Prediction method of marine atmosphere correction refractive index profile

Abstract

The invention relates to the field of marine atmospheric waveguide environment detection, and discloses a method for predicting a marine atmospheric correction refractive index profile, which comprises the following steps of: acquiring meteorological observation data at different heights; calculating an evaporation waveguide height prediction value and a reference atmosphere correction refractive index profile; constructing a fusion relation model of the critical gradient of the bit refractive index and meteorological observation data; inputting meteorological observation data into the fusion relation model to obtain a critical gradient of a refractive index of a prediction bit, calculating a critical gradient of a refractive index of a reference bit, and obtaining an optimal weight coefficient of the model through iteration by using a particle swarm intelligent optimization algorithm; and inputting meteorological observation data into the optimized fusion relation model, calculating an optimized bit refractive index critical gradient, and finally obtaining an optimized atmosphere correction refractive index profile. According to the method disclosed by the invention, the influence of different meteorological conditions on the critical gradient of the refractive index is considered in the calculation process of the atmospheric correction refractive index profile, and the prediction precision of the atmospheric correction refractive index profile is improved.

Description

Technical Field

[0001] The invention relates to the field of ocean atmosphere waveguide environment detection, and in particular to a method for predicting a marine atmosphere corrected refractive index profile. Background Art

[0002] In the marine environment, evaporation and turbulent movement of the sea surface will induce the formation of evaporation ducts, and the factors such as sea waves, large-scale frontal subsidence, sea-land dry and warm air advection, and nighttime land radiation cooling will have a complex impact on the characteristics of the evaporation duct. The characteristic transformation of the evaporation duct will change the atmospheric correction refractive index profile, which will have a direct impact on the propagation of electromagnetic waves, causing abnormal refraction of electromagnetic waves near the sea surface. When electromagnetic waves propagate in the atmospheric stratification of the evaporation duct, the electromagnetic waves will be trapped in a narrow area for propagation, and a detection blind area will be formed in the upper part of the duct layer. Therefore, obtaining accurate atmospheric correction refractive index profiles near the sea surface is crucial for radar to accurately detect long-distance marine targets and predict detection blind areas in advance, so as to achieve over-the-horizon target detection.

[0003] In recent years, there have been many studies on the generation mechanism, detection means, diagnostic methods and influence of the sea surface atmospheric correction refractive index profile on radio systems. At present, the main detection method for the sea surface atmospheric correction refractive index profile is the evaporation duct model method based on the Morning-Ophoff similarity theory. By inputting the temperature, humidity, wind speed, air pressure and sea surface temperature at a certain height above the sea surface, the stability correction function and roughness parameterization method in the evaporation duct model are used to obtain the temperature, humidity and air pressure profiles and then obtain the atmospheric correction refractive index profile. In the model calculation process, the critical gradient of the potential refractive index adopts the value under the condition of neutral stability, that is, 0.125 , but failed to take into account the fact that the critical gradient of potential refractivity is affected by the meteorological conditions near the sea surface, and that the value of the critical gradient of potential refractivity is different under different meteorological conditions in different sea areas, which affects the accurate diagnosis of the corrected refractive index profile of the sea surface atmosphere. Summary of the invention

[0004] In order to solve the above technical problems, the present invention provides a prediction method for the atmospheric corrected refractive index profile at sea. In the process of calculating the atmospheric corrected refractive index profile, the influence of different meteorological conditions on the critical refractive index gradient is considered to achieve the purpose of improving the diagnostic accuracy of the atmospheric corrected refractive index profile in the target sea area.

[0005] To achieve the above object, the technical solution of the present invention is as follows: A method for predicting a marine atmospheric corrected refractive index profile comprises the following steps: Step 1: Use the gradient meteorological and hydrological sensors installed on the sea surface observation platform to obtain meteorological observation data at different heights; Step 2: Input the acquired meteorological observation data into the evaporation duct model to obtain the predicted value of the evaporation duct height, and fit the atmospheric corrected refractive index value at this height obtained from the meteorological observation data at different heights to obtain a reference atmospheric corrected refractive index profile; Step 3: Set the range of simulated meteorological parameters, input the simulated meteorological parameters into the evaporation duct model, calculate the simulated atmospheric correction refractive index profile and the simulated value of the evaporation duct height, input the simulated atmospheric correction refractive index profile and the simulated value of the evaporation duct height into the Gerstoft multi-parameter model to obtain the simulated value of the critical gradient of the potential refractive index, and construct a fusion relationship model between the critical gradient of the potential refractive index and the meteorological observation data; Step 4: input meteorological observation data into the fusion relationship model to obtain the predicted refractive index critical gradient, calculate the reference refractive index critical gradient by using the Gerstoft multi-parameter model for the predicted value of the evaporation duct height and the reference atmospheric corrected refractive index profile, calculate the fitness by taking the root mean square error between the predicted refractive index critical gradient and the reference refractive index critical gradient as the fitness function, and obtain the optimal weight coefficient of the fusion relationship model through iteration using the particle swarm intelligent optimization algorithm to obtain the optimized fusion relationship model; Step 5: Input the meteorological observation data into the optimized fusion relationship model, calculate the optimized potential refractive index critical gradient, input the optimized potential refractive index critical gradient and the evaporation duct height prediction value into the Gerstoft multi-parameter model to obtain the optimized atmospheric corrected refractive index profile.

[0006] In the above scheme, the meteorological and hydrological sensors include temperature and humidity sensors, air pressure sensors, wind speed sensors and infrared temperature sensors installed at different heights to obtain temperature and humidity, air pressure, wind speed and sea surface temperature data respectively.

[0007] In the above scheme, the evaporation waveguide model includes the PJ model, the NPS model, the BYC model, the MGB model and the pseudo-refractive index model.

[0008] In the above scheme, in step 2, the specific method for calculating the atmospheric correction refractive index values ​​at different altitudes according to the meteorological observation data at different altitudes is as follows: Height above sea level The atmospheric corrected refractive index value at Calculated by the following formula: ; ; ; in, is the sensor installation height, is the atmospheric temperature at the sensor installation height, is the air pressure at the sensor installation height, is the specific humidity at the sensor installation height, which is determined by the atmospheric humidity at the sensor installation height. It is calculated; is the ratio of the gas constant of dry air to the gas constant of water vapor, is the potential water vapor pressure, is the refractive index.

[0009] In the above scheme, in step 2, the reference atmospheric correction refractive index profile formula is obtained by fitting the atmospheric correction refractive index values ​​at different altitudes as follows: ; in, is the height, a, b, c are fitting coefficients, For height The reference atmospheric corrected refractive index profile obtained by fitting at .

[0010] In the above scheme, in step three, the simulated value of the critical gradient of the refractive index is simulated and analyzed with the temperature, humidity, wind speed, air pressure and sea surface temperature respectively to obtain a scatter plot of the simulated value of the critical gradient of the refractive index as the meteorological parameters change; a regression fitting method is used to construct a relationship model with the meteorological parameter as input and the simulated value of the critical gradient of the refractive index as output one by one, and the relationship model of the critical gradient of the refractive index with the temperature, humidity, wind speed, air pressure and sea surface temperature is fused by assigning weight coefficients to obtain a fused relationship model.

[0011] In the above scheme, the formula of the Gerstoft multi-parameter model is as follows: ; in, is the height, is the critical refractive index gradient, is the waveguide strength of the evaporated waveguide or the surface waveguide without the base layer, is the atmospheric-corrected refractive index at the sea surface, To correct the slope of the refractive index change, is the evaporation duct height, is the aerodynamic roughness factor.

[0012] In a further technical solution, in step 3, the critical gradient simulation value of the refractive index is obtained by calculating the second-order partial derivative of the Gerstoft multi-parameter model, which is calculated as follows: ; in, is the simulated value of the critical gradient of the refractive index, is the simulated value of the evaporation waveguide height.

[0013] In the above scheme, in step 3, the Monte Carlo method is used for simulation analysis to obtain a scatter plot of the change of the critical gradient simulation value of the refractive index with the change of each meteorological parameter; the obtained fusion relationship model is as follows: ; in, are weight coefficients, is the atmospheric temperature at the sensor installation height, is the atmospheric humidity at the sensor installation height, is the wind speed at the sensor installation height, is the air pressure at the sensor installation height, is the sea surface temperature.

[0014] In the above scheme, in step 4, the method of using the particle swarm intelligent optimization algorithm to iteratively obtain the optimal weight coefficient of the fusion relationship model is as follows: S1: Initialization. In the fusion relationship model, the number of weight coefficients is 5, so the number of parameters to be optimized is determined to be 5, and N 5-dimensional particles are generated as the initial particle swarm; S2: Calculate particle fitness: ; in, For the The critical gradient value of the refractive index of the reference position of each particle is For the The predicted critical refractive index gradient value for each particle, is the number of particles, For fitness; S3: Update particle speed and particle position according to particle fitness. The position of a particle is , Indicates The position of a particle in the first dimension; The speed of a particle is , Indicates The velocity of a particle in the first dimension; The optimal position of a particle is , Indicates The optimal position of a particle in the first dimension; the global optimal position searched by the particle swarm is , Indicates the optimal position of the particle swarm in the first dimension; The position and velocity of the particle after each iteration are calculated by the following formula: ; ; in, is the individual learning factor, is the group learning factor, and its value range is [0, 2]; is the inertia weight; is the number of iterations; and is a random number, ranging from [0, 1], For the The particle in The position of the iteration; No. The particle in The position of the iteration; For the The particle in The speed of iterations; No. The particle in The speed of iterations; For the The particle in The optimal position of the iteration; For the particle group The optimal position searched in the iteration; S4: Repeat steps S2 and S3 until the maximum number of iterations is reached and the The optimal position searched by the particles and the optimal position searched by the particle swarm , In The five optimal weight coefficients corresponding to the fusion relationship model.

[0015] Through the above technical solution, the method for predicting the marine atmospheric corrected refractive index profile provided by the present invention has the following beneficial effects: The present invention is based on the Monning-Ophoff similarity theory. On the basis of the traditional evaporation duct model calculation, the relationship between the critical gradient of the potential refractive index and different meteorological parameters is introduced. Compared with the traditional calculation method that takes the critical gradient of the potential refractive index as a fixed value, it is more suitable for the calculation of the atmospheric corrected refractive index profile in a complex sea surface environment. This makes the atmospheric corrected refractive index profile calculated by this method closer to the real atmospheric corrected refractive index profile in the near sea surface where the interaction between the sea and air interface is extremely complex.

[0016] The present invention takes into account all possible meteorological conditions that may actually occur in the sea area in the calculation of the critical gradient of the refractive index, and the data range is comprehensive. The relationship between the constructed critical gradient of the refractive index and different meteorological parameters is reliable. The real meteorological data of the target sea area is collected by meteorological and hydrological sensors, and the constructed fusion relationship model is optimized by an intelligent optimization algorithm related to artificial intelligence, which has strong applicability to the target sea area. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art are briefly introduced below.

[0018] Figure 1 A schematic flow chart of a method for predicting a corrected refractive index profile of the marine atmosphere disclosed in an embodiment of the present invention; Figure 2 is a scatter plot of the critical gradient of refractive index changing with humidity; Figure 3 Schematic diagram of the particle swarm intelligent optimization algorithm process. DETAILED DESCRIPTION

[0019] The technical solutions in the embodiments of the present invention will be described clearly and completely below in conjunction with the accompanying drawings in the embodiments of the present invention.

[0020] The present invention provides a method for predicting the atmospheric correction refractive index profile at sea. Figure 1 As shown, the following steps are included: Step 1: Use the gradient meteorological and hydrological sensors installed on the sea surface observation platform to obtain meteorological observation data at different heights, and record the installation heights of the temperature and humidity sensors, wind speed sensors, and air pressure sensors.

[0021] Meteorological and hydrological sensors include temperature and humidity sensors, air pressure sensors, wind speed sensors and infrared temperature sensors installed at different heights to obtain temperature and humidity respectively. ,humidity , air pressure , wind speed and sea surface temperature data.

[0022] Observation platforms include: offshore meteorological observation towers, ship observation platforms, floating observation platforms and other offshore platforms that meet the conditions for sensor installation.

[0023] The installation position of the sensor on the observation platform should avoid being blocked by objects on the observation platform as much as possible. The air pressure sensor and wind speed sensor are installed at the same height. An infrared temperature sensor is installed on each side of the observation platform to avoid the sun shining on the observation platform to produce shadows and affect the accuracy of the measurement data. To ensure the validity of the data, the infrared sensor is installed within 20 meters of the sea level as much as possible. The air pressure sensor is installed at 6m to obtain the air pressure, and the wind speed sensor is installed at 6m to obtain the wind speed.

[0024] Step 2: Input the acquired meteorological observation data into the evaporation duct model to obtain the predicted value of the evaporation duct height, and fit the atmospheric corrected refractive index value at this height obtained according to the meteorological observation data at different heights to obtain the reference atmospheric corrected refractive index profile.

[0025] The evaporation duct model is the main method for calculating the evaporation duct height and the atmospheric correction refractive index profile, including the PJ model, NPS model, BYC model, MGB model and pseudo-refractive index model. The existing evaporation duct prediction model is mainly based on the Morning-Obukhov near-sea surface atmospheric similarity theory. In the NPS model calculation, the sea surface temperature and the temperature, humidity, air pressure and wind speed at a certain height above the sea surface are used to calculate the profile of the near-sea surface temperature, specific humidity and air pressure changing with height. Then, the predicted atmospheric correction refractive index profile is obtained based on the relationship between the atmospheric correction refractive index and the temperature, specific humidity and air pressure. The predicted value of the evaporation duct height is then determined using the position of the minimum value of the predicted atmospheric correction refractive index profile.

[0026] The specific method for calculating the atmospheric correction refractive index values ​​at different altitudes based on meteorological observation data at different altitudes is as follows: Height above sea level The atmospheric corrected refractive index at Calculated by the following formula: ; ; ; in, is the sensor installation height, is the atmospheric temperature at the sensor installation height, is the air pressure at the sensor installation height, is the specific humidity at the sensor installation height, which is determined by the atmospheric humidity at the sensor installation height. It is calculated; is the ratio of the gas constant of dry air to the gas constant of water vapor, is the potential water vapor pressure, is the refractive index.

[0027] The reference atmospheric corrected refractive index profile formula is obtained by fitting the atmospheric corrected refractive index values ​​at different altitudes as follows: ; in, is the height, a, b, c are fitting coefficients, For height The reference atmospheric corrected refractive index profile obtained by fitting at .

[0028] Step 3: Set the range of simulated meteorological parameters, input the simulated meteorological parameters into the evaporation duct model, calculate the simulated atmospheric correction refractive index profile and the simulated value of the evaporation duct height, input the simulated atmospheric correction refractive index profile and the simulated value of the evaporation duct height into the Gerstoft multi-parameter model to obtain the simulated value of the critical gradient of the potential refractive index, and construct a fusion relationship model between the critical gradient of the potential refractive index and the meteorological observation data; Nonlinear least squares fitting is a data fitting method used to estimate nonlinear model parameters. The critical gradient simulation value of the potential refractive index is obtained by fitting the Gerstoft multi-parameter model and the simulated atmospheric corrected refractive index profile.

[0029] The formula of the Gerstoft multi-parameter model is as follows: ; in, is the height, is the critical refractive index gradient, is the waveguide strength of the evaporated waveguide or the surface waveguide without the base layer, is the atmospheric-corrected refractive index at the sea surface, To correct the slope of the refractive index change, is the evaporation duct height, is the aerodynamic roughness factor, usually assumed to be 0.00015.

[0030] The critical gradient simulation value of the refractive index is obtained by calculating the second-order partial derivative of the Gerstoft multi-parameter model, which is calculated as follows: ; in, is the simulated value of the critical gradient of the refractive index, is the simulated value of the evaporation waveguide height.

[0031] The method for constructing the fusion model is as follows: the simulated value of the critical gradient of the refractive index is simulated and analyzed with the temperature, humidity, wind speed, air pressure and sea surface temperature respectively to obtain a scatter plot of the simulated value of the critical gradient of the refractive index with the change of each meteorological parameter; a relationship model with the meteorological parameter as input and the simulated value of the critical gradient of the refractive index as output is constructed one by one using the regression fitting method; the relationship model with the simulated value of the critical gradient of the refractive index and the temperature, humidity, wind speed, air pressure and sea surface temperature is fused by assigning weight coefficients to obtain a fusion relationship model.

[0032] Specifically, the Monte Carlo method is used for simulation analysis, and the input range of each meteorological and hydrological parameter is set. The temperature parameter range is set to [20, 40], the unit is ℃, the parameter spacing is 2, the humidity range is set to [50, 90], the unit is %, the parameter spacing is 10, the air pressure parameter range is set to [900, 1100], the unit is hPa, the parameter spacing is 100, the wind speed parameter range is set to [0, 16], the unit is m / s, the parameter spacing is 1, and the air-sea temperature difference parameter range is set to [-5, 5], the unit is ℃, and the parameter spacing is 1. Considering the errors of the sensors measuring meteorological and hydrological parameters, different degrees of disturbance are added to the input meteorological and hydrological parameters. The temperature input is 0.1°C, the humidity input is 1%, the wind speed input is 0.5m / s, the air pressure input is 0.2hpa, and the sea surface temperature input is 1°C; a scatter plot of the simulated value of the potential refractive index critical gradient with the change of each meteorological parameter is obtained. As an example, the critical gradient of refractive index is The scatter plot of humidity changes is as follows: Figure 2 As shown, the relationship function is constructed by fitting the function curve according to the trend of scatter point changes.

[0033] The fusion relationship model finally obtained by regression fitting method is as follows: ; in, are weight coefficients, is the atmospheric temperature at the sensor installation height, is the atmospheric humidity at the sensor installation height, is the wind speed at the sensor installation height, is the air pressure at the sensor installation height, is the sea surface temperature.

[0034] Step 4: Input the meteorological observation data into the fusion relationship model to obtain the predicted refractive index critical gradient. The predicted value of the evaporation duct height and the reference atmospheric corrected refractive index profile are calculated through the Gerstoft multi-parameter model to obtain the reference refractive index critical gradient. The root mean square error between the predicted refractive index critical gradient and the reference refractive index critical gradient is used as the fitness function to calculate the fitness. The particle swarm intelligent optimization algorithm is used to iteratively obtain the optimal weight coefficient of the fusion relationship model to obtain the optimized fusion relationship model.

[0035] The particle swarm optimization algorithm is a type of evolutionary algorithm and a parallel computing algorithm. The algorithm starts from a random solution, searches for the optimal solution through an iterative method, evaluates the quality of the solution found through fitness, and searches for the global optimal value by following the currently searched optimal value. The algorithm rules are relatively simple and are suitable for finding multiple optimal weight coefficients in this fusion relationship model.

[0036] like Figure 3 As shown in the figure, the method of obtaining the optimal weight coefficient of the fusion relationship model through iteration using the particle swarm intelligent optimization algorithm is as follows: S1: Initialization. In the fusion relationship model, the number of weight coefficients is 5, so the number of parameters to be optimized is determined to be 5, and N 5-dimensional particles are generated as the initial particle swarm; S2: Calculate particle fitness: ; in, For the The critical gradient value of the refractive index of the reference position of each particle is For the The predicted critical refractive index gradient value for each particle, is the number of particles, For fitness; S3: Update particle speed and particle position according to particle fitness. The position of a particle is , Indicates The position of a particle in the first dimension; The speed of a particle is , Indicates The velocity of a particle in the first dimension; The optimal position of a particle is , Indicates The optimal position of a particle in the first dimension; the global optimal position searched by the particle swarm is , Indicates the optimal position of the particle swarm in the first dimension; The position and velocity of the particle after each iteration are calculated by the following formula: ; ; in, is the individual learning factor, is the group learning factor, and its value range is [0, 2]; is the inertia weight; is the number of iterations; and is a random number, ranging from [0, 1], For the The particle in The position of the iteration; No. The particle in The position of the iteration; For the The particle in The speed of iterations; No. The particle in The speed of iterations; For the The particle in The optimal position of the iteration; For the particle group The optimal position searched in the iteration; S4: Repeat steps S2 and S3 until the maximum number of iterations is reached and the The optimal position searched by the particles and the optimal position searched by the particle swarm , In The five optimal weight coefficients corresponding to the fusion relationship model.

[0037] Step 5: Input the meteorological observation data into the optimized fusion relationship model, calculate the optimized potential refractive index critical gradient, input the optimized potential refractive index critical gradient and the evaporation duct height prediction value into the Gerstoft multi-parameter model to obtain the optimized atmospheric corrected refractive index profile.

[0038] Meteorological observation data are collected through meteorological and hydrological sensors, and the optimal weight coefficient obtained by the particle swarm optimization algorithm is verified by measured data to determine the critical gradient of the refractive index after the particle swarm optimization algorithm is optimized. After the fusion relationship model associated with different meteorological parameters meets the actual application requirements, the critical gradient of the refractive index is calculated according to the optimized fusion relationship model through the meteorological observation data collected by the sensor. , the critical gradient of the refractive index calculated by the fusion relationship model The predicted value of the evaporation duct height calculated by the evaporation duct model is input into the Gerstoft multi-parameter model to obtain a more accurate atmospheric corrected refractive index profile.

[0039] In the embodiment, according to the actual observation data, the atmospheric correction refractive index profile prediction method based on the dynamic potential refractive index critical gradient proposed in the present invention is used to construct the potential refractive index critical gradient. The fusion relationship model with different meteorological parameters is used to obtain the critical gradient of the potential refractive index under different meteorological conditions. , theoretically related to the critical refractive index gradient Compared with a fixed value, the calculation is more reasonable in complex sea surface environments.

[0040] The above description of the disclosed embodiments enables one skilled in the art to implement or use the present invention. Various modifications to these embodiments will be apparent to one skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the present invention. Therefore, the present invention will not be limited to the embodiments shown herein, but rather to the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. A method for predicting the atmospheric corrected refractive index profile at sea, characterized in that: The steps include: Step 1: Use the gradient meteorological and hydrological sensors installed on the sea surface observation platform to obtain meteorological observation data at different heights; Step 2: Input the acquired meteorological observation data into the evaporation duct model to obtain the predicted value of the evaporation duct height, and fit the atmospheric corrected refractive index value at this height obtained from the meteorological observation data at different heights to obtain a reference atmospheric corrected refractive index profile; Step 3: Set the range of simulated meteorological parameters, input the simulated meteorological parameters into the evaporation duct model, calculate the simulated atmospheric correction refractive index profile and the simulated value of the evaporation duct height, input the simulated atmospheric correction refractive index profile and the simulated value of the evaporation duct height into the Gerstoft multi-parameter model to obtain the simulated value of the critical gradient of the potential refractive index, and construct a fusion relationship model between the critical gradient of the potential refractive index and the meteorological observation data; Step 4: input meteorological observation data into the fusion relationship model to obtain the predicted refractive index critical gradient, calculate the reference refractive index critical gradient by using the Gerstoft multi-parameter model for the predicted value of the evaporation duct height and the reference atmospheric corrected refractive index profile, calculate the fitness by taking the root mean square error between the predicted refractive index critical gradient and the reference refractive index critical gradient as the fitness function, and obtain the optimal weight coefficient of the fusion relationship model through iteration using the particle swarm intelligent optimization algorithm to obtain the optimized fusion relationship model; Step 5: Input the meteorological observation data into the optimized fusion relationship model, calculate the optimized potential refractive index critical gradient, input the optimized potential refractive index critical gradient and the evaporation duct height prediction value into the Gerstoft multi-parameter model to obtain the optimized atmospheric corrected refractive index profile.

2. The method for predicting a marine atmospheric corrected refractive index profile according to claim 1, characterized in that: The meteorological and hydrological sensors include temperature and humidity sensors, air pressure sensors, wind speed sensors and infrared temperature sensors installed at different heights to obtain temperature and humidity, air pressure, wind speed and sea surface temperature data respectively.

3. The method for predicting a marine atmospheric corrected refractive index profile according to claim 1, characterized in that: The evaporation waveguide models include PJ model, NPS model, BYC model, MGB model and pseudo-refractive index model.

4. The method for predicting a marine atmospheric corrected refractive index profile according to claim 1, characterized in that: In step 2, the specific method for calculating the atmospheric correction refractive index values ​​at different altitudes based on the meteorological observation data at different altitudes is as follows: Height above sea level The atmospheric corrected refractive index value at Calculated by the following formula: ； ； ； in, is the sensor installation height, is the atmospheric temperature at the sensor installation height, is the air pressure at the sensor installation height, is the specific humidity at the sensor installation height, which is determined by the atmospheric humidity at the sensor installation height. It is calculated; is the ratio of the gas constant of dry air to the gas constant of water vapor, is the potential water vapor pressure, is the refractive index.

5. The method for predicting a marine atmospheric corrected refractive index profile according to claim 1, characterized in that: In step 2, the reference atmospheric corrected refractive index profile formula is obtained by fitting the atmospheric corrected refractive index values ​​at different altitudes as follows: ； in, is the height, a, b, c are fitting coefficients, For height The reference atmospheric corrected refractive index profile obtained by fitting at .

6. The method for predicting the marine atmospheric corrected refractive index profile according to claim 2, characterized in that: In step three, the simulated value of the critical gradient of the refractive index is simulated and analyzed with temperature, humidity, wind speed, air pressure and sea surface temperature respectively to obtain a scatter plot of the simulated value of the critical gradient of the refractive index changing with the change of each meteorological parameter; the regression fitting method is used to construct a relationship model with the meteorological parameter as input and the simulated value of the critical gradient of the refractive index as output one by one, and the relationship model of the simulated value of the critical gradient of the refractive index with temperature, humidity, wind speed, air pressure and sea surface temperature is fused by assigning weight coefficients to obtain a fused relationship model.

7. The method for predicting the marine atmospheric corrected refractive index profile according to claim 1, characterized in that: The formula of the Gerstoft multi-parameter model is as follows: ； in, is the height, is the critical refractive index gradient, is the waveguide strength of the evaporated waveguide or the surface waveguide without the base layer, is the atmospheric-corrected refractive index at the sea surface, To correct the slope of the refractive index change, is the evaporation duct height, is the aerodynamic roughness factor.

8. The method for predicting the marine atmospheric corrected refractive index profile according to claim 7, characterized in that: In step 3, the critical gradient simulation value of the refractive index is obtained by taking the second-order partial derivative of the Gerstoft multi-parameter model, which is calculated as follows: ； in, is the simulated value of the critical gradient of the refractive index, is the simulated value of the evaporation waveguide height.

9. The method for predicting the marine atmospheric corrected refractive index profile according to claim 6, characterized in that: In step 3, the Monte Carlo method is used for simulation analysis to obtain a scatter plot of the simulated value of the critical gradient of the refractive index as the meteorological parameters change; the obtained fusion relationship model is as follows: ； in, are weight coefficients, is the atmospheric temperature at the sensor installation height, is the atmospheric humidity at the sensor installation height, is the wind speed at the sensor installation height, is the air pressure at the sensor installation height, is the sea surface temperature.

10. A method for predicting the marine atmospheric corrected refractive index profile according to claim 9, characterized in that: In step 4, the method of using the particle swarm intelligent optimization algorithm to iteratively obtain the optimal weight coefficient of the fusion relationship model is as follows: S1: Initialization. In the fusion relationship model, the number of weight coefficients is 5, so the number of parameters to be optimized is determined to be 5, and N 5-dimensional particles are generated as the initial particle swarm; S2: Calculate particle fitness: ； in, For the The critical gradient value of the refractive index of the reference position of each particle is For the The predicted critical refractive index gradient value for each particle, is the number of particles, For fitness; S3: Update particle speed and particle position according to particle fitness. The position of a particle is , Indicates The position of a particle in the first dimension; The speed of a particle is , Indicates The velocity of a particle in the first dimension; The optimal position of a particle is , Indicates The optimal position of a particle in the first dimension; the global optimal position searched by the particle swarm is , Indicates the optimal position of the particle swarm in the first dimension; The position and velocity of the particle after each iteration are calculated by the following formula: ； ； in, is the individual learning factor, is the group learning factor, and its value range is [0, 2]; is the inertia weight; is the number of iterations; and is a random number, ranging from [0, 1], For the The particle in The position of the iteration; No. The particle in The position of the iteration; For the The particle in The speed of iterations; No. The particle in The speed of iterations; For the The particle in The optimal position of the iteration; For the particle group The optimal position searched in the iteration; S4: Repeat steps S2 and S3 until the maximum number of iterations is reached and the The optimal position searched by the particles and the optimal position searched by the particle swarm , In The five optimal weight coefficients corresponding to the fusion relationship model.

Images