Ocean ambient noise prediction model for ship abnormal behavior monitoring

Abstract

The application provides a marine environment noise prediction method for ship abnormal behavior monitoring, and comprises the following steps: establishing a marine environment noise prediction model; determining the type of the marine environment noise numerical prediction model; and performing marine environment noise prediction according to the determined type of the prediction model. The application constructs a fast and refined marine environment noise prediction model, the fast model considers three noise sources of wind, rain and ships, and realizes the purpose of real-time fast prediction of marine environment noise according to the source level empirical formula of different noise sources summarized according to the measured data; the refined model considers two noise sources of wind and ships, and realizes the purpose of refined prediction of marine environment noise by using the source level empirical formula of different noise sources, combining real-time gridded wind speed and ship AIS data to complete the construction of wind and ship noise source models, and comprehensively considering the factors of gridded hydrological information and ocean depth database and calculation frequency.

Description

Technical Field

[0001] This invention belongs to the field of marine environmental noise forecasting technology, specifically relating to a marine environmental noise forecasting model for monitoring abnormal ship behavior. Background Technology

[0002] Marine environmental noise is the background sound field in the ocean, a combined contribution from various noise sources (wind, ships, rainfall, organisms, etc.) transmitted through the marine waveguide environment along different paths to the receiving point. Numerical prediction of marine environmental noise mainly involves three parts: marine environmental noise source models, marine waveguide environmental data, and sound propagation models adapted for numerical calculations. The marine environmental noise source models involve the spatial components of different noise sources and their sound source levels; the marine waveguide environmental data includes sea surface and seabed interface roughness, other sound propagation information, and changes in hydrological information within the ocean; and the sound propagation models matched to different marine waveguide environments need to comprehensively consider changes in the marine waveguide environment, calculation frequencies, and other information to ensure simulation accuracy and reduce computation time.

[0003] Currently, most marine environmental noise forecasting models focus on a single marine environmental noise source, such as wind-induced noise. Numerical forecasting combining wind, ship, and rainfall noise is supported by the RANDI model, developed in the United States decades ago. This model utilizes historical ship density and wind speed databases for noise sources; the ETOPO5 and DBDBC seabed topography databases for marine environmental information; the HOP and SWSS databases for sound velocity profiles; and the LFBL and BLUG databases for seabed acoustic parameters. A parabolic equation model is used for sound propagation. However, because the parabolic equation sound propagation model is not suitable for deep-sea and high-frequency environments, the RANDI model has limitations in its application in deep-sea high-frequency environments.

[0004] For large-scale marine environmental noise forecasting, due to the wide forecast range and large number of calculation points, there is an urgent need for a large-scale marine environmental noise forecasting model that can meet different application purposes. Summary of the Invention

[0005] To address the problems existing in the prior art, this application proposes a marine environmental noise prediction method for monitoring abnormal ship behavior, comprising the following steps:

[0006] Establish a marine environmental noise forecasting model;

[0007] Determine the type of marine environmental noise numerical prediction model to be used;

[0008] Marine environmental noise forecasting is performed based on the determined forecasting model type.

[0009] Furthermore, the establishment of a marine environmental noise forecasting model specifically includes:

[0010] A marine environmental noise source model is established, and empirical formulas for the sound source levels of wind noise, ship noise, and rainfall noise are given.

[0011] By combining gridded wind speed, rainfall information, and ship AIS information, the sound source level and location of different noise sources are determined;

[0012] Based on marine waveguide environmental data, information such as sound velocity profiles and seabed topography around the receiving point is provided.

[0013] Furthermore, the formula for the noise source level of the wind gate is:

[0014]

[0015] Where: f is the frequency; U is the wind speed (knot).

[0016] Furthermore, the formula for the noise source level of a ship is:

[0017]

[0018] Where: ls is the length of the ship (ft), and v is the speed (knot);

[0019] ;

[0020] ;

[0021] ;

[0022] ;

[0023] ;

[0024] .

[0025] Furthermore, the empirical formula for the noise source level of rainfall is:

[0026] ;

[0027] in: V The speed at which raindrops enter the water. D Where is the radius of the raindrop. n This refers to the number of raindrops per unit area and per unit time. S r0 It is the underwater average power spectrum generated by raindrops;

[0028] Then, the Logistic and Log-linear models are used to fit the frequency bands of the rain noise power spectrum to achieve the regression of the rain noise power spectrum:

[0029] ;

[0030] ;

[0031] Among them: A, B, , ,and These are model parameters.

[0032] Furthermore, marine environmental noise prediction models include:

[0033] Rapid prediction model for marine environmental noise;

[0034] A refined forecasting model for marine environmental noise.

[0035] Furthermore, the calculation process for the rapid forecasting model of marine environmental noise includes:

[0036] Calculate the sound propagation loss of the windproof noise;

[0037] Calculate the sound propagation loss of ship noise;

[0038] After calculation, the wind noise contribution from each receiving point is superimposed to obtain the wind noise and ship noise spectral levels at different frequencies.

[0039] By superimposing the regression spectrum of rainfall noise, the noise level result for rapid forecasting of marine environmental noise is obtained.

[0040] Furthermore, the formulas involved in the rapid forecasting model for marine environmental noise include:

[0041] The noise field generated by the wind gate and the ship is represented as follows:

[0042] ;

[0043] in: This is considered a wind-induced noise source; based on the statistical characteristics of ship distribution, it is assumed that... Distributed on a plane N One noise source (Different types of ships) Zs (2 different), each noise source is independent of the others, and different types of ships have different noise source levels. The location is provided by the modern ship radiated noise source level model and the location is provided by the AIS ship distribution statistics. for With then Green's function among noise sources in each region;

[0044] For fast forecasting models, in order to improve the forecasting calculation speed, the acoustic propagation loss (logarithmic scale of the Green's function) is calculated using an empirical formula;

[0045] For shallow sea environments, the empirical formula for propagation loss is expressed piecewise according to the propagation distance as follows:

[0046] when At that time, adopt ;

[0047] when At that time, adopt ;

[0048] when At that time, adopt ;

[0049] in: r For transmission distance, H Because the sea is deep, k For wave number, The medium absorption coefficient, , , , , These are the density of seawater and the speed of sound of the medium on the seabed, respectively.

[0050] For deep-sea environments, the empirical formula for propagation loss is expressed as follows based on depth:

[0051] Fresnel District: ;

[0052] Fang and Fei District ( ): ;

[0053] in, z 1. z These are the source depth and the receiving depth, respectively.

[0054] Furthermore, the calculation process for the refined marine environmental noise forecasting model includes:

[0055] Input the latitude and longitude grid, wind speed grid data, valid AIS ship data, and rainfall grid data for the sea area;

[0056] Calculate the frequency points, depth, and number of horizontal partitions;

[0057] By summing the energy of wind noise and ship noise at the same frequency at that location, the noise level for a detailed forecast of marine environmental noise can be obtained.

[0058] Furthermore, the formulas involved in fine-grained forecasting of marine environmental noise include:

[0059] Let the source level per unit area of ​​the noise source in the wind be... (dB, reference value is 1m) Then the wind noise source intensity per unit area ,use Indicates the depth Horizontal distance (in (Horizontal partition number), orientation (in The sound source at the azimuth zone number reaches the receiving position. The complex sound pressure, then the receiving point The wind-driven environmental noise field can be obtained by superimposing the sound fields generated by all noise sources on the sea surface considering random phases:

[0060] ;

[0061] in: and It is a random number uniformly distributed in [0, 2π], representing random phase information in both distance and orientation directions; area factor ;

[0062] The spatial distribution characteristics of a noise field are usually represented by the ensemble average of the complex conjugate product of the sound fields between two points, i.e., the cross-spectral density. The cross-spectral density represents the spatial characteristics of the noise field, and its definition is:

[0063] ;

[0064] For ship noise, still using the vertical receiver array as the z-axis, it is assumed that there are N independent ship noise sources distributed in the calculation area, each with a different noise source intensity. So at the receiving point The noise field generated by N ships can be represented as:

[0065] ;

[0066] in: For the noise source of the nth ship reaching the receiving point The complex sound pressure, Let the depth of the sound source of the nth ship be . Let n be the horizontal distance of the nth ship from the receiving point. Let be the bearing angle of the nth ship;

[0067] By summing the energy of wind noise and ship noise at the same frequency at a certain receiving location, the noise level of a refined forecast of marine environmental noise can be obtained.

[0068] Compared with the prior art, the advantages of this invention are as follows:

[0069] This invention constructs rapid and refined marine environmental noise forecasting models. The rapid model considers three noise sources: wind, rain, and ships. It summarizes source-level empirical formulas for different noise sources based on measured data and utilizes real-time gridded wind speed, rainfall data, and ship AIS distribution data for a specified sea area to complete the noise source model. Sound propagation is represented by empirical formulas instead of a sound propagation model, enabling real-time and rapid forecasting of marine environmental noise. The refined model considers two noise sources: wind and ships. It also utilizes source-level empirical formulas for different noise sources combined with real-time gridded wind speed and ship AIS data to complete the wind and ship noise source models. By integrating gridded hydrological information and ocean depth databases with factors such as calculation frequency, different sound propagation models are adapted to different waveguide environments and frequencies, enabling refined forecasting of marine environmental noise. Attached Figure Description

[0070] Figure 1 This invention presents a hybrid non-uniformly distributed noise source model.

[0071] Figure 2 This invention proposes a rapid forecasting model process for marine environmental noise;

[0072] Figure 3 This is a schematic diagram of the partition calculation of the model of the present invention;

[0073] Figure 4 This invention provides a flowchart for a refined marine environmental noise forecasting model.

[0074] Figure 5 The depth of the South China Sea;

[0075] Figure 6 This is the marine environmental noise forecast result I calculated by the rapid forecasting model of this invention (frequency 100Hz, receiving depth distribution of 5m).

[0076] Figure 7 This is the marine environmental noise forecast result II calculated by the rapid forecasting model of this invention (frequency 100Hz, receiving depth distribution of 50m).

[0077] Figure 8 This is the marine environmental noise forecast result III calculated by the rapid forecasting model of this invention (frequency 3200Hz, receiving depth distribution of 5m).

[0078] Figure 9 This is the marine environmental noise forecast result IV (frequency 3200Hz, receiving depth distribution of 50m) calculated by the rapid forecasting model of this invention.

[0079] Figure 10 This is the refined marine environmental noise forecast result I of the present invention (N=3, latitude and longitude grid 0.25°×0.25°, receiving depth 50m);

[0080] Figure 11 This is the refined marine environmental noise forecast result II of the present invention (N=3, latitude and longitude grid 0.25°×0.25°, receiving depth 70m). Detailed Implementation

[0081] To enable those skilled in the art to better understand the technical solution of the present invention, the present invention will be further described below.

[0082] This application proposes a marine environmental noise forecasting method for monitoring abnormal ship behavior, which mainly includes the following steps:

[0083] Establish a marine environmental noise forecasting model;

[0084] Determine the type of marine environmental noise numerical prediction model to be used;

[0085] Marine environmental noise forecasting is performed based on the determined forecasting model type.

[0086] Numerical prediction of marine environmental noise mainly involves three parts: marine environmental noise source model, marine waveguide environmental data, and sound propagation model adapted to numerical calculation.

[0087] 1. Marine environmental noise source model

[0088] The marine environmental noise source model mainly considers three noise sources: wind, ships, and rainfall.

[0089] (1) Noise from the wind gate

[0090] The primary frequency band for wind-block noise is generally considered to be 500Hz-25kHz. Its noise mechanism mainly stems from the shear force generated by wind at the sea surface, which can cause bubbles of various diameters within the waves to break, producing noise at different frequencies. The wind-block noise source level formula is based on measured environmental noise data from the South my country Sea and corrected accordingly.

[0091] (1)

[0092] Where f is the frequency, U is the wind speed, and the unit is knot.

[0093] (2) Ship noise

[0094] Shipping noise is a major contributor to noise pollution in both shallow and deep-sea areas, with the 20Hz-300Hz frequency band being particularly prominent. Below 1kHz, ship noise sources exhibit a continuous broadband spectrum and multi-line spectrum superposition caused by rotating machinery. Fishing vessel noise is also a primary source of near-shore, non-channel marine environmental noise. Typical ship noise modeling utilizes the ship noise source level formula summarized by Ross:

[0095] (2)

[0096] Where ls is the length of the ship (ft) and v is the speed (knot).

[0097] (3)

[0098] (4)

[0099] (5)

[0100] (6)

[0101] Ships are categorized into five length classes: greater than 300 meters, 200-300 meters, 100-200 meters, 50-100 meters, and less than 50 meters. Considering that the data samples for Ross's ship noise source level formula are mostly from ships of the mid-20th century, and with advancements in shipbuilding technology, especially for large ships longer than 200 meters which often employ two-stroke engines to effectively reduce propeller shaft frequency and radiated noise, this formula may no longer be applicable to modern ships. For large ships longer than 200 meters, a ship noise spectrum source level model is established using ship radiated noise data measured in the Yellow Sea channel in recent years.

[0102] (7)

[0103] (8)

[0104] (3) Rainfall noise

[0105] Rainfall at sea can significantly increase the marine environmental noise level in the 1kHz to 100kHz frequency band. From a microscopic perspective, rain noise is the sum of sound wave energy radiated by a large number of statistically independent raindrops as point sources. For a single raindrop, the underwater noise intensity is mainly related to factors such as raindrop size, shape, entry velocity, and entry angle. It can be described using the raindrop diameter, entry velocity, and underwater power spectrum generated by a single raindrop, expressed as:

[0106] (9)

[0107] inV The speed at which raindrops enter the water. D Where is the radius of the raindrop. n This refers to the number of raindrops per unit area and per unit time. Sr0 It is the underwater average power spectrum generated by raindrops.

[0108] Considering that some parameters in equation (9) are still difficult to observe and quantify effectively, this project utilizes the statistical relationship between macroscopic rainfall and underwater noise level to predict rain-induced noise. By fitting the rain noise power spectrum to frequency bands using the Logistic model and the log-linear model, the regression of the rain noise power spectrum can be achieved.

[0109] (10)

[0110] (11)

[0111] Among them, A, B, , ,and These are model parameters.

[0112] 2. Rapid forecasting model for marine environmental noise

[0113] Based on the statistical characteristics of ship density distribution at sea, and considering the existence of wind-induced noise sources such as wind-driven breaking waves, which are usually considered to be uniformly distributed below the sea surface, the rapid forecasting of marine environmental noise adopts a hybrid non-uniformly distributed noise source model. This model assumes that countless independent wind-induced noise sources of equal intensity are uniformly distributed on an infinitely large horizontal plane below the sea surface, and that there are also wind-induced noise sources at another depth below the sea surface. N There are several ship noise sources of varying intensities. For ease of description, we consider the wind noise sources to be distributed in... Ship noise sources are distributed in On a plane, without loss of generality, it can be extended to the case where there is a noise source on a plane at any depth. In fact, based on the ship classification results, it is assumed that the noise sources of the same type of ship are at the same depth.

[0114] So, The noise field generated by the wind gate and the ship can be represented as:

[0115] (12)

[0116] in, This is considered a wind-induced noise source; based on the statistical characteristics of ship distribution, it is assumed that... Distributed on a plane N One noise source (Different types of ships) Zs(2 different), each noise source is independent of the others, and different types of ships have different noise source levels. The location is provided by the modern ship radiated noise source level model and the location is provided by the AIS ship distribution statistics. for With the n Green's function between noise sources in each region.

[0117] For rapid forecasting models, to improve forecasting speed, the acoustic propagation loss (logarithmic scale of the Green's function) is calculated using an empirical formula. For shallow sea environments, the empirical formula for propagation loss is expressed piecewise according to the propagation distance:

[0118] when At that time, adopt ;

[0119] when At that time, adopt ;

[0120] when At that time, adopt .

[0121] in, r For transmission distance, H Because the sea is deep, k For wave number, The medium absorption coefficient, , , , , These are the density of seawater and the speed of sound in the seabed, respectively.

[0122] For deep-sea environments:

[0123] Fresnel District:

[0124] Fang and Fei District ( ):

[0125] in, z 1. z These are the source depth and the receiving depth, respectively.

[0126] When implementing the rapid model in programming, the latitude and longitude grid of the receiving point is first provided based on the latitude and longitude of the calculation area and the required calculation accuracy (model calculation input parameter). Sea surface wind grid data and rainfall grid data are obtained from meteorological data acquired in the project and used as input parameters for wind noise and rainfall noise calculation. The distribution of ships in the calculation area is obtained from AIS data, and AIS data with complete ship information (including ship MMSI number, length, speed, draft, latitude and longitude) is retained as input information for ship noise calculation. Ocean depth data is used as a model input parameter to determine whether there are any receiving points located on land or islands. Other input parameters also require the calculation of frequency points and receiving depth points.

[0127] 3. Fine-grained prediction model for marine environmental noise

[0128] In reality, marine environmental noise propagates from different types of noise sources to the receiving location. In addition to the temporal and spatial changes of the noise sources, changes in the marine waveguide environment (such as sound velocity profile and seabed topography) also affect marine environmental noise. Therefore, the sound propagation model is used in the fine prediction model of marine environmental noise to calculate the sound propagation loss from the sound source to the receiving point.

[0129] The noise sources in the refined marine environmental noise forecasting model mainly consider wind noise and ship noise. The refined forecasting model uses the same source model as the rapid forecasting model.

[0130] After establishing the aforementioned wind gate and ship noise source models, the refined marine environmental noise forecasting model also needs to consider environmental parameters of the calculation area, such as seabed topography and water sound velocity profiles, and select an appropriate sound propagation model based on factors such as calculation frequency to ensure the accuracy of the model calculations. Currently, the refined marine environmental noise forecasting model uses global ocean depth grid data to obtain seabed topography, with a spatial resolution of 30 arcseconds × 30 arcseconds. Parameters such as water sound velocity profiles are based on empirical data.

[0131] The ray method has unique advantages in handling high-frequency and deep-sea sound propagation calculations. Compared with other sound propagation models, the ray method is computationally simple, suitable for solving distance-dependent sound field environments, and provides intuitive graphics with clear physical meaning. The parabolic equation method has the advantages of speed and flexibility in handling distance-dependent low-frequency sound propagation. In the fine-grained prediction modeling of marine environmental noise, wind noise mainly adopts the N×2D three-dimensional approximation algorithm, using the low-frequency parabolic equation sound propagation model and the high-frequency ray method sound propagation model. Ship noise only calculates frequencies below 1000Hz, also using the low-frequency parabolic equation sound propagation model and the high-frequency ray method sound propagation model.

[0132] The model uses cylindrical coordinates. For wind-induced noise, taking a vertical receiving array at a certain location as an example, the location of the vertical receiving array is set as the Z-axis. Horizontal partitions are divided with a certain azimuth step size. The seabed topography at the center of the angle of each azimuth partition is given according to the ocean depth grid data. The sound field from the sound source to the receiving point at the center angle of each azimuth sector is calculated using the sound propagation model. The total sound field of the receiving point is obtained by summing up the data from each sector.

[0133] Let the source level per unit area of ​​the noise source in the wind be... (dB, reference value is 1m) Then the wind noise source intensity per unit area .use Indicates the depth Horizontal distance (in (Horizontal partition number), orientation (in The sound source at the azimuth zone number reaches the receiving position. The complex sound pressure, then the receiving point The wind-driven environmental noise field can be obtained by superimposing the sound fields generated by all noise sources on the sea surface considering random phases:

[0134]

[0135] in and It is a random number uniformly distributed in [0, 2π], representing random phase information in both distance and orientation directions; area factor ;

[0136] The spatial distribution characteristics of a noise field are usually represented by the ensemble average of the complex conjugate product of the sound fields between two points, i.e., the cross-spectral density. The cross-spectral density represents the spatial characteristics of the noise field, and its definition is:

[0137]

[0138] For ship noise, still using the vertical receiver array as the z-axis, it is assumed that there are N independent ship noise sources distributed in the calculation area, each with a different noise source intensity. So at the receiving point The noise field generated by N ships can be represented as:

[0139]

[0140] In the formula For the noise source of the nth ship reaching the receiving point The complex sound pressure, Let the depth of the sound source of the nth ship be . Let n be the horizontal distance of the nth ship from the receiving point. Let be the bearing angle of the nth ship.

[0141] The model input parameters also need to include the latitude and longitude grid of the calculated sea area, wind speed grid data, valid AIS ship data, rainfall grid data, calculation frequency points, calculation depth, and the number of horizontal partitions.

[0142] By summing the energy of wind noise and ship noise at the same frequency at a certain receiving location, the noise level for a refined forecast of marine environmental noise can be obtained. Following the same steps, marine environmental noise forecasts for other receiving locations can be generated.

[0143] When the frequency changes, and the locations of the sound source and receiver, as well as the marine waveguide environment between them, change, the sound propagation loss needs to be recalculated. Therefore, the fine prediction model for marine environmental noise requires more computation time than the fast prediction model.

[0144] The above description is only the best specific embodiment of the present invention, but the protection scope of the present invention is not limited thereto. Any changes or substitutions that can be easily conceived by those skilled in the art within the technical scope disclosed in the present invention should be included within the protection scope of the present invention.

Claims

1. A method for predicting marine environmental noise for monitoring abnormal ship behavior, characterized in that, Includes the following steps: Establish a marine environmental noise forecasting model; Determine the type of marine environmental noise numerical prediction model to be used; Marine environmental noise forecasting is performed based on the determined forecasting model type; The establishment of a marine environmental noise forecasting model specifically includes: A marine environmental noise source model is established, and empirical formulas for the sound source levels of wind noise, ship noise, and rainfall noise are given. By combining gridded wind speed, rainfall information, and ship AIS information, the sound source level and location of different noise sources are determined; Based on marine waveguide environmental data, the sound velocity profile and seabed topography information around the receiving point are provided. Marine environmental noise prediction models include: Rapid prediction model for marine environmental noise; A refined forecasting model for marine environmental noise; The calculation process of the rapid prediction model for marine environmental noise includes: Calculate the sound propagation loss of the windproof noise; Calculate the sound propagation loss of ship noise; After calculation, the wind noise contribution from each receiving point is superimposed to obtain the wind noise and ship noise spectral levels at different frequencies. By superimposing the regression spectrum of rainfall noise, the noise level result for rapid forecasting of marine environmental noise is obtained; The calculation process for the refined marine environmental noise prediction model includes: Input the latitude and longitude grid, wind speed grid data, valid AIS ship data, and rainfall grid data for the sea area; Calculate the frequency points, depth, and number of horizontal partitions; By summing the energy of wind noise and ship noise at the same frequency at that location, the noise level for a detailed forecast of marine environmental noise can be obtained.

2. The marine environmental noise forecasting method for monitoring abnormal ship behavior according to claim 1, characterized in that, The formula for the noise source level of a windproof door is: Where: f is the frequency; U is the wind speed (knot).

3. The marine environmental noise forecasting method for monitoring abnormal ship behavior according to claim 2, characterized in that, The formula for the noise source level of a ship is: Where: ls is the length of the ship (ft), and v is the speed (knot); ； ； ； ； ； 。 4. The marine environmental noise forecasting method for monitoring abnormal ship behavior according to claim 3, characterized in that, The empirical formula for the noise source level of rainfall is: ； in: V The speed at which raindrops enter the water. D Where is the radius of the raindrop. n This refers to the number of raindrops per unit area and per unit time. S r0 It is the underwater average power spectrum generated by raindrops; Then, the Logistic and Log-linear models are used to fit the frequency bands of the rain noise power spectrum to achieve the regression of the rain noise power spectrum: Among them: A, B, , ,and These are model parameters.

5. A method for predicting marine environmental noise based on monitoring abnormal ship behavior according to claim 1, characterized in that, The formulas involved in the rapid prediction model for marine environmental noise include: The noise field generated by the wind gate and the ship is represented as follows: ； in: This is considered a wind-induced noise source; based on the statistical characteristics of ship distribution, it is assumed that... Distributed on a plane N One noise source (Different types of ships) Zs (2 different), each noise source is independent of the others, and different types of ships have different noise source levels. The location is provided by the modern ship radiated noise source level model and the location is provided by the AIS ship distribution statistics. for With the n Green's function among noise sources in each region; For fast forecasting models, in order to improve the forecasting calculation speed, the acoustic propagation loss (logarithmic scale of the Green's function) is calculated using an empirical formula; For shallow sea environments, the empirical formula for propagation loss is expressed piecewise according to the propagation distance as follows: when At that time, adopt ; when At that time, adopt ; when At that time, adopt ; in: r For transmission distance, H Because the sea is deep, k For wave number, The medium absorption coefficient, , , , , These are the density of seawater and the speed of sound of the medium on the seabed, respectively. For deep-sea environments, the empirical formula for propagation loss is expressed as follows based on depth: Fresnel District: ; Fang and Fei District ( ): ; in, z 1. z These are the source depth and the receiving depth, respectively.

6. A method for predicting marine environmental noise based on monitoring abnormal ship behavior according to claim 1, characterized in that, The formulas involved in fine-grained forecasting of marine environmental noise include: Let the source level per unit area of ​​the noise source in the wind be... (dB, reference value is 1m) Then the wind noise source intensity per unit area ,use Indicates the depth Horizontal distance (in (Horizontal partition number), orientation (in The sound source at the azimuth zone number reaches the receiving position. The complex sound pressure, then the receiving point The wind-driven environmental noise field can be obtained by superimposing the sound fields generated by all noise sources on the sea surface considering random phases: ； in: and It is a random number uniformly distributed in [0, 2π], representing random phase information in both distance and orientation directions; area factor ; The spatial distribution characteristics of a noise field are usually represented by the ensemble average of the complex conjugate product of the sound fields between two points, i.e., the cross-spectral density. The cross-spectral density represents the spatial characteristics of the noise field, and its definition is: ； For ship noise, still using the vertical receiver array as the z-axis, it is assumed that there are N independent ship noise sources distributed in the calculation area, each with a different noise source intensity. So at the receiving point The noise field generated by N ships can be represented as: ； in: For the noise source of the nth ship reaching the receiving point The complex sound pressure, Let the depth of the sound source of the nth ship be . Let n be the horizontal distance of the nth ship from the receiving point. Let be the bearing angle of the nth ship; By summing the energy of wind noise and ship noise at the same frequency at a certain receiving location, the noise level of a refined forecast of marine environmental noise can be obtained.

Images