Method, device, medium and program for rapidly forecasting noise of underwater vehicle propeller based on excitation source reconstruction

Abstract

The invention discloses an underwater vehicle propeller noise rapid forecasting method and device based on excitation source reconstruction, a medium and a program, and belongs to the technical field of noise forecasting. According to the method, based on basic theories such as invariance of a vibration transfer function relation and a sound wave energy superposition principle, a vibration source reconstruction technology is introduced into propeller noise forecasting, and vibration responses obtained through measurement of a limited number of vibration sensors on a propeller stern structure are used as input; reconstructing excitation source matching structure surface vibration based on a vibration transfer function, and solving radiation noise caused by propeller shaft boat structure vibration generated by the structure vibration through the reconstructed excitation source; and establishing a rapid prediction method for direct sound production of the propeller under different coherences, and realizing rapid prediction and evaluation of propeller shaft boat coupling noise and direct sound production propeller noise based on an energy superposition principle. The method is suitable for ship real-time monitoring of propeller shafting vibration and radiation noise.

Description

Technical Field

[0001] The present invention belongs to the technical field of noise prediction, and particularly relates to a method, device, medium and program for rapidly predicting the noise of an underwater vehicle propeller based on excitation source reconstruction. Background Technique

[0002] The vibration and acoustic radiation noise of an underwater vehicle are important indicators for evaluating its performance. The increase in vibration noise not only increases the risk of being detected by the opponent, but also interferes with the accuracy of its own navigation and guidance systems and other equipment, thereby reducing its working ability and survivability. Therefore, we must attach importance to the vibration noise problem of underwater vehicles. A large number of studies have shown that the noise generated by the propeller device is one of the main sources of the vibration noise of underwater vehicles, and the propeller noise mainly includes the noise generated by the propeller pulsating pressure exciting the hull through the shafting or directly exciting the hull, the propeller rotation noise and the cavitation noise. The rotation noise of the propeller is the noise generated due to its operation in an uneven flow field, and its frequency is mainly determined by the rotation frequency of the blade. Once the shafting reaches the cavitation critical speed, the propeller noise becomes the main noise source of the underwater vehicle, seriously affecting its stealth. Therefore, the vibration reduction and noise reduction design of underwater vehicles is crucial for their development.

[0003] Understanding the vibration and acoustic characteristics of underwater vehicles is the premise for their vibration reduction and noise reduction design. At present, the research on the noise of underwater vehicle propellers is mainly based on computer simulation calculations, which require a lot of time for modeling and calculation, and the prediction results can only reflect the vibration and acoustic characteristics of the propeller-shaft-hull to a certain extent, lacking the verification of experimental data and being difficult to obtain strong support. Therefore, inventing a method for testing the vibration of the propeller-shaft-hull of an underwater vehicle and predicting the propeller noise is crucial for understanding the propeller noise of underwater vehicles, thus providing support for the development of the vibration reduction and noise reduction design of underwater vehicles. Summary of the Invention

[0004] The purpose of the present invention is to provide a method, device, medium and program for rapidly predicting the noise of an underwater vehicle propeller based on excitation source reconstruction, aiming to provide a method and application device for rapidly predicting the noise of ship propellers and solve the problems of real-time monitoring of the vibration of the propeller shafting of ships and rapid assessment of the radiated noise.

[0005] The purpose of the present invention is achieved by the following technical solutions:

[0006] A method for rapidly predicting the noise of an underwater vehicle propeller based on excitation source reconstruction includes the following steps:

[0007] Step 1: For the noise caused by the propeller exciting the stern structure, determine the positions of the vibration sensing measurement points, the positions of the reconstructed sources, and the positions of the acoustic field assessment points;

[0008] Step 2: According to the positions of the vibration perception measurement points and the reconstruction source obtained in Step 1, calculate the vibration transfer function from the reconstruction source to the monitoring points on the stern structure through simulation;

[0009] Step 3: According to the positions of the acoustic field assessment points obtained in Step 1, obtain the acoustic-vibration transfer function from the reconstruction source to the acoustic field assessment points based on the acoustic finite element method;

[0010] Step 4: Based on the transfer functions obtained in Steps 2 and 3, establish an acoustic field reconstruction model based on the reconstruction source according to the invariance of the transfer function;

[0011] Step 5: By measuring the response at the actual shell vibration measurement points, based on the vibration transfer function from the reconstruction source to the outer surface of the main body, estimate the excitation magnitude of the reconstruction source, obtain the vibration response of the stern structure excited by the thruster, and calculate the noise of the stern structure excited by the thruster;

[0012] Step 6: Establish a direct sound calculation model of the thruster, and obtain the direct sound noise of the thruster according to the ship speed and rotational speed;

[0013] Step 7: Perform energy superposition on the noise of the stern structure excited by the thruster obtained in Step 5 and the direct sound noise obtained in Step 6 to complete the thruster noise calculation.

[0014] Furthermore, for the vibration transfer function from the main vibration source to the stern structure obtained through simulation calculation in Step 1, establish an inverse sensing network model based on the Bi-LSTM neural network;

[0015] Establish the vibration transfer relationship between the vibration perception points on the shafting and the surface of the stern shell through the Bi-LSTM neural network, realize the vibration reconstruction of the stern structure, and determine the number, position and magnitude of the reconstruction excitation sources;

[0016] The magnitude of the reconstruction excitation source is specifically:

[0017] Regard the thruster structure as a linear system, and establish the reconstruction excitation source p j The excitation vector of is a P (p j , ω) and the vibration acceleration response A i of the measurement point r P (r i , ω) The mutual relationship between them;

[0018] A P (r i , ω) = H vP (r i p j , ω) * a P (p j , ω)

[0019] Among them, H vP (r i p j , ω) is the vibration mapping relationship matrix of dimension M*K between the reconstruction excitation source p j and the measuring point r of the stern structure i ;

[0020] Introduce the generalized inverse matrix to obtain the excitation source vector:

[0021] a P (p j , ω) = [[H vP (r i p j , ω)] H [H vP (r i p j , ω)] -1 [H vP (r i p j , ω)] H *A P (r i , ω).

[0022] Furthermore, in step 2, establish the geometric model of the underwater vehicle, including the propeller, the propeller shaft and the stern structure of the hull; place the geometric model of the underwater vehicle in the fluid domain, and establish the fluid domain grid model and the structural finite element model according to the geometric model of the underwater vehicle respectively;

[0023] According to the reconstruction source and the vibration sensing measuring point positions established in step 1, evenly distribute the monitoring points on the stern structure of the hull. By applying a unit force at the reconstruction source position, obtain the vibration responses of each vibration sensing point. The vibration responses are the vibration transfer functions from the reconstruction source to each monitoring point of the stern structure.

[0024] Furthermore, in step 3, according to the sound field assessment points established in step 1, set the sound field assessment points in the model flow field domain. By applying a unit force at the reconstruction source position, obtain the sound pressure magnitude at the sound field assessment points based on the acoustic finite element method. At this time, the sound pressure at the assessment points is the sound-vibration transfer function from the reconstruction source to the sound field assessment points.

[0025] Furthermore, in step 4, convert the radiation sound power calculation formula into the sound pressure calculation formula, establish the relationship between the sound pressure and the transfer function, and obtain the radiation sound pressure calculation formula under the excitation of the propeller on the stern structure:

[0026] P(Q) = H 1 pa |a 1 | + H 2pa |a 2 |+...H j pa |a P (p j ,ω)|

[0027] where P(Q) is the sound pressure of the noise generated by the thruster exciting the hull structure, and H j pa is the vibro-acoustic transfer function from the encoded reconstructed source to the sound field assessment point, and thus a sound field reconstruction model based on the reconstructed source is established; a P (p j ,ω) is the excitation source vector.

[0028] Further, the step 6 is as follows:

[0029] Determine the rotational speed of the underwater vehicle thruster;

[0030]

[0031] In the formula, n max is the maximum rotational speed of the thruster, v is the linear velocity of the thruster, and v max is the maximum linear velocity of the thruster; the f q value of the propeller blade frequency q of the thruster and its second and third tuning frequency values are calculated according to the following formula:

[0032] f q = qnZ, Hz, q = 1, 2, 3

[0033] In the formula, f q is the thruster frequency, q is the propeller blade frequency of the thruster, n is the rotational speed of the thruster, and Z is the number of revolutions;

[0034] The radiation levels of the direct sound P Bq of the thruster and its second and third tuning frequencies are calculated according to the following formula:

[0035]

[0036] In the formula, F xq is the propeller blade frequency of the thruster in the X direction, F xq is the propeller blade frequency of the thruster in the Y direction, F xq is the propeller blade frequency of the thruster in the Z direction, R 1 is the maximum radius of the thruster, and P Bq is the direct sound of the thruster.

[0037] Further, the step 7 is as follows:

[0038] Based on the principle of energy superposition, the thruster noise considering the direct sound is obtained according to the following formula:

[0039] P all = P(Q) + P Bq

[0040] where P(Q) is the sound pressure of the noise generated by the propeller exciting the hull structure, and P Bq is the direct sound of the propeller.

[0041] A rapid prediction device for the propeller noise of an underwater vehicle based on excitation source reconstruction, comprising a data acquisition system and a prediction system;

[0042] The data acquisition system is used to send the acquired data to the prediction system;

[0043] The prediction system is used to obtain the real-time propeller noise level and form a storage log.

[0044] A computer-readable storage medium, on which a computer program / instructions are stored, and when the computer program / instructions are executed by a processor, the steps of a rapid prediction method for the propeller noise of an underwater vehicle based on excitation source reconstruction are implemented.

[0045] A computer program product, comprising a computer program / instructions, and when the computer program / instructions are executed by a processor, the steps of a rapid prediction method for the propeller noise of an underwater vehicle based on excitation source reconstruction are implemented.

[0046] The beneficial effects of the present invention are as follows:

[0047] The method of the present invention is based on basic theories such as the invariance of the vibration transfer function relationship and the principle of acoustic wave energy superposition. The vibration source reconstruction technology is introduced into the propeller noise prediction. Without the need to know the information of the real excitation source, the vibration response measured by a limited number of vibration sensors on the stern structure of the propeller is used as the input. Based on the vibration transfer function, the excitation source is reconstructed to match the surface vibration of the structure. The superposition of the reconstructed excitation source vibrations exactly matches the above-mentioned measured structure vibrations. Through the reconstructed excitation source, the radiation noise caused by the structural vibration generated by the propeller shaft hull structure vibration can be solved. With a small number of monitoring sensors in a wide frequency band, the surface vibration of the structure can be matched, and the problem of large demand for the number of sensors is solved. Considering the influence of coherence, a rapid prediction method for the direct sound of the propeller under different coherences is established, and based on the principle of energy superposition, a rapid prediction and evaluation of the propeller noise including the propeller shaft hull coupling noise and the direct sound of the propeller are realized.

[0048] Based on the above method, the present invention also forms a set of devices for quickly predicting the vibration of the stern structure of a thruster and the noise of the thruster. The prediction device consists of two parts: a prediction system and a data acquisition system. The data acquisition system mainly consists of a signal acquisition system and a high-precision vibration acceleration sensor. The exterior of the prediction system is made of aluminum alloy material, and inside it is equipped with signal acquisition channels, a signal source card, an industrial control mainboard, and a storage hard disk.

[0049] The present invention is applicable to real-time monitoring of the vibration of the thruster shafting and the radiated noise of a ship. BRIEF DESCRIPTION OF THE DRAWINGS

[0050] Figure 1 It is a schematic diagram of the calculation process of a method for predicting thruster noise based on excitation source reconstruction in an embodiment of the present invention;

[0051] Figure 2 It is a schematic diagram of the specific execution of a method for predicting thruster noise based on excitation source reconstruction and its application in an embodiment of the present invention;

[0052] Figure 3 It is a schematic diagram of the evaluation result of the thruster noise of an underwater vehicle in an embodiment of the invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0053] The present invention will be further described below with reference to the accompanying drawings.

[0054] Embodiment 1, this embodiment is an embodiment of a method for quickly predicting the thruster noise of an underwater vehicle based on excitation source reconstruction, specifically including:

[0055] Figure 1 It is a flowchart of a method for predicting thruster noise based on excitation source reconstruction in an embodiment of the present invention.

[0056] As Figure 1 shown, the method for predicting thruster noise based on excitation source reconstruction includes the following steps:

[0057] In step S1, for the noise caused by the excitation of the thruster on the stern structure, first, clarify the positions of the vibration sensing measurement points, the positions of the reconstructed sources, and the positions of the acoustic field assessment points.

[0058] In step S101, for the thruster bearing force, the propulsion shafting and its supporting structure are the transmission paths of the excitation force, and the hull is the acoustic radiator that finally generates radiated noise. Therefore, evenly distribute the vibration sensing points on the surface of the stern structure, and establish a vibration sensing layout scheme for the thruster with limited points based on gradually reducing the sensing points.

[0059] In step S102, acceleration sensors are arranged on the surface of the stern structure of the underwater vehicle to measure the vibration responses of the vehicle at different speeds and rotational speeds. The vibration measurement points are numbered in sequence, i = 1:M, and the measured vibration acceleration is set as a P (p j , ω), which is used as the data input for the prediction system.

[0060] In step S103, in the problem of coupled vibration and acoustic radiation of the propeller, propeller shafting and hull, the bearing force of the propeller is the excitation source for generating radiation noise, the propeller shafting and its supporting structure are the transmission paths of the excitation force, and the hull is the acoustic radiator that finally generates radiation noise. Therefore, the positions of the stern bearing, thrust bearing, die head, etc. are the main vibration sources. The vibration transfer function from the main vibration sources obtained through simulation calculation to the stern structure is used to establish an inverse sensing network model based on the Bi-LSTM neural network. The vibration transfer relationship between the vibration sensing points on the shafting and the surface of the stern shell is established through the Bi-LSTM neural network, and the computer is trained to realize the vibration reconstruction of the stern structure. Thus, the number and position of the reconstruction sources are determined through machine learning. The number of reconstruction sources is set as j and numbered, j = 1:K.

[0061] In step S104, the magnitude of the reconstruction excitation source is determined according to the following steps. The propeller structure is regarded as a linear system, and the acceleration of the excitation source pj on the structure of the vibration transfer function from the reconstruction source to the hull structure surface is a P (p j , ω) and the vibration acceleration response A i of the measurement point r P (r i , ω) on the stern structure.

[0062] A P (r i , ω) = H vP (r i p j , ω) * a P (p j , ω)

[0063] H vP (r i p j , ω) is the vibration mapping relationship matrix of dimension M*K between the excitation point p j and the measurement point r i on the stern structure.

[0064] Furthermore, in practical engineering applications, to meet the unique solution condition, a generalized inverse matrix is introduced to obtain the excitation source vector:

[0065] a P (p j, ω) = [[H vP (r i p j , ω)] H [H vP (r i p j , ω)]] -1 [H vP (r i p j , ω)] H *A P (r i , ω)

[0066] Based on the established vibration mapping relationship, a generalized inverse matrix is introduced to obtain the excitation source vector a P (p j , ω). In order to obtain a P (p j , ω) as accurately as possible, keep the number of measurement points i on the shafting system greater than or equal to the number of reconstructed main excitation sources to avoid the occurrence of ill-conditioned equations, and introduce the least squares theory to obtain the best approximation solution of the system excitation source.

[0067] In step S105, according to actual requirements, the positions of the sound pressure assessment points are generally selected at the transverse and circumferential positions of the vehicle. The circumferential radius is set to 10 meters away from the outside of the hull, and the sound pressure assessment points are arranged in the athwartship direction on the starboard side of the vehicle.

[0068] To sum up, in step S1 for the coupled vibration and sound radiation problems of the propeller - propeller shafting - hull, the propeller bearing force is the excitation source generating the radiation noise, the propeller shafting and its supporting structure are the transmission paths of the excitation force, and the hull is the sound radiator finally generating the radiation noise. Therefore, vibration sensors are evenly arranged on the surface of the stern structure, the vibration transfer function from the main vibration source to the hull structure is obtained based on numerical simulation, and an inverse sensing network model is established based on the Bi-LSTM neural network. The number and distribution of the excitation sources are reconstructed through machine learning to complete the establishment of the reconstructed source model.

[0069] In step S2, the vibration transfer function from the reconstructed source to the monitoring points on the stern structure is obtained through simulation calculation.

[0070] In step S201, a structural geometric model of the propeller-shaft-stern part (i.e., including the propeller, propeller shaft, and hull stern) of the underwater vehicle is established according to the preset drawing data. The structural geometric model of the propeller-shaft-stern part of the underwater vehicle is placed in the fluid domain, and a fluid domain grid model and a structural finite element model are respectively established based on the structural geometric model of the propeller-shaft-stern part of the underwater vehicle. It should be noted that the distance between the inlet of the fluid domain and the bow of the structural geometric model of the propeller-shaft-stern part of the underwater vehicle is greater than 6L, the distance between the outlet of the fluid domain and the stern of the structural geometric model of the propeller-shaft-stern part of the underwater vehicle is greater than 10L, and the distances between the remaining boundaries of the fluid domain and the geometric model of the underwater vehicle are greater than 3L, where L is the total length of the propeller-shaft-stern structure of the underwater vehicle.

[0071] In step S202, monitoring points are evenly distributed on the stern structure of the model. By applying a unit force at the reconstructed source position established in S1, the vibration response conditions of each vibration sensing point are obtained. Since a unit force is applied, the vibration response at this time is also the vibration transfer function from the reconstructed source to each monitoring point of the stern structure, and they are numbered in sequence.

[0072] In step S3, the vibro-acoustic transfer function from the reconstructed source to the sound field assessment point is obtained based on the acoustic finite element method.

[0073] Specifically, according to the sound field assessment points established in S1, sound field assessment points are set in the model fluid field domain. By applying a unit force at the reconstructed source position, the sound pressure magnitude at the sound field assessment points is obtained based on the acoustic finite element method. Since a unit force is applied, the sound pressure at the assessment points at this time is the vibro-acoustic transfer function from the reconstructed source to the sound field assessment points.

[0074] In step S4, a sound field reconstruction model based on the reconstructed source is established based on the invariance of the transfer function.

[0075] In step S401, according to the acoustic principle of the underwater vehicle, there is the following relationship between the vibration velocity on the surface of the propeller main body and its radiated sound power:

[0076]

[0077] In the formula, W rad is the radiated sound power of the propeller, ρc is the acoustic impedance of the propeller structure, S is the underwater radiation surface area of the structure, i.e., the wetted surface area of the propeller; σ ra d is the radiation efficiency of the underwater vehicle; is the root mean square vibration velocity of the wetted surface area of the propeller. By converting the reconstructed vibration response A p (r i , ω), the root mean square vibration velocity of the propeller main body structure is reconstructed.

[0078] In S402, there is the following relationship between the reconstructed excitation source F and the root mean square vibration velocity of the outer surface of the main body: There is the following relationship:

[0079]

[0080] Substituting Equation (1) into the above Equation (2), we can obtain:

[0081]

[0082] In the formula, is the vibration transfer function from the reconstructed source to the outer surface of the main body, and a is the excitation magnitude of the reconstructed source.

[0083] In step S403, the radiation sound power calculation formula is converted into the sound pressure calculation formula, the relationship between the sound pressure and the transfer function is established, and the radiation sound pressure calculation formula under the excitation of the thruster on the stern structure is obtained.

[0084] P(Q) = H 1 pa |a 1 | + H 2 pa |a 2 | +... H j pa |a P (p j , ω)|

[0085] where P(Q) is the noise sound pressure generated by the thruster exciting the hull structure, and H j pa is the acoustic-vibration transfer function from the encoded reconstructed source to the sound field assessment point.

[0086] Thus, an acoustic field reconstruction model based on the reconstructed source is established.

[0087] In step S4, according to a series of drawings such as the offset table, general arrangement plan, thruster model drawing, and weight distribution of a specific ship, an acoustic-vibration prediction model of the thruster - shaft - stern structure is established. Unit vibration acceleration data is directly loaded at the measuring point position of the reconstructed excitation source. Through numerical simulation, the sound pressure at the underwater assessment points of the propeller - shaft - ship is calculated, and the acoustic-vibration transfer function of the ship is directly obtained. These acoustic-vibration transfer functions are stored in the computer storage module in the form of a matrix; based on the invariance of the acoustic-vibration transfer function, the numerical solution of the reconstructed excitation source is multiplied by the acoustic-vibration transfer function to quickly obtain the sound pressure of the acoustic field to be analyzed. As long as the structural model remains unchanged, the pre-stored acoustic-vibration transfer function will not become invalid, which has important application value in engineering and solves to a certain extent the problem of online monitoring of the vibration acceleration data of the propeller - shaft - ship of a specific ship and online predicting the underwater radiation acoustic field of the propeller - shaft - ship of the ship.

[0088] In step S5, using the response at the actual hull vibration measurement point as the input, by reconstructing the vibration transfer function from the source to the outer surface of the main body, the excitation magnitude of the reconstruction source is obtained. Calculate the vibration response of the thruster excitation stern structure and calculate the noise of the thruster excitation stern structure.

[0089] In step S6, establish a direct sound calculation model of the thruster, and obtain the direct sound noise of the thruster according to the ship speed and rotational speed.

[0090] The thruster noise mainly includes thruster rotational noise and cavitation noise. Rotational noise is the noise caused by the thruster operating in an uneven flow field (its frequency is mainly determined by the blade frequency). Once the shafting reaches the cavitation critical speed, the thruster noise becomes the main noise source of the underwater vehicle. The characteristics of cavitation noise are related to factors such as blade shape, area, blade pitch distribution, etc. At a specific rotational speed, the vortex frequency generated by the thruster blade is close to the natural frequency of the blade, and blade singing will occur.

[0091] Specifically, determine the rotational speed of the underwater vehicle thruster;

[0092]

[0093] Where n max is the maximum rotational speed of the thruster, v is the linear velocity of the thruster, v max is the maximum linear velocity of the thruster

[0094] Z is the number of revolutions, the f value of the thruster blade frequency (q = 1) q and its second (q = 2) and third (q = 3) tuned frequency values are calculated according to the following formula:

[0095] f q = qnZ, Hz, q = 1, 2, 3

[0096] The sound radiation level P at the blade frequency Bq and the radiation levels at its second and third tuned frequencies are calculated according to the following formula:

[0097]

[0098] Where F iq is the thruster blade frequency in three directions, R 1 is the maximum radius of the thruster.

[0099] In step S7, perform energy superposition on the noise of the thruster excitation stern structure and the direct sound to complete the thruster noise calculation.

[0100] Based on the principle of energy superposition, the thruster noise considering the direct sound is obtained according to the following formula.

[0101] P all = P(Q) + PBq 。

[0102] In step S7, based on the superposition principle of sound waves, ignoring the phase relationship between two waves and only superimposing the sound wave amplitudes, the incoherent sound wave superposition result, that is, the sound wave energy superposition result, is given to obtain the propeller noise considering the direct sound, as Figure 2 shown.

[0103] Embodiment 2: This embodiment is directed to a device for a fast prediction method of an underwater vehicle propeller noise based on excitation source reconstruction as described above. Based on the above method, a set of devices suitable for the vibration of the propeller stern structure and the fast prediction of propeller noise is also formed, specifically including:

[0104] The fast prediction device is composed of two parts: a data acquisition system and a prediction system.

[0105] The data acquisition system is mainly composed of a signal acquisition system and a high-precision vibration acceleration sensor. The signal acquisition system will transmit the collected vibration signals into the prediction system through optical fibers. Specifically, the measurement signal acquisition of the data acquisition system and the prediction system adopts the following method: the acceleration sensor is first connected to an adapter amplifier, and the measurement signals are transmitted to the acquisition control end of the prediction system through a multi-channel signal acquisition system.

[0106] The exterior of the prediction system is made of aluminum alloy material, and inside it is equipped with signal acquisition channels, signal source cards, an industrial control main board, and a storage hard disk. Specifically, the industrial control main board stores data such as the magnitude and distribution of the reconstructed source and the acoustic-vibration transfer function between the reconstructed source and the sound pressure assessment point. The vibration data obtained by the signal acquisition system will be input into the prediction system through the signal acquisition channels. Through the fast prediction module of the propeller noise in the prediction system, the real-time propeller noise level can be obtained with one key, and a storage log is formed.

[0107] As Figure 3 shown, the propeller noise calculated in this application and the simulation calculation results are in good agreement both in terms of trend and peak value, verifying the feasibility of this application.

[0108] The above are only the preferred embodiments of the present invention and are not used to limit the present invention. For those skilled in the art, the present invention can have various changes and modifications. Any modifications, equivalent replacements, improvements, etc. made within the spirit and principle of the present invention shall be included within the protection scope of the present invention.

Claims

1. A method for rapid prediction of underwater vehicle propeller noise based on excitation source reconstruction, characterized in that: The following steps are involved: Step 1: For the noise caused by the propeller exciting the stern structure, determine the location of the vibration sensing point, the location of the reconstruction source and the location of the sound field assessment point; Step 2: Based on the vibration sensing point position and the position of the reconstructed source obtained in step 1, the vibration transfer function from the reconstructed source to the monitoring point of the stern structure is obtained through simulation calculation; Step 3: According to the position of the sound field test point obtained in step 1, the acoustic vibration transfer function from the reconstructed source to the sound field test point is obtained based on the acoustic finite element method; Step 4: Based on the transfer functions obtained in steps 2 and 3, and based on the invariance of the transfer functions, a sound field reconstruction model based on the reconstruction source is established; Step 5: By measuring the response at the actual shell vibration measurement point, based on the vibration transfer function from the reconstructed source to the outer surface of the main body, the excitation magnitude of the reconstructed source is calculated, the vibration response of the propeller-excited stern structure is obtained, and the noise of the propeller-excited stern structure is calculated; Step 6: Establish a propeller direct sound calculation model and obtain the propeller direct sound noise according to the ship speed and rotation speed; Step 7: Superimpose the energy of the noise of the propeller-excited stern structure obtained in step 5 and the direct sound noise obtained in step 6 to complete the propeller noise calculation.

2. The method for rapid prediction of underwater vehicle propeller noise based on excitation source reconstruction according to claim 1 is characterized by: In the step 1, the vibration transfer function from the main vibration source to the stern structure is obtained by simulation calculation, and an inverse sensing network model is established based on the Bi-LSTM neural network; The Bi-LSTM neural network is used to establish the vibration transmission relationship between the shaft system vibration sensing point and the stern shell surface, realize the stern structure vibration reconstruction, and determine the number, position and size of the reconstruction excitation source; The size of the reconstructed excitation source is specifically: Considering the propeller structure as a linear system, the reconstruction excitation source p is established. j The excitation vector is a P (p j ,ω) and the measuring point r on the stern structure i Vibration acceleration response A P (r i ,ω) between each other; A P (r i ,ω)=H vP (r i p j ,ω)*a P (p j ,oh) Among them, H vP (r i p j ,ω) is the reconstructed excitation source p j And the stern structure measurement point i The vibration mapping relationship matrix with dimension M*K between them; Introduce the generalized inverse matrix to obtain the excitation source vector: a P (p j ,ω)=[[H vP (r i p j ,oh)] H [H vP (r i p j ,oh)]] -1 [H vP (r i p j ,oh)] H *A P (r i ,oh)。 3. A method for rapid prediction of underwater vehicle propeller noise based on excitation source reconstruction according to claim 2, characterized in that: The step 2 includes establishing a geometric model of the underwater vehicle, including a propeller, a propeller shaft and a stern structure of the hull; placing the geometric model of the underwater vehicle in a fluid domain, and establishing a fluid domain mesh model and a structural finite element model according to the geometric model of the underwater vehicle; According to the reconstructed source and vibration sensing point positions established in step 1, monitoring points are evenly distributed on the stern structure of the hull. By applying a unit force at the position of the reconstructed source, the vibration response of each vibration sensing point is obtained. The vibration response is the vibration transfer function from the reconstructed source to each monitoring point of the stern structure.

4. A method for rapid prediction of underwater vehicle propeller noise based on excitation source reconstruction according to claim 2, characterized in that: The step 3 sets the sound field assessment point in the model flow field according to the sound field assessment point established in step 1, and obtains the sound pressure at the sound field assessment point based on the acoustic finite element method by applying a unit force at the reconstructed source position. At this time, the sound pressure at the assessment point is the acoustic vibration transfer function from the reconstructed source to the sound field assessment point.

5. A method for rapid prediction of underwater vehicle propeller noise based on excitation source reconstruction according to claim 1, characterized in that: The step 4 converts the radiated sound power calculation formula into the sound pressure calculation formula, establishes the relationship between the sound pressure and the transfer function, and obtains the radiated sound pressure calculation formula under the propeller excitation stern structure: P(Q)=H1 pa |a1|+H2 pa |a2|+...H j pa |a P (p j ,ω)| Where P(Q) is the noise pressure generated by the propeller exciting the hull structure, H j pa is the acoustic vibration transfer function from the encoded reconstructed source to the sound field test point, thereby establishing a sound field reconstruction model based on the reconstructed source; a P (p j ,ω) is the excitation source vector.

6. A method for rapid prediction of underwater vehicle propeller noise based on excitation source reconstruction according to claim 1, characterized in that: The step 6 is: Determine the rotational speed of underwater vehicle propellers; Where n max is the maximum speed of the propeller, v is the linear speed of the propeller, and v max is the maximum linear velocity of the propeller; Propeller blade frequency qf q The value and its second and third tuning frequency values ​​are calculated as follows: f q =qnZ,Hz,q=1,2,3 In the formula, f q is the propeller frequency, q is the propeller blade frequency, n is the propeller speed, and Z is the number of revolutions; Propeller direct sound P Bq The radiation levels of the second and third tuning frequencies are calculated as follows: In the formula, F xq is the propeller blade frequency in the X direction, F xq is the propeller blade frequency in the Y direction, F xq is the propeller blade frequency in the Z direction, R1 is the maximum radius of the propeller, P Bq For thruster straight sound.

7. A method for rapid prediction of underwater vehicle propeller noise based on excitation source reconstruction according to claim 1, characterized in that: The step 7 is: Based on the principle of energy superposition, the thruster noise including direct sound is obtained as follows: P all =P(Q)+P Bq Where P(Q) is the noise pressure generated by the propeller exciting the hull structure, P Bq For thruster straight sound.

8. The device for rapid prediction of underwater vehicle propeller noise based on excitation source reconstruction according to any one of claims 1 to 7, characterized in that: Including data acquisition system and forecasting system; The data acquisition system is used to send the collected data to the forecasting system; The prediction system is used to obtain the real-time propeller noise level and form a storage log.

9. A computer-readable storage medium having a computer program / instruction stored thereon, characterized in that: When the computer program / instructions are executed by a processor, the steps of the method according to any one of claims 1 to 7 are implemented.

10. A computer program product comprising a computer program / instructions, characterized in that: When the computer program / instructions are executed by a processor, the steps of the method according to any one of claims 1 to 7 are implemented.

Images