. Since the second industrial revolution in the 1960s, with the sharp increase in the consumption of fossil energy in human production and life, more and more particulate matter such as PM2.5 and PM10 have been emitted into the atmosphere, and the aerosols formed by condensation nuclei of the...
Claims
the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to View More Application Information
Patent Timeline
Login to View More Patent Type & Authority Applications(China)
IPC IPC(8): G06F30/27G06N3/04G06N3/08G06K9/62
CPCG06F30/27G06N3/049G06N3/08G06N3/045G06F18/2411
Inventor 陆冰鉴王兴詹少伟苗春生周可薛丰昌张越
Owner 南京信大气象科学技术研究院有限公司