In-situ evaluation system for sound insulation efficiency of sound barrier in all working conditions of ship dome

Abstract

This invention provides an in-situ evaluation system for the sound insulation performance of a ship's acoustic baffle under all operating conditions. The system includes: multiple acoustic sensor arrays, mirror-mounted on the front and back surfaces of the acoustic baffle, for receiving sound field data from both surfaces; a data acquisition module connected to the acoustic sensor arrays, for receiving and storing the sound field data transmitted by the arrays and acquiring ship operating parameters at corresponding times; and a data processing module for evaluating the sound insulation performance of the acoustic baffle under corresponding operating conditions based on the sound field data and the ship operating parameters at corresponding times, obtaining the evaluation results. By implementing this solution, the sound field difference between the front and back surfaces of the acoustic baffle can be captured in real time, and the sound insulation performance can be accurately evaluated in conjunction with the corresponding operating parameters, making the evaluation results more consistent with actual usage requirements.

Description

Technical Field

[0001] This invention belongs to the field of ship sound insulation performance evaluation technology, specifically involving an in-situ evaluation system for the sound insulation performance of the sound baffle inside the ship's fairing under all working conditions. Background Technology

[0002] As a front-end component of sonar and other underwater acoustic systems, the core function of a ship's fairing is to provide hydrodynamic protection for the underwater acoustic transducer elements while minimizing interference from hydrodynamic noise. However, under actual ship navigation conditions, broadband self-noise generated by mechanical vibration, propeller operation, and end-flow boundary layers can enter the fairing through both structural transmission and fluid-acoustic coupling, severely limiting the detection range and resolution of the sonar system. To improve the acoustic environment of the sonar compartment, acoustic baffles are typically installed inside the fairing to isolate and attenuate noise from the direction of the ship's hull; their performance directly determines the final signal-to-noise ratio of the underwater acoustic system.

[0003] Currently, the evaluation of sound baffle performance generally relies on laboratory measurements, such as measuring insertion loss using a standard sound source in an anechoic pool. However, such methods are difficult to reproduce the complex conditions of real-world navigation and cannot assess the full-condition sound insulation performance of sound baffles under real-world conditions, leading to a disconnect between evaluation results and actual usage scenarios. Summary of the Invention

[0004] In view of this, the purpose of the present invention is to provide an in-situ evaluation system for the sound insulation performance of the acoustic baffle inside the ship's fairing under all operating conditions, so as to meet the need to evaluate the sound insulation performance of the acoustic baffle under real conditions under all operating conditions.

[0005] To achieve the above objectives, the present invention provides the following technical solution:

[0006] This invention provides an in-situ evaluation system for the sound insulation performance of a ship's acoustic baffle under all operating conditions. The system includes: multiple acoustic sensor arrays, mirror-mounted on the front and back surfaces of the acoustic baffle, for receiving sound field data from both surfaces; a data acquisition module connected to the multiple acoustic sensor arrays, for receiving and storing the sound field data transmitted by the acoustic sensor arrays and acquiring ship operating parameters at corresponding times; and a data processing module for evaluating the sound insulation performance of the acoustic baffle under corresponding operating conditions based on the sound field data and the ship operating parameters at corresponding times, thereby obtaining the evaluation results.

[0007] This embodiment provides an in-situ evaluation system for the sound insulation performance of the acoustic baffle inside a ship's fairing under all operating conditions. Through the architecture of mirrored deployment of multiple acoustic sensor arrays, synchronous acquisition of sound field and operating condition parameters by a data acquisition module, and correlation evaluation by a data processing module, it can directly cover all operating conditions such as anchored stop, low-speed cruise, medium-speed navigation, full-speed navigation, extreme speed, acceleration / deceleration, etc., under the actual navigation state of the ship. It can capture the sound field difference between the front and back surfaces of the acoustic baffle in real time, and accurately evaluate the sound insulation performance in combination with the corresponding operating condition parameters. The evaluation results are more in line with actual use requirements.

[0008] Other advantages, objectives, and features of the invention will be set forth in the following description and will be apparent to those skilled in the art in some respects, or may be learned by practice of the invention. The objectives and other advantages of the invention can be realized and obtained through the following description. Attached Figure Description

[0009] To make the objectives, technical solutions, and beneficial effects of this invention clearer, the following figures are provided for illustration:

[0010] Figure 1 This is a flowchart illustrating a specific example of the in-situ evaluation system for the sound insulation performance of the acoustic baffle inside the ship's fairing under all operating conditions, as described in this invention. Detailed Implementation

[0011] The technical solution of the present invention will now be clearly and completely described with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of the present invention. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0012] In the description of this invention, it should be noted that, unless otherwise explicitly specified and limited, the terms "installation," "connection," and "linking" should be interpreted broadly. For example, they can refer to a fixed connection, a detachable connection, or an integral connection; they can refer to a mechanical connection or an electrical connection; they can refer to a direct connection or an indirect connection through an intermediate medium; they can also refer to the internal connection of two components; and they can refer to a wireless connection or a wired connection. Those skilled in the art can understand the specific meaning of the above terms in this invention based on the specific circumstances.

[0013] Furthermore, the technical features involved in the different embodiments of the present invention described below can be combined with each other as long as they do not conflict with each other.

[0014] This invention provides an in-situ evaluation system for the sound insulation performance of the acoustic baffle inside a ship's fairing under all operating conditions, such as... Figure 1 As shown, it includes:

[0015] Multiple acoustic sensor arrays 101 are respectively mirrored on the front and back sides of the sound baffle to receive sound field data from the front and back sides of the sound baffle.

[0016] The data acquisition module 102 is connected to multiple acoustic sensor arrays and is used to receive and store sound field data sent by the acoustic sensor arrays and to acquire ship operating parameters at the corresponding time.

[0017] The data processing module 103 is used to evaluate the sound insulation performance of the sound baffle under the corresponding operating conditions based on the sound field data and the ship's operating condition parameters at the corresponding time, and obtain the evaluation results.

[0018] For example, the acoustic sensor array 101 can be composed of high-precision vector hydrophones, and the hydrophones can be arranged in a grid at preset intervals on the front and back sides of the sound baffle, for example, according to... Regarding the grid setup, this embodiment only provides an example of the layout method. The specific setup can be determined based on the shape of the acoustic baffle, and this embodiment is not limited to this. For mirror setup, laser positioning can be used to determine the positional deviation of the placement points on the front and back acoustic surfaces, and then calibrate. In this embodiment, each hydrophone can be connected to a watertight junction box, and then connected to the data acquisition module through the watertight junction box. Differential signal transmission is used to reduce the impact of ship electromagnetic interference on the data.

[0019] The data acquisition module 102 has a built-in signal conditioning circuit, which includes a preamplifier, a filter, and an AD converter. Each signal conditioning circuit receives the sound field data from the corresponding hydrophone and converts the analog signal into a digital signal. In addition, the data acquisition module 102 communicates with the ship's central control system to collect real-time ship operating parameters, including speed, draft, and main engine load. The data acquisition module 102 needs to synchronously store the sound field data and operating parameters; therefore, a GPS timing module can be used to provide a synchronization clock for the data acquisition module, with both types of data carrying corresponding timestamps.

[0020] The data processing module 103 communicates with the data acquisition module 102 and can be a terminal with computing capabilities. The method for evaluating the sound insulation performance of the sound baffle under corresponding operating conditions based on sound field data and the ship's operating parameters at the corresponding time can be as follows: First, classify the operating conditions based on the operating parameters, as shown in Table 1 below:

[0021] Table 1

[0022]

[0023] The above classification of working conditions is for illustrative purposes only; the specific classification method can be determined based on the actual situation.

[0024] After obtaining the different operating conditions, one optional implementation method is to evaluate the sound insulation performance of the sound baffle under the corresponding operating conditions based on the sound field data and the ship's operating condition parameters at the corresponding time. The evaluation results are obtained in the following way:

[0025] For each pair of mirror hydrophones, calculate the sound pressure level difference at 1 / 3 octave band to obtain the initial sound insulation. The formula is:

[0026] ;

[0027] in, For frontal hydrophones at frequencies f sound pressure level at ， For back-facing hydrophones at frequencies f Sound pressure level at; frequency f It covers 10-2000Hz, divided into 1 / 3 octaves, such as 10Hz, 12.5Hz, 16Hz...2000Hz.

[0028] This allows us to obtain the sound insulation corresponding to different frequencies of hydrophones under different operating conditions. Based on this, we can perform longitudinal and lateral comparisons. Longitudinal comparison is defined as comparing the trend of sound insulation changing with operating conditions at the same frequency. Specifically, firstly, we select the sensitive frequency band of ship acoustic equipment (100-2000Hz) and choose six representative frequency points, such as 125Hz, 250Hz, 500Hz, 1000Hz, 1600Hz, and 2000Hz, covering low, medium, and high frequencies. We then select five steady-state operating conditions from low to high speed: anchored stop, low-speed cruising, medium-speed navigation, full-speed navigation, and maximum speed. To avoid the transitional effects of operating condition switching, transient operating conditions such as acceleration / deceleration are excluded. From the three-dimensional data matrix of the five steady-state operating conditions, we extract the sound insulation data of the six representative frequency points and calculate the average sound insulation of each frequency point under a single operating condition. Next, with the operating speed as the horizontal axis and the anchor stop = 0 knots, low speed = 8 knots, medium speed = 15 knots, full speed = 25 knots, and maximum speed = 32 knots, and the average sound insulation as the vertical axis, a speed-average sound insulation curve was plotted for each frequency point.

[0029] Next, the sensitivity coefficient K for each frequency point is calculated using the following formula:

[0030] ;

[0031] Based on the above coefficients, the variation of sound insulation with flight speed is quantified as follows: K -0.1dB / section, high sensitivity frequency; -0.1≤K≤0.05dB / section, medium sensitivity frequency; K 0.05dB / section, low sensitivity / positive sensitivity frequency.

[0032] Horizontal comparison is defined as comparing the sound insulation curves across the entire frequency band (10-2000Hz, 24 1 / 3 octave bands) to identify the weak frequency bands where sound insulation performance significantly decreases under a certain operating condition. Specifically, for each operating condition, the average sound insulation of the 24 frequency bands is calculated, and a frequency-average sound insulation curve is generated with the center frequency of the frequency band as the x-axis and the average sound insulation as the y-axis. There are two methods for determining the weak frequency bands. The first method, when comparing between operating conditions, involves comparing the full-frequency band curves of any two operating conditions to determine the difference in sound insulation between the frequency bands. TL = Average sound insulation of working condition A - Average sound insulation of working condition B, when TL -3dB, and continued Three adjacent frequency bands constitute a weak frequency band; the second method involves a horizontal comparison between operating conditions and threshold values, comparing the single operating condition curve with the design threshold curve to determine the difference between the average sound insulation of the frequency band and the threshold value. TL = Average sound insulation under operating conditions - Threshold, when TL 0dB, ​​and continuously Two adjacent frequency bands are considered weak frequency bands.

[0033] Finally, by combining the results of horizontal and vertical comparisons, the sound insulation performance of the sound baffle is comprehensively evaluated. Specifically, from the horizontal comparison, a list of frequency bands identified as weak points under each evaluated operating condition is extracted. From the vertical comparison, the operating condition sensitivity coefficient K and its classification are obtained for each analyzed frequency band, including: high sensitivity, medium sensitivity, and low sensitivity / positive sensitivity. Next, a two-dimensional matrix is ​​created, with the horizontal axis representing all operating conditions to be evaluated and the vertical axis representing the complete analyzed frequency bands.

[0034] Each cell in the matrix contains two core types of information: performance status and sensitivity attributes. Performance status represents the average sound insulation of the frequency band under that operating condition, and clearly indicates whether it belongs to a weak frequency band. Sensitivity attributes are determined by the inherent operating condition sensitivity coefficient K and sensitivity type label for that frequency band. Based on this, through cross-analysis of the matrix data, the causes of each identified weak frequency band are diagnosed. Specifically, for each frequency band marked as weak in the horizontal comparison, such as the combination of frequency band F and operating condition G, the following diagnostic logic is executed: First, the vertical sensitivity classification of frequency band F is found in the matrix, and then a correlation analysis is performed to obtain the evaluation results. An example of the correlation analysis is as follows:

[0035] When frequency band F falls into the high-sensitivity or medium-sensitivity category and exhibits weakness under operating condition G (typically high speed and high load), its sound insulation performance is inherently sensitive to changes in operating condition parameters such as speed and load. When the ship enters a high-excitation operating condition, the energy of excitations such as water flow noise and main engine vibration significantly increases in this frequency band, exceeding the inherent sound insulation capacity of the sound baffle in that band, leading to performance degradation. This indicates that the root cause of the problem lies primarily in the deterioration of the external excitation environment; therefore, this weak frequency band is highly likely to be dominated by operating condition excitation.

[0036] If frequency band F belongs to the low-sensitivity or positive-sensitivity category, but exhibits weakness under a certain operating condition G, then the sound insulation performance of that frequency band should not significantly deteriorate with the operating condition. The weakness is more likely due to design or manufacturing defects in the sound baffle itself at that frequency band, such as insufficient sound insulation quality, poor sealing, or structural resonance occurring at the excitation frequency of the specific operating condition, leading to a sharp decrease in sound insulation. This indicates that the root cause of the problem lies primarily in the sound baffle itself. This weak frequency band is highly suspected to be due to structural defects or resonance.

[0037] If a certain frequency band is weak under multiple operating conditions, then regardless of the changes in operating conditions, the sound insulation performance of that frequency band cannot meet the requirements. This indicates that the inherent sound insulation design of the sound barrier in that frequency band is inadequate and is an inherent weakness that urgently needs to be addressed. Therefore, this weak frequency band is basically determined to be of the type with inherently insufficient performance.

[0038] Finally, the above assessment results are summarized to form a weak frequency band cause diagnosis table, listing each weak frequency band, its corresponding operating condition, sensitivity attributes, and assessment type. It should be noted that the above implementation method, which uses a comprehensive assessment of horizontal and vertical comparison results, is only an optional example and is not intended to limit this solution.

[0039] This invention provides an in-situ evaluation system for the sound insulation performance of the acoustic baffle inside a ship's fairing under all operating conditions. Through the architecture of mirrored deployment of multiple acoustic sensor arrays, synchronous acquisition of sound field and operating condition parameters by a data acquisition module, and correlation evaluation by a data processing module, it can directly cover all operating conditions such as anchored stop, low-speed cruise, medium-speed navigation, full-speed navigation, extreme speed, acceleration / deceleration, etc., under the actual navigation state of the ship. It can capture the sound field difference between the front and back surfaces of the acoustic baffle in real time, and accurately evaluate the sound insulation performance in combination with the corresponding operating condition parameters. The evaluation results are more in line with actual use requirements.

[0040] As an optional implementation method, the in-situ evaluation system for the all-condition sound insulation performance of the acoustic baffle inside the ship's fairing also includes:

[0041] Vibration sensors are installed at the target locations to acquire vibration parameters at the corresponding target locations. The data processing module also includes: evaluating the sound insulation performance of the sound baffle under the corresponding operating conditions based on the sound field data and the ship's operating condition parameters and vibration parameters at the corresponding time.

[0042] For example, the vibration sensor can be a piezoelectric accelerometer. Based on the vibration transmission characteristics of the baffle, vibration sensors are deployed at key points in the actual excitation transmission path, such as the connection point between the baffle and the ship hull, the support structure near the main engine, and the edge position where the water flow impact is significant, to ensure that the vibration signal of the main distributed excitation can be captured.

[0043] As an optional implementation method, the sound insulation performance of the sound baffle under the corresponding operating conditions is evaluated based on the sound field data and the ship's operating parameters and vibration parameters at the corresponding time, including:

[0044] Based on the sound field data of the front and back surfaces, the initial sound insulation of the sound baffle is calculated. A pre-stored operating condition-vibration-propagation coupling weight matrix is ​​then invoked. This coupling weight matrix is ​​trained using historical flight data, which includes at least operating condition parameters, vibration parameters, and corresponding actual sound insulation deviation data. The coupling weight matrix characterizes the coupling effect among operating condition, vibration, and sound propagation. The operating condition parameters and vibration parameters are input into the coupling weight matrix to calculate the comprehensive correction. Based on the initial sound insulation and the comprehensive correction, the operating condition sound insulation is calculated, and a frequency response curve for the operating condition sound insulation is generated. Based on the frequency response curve for the operating condition sound insulation and a preset design threshold, the weak frequency bands where the sound insulation is below the design threshold are output.

[0045] The training process of the coupling weight matrix includes: collecting multiple sets of historical real-time navigation data. Each set of historical real-time navigation data includes: the current operating speed, load, and other state parameters of the actual ship; the vibration acceleration frequency characteristic data of the sound baffle and surrounding structures under the corresponding operating conditions; the sound pressure data of the frontal and rear acoustic surfaces collected during the normal operation of the actual ship; and the sound pressure data of the incident sound from the frontal surface and the transmitted sound from the rear acoustic surface collected under the same operating conditions using standard testing methods. Among these, the conventional sound field data is used to calculate the initial sound insulation, and the calibration sound field data is used to obtain the calibration sound insulation. The two sets of data are collected independently and the collection paths do not overlap.

[0046] Based on the conventional sound field data in each set of historical navigation data, the initial sound insulation for the corresponding historical scenario is calculated. Based on the calibration sound field data in each set of historical navigation data, the calibration sound insulation is obtained. The calibration sound insulation is defined as the actual sound insulation capability of the sound baffle against incident sound waves under actual ship operating conditions. The methods for obtaining this value include:

[0047] A standard sound source is set up on the frontal surface of the sound baffle to emit a test signal with a known sound power spectrum, ensuring that the incident sound wave is consistent with the actual sound source characteristics of the ship. The incident sound pressure level on the frontal surface is collected by an array of microphones, and the distance attenuation and reflection interference between the standard sound source and the sound baffle are deducted to obtain the corrected true incident sound pressure level. The transmitted sound pressure level is collected by a microphone on the back surface, and the background noise of the ship is deducted. The background noise of the ship is obtained through a test under the condition of an empty cabin without a sound source to obtain the corrected true transmitted sound pressure level. Then, the actual sound insulation deviation of each group of historical scenarios is calculated. The deviation is used to characterize the difference between the ideal calculated value (initial sound insulation) and the true value (calibrated sound insulation) caused by the coupling effect of operating conditions, vibration and sound propagation. That is, it is the difference between the two.

[0048] Finally, using operating condition parameters and vibration parameters as input variables, and actual sound insulation deviation as output variable, a machine learning model is used to train the operating condition-vibration-propagation coupling weight matrix. Through machine learning, a mapping relationship is established using calibration data under certain operating conditions, thereby predicting the correction amount for other operating conditions.

[0049] As an optional implementation method, under the target operating condition, the sound insulation performance of the sound baffle under the corresponding operating condition is evaluated based on the sound field data and the ship's operating condition parameters and vibration parameters at the corresponding time. This includes: determining the initial sound insulation of the sound baffle based on the sound field data of the front and back surfaces; correcting the initial sound insulation for vibration effects based on the vibration parameters to obtain the vibration-corrected sound insulation; calling the pre-stored operating condition influence weight matrix, correcting the vibration-corrected sound insulation for operating condition propagation deviation based on the operating condition parameters, and outputting the operating condition sound insulation frequency response curve. The operating condition influence weight matrix is ​​trained from historical real-time navigation data; and outputting the weak frequency bands where the sound insulation is lower than the design threshold based on the operating condition sound insulation frequency response curve and the preset design threshold.

[0050] In actual use, when the ship is actually sailing under the target operating condition, which is a low-speed cruise condition, the coupling strength between vibration and operating condition is low. This embodiment proposes to use a step-by-step correction method to complete the correction for the operating condition with low coupling strength. Specifically, three types of data are collected simultaneously: real-time vibration data of all vibration sensors (covering multiple excitation paths); sound pressure data of all hydrophones on the front / back surface; and ship operating parameters, such as speed and main engine speed, to help determine the current main excitation source, such as enhanced water flow excitation at high speed.

[0051] Calculate the multicoherence coefficients of all current vibration sensor combinations and the sound pressure of the hydrophone at various frequencies; compare with the offline coherence reference range to determine the dominant frequency band of structural sound under the current operating conditions: if the coherence coefficient at a certain frequency is >0.7 (and consistent with the characteristics of the offline dominant frequency band of structural sound), it is determined to be the dominant frequency band of structural sound, and the structural sound contribution needs to be estimated using the MISO model; if the coherence coefficient is <0.4, it is determined to be the dominant frequency band of airborne sound, and no additional correction is required.

[0052] The MISO model estimates the structural acoustic contribution by performing the following operations only in the dominant structural acoustic frequency band: extracting the online vibration velocities of all vibration sensors in this frequency band (as real-time input to the MISO model); substituting the real-time input into the offline-established MISO basic model, the model will directly output the structural acoustic radiation sound pressure level generated by the current multiple vibration sources at the hydrophone according to the contribution weights of each vibration source; if the difference between the estimated structural acoustic pressure level and the total sound pressure level of the back-side hydrophone minus the airborne sound penetration sound pressure level at the front-side is <2dB, the estimation is valid; if the difference is too large, the model weights are fine-tuned based on the current operating parameters to ensure estimation accuracy.

[0053] In the structural sound dominant frequency band, the structural sound radiation sound pressure level estimated by the MISO model is subtracted from the total sound pressure level of the back acoustic surface hydrophone to obtain the corrected airborne sound pressure level on the back acoustic surface. The vibration correction sound insulation is obtained by subtracting this correction value from the sound pressure level on the front acoustic surface. In the airborne sound dominant frequency band, the initial sound insulation, that is, the difference between the sound pressure levels on the front and back acoustic surfaces, is directly used as the vibration correction sound insulation without deducting the structural sound contribution.

[0054] It should be noted that the model used in the above process was prepared offline in advance. Specifically, multiple sets of exciters were used to apply simulated distributed excitation: for example, one set of exciters was used to simulate the vibration of the host machine and another set of exciters was used to simulate the impact of water flow. The excitation signal was a wideband random signal. The vibration parameters of all vibration sensors and the sound pressure data of all hydrophones were collected simultaneously to form an offline dataset with multiple vibration inputs and a single sound pressure output. Each hydrophone corresponds to a set of multiple vibration inputs and sound pressure output data.

[0055] For each hydrophone, the multiphase coherence coefficient of the sound pressure at each frequency is calculated for all vibration sensor combinations and the hydrophone's sound pressure, with values ​​ranging from 0 to 1. The multiphase coherence coefficient can be determined by first defining the input as the time-domain vibration signals of N vibration sensors and the output as the time-domain sound pressure signal of the hydrophone at a certain position on the back acoustic surface. Therefore, this embodiment uses the Welch average periodogram method to calculate the auto-power spectral density of the hydrophone's sound pressure signal. That is, the sound pressure signal is divided into multiple overlapping segments, each segment is windowed and subjected to a Fourier transform, the square of the spectral amplitude of each segment is calculated to obtain a power spectrum estimate, and then all estimates are averaged to obtain the energy distribution of the sound pressure at each frequency and the corresponding frequency vector. Next, for each vibration sensor, its own power spectral density and cross-power spectral density with the sound pressure signal are calculated, and then the cross-power spectral density between each pair of different sensors is calculated. Then, for each frequency point, an N×N input power spectral density matrix is ​​constructed, with the diagonal representing the sensor's own power spectrum and the off-diagonal representing the cross-power spectrum between sensors, and it is a conjugate symmetric Hermitian matrix. An N×1 input-output cross-power spectral density vector is also constructed, with elements representing the cross-power spectrum between the sensor and the sound pressure. Then, the input power spectral density matrix, the input-output cross-power spectral density vector, and the sound pressure power spectral density scalar are extracted for that frequency point. Because vibration sensor signals are prone to high correlation leading to ill-conditioned matrices, regularization is performed by diagonal loading (taking a minimal proportion of the matrix trace as a regularization parameter, adding it to the identity matrix, and then superimposing it with the original matrix), and then the inverse of the regularization matrix is ​​calculated. Finally, based on the calculated values, the multicoherence function value is calculated using the following formula. In this embodiment, the multicoherence function is defined as the ratio of explained power to total power; therefore, the formula is as follows:

[0056] ;

[0057] in, Represents frequency The multiphase function value at that point, The input and output cross power spectral density vectors. for The conjugate transpose of . The input is the power spectral density matrix. yes The inverse matrix, The output is a power spectral density scalar.

[0058] The closer the multicoherence coefficient is to 1, the stronger the combined influence of these vibration sources on the sound pressure of the hydrophone in that frequency band; a coefficient close to 0 indicates that the sound pressure is mainly dominated by airborne sound penetration. Record the coherence reference range of each hydrophone in different frequency bands, such as the multicoherence coefficient of a certain hydrophone in the 200-800Hz range (the main vibration frequency of the main unit). 0.8 (structure sound dominant), above 1000Hz 0.3 (Airborne sound dominates).

[0059] A basic MISO model is established for each hydrophone, with the vibration parameters of all vibration sensors as input variables and the structure-radiated sound pressure measured by the hydrophone (in offline testing, airborne sound has been eliminated through a sealed soundproof box, and only vibration-radiated sound pressure is retained) as the output variable. The basic MISO model is established by modeling through linear regression or frequency domain transfer function: quantifying the contribution weight of each vibration source to the sound pressure of the hydrophone when multiple vibration sources act simultaneously, avoiding the simplified estimation of a single vibration source.

[0060] In actual navigation environments, the measured noise on the back surface of the sound baffle is a superposition of airborne and structural sound. Furthermore, variations in sound field propagation characteristics under different operating conditions can further interfere with the calculated sound insulation. Therefore, in this embodiment, after obtaining the vibration-corrected sound insulation based on the above process, the vibration-corrected sound insulation is further corrected for propagation deviations based on a pre-determined operating condition influence weight matrix. Specifically, operating condition parameter matching and matrix retrieval: the data acquisition module acquires ship operating condition parameters in real time, including speed. Draft Host load The current operating condition type is determined according to the operating condition classification rules, and the operating condition influence weight matrix for the corresponding frequency band is called. This matrix has dimensions. The coefficient matrix, where, For the number of working condition characteristics, For each element, the number of frequency bands is... Representing the The first working condition characteristic for the first The influence weight of sound insulation in each frequency band.

[0061] Generation of sound insulation frequency response curve under operating conditions: Correcting the sound insulation for vibration Weighting matrix of operating conditions Perform matrix operations to obtain the corrected sound insulation value under operating conditions. The formula is as follows:

[0062] ;

[0063] For example, in the 200-500Hz frequency band (the frequency band sensitive to main engine vibration), the weighting coefficient under extreme speed conditions This indicates that under this operating condition, the sound field propagation deviation will cause the measured sound insulation value to be 15% lower, requiring weight correction to restore the true performance. (Based on the full frequency band) This generates a frequency response curve of sound insulation under working conditions, which intuitively reflects the distribution of sound insulation performance of the sound baffle under the current working conditions.

[0064] Finally, the operating condition sound insulation frequency response curve is compared with the preset design threshold curve, such as the minimum sound insulation required by the acoustic equipment. By comparing dB, it can be determined whether the frequency band is a weak sound insulation band.

[0065] This invention provides an in-situ evaluation system for the sound insulation performance of a ship's acoustic baffle under all operating conditions. During ship navigation, the acoustic baffle is affected by vibrations from the ship's hull, main engine, and water flow. These vibrations are transmitted to the back acoustic surface through the structure, resulting in back acoustic surface noise comprising airborne sound penetration noise and structural vibration radiated noise. Traditional evaluations based solely on sound field data may mistakenly include vibration radiated noise in the airborne sound penetration noise, and vibration is a core source of interference, severely underestimating the actual sound insulation capability of the acoustic baffle. This embodiment adds vibration sensors to acquire vibration parameters at key locations. The data processing module combines sound field data, operating condition parameters, and vibration parameters for evaluation, separating the influence of vibration radiated noise on the back acoustic surface sound field, eliminating vibration interference, and making the evaluation results closer to the true airborne sound insulation performance of the acoustic baffle. This avoids misjudgments caused by vibration factors, ensuring that the evaluation results meet the requirements of actual use scenarios while minimizing core interference.

[0066] As an optional implementation, the process of determining the operating condition influence weight matrix includes: acquiring historical flight data, vibration data correction relationships for sound insulation, and a reference sound insulation frequency curve. The historical flight data includes operating condition parameters, vibration data, and sound field data of the front and back surfaces of the sound baffle. The vibration data correction relationships for sound insulation and the reference sound insulation frequency curve are all determined based on laboratory vibration tests. Based on pre-divided frequency bands, the following operations are performed within each frequency band: based on the operating condition parameters, extract features under each operating condition to obtain the fused feature matrix for that frequency band; based on the historical flight data... Based on the vibration data and the correction relationship between the vibration data and the sound insulation, the vibration sound insulation correction amount in this frequency band is determined; based on the sound field data of the front and back surfaces of the sound baffle, the vibration sound insulation correction amount in this frequency band, and the reference sound insulation frequency curve, the operating condition propagation deviation amount in this frequency band is determined; using the fusion feature matrix in this frequency band as input and the operating condition propagation deviation amount in this frequency band as label, a gradient boosting tree model is used for training to obtain the trained target model; the influence coefficients of each feature on the target variable are extracted from the trained target model to form the operating condition influence weight matrix in this frequency band.

[0067] For example, the operating condition influence weight matrix is ​​essentially a mapping model of operating condition characteristics and frequency band sound insulation deviation obtained by training with historical flight data. The construction of the matrix involves four stages: data acquisition, feature extraction, model training, and weight extraction, to ensure the accuracy and generalization ability of the matrix. The specific process is as follows:

[0068] First, basic data collection and preprocessing include collecting historical navigation data from at least three ships of the same type, covering 12 typical operating conditions, including five steady-state conditions, four transient conditions, and three extreme conditions. Under each condition, three types of data are collected simultaneously: operating condition parameters: speed. Draft Host load Seawater temperature Sound field data: sound pressure level data of the acoustic sensor array on the front / back surface; vibration data: acceleration and velocity data of each vibration sensor.

[0069] The reference sound insulation frequency curve of the sound baffle under standard environment was obtained through laboratory vibration testing. The relationship between vibration data and sound insulation is used as a reference benchmark for calculating deviations in actual flight data. The 10-2000Hz full frequency band is subdivided into multiple core sub-bands based on acoustic sensitivity characteristics. For each sub-band, the operating condition propagation deviation in historical flight data is calculated. The formula is as follows:

[0070] ;

[0071] in, This is the vibration correction for a specific operating condition based on historical data. This represents the sound insulation deviation caused by propagation deviation under this working condition. A positive value indicates that the measured value is too high, and a negative value indicates that it is too low.

[0072] For each sub-band, a fusion feature matrix is ​​extracted based on the standardized operating parameters. The fusion process employs principal component analysis for dimensionality reduction, transforming the four original operating condition features into three principal component features, such as... Speed-load fusion characteristics, contributing 70%; : Draft depth characteristics, contributing 20%; Seawater temperature characteristics contribute 10% to the final dimension. fusion feature matrix , This represents the number of historical data samples.

[0073] The fusion feature matrix of each sub-band As input variables, propagate the deviation amount according to the corresponding operating conditions. To output labels, a training dataset is constructed. Gradient Boosting Tree (XGBoost) is chosen as the training model because it can handle non-linear feature relationships and reduces the risk of overfitting through ensemble learning. The model parameters can be set to a learning rate of 0.1, a tree depth of 5, and 100 iterations. The parameters are optimized using 5-fold cross-validation.

[0074] For the trained target model, extract the output bias of each principal component feature. The influence coefficient is obtained through the feature importance evaluation function built into the model. For example, in the low-to-medium frequency range, The influence coefficient of the (speed-load fusion characteristic) is 0.65. (Draft) is 0.25. The seawater temperature is 0.10, indicating that speed and load are the main factors affecting the propagation deviation in this frequency band.

[0075] The influence coefficients of each principal component feature are inversely calculated with the contribution rates of the original operating condition parameters to restore the influence weight of the original operating condition features on each frequency band. For example, assuming that the contribution rate of speed is 60% and the contribution rate of load is 40% in PC1, the influence weight of speed on the mid-to-low frequency band is... The impact weight of load Draft Seawater temperature Ultimately, dimensions are formed. The working condition influence weight matrix.

[0076] This invention provides an in-situ evaluation system for the sound insulation performance of acoustic baffles within ship fairings under all operating conditions. It utilizes dual data sources: historical sea trial data and laboratory benchmark data, combined with a gradient boosting tree model for training. Based on historical sea trial data, operating condition characteristics are extracted. These characteristics are then combined with vibration correction relationships determined in the laboratory and benchmark sound insulation curves to quantify the impact of each operating condition characteristic on the propagation deviation of sound insulation at different frequency bands. The gradient boosting tree model can handle the nonlinear relationship between operating condition characteristics and propagation deviation. The trained target model can accurately output the influence coefficients of each characteristic, forming a frequency-band-specific operating condition influence weight matrix. This matrix, trained based on actual data, can cover different sea trial scenarios for the same type of ship, exhibiting strong adaptability. Furthermore, the frequency-band design can accurately correct for the acoustic characteristics of different frequency bands, ensuring the accuracy and consistency of sound insulation assessment under various operating conditions.

[0077] As an optional implementation, the acoustic sensor array consists of multiple hydrophones, and the system also includes:

[0078] The sound source is positioned at the central axis of the hydrophone array and is used to transmit target audio signals when the ship is not in operation.

[0079] A standard reference hydrophone is set in a preset stable sound field area to collect reference sound field response data of the target audio signal;

[0080] Acoustic sensor arrays are also used to collect target sound field response data of target audio signals on the front and back sides of the sound baffle;

[0081] The data processing module is also used to calculate the amplitude ratio and phase difference between each hydrophone in the acoustic sensor array and the standard reference hydrophone based on the reference sound field response data and the target sound field response data.

[0082] The data processing module is also used to correct the sound field data based on the amplitude ratio and phase difference.

[0083] For example, a broadband programmable sound source is arranged along the central axis of the acoustic sensor array on the sound-facing side of the sound baffle. The sound source supports multiple signal types such as sine wave, white noise, and pink noise. A standard reference hydrophone is installed in a preset stable sound field area. This sound field area is the area inside the shroud that is far away from the sound source, the sound baffle, and the vibration source. The uniformity error of the sound field in this area has been confirmed by previous tests to be ≤0.5dB.

[0084] When the ship is stationary and anchored, a trigger signal is sent to the sound source via the I / O port of the data acquisition module. After receiving the signal, the sound source begins to transmit the target audio signal after a 1-second delay. To improve data reliability, the same set of target audio signals is transmitted three times, and the average of the three acquisitions is taken as the final data. The standard reference hydrophone and the acoustic sensor array share a GPS timing module, with a time synchronization error ≤10μs. The data acquisition module simultaneously receives the reference sound field response data from the standard reference hydrophone. Target acoustic field response data of acoustic sensor array , where i,j are the hydrophone coordinate indices and time-domain signals.

[0085] Perform a Fourier transform on the time-domain signal to convert it into a frequency-domain signal. and Next, the reference amplitude is extracted. Reference phase Target amplitude Target phase Then, calculate the amplitude ratio and phase difference. The amplitude ratio is:

[0086] ;

[0087] The amplitude ratio reflects the difference in amplitude response between the hydrophone ij and the standard reference hydrophone.

[0088] The phase difference is:

[0089] ;

[0090] The phase difference reflects the difference in phase response between the hydrophone ij and the standard reference hydrophone.

[0091] Finally, the sound field data is corrected based on the amplitude ratio and phase difference; specifically, the sound pressure level in the original sound field data is adjusted. Based on the factory sensitivity of the hydrophone ij calculate:

[0092] ;

[0093] in, This is the output voltage of the hydrophone. .

[0094] Considering sensitivity drift in actual use, the amplitude ratio is used. Revised to:

[0095] .

[0096] Phase information in sound field data is used to analyze the sound wave propagation path and phase difference. This is mainly caused by the phase drift of the hydrophone itself and differences in installation position; the phase difference caused by differences in installation position can be calculated using geometric distance: The distance difference between hydrophone ij and the standard reference hydrophone is given by , where c is the speed of sound; the hydrophone's own phase drift correction is also included.

[0097] ;

[0098] The corrected phase difference used for sound insulation calculation is:

[0099]

[0100] in, For the phase correction of the frontal hydrophone, The phase correction for the back acoustic surface hydrophone.

[0101] This invention provides an in-situ evaluation system for the sound insulation performance of the acoustic baffle inside a ship's fairing under all operating conditions. Hydrophones in the acoustic sensor array may experience sensitivity drift and phase shift during long-term use, and the uneven sound field distribution within the fairing (e.g., high sound pressure level near the sound source and low sound pressure level further away) can lead to deviations in the sound field data. This embodiment, under ship shutdown conditions, transmits a standard target audio signal through a central axis sound source. A standard reference hydrophone collects baseline sound field response data, and the acoustic sensor array collects target sound field response data. Based on the amplitude ratio and phase difference of the two, the sensitivity drift and phase shift of each hydrophone are corrected, while simultaneously eliminating measurement deviations caused by uneven sound field within the fairing. The corrected sound field data is closer to the actual sound field state, providing a reliable basis for subsequent sound insulation calculations and avoiding evaluation deviations caused by sensor errors or uneven sound field.

[0102] As an optional implementation, the in-situ evaluation system for the sound insulation performance of the acoustic baffle inside the ship's fairing under all operating conditions, including the data processing module, is also used to: collect ship operating condition parameters; determine the number of operating condition mutation parameters based on the ship operating condition parameters and the operating condition mutation algorithm; and determine the data acquisition frequency of the data acquisition module based on the number of operating condition mutation parameters.

[0103] For example, to improve data acquisition accuracy during sudden changes in operating conditions (such as a sudden increase in speed or a sudden change in main engine load), the data processing module can dynamically adjust the data acquisition frequency based on the operating condition change algorithm, as follows: a speed change rate > 0.3 knots / second is considered a speed change; a load change rate > 5% / second is considered a load change; and a draft change rate > 0.5 m / min is considered a draft change. When any of the above changes occurs: the acquisition frequency is increased to a first frequency, such as 5 kHz, and acquisition continues until the operating condition stabilizes, then returns to normal, such as 1 kHz; when multiple changes occur simultaneously, the acquisition frequency is increased to a second frequency, which is higher than the first frequency.

[0104] Finally, it should be noted that the above preferred embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit it. Although the present invention has been described in detail through the above preferred embodiments, those skilled in the art should understand that various changes can be made to it in form and detail without departing from the scope defined by the claims of the present invention.

Claims

1. A system for in-situ evaluation of sound insulation efficiency of a sound barrier in a full working condition in a ship's fairwater dome, characterized in that, include: Multiple acoustic sensor arrays are mirror-mounted on the front and back sides of the sound baffle to receive sound field data from the front and back sides of the sound baffle. The data acquisition module is connected to multiple acoustic sensor arrays and is used to receive and store sound field data sent by the acoustic sensor arrays and to acquire ship operating parameters at the corresponding time. The data processing module is used to evaluate the sound insulation performance of the sound baffle under the corresponding operating conditions based on the sound field data and the ship's operating parameters at the corresponding time, and obtain the evaluation results. The system also includes: Vibration sensors are installed at the target locations to acquire vibration parameters at the corresponding target locations; The data processing module also includes: Based on the sound field data and the ship's operating parameters and vibration parameters at the corresponding time, the sound insulation performance of the sound baffle under the corresponding operating conditions is evaluated. Based on sound field data and the ship's operating parameters and vibration parameters at the corresponding time, the sound insulation performance of the sound baffle under the corresponding operating conditions is evaluated, including: Based on the sound field data of the front and back surfaces, the initial sound insulation of the sound baffle is calculated. The pre-stored operating condition-vibration-propagation coupling weight matrix is ​​invoked. The coupling weight matrix is ​​trained from historical flight data. The historical flight data includes at least operating condition parameters, vibration parameters and corresponding actual sound insulation deviation data. The coupling weight matrix is ​​used to characterize the coupling correlation effect among operating condition, vibration and sound propagation. The operating parameters and vibration parameters are input into the coupling weight matrix to calculate the comprehensive correction amount. Based on the initial sound insulation and the comprehensive correction, the sound insulation under working conditions is calculated, and the frequency response curve of the sound insulation under working conditions is generated. Based on the operating condition sound insulation frequency response curve and the preset design threshold, the weak frequency band with sound insulation below the design threshold is output. or, Under the target operating conditions, based on the sound field data and the ship's operating parameters and vibration parameters at the corresponding time, the sound insulation performance of the sound baffle under the corresponding operating conditions is evaluated, including: Based on the sound field data of the front and back surfaces, the initial sound insulation of the sound baffle is determined. Based on the vibration parameters, the initial sound insulation is corrected for vibration effect to obtain the vibration-corrected sound insulation. The pre-stored operating condition influence weight matrix is ​​called, and the operating condition propagation deviation is corrected for the vibration correction sound insulation amount based on the operating condition parameters. The operating condition sound insulation amount frequency characteristic curve is output. The operating condition influence weight matrix is ​​trained from historical flight data. Based on the operating condition sound insulation frequency response curve and the preset design threshold, the output sound insulation is lower than the design threshold in the weak frequency band.

2. The in-situ evaluation system for the all-condition sound insulation performance of the acoustic baffle inside the ship's fairing according to claim 1, characterized in that, The process of determining the weight matrix of operating conditions includes: The correction relationship between historical flight data, vibration data and sound insulation, and the frequency curve of the reference sound insulation are obtained. The historical flight data includes operating parameters, vibration data and sound field data of the front and back surfaces of the sound baffle. The correction relationship between vibration data and sound insulation and the frequency curve of the reference sound insulation are all determined based on laboratory vibration tests. Based on the pre-allocated frequency bands, perform the following operations within each frequency band: Based on the operating parameters, features under each operating condition are extracted to obtain the fusion feature matrix for that frequency band; Based on the vibration data from historical flight data and the relationship between vibration data and sound insulation, the vibration sound insulation correction amount for this frequency band is determined. Based on the sound field data of the front and back surfaces of the sound baffle, the vibration sound insulation correction amount in this frequency band, and the reference sound insulation frequency curve, the working condition propagation deviation amount in this frequency band is determined. Using the fusion feature matrix of this frequency band as input and the working condition propagation deviation of this frequency band as label, the gradient boosting tree model is used for training to obtain the trained target model. The influence coefficients of each feature on the target variable are extracted from the trained target model to form the operating condition influence weight matrix for that frequency band.

3. The in-situ evaluation system for the all-condition sound insulation performance of the acoustic baffle inside the ship's fairing according to claim 1, characterized in that, The acoustic sensor array consists of multiple hydrophones, and the system also includes: The sound source is positioned at the central axis of the hydrophone array and is used to transmit target audio signals when the ship is not in operation. A standard reference hydrophone is set in a preset stable sound field area to collect reference sound field response data of the target audio signal; Acoustic sensor arrays are also used to collect target sound field response data of target audio signals on the front and back sides of the sound baffle; The data processing module is also used to calculate the amplitude ratio and phase difference between each hydrophone in the acoustic sensor array and the standard reference hydrophone based on the reference sound field response data and the target sound field response data. The data processing module is also used to correct the sound field data based on the amplitude ratio and phase difference.

4. The in-situ evaluation system for the all-condition sound insulation performance of the acoustic baffle inside the ship's fairing according to claim 1, characterized in that, The data processing module is also used for: Collect ship operating condition parameters to determine the number of parameters that cause sudden changes in operating conditions; The data acquisition frequency of the data acquisition module is determined based on the number of parameters that cause sudden changes in operating conditions.

5. The in-situ evaluation system for the all-condition sound insulation performance of the acoustic baffle inside the ship's fairing according to claim 4, characterized in that, Based on the number of parameters indicating sudden changes in operating conditions, the data acquisition frequency of the data acquisition module is determined, including: When the number of parameters that cause sudden changes in operating conditions is less than or equal to the target threshold, the data acquisition frequency is determined to be the first frequency. When the number of parameters that cause sudden changes in operating conditions exceeds the target threshold, the data acquisition frequency is determined to be the second frequency, which is higher than the first frequency.

Images