CN113011600A - Method for correcting evaporation waveguide prediction model based on machine learning - Google Patents

Method for correcting evaporation waveguide prediction model based on machine learning Download PDF

Info

Publication number
CN113011600A
CN113011600A CN202110317266.9A CN202110317266A CN113011600A CN 113011600 A CN113011600 A CN 113011600A CN 202110317266 A CN202110317266 A CN 202110317266A CN 113011600 A CN113011600 A CN 113011600A
Authority
CN
China
Prior art keywords
refractive index
waveguide
evaporation
evaporation waveguide
height
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
Application number
CN202110317266.9A
Other languages
Chinese (zh)
Inventor
石广亮
王健
马建国
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.)
Tianjin University
Original Assignee
Tianjin University
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 Tianjin University filed Critical Tianjin University
Priority to CN202110317266.9A priority Critical patent/CN113011600A/en
Publication of CN113011600A publication Critical patent/CN113011600A/en
Pending legal-status Critical Current

Links

Images

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
    • G06N20/00—Machine learning
    • 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"

Landscapes

  • Engineering & Computer Science (AREA)
  • Business, Economics & Management (AREA)
  • Theoretical Computer Science (AREA)
  • Software Systems (AREA)
  • General Physics & Mathematics (AREA)
  • Strategic Management (AREA)
  • Human Resources & Organizations (AREA)
  • Physics & Mathematics (AREA)
  • Economics (AREA)
  • Development Economics (AREA)
  • Medical Informatics (AREA)
  • Mathematical Physics (AREA)
  • Computer Vision & Pattern Recognition (AREA)
  • Artificial Intelligence (AREA)
  • Evolutionary Computation (AREA)
  • Computing Systems (AREA)
  • Game Theory and Decision Science (AREA)
  • Data Mining & Analysis (AREA)
  • General Engineering & Computer Science (AREA)
  • Entrepreneurship & Innovation (AREA)
  • Marketing (AREA)
  • Operations Research (AREA)
  • Quality & Reliability (AREA)
  • Tourism & Hospitality (AREA)
  • General Business, Economics & Management (AREA)
  • Feedback Control In General (AREA)

Abstract

The invention discloses a method for correcting an evaporation waveguide prediction model based on machine learning, which comprises the following steps: step A: collecting atmospheric refractive index profile data, and analyzing evaporation waveguide parameters; and B: determining a multi-parameter evaporation waveguide prediction empirical model; and C: modifying the model by using a machine learning method, and determining optimal parameters; step D: the height of the evaporation waveguide is predicted. The invention provides a multi-parameter evaporation waveguide prediction empirical model correction method based on machine learning and capable of determining optimal parameters according to time distribution. The method organically combines a multi-parameter evaporation waveguide prediction model, an atmospheric refractive index profile and machine learning, and adopts actually measured atmospheric refractive index profile data to construct a training set and a test set, so that correction of the multi-parameter evaporation waveguide prediction empirical model is realized.

Description

Method for correcting evaporation waveguide prediction model based on machine learning
Technical Field
The invention relates to an evaporation waveguide model and the field of machine learning, in particular to a method for correcting an evaporation waveguide prediction model based on machine learning.
Background
The evaporation waveguide is a relatively common form of atmospheric waveguide, which is generated under specific meteorological hydrological conditions, and accompanied with evaporation and diffusion of sea surface water vapor, atmospheric humidity above the sea surface forms a larger humidity vertical gradient change along with height sudden reduction, and a corresponding atmospheric refractive index reduces along with height increase to generate a negative gradient change trend. The existence of the evaporative waveguide has great influence on the performance of sea surface radio communication, so that the over-the-horizon transmission can be realized. When the microwave is transmitted in the evaporation waveguide environment, the wave front is converted from spherical expansion to approximate cylindrical expansion, so that the path loss of the electric wave transmission is greatly reduced, and the forward ultra-long distance transmission is realized. Evaporative waveguides are a special atmospheric structure that often occurs at sea and have a significant effect on electromagnetic wave propagation. Meanwhile, researches prove that the evaporation waveguide can be considered to exist continuously in most sea areas, the height of the evaporation waveguide is a key parameter for determining the refractive index profile of the low-layer atmosphere, and the height indicates the strength and the thickness of the evaporation waveguide, so that the method has important practical significance for predicting the height of the evaporation waveguide.
The evaporative waveguide model is essentially an empirical model obtained based on boundary layer theory and land and sea waveguide observation experiments. With the vigorous development of the field of artificial intelligence, machine learning methods are becoming mature, the methods are effective means for extracting potential rules from data, and the purpose of machine learning is to extract the experience and rules from the data sets for predicting the future state of the data. Machine learning is widely applied in the fields of data calculation, Internet of things engineering, economics and the like, but is less applied in the research of evaporation waveguide prediction models.
In recent years, with the development of observation equipment and the progress of observation methods, scholars at home and abroad develop a plurality of large-scale marine observation experiments, and summarize novel empirical relations among similar variables of a plurality of boundary layers based on the experiments. In combination with the boundary layer similarity theory, a plurality of evaporative waveguide diagnostic models have been proposed. However, the existing modeling method of the evaporation waveguide diagnostic model is based on the boundary layer similarity theory and is also constrained by a plurality of assumptions and approximations in the boundary layer similarity theory, so that inherent errors exist in the diagnosed evaporation waveguide characteristic parameters and the atmospheric refractive index profile characteristics. The accuracy of conventional diagnostic models has its own limitations based on the influence of boundary layer theory. Machine learning is an effective way to obtain underlying laws that are completely data-based and not limited by the fundamental assumptions in the boundary layer theory.
In summary, it is necessary to provide a method for correcting an evaporation waveguide prediction model based on machine learning.
Disclosure of Invention
The application provides a method for correcting an evaporation waveguide prediction model based on machine learning, and the method can be used for obtaining the optimal parameters of a multi-parameter evaporation waveguide prediction empirical model, so that the evaporation waveguide prediction error of the multi-parameter evaporation waveguide prediction empirical model is greatly reduced.
Specifically, the method for correcting the evaporation waveguide prediction model based on machine learning provided by the invention comprises the following steps:
step A: collecting atmospheric refractive index profile data, namely data of atmospheric refractive index changing along with height, wherein the independent variable is the height, the dependent variable is the atmospheric refractive index, and accordingly, the strength and the thickness of the evaporation waveguide are analyzed according to the inflection point when the refractive index gradient is changed from negative to positive;
and B: determining a multi-parameter evaporation waveguide prediction empirical model by using an expert system;
and C: modifying the model by using a machine learning method, and determining optimal parameters;
step D: and predicting the height of the evaporation waveguide, and analyzing the intensity and the thickness of the evaporation waveguide.
Further, the air conditioner is provided with a fan,
the step A comprises the following steps:
step A1: analyzing known parameters including the atmospheric refractive index and the corresponding height thereof in the collected atmospheric refractive index profile data;
step A2: analyzing the height of the evaporation waveguide, wherein the height corresponding to the inflection point when the atmospheric refractive index gradient is changed from negative in the atmospheric refractive index profile is the height of the evaporation waveguide;
step A3: and analyzing the parameters of the evaporation waveguide according to the height of the evaporation waveguide.
Further, the air conditioner is provided with a fan,
in step a3, the evaporation waveguide parameters are intensity and thickness parameters of the evaporation waveguide.
Further, the air conditioner is provided with a fan,
the step B comprises the following steps:
step B1: carrying out preliminary characteristic analysis on the atmospheric refractive index profile data set to know known parameters and target parameters;
step B2: and selecting a multi-parameter evaporation waveguide prediction empirical model.
Further, the air conditioner is provided with a fan,
the step C comprises the following steps:
step C1: dividing the atmospheric refractive index profile data into a training data set and a testing data set, and dividing the data into n observation groups, wherein the value of n is 12 and corresponds to 12 months;
step C2: determining an evaluation index using the root mean square error as model training;
step C3: determining target variables, thickness t of evaporation waveguide, gradient k of mixed layer, and height h of evaporation waveguide topt；
Step C4: setting necessary parameters, measuring the height h of the position and the atmospheric refractive index M;
step C5: introducing atmospheric refractive index profile data for training, and obtaining an optimal parameter set by using a regression method, wherein the optimal parameter set comprises the thickness t of the evaporation waveguide, the gradient k of the mixed layer and the height h of the top of the evaporation waveguidet；
Step C6: determining optimal parameters by using the trained machine learning model; testing the corrected multi-parameter equation evaporation waveguide prediction model, and calculating the obtained atmospheric refractive index model and the height value of the evaporation waveguide; and finally obtaining n corrected multi-parameter equation evaporation waveguide prediction models in total for n time periods.
Step C7: the modified multiparameter equation evaporative waveguide prediction model is evaluated using the root mean square error.
Further, the air conditioner is provided with a fan,
the step D comprises the following steps:
step D1: drawing an atmospheric refractive index profile for the obtained n groups of parameters;
step D2: and predicting the heights of the evaporation waveguides at different times, and analyzing the intensity and the thickness of the evaporation waveguides.
Compared with the prior art, the method has the beneficial effects that the method for correcting the multi-parameter evaporation waveguide prediction empirical model based on machine learning and capable of determining the optimal parameters according to time distribution is provided. The method organically combines a multi-parameter evaporation waveguide prediction model, an atmospheric refractive index profile and machine learning, and adopts actually measured atmospheric refractive index profile data to construct a training set and a test set, so that correction of the multi-parameter evaporation waveguide prediction empirical model is realized.
Drawings
FIG. 1 is a flow chart of a method of the present application;
FIG. 2 is a flow chart of a method for collecting atmospheric refractive index profile data and analyzing evaporative waveguide parameters in accordance with the present application;
FIG. 3 is a flow chart of a method for determining optimal parameters by modifying a model using machine learning in the present application.
Detailed Description
It should be noted that the embodiments and features of the embodiments in the present application may be combined with each other without conflict.
The invention is described in further detail below with reference to the figures and specific examples. It should be understood that the specific embodiments described herein are merely illustrative of the invention and are not intended to limit the invention.
Electromagnetic wave propagation in the atmospheric waveguide is a hot point of research in recent two years, so that the method has extremely strong practical significance for analyzing atmospheric factors influencing the propagation. At present, no method based on machine learning is applied to an evaporation waveguide prediction model. The invention provides a method based on machine learning, which organically combines the machine learning and a multi-parameter evaporation waveguide prediction empirical model to deduce the optimal parameters of the multi-parameter evaporation waveguide prediction empirical model, so that the error of the multi-parameter evaporation waveguide prediction empirical model for predicting the evaporation waveguide is greatly reduced.
As shown in fig. 1, the embodiment of the present invention includes the following steps:
step A: collecting atmospheric refractive index profile data, namely data of atmospheric refractive index changing along with height, wherein the independent variable is the height, the dependent variable is the atmospheric refractive index, and accordingly, the strength and the thickness of the evaporation waveguide are analyzed according to the inflection point of the refractive index gradient changing from negative to positive;
specifically, as shown in fig. 2, step a includes the following steps:
step A1: and analyzing known parameters in the data of the marine atmospheric refractive index profile, including the atmospheric refractive index and the corresponding height thereof. For months 1-12, at the height of 0-40 m, the point interval is 0.1 m, and the corresponding atmospheric refractive index data obtains the atmospheric refractive index profile of different months;
step A2: the height corresponding to the positive inflection point when the atmospheric refractive index gradient is changed from negative in the atmospheric refractive index profile is the height of the evaporation waveguide;
step A3: analyzing the height of the evaporation waveguide to obtain the strength and thickness parameters of the evaporation waveguide;
and B: determining a multi-parameter evaporation waveguide prediction empirical model, wherein the parameters of the multi-parameter model are expressed by vectors as follows:
M(h)＝(h,t,k,ht)
where h is the height of the measurement site, t is the thickness of the evaporation waveguide, k is the slope of the mixed layer, htIs the height of the top of the evaporation waveguide;
the model can be used to express the evaporation waveguide profile and the modified index value as a function of height can be given as follows:
Figure BDA0002991658780000051
Figure BDA0002991658780000052
Figure BDA0002991658780000053
wherein the range of the height d of the evaporation waveguide is (0, 40) m, the range of the gradient k of the mixed layer is (-1, 0.4), and the height h of the top of the evaporation waveguidetThe value range is (0, 500) m.
The step B comprises the following steps:
step B1: carrying out preliminary characteristic analysis on the atmospheric refractive index profile data set to determine known parameters and target parameters;
step B2: selecting a multi-parameter evaporation waveguide prediction empirical model;
and C: correcting the model by using a machine learning method, and determining the thickness t of the evaporation waveguide, the refractive index k of the mixed layer and the top height h of the evaporation waveguidetThe optimum parameter of (2);
as shown in fig. 3, step C includes the following steps:
step C1: the atmospheric refractive index profile data is divided into a training data set and a test data set. Dividing the atmospheric refractive index profile data into observation groups of 12 different months;
wherein the 80% atmospheric refractive index profile data is a training data set, and the 20% annual atmospheric refractive index profile data is a testing data set;
step C2: determining an evaluation index using the root mean square error as model training;
step C3: determining target variables, thickness d of evaporation waveguide, gradient k of mixed layer, and height h of evaporation wave crestt；
Step C4: setting the necessary parameter h as the height of the measuring position, and setting M as the corresponding value of the atmospheric refractive index, namely, for 1-12 months, at the height of 0-40 meters, the corresponding data of the atmospheric refractive index with the point interval of 0.1 meter;
step C5: importing atmospheric refractive index profile data for training, adjusting parameters, obtaining an optimal parameter set, the thickness d of the evaporation waveguide, the gradient k of the mixed layer and the height h of the top of the evaporation waveguide by using a regression method for observation groups of 12 different monthst；
Step C6: and determining the optimal parameters by using the trained machine learning model. And for different values of h between 0 and 40 meters and at an interval of 0.1 meter, testing the corrected multi-parameter equation evaporation waveguide prediction model, and comparing the calculated height values of the atmospheric refractive index model and the evaporation waveguide with the atmospheric refractive index profile of the test data set. For 12 months, 12 corrected multi-parameter equation evaporation waveguide prediction models are finally obtained in total.
Step C7: the modified multiparameter equation evaporative waveguide prediction model is evaluated using the root mean square error.
Step D: and predicting the height of the evaporation waveguide, and analyzing the intensity and the thickness of the evaporation waveguide.
The step D specifically comprises the following steps:
step D1: for the obtained 12 groups of parameters of different months, an atmospheric refractive index profile is drawn;
step D2: and predicting the heights of the evaporation waveguides at different times, and analyzing the intensity and the thickness of the evaporation waveguides.
The technical means not described in detail in the present application are known techniques.
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 decorations can be made without departing from the principle of the present invention, and these modifications and decorations should also be regarded as the protection scope of the present invention.

Claims (6)

1. A method for correcting an evaporation waveguide prediction model based on machine learning is characterized in that,
the method comprises the following steps:
step A: collecting atmospheric refractive index profile data, namely data of atmospheric refractive index changing along with height, wherein the independent variable is the height, the dependent variable is the atmospheric refractive index, and accordingly, the strength and the thickness of the evaporation waveguide are analyzed according to the inflection point when the refractive index gradient is changed from negative to positive;
and B: determining a multi-parameter evaporation waveguide prediction empirical model by using an expert system;
and C: training a model by using a machine learning method, and determining optimal parameters;
step D: and predicting the height of the evaporation waveguide, and analyzing the intensity and the thickness of the evaporation waveguide.
2. The method for modifying an evaporative waveguide prediction model based on machine learning of claim 1,
the step A comprises the following steps:
step A1: analyzing known parameters including the atmospheric refractive index and the corresponding height thereof in the collected atmospheric refractive index profile data;
step A2: analyzing the height of the evaporation waveguide, wherein the height corresponding to an inflection point when the atmospheric refractive index gradient is changed from negative in the atmospheric refractive index profile is the height of the evaporation waveguide;
step A3: and analyzing the parameters of the evaporation waveguide according to the height of the evaporation waveguide.
3. The method for modifying an evaporative waveguide prediction model based on machine learning of claim 2,
in step a3, the evaporation waveguide parameters are intensity and thickness parameters of the evaporation waveguide.
4. The method for modifying an evaporative waveguide prediction model based on machine learning of claim 1,
the step B comprises the following steps:
step B1: carrying out preliminary characteristic analysis on the atmospheric refractive index profile data set to determine known parameters and target parameters;
step B2: and selecting a multi-parameter evaporation waveguide prediction empirical model.
5. The method for modifying an evaporative waveguide prediction model based on machine learning of claim 1,
the step C comprises the following steps:
step C1: dividing atmospheric refractive index profile data into a training data set and a testing data set, and dividing the data into n observation groups;
step C2: determining an evaluation index using the root mean square error as model training;
step C3: determining target variables, thickness t of evaporation waveguide, gradient k of mixed layer, and height h of evaporation waveguide topt；
Step C4: setting necessary parameters, measuring the height h of the position and the atmospheric refractive index M;
step C5: introducing atmospheric refractive index profile data for training, and obtaining an optimal parameter set by using a regression method, wherein the optimal parameter set comprises the thickness t of the evaporation waveguide, the gradient k of the mixed layer and the height h of the top of the evaporation waveguidet；
Step C6: determining optimal parameters by using the trained machine learning model; testing the corrected multi-parameter equation evaporation waveguide prediction model, and calculating the obtained atmospheric refractive index model and the height value of the evaporation waveguide; for n time periods, finally obtaining n corrected multi-parameter equation evaporation waveguide prediction models in total;
step C7: the modified multiparameter equation evaporative waveguide prediction model is evaluated using the root mean square error.
6. The method for modifying an evaporative waveguide prediction model based on machine learning of claim 1,
the step D comprises the following steps:
step D1: drawing an atmospheric refractive index profile for the obtained n groups of parameters;
step D2: and predicting the heights of the evaporation waveguides at different times, and analyzing the intensity and the thickness of the evaporation waveguides.
CN202110317266.9A 2021-03-25 2021-03-25 Method for correcting evaporation waveguide prediction model based on machine learning Pending CN113011600A (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202110317266.9A CN113011600A (en) 2021-03-25 2021-03-25 Method for correcting evaporation waveguide prediction model based on machine learning

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202110317266.9A CN113011600A (en) 2021-03-25 2021-03-25 Method for correcting evaporation waveguide prediction model based on machine learning

Publications (1)

Publication Number Publication Date
CN113011600A true CN113011600A (en) 2021-06-22

Family

ID=76406796

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202110317266.9A Pending CN113011600A (en) 2021-03-25 2021-03-25 Method for correcting evaporation waveguide prediction model based on machine learning

Country Status (1)

Country Link
CN (1) CN113011600A (en)

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN117421601A (en) * 2023-12-19 2024-01-19 山东省科学院海洋仪器仪表研究所 Sea surface evaporation waveguide near-future rapid forecasting method

Citations (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN102540162A (en) * 2011-12-12 2012-07-04 中国船舶重工集团公司第七二四研究所 Method for estimating low-altitude electromagnetic wave propagation characteristic on basis of sea clutter
CN111310889A (en) * 2020-01-16 2020-06-19 西北工业大学 Evaporation waveguide profile estimation method based on deep neural network

Patent Citations (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN102540162A (en) * 2011-12-12 2012-07-04 中国船舶重工集团公司第七二四研究所 Method for estimating low-altitude electromagnetic wave propagation characteristic on basis of sea clutter
CN111310889A (en) * 2020-01-16 2020-06-19 西北工业大学 Evaporation waveguide profile estimation method based on deep neural network

Non-Patent Citations (2)

* Cited by examiner, † Cited by third party
Title
廖麒翔: ""对流层大气波导反演算法研究"", 《中国优秀博硕士学位论文全文数据库(博士)基础科学辑》 *
朱啸宇: ""基于机器学习的蒸发波导预测研究"", 《中国优秀博硕士学位论文全文数据库(硕士)工程科技Ⅱ辑》 *

Cited By (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN117421601A (en) * 2023-12-19 2024-01-19 山东省科学院海洋仪器仪表研究所 Sea surface evaporation waveguide near-future rapid forecasting method
CN117421601B (en) * 2023-12-19 2024-03-01 山东省科学院海洋仪器仪表研究所 Sea surface evaporation waveguide near-future rapid forecasting method

Similar Documents

Publication Publication Date Title
CN112711899B (en) Fusion prediction method for height of evaporation waveguide
CN110245443B (en) Missile simulation model credibility comprehensive evaluation method based on mapping function
CN109299812B (en) Flood prediction method based on deep learning model and KNN real-time correction
CN108428017A (en) Wind power interval prediction method based on core extreme learning machine quantile estimate
CN111680870B (en) Comprehensive evaluation method for quality of target motion trail
CN113255986B (en) Multi-step daily runoff forecasting method based on meteorological information and deep learning algorithm
CN110213003B (en) Wireless channel large-scale fading modeling method and device
Qiu et al. Modeling the mean wind loads on cylindrical roofs with consideration of the Reynolds number effect in uniform flow with low turbulence
CN102693451A (en) Method for predicting ammonia process flue gas desulphurization efficiency based on multiple parameters
CN107607692B (en) Soil moisture monitoring and optimizing point distribution method based on maximum water storage capacity of soil
CN106845437A (en) Leakage of City Gas Pipeline localization method based on Support vector regression
CN108197394A (en) A kind of wind speed curve emulation mode
CN107657110A (en) The fatigue damage evaluation method of large scale wind power machine blade
CN108776340A (en) Forward-orbit interference synthetic aperture radar sea surface flow field inversion method based on genetic algorithm
CN104657584A (en) Lorenz-system-based wind speed prediction method
CN113011600A (en) Method for correcting evaporation waveguide prediction model based on machine learning
CN113435630A (en) Basin hydrological forecasting method and system with self-adaptive runoff yield mode
CN108736990A (en) A method of detection multi-source passive microwave data radio frequency interference
CN116029465A (en) Method, device, medium and equipment for predicting wind power before day in turning weather
CN115097451A (en) Sea wave multi-parameter inversion method and system based on SAR satellite data
WO2022032873A1 (en) Adversarial neural network-based hydrological parameter calibration method for data-lacking region
CN113868926A (en) Method for constructing spatial distribution model of water quality parameters of culture pond
WO2022032874A1 (en) Adversarial neural network-based hydrological parameter calibration method for data region
CN116541658B (en) Urban near-ground kilometer height wind profile measurement analysis method and device
CN115907178B (en) Clean ecosystem CO 2 Exchange amount prediction method

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