Unmanned ship navigation noise measurement system and data processing method

Abstract

The invention provides an unmanned ship navigation noise measurement system and a data processing method, which are used for solving the problems that the adaptability of an existing standard to an unmanned ship point sound source is poor, the laying and the recovery of a measurement system are tedious, and the deviation of a data processing result is large. Specifically, the measurement system is composed of a computer, a data acquisition module and the like, and the attitude and stability of the hydrophone are ensured through a single-point suspension design; according to data processing, 1-2 s time domain data about the maximum value of the broadband sound pressure level is selected, a line spectrum and a 1 / 3Oct spectrum are obtained through FIR band-pass filtering, framing windowing, FFT and the like, and the total sound pressure level and the sound source level are calculated. The noise of the unmanned ship is conveniently and accurately measured and processed, the test cost is saved, and the real level of the noise is accurately reflected.

Description

Technical Field

[0001] This invention belongs to the field of underwater radiated noise testing technology. Background Technology

[0002] Currently, there are corresponding national standards for the testing and data processing of ship noise, and the specific technical specifications are as follows: Test operation requirements: During the measurement process, the specified method must be followed (as shown in the attached document). Figure 1 (As shown) Deploy the underwater acoustic system; when the signal-to-noise ratio of the maximum broadband sound pressure level recorded by the radiated noise test system is greater than 6dB, it is determined to be a valid single voyage. Two valid single voyages need to be obtained under the same operating condition, and the difference between the broadband sound source levels of the two should be less than 3dB. Otherwise, it needs to be measured again.

[0003] Basic data processing requirements: After the test is completed, the radiated noise spectrum sound pressure level and the 1 / 3 Oct band sound pressure level need to be obtained through data analysis. The core process includes: (1) Selection of valid signals: To ensure data reliability, the selection of valid signals must meet the following four core requirements: Signal validity determination: Only when the signal output signal-to-noise ratio is greater than 6dB can the signal be included as a valid signal in subsequent analysis; Data selection rules: The measurement results must be based on the measurement data of all valid single voyages, and the average sound pressure level of two valid single voyages shall be taken as the final basic data; 1 / 3 Oct band analysis signal selection: When analyzing the radiated noise 1 / 3 Oct band sound pressure level spectrum, the signal that passes through the characteristic curve for 2-8 seconds to the left and right of the maximum value of the broadband sound pressure level should be selected as the analysis object. Signal consistency requirement: The signal selected for line spectrum analysis must be completely consistent with the signal for 1 / 3 Oct band sound pressure level analysis to ensure that the two types of analysis data are from the same source.

[0004] (2) Spectrum analysis requirements: Existing standards specify the following core parameters for spectrum analysis. 1 / 3 Oct band sound pressure level analysis: Play back the selected radiated noise signal and use a 1 / 3 Oct digital spectrum analyzer. The analysis process adopts the exponential averaging method. The averaging time is set to 2-8 seconds. The specific duration needs to be adapted according to the lower limit frequency of the signal to be analyzed and the actual distance between the hydrophone and the equivalent sound center of the object under test. Line spectrum sound pressure level analysis: Play back the selected radiated noise signal and analyze it using a signal analyzer. The sampling frequency must be no less than three times the highest analysis frequency. The window function can be selected from either the Hanning window function or the rectangular window function. The averaging method can be exponential averaging or linear averaging, and the averaging time is set to 4-8 seconds.

[0005] Although existing national standards provide a basic framework for implementation, the following key technical issues exist in the practical application of unmanned surface vessels: The measurement system is not practical enough: the existing sound field measurement system has a large number of instruments that are scattered, which makes the assembly time long. At the same time, the deployment and retrieval of hydrophones are difficult, which affects the testing efficiency.

[0006] Poor standard adaptability: Existing sound field testing standards are mainly applicable to equipment with large volume and many mechanical noise sources. Such equipment cannot be regarded as point sound sources in sound field measurement. However, unmanned surface vessels have the core characteristics of small size and simple power equipment. They should be regarded as point sound sources in sound field measurement. Moreover, the slope of their characteristic curve changes rapidly. Existing standards have not been adapted to the noise characteristics of such point sound sources.

[0007] Large deviations in data processing results: Unmanned surface vessels (USVs) are characterized by their small size and simple power equipment, and can be considered as point sources in sound field measurements. Furthermore, their high speed results in rapid changes in the slope of their characteristic curves. When analyzing the radiated noise of USVs, if the data duration used for signal processing is too long, the results obtained through Fourier transform and other data processing methods will deviate significantly from the actual total sound source level of the USV's navigation noise. Summary of the Invention

[0008] This invention provides an unmanned surface vessel (USV) navigation noise measurement system and data processing method, applicable to the measurement and processing of navigation noise from small USVs with simple power equipment that can be considered point sound sources. The specific technical solution is as follows: An unmanned surface vessel (USV) navigation noise measurement system is disclosed, which uses a single-point suspended hydrophone to measure the navigation noise of the USV. The system comprises a computer, a data acquisition module, a hydrophone, a lead weight, a nylon rope, and a rubber band. In use, the computer, data acquisition module, and hydrophone are connected together, and the nylon rope and hydrophone cable are fixed together. The lead weight is suspended from the end of the nylon rope to ensure the hydrophone is vertical. One end of the rubber band is connected to the nylon rope, and the other end is fixed to the measurement vessel to reduce the influence of the measurement vessel's sway on the hydrophone measurement results.

[0009] Furthermore, the system's acquisition parameters are set as follows: the sampling frequency is more than three times higher than the upper limit of the target analysis frequency; when acquiring line spectra, if the upper limit of the target analysis frequency is lower than 5kHz, the number of analysis lines is not less than 1600 lines, and if the upper limit of the target analysis frequency is higher than 5kHz, the number of analysis lines is not less than 6400 lines; when setting the 1 / 3 Octave spectrum, the broadband radiated noise level is set and displayed in real time for preliminary recording of test results and adjustment of the transverse distance between the hydrophone and the unmanned surface vessel under test.

[0010] A method for processing unmanned surface vessel (USV) navigation noise data, including 1) Select effective unmanned surface vessel (USV) navigation noise time-domain data within 1-2 seconds to the left and right of the maximum value of the broadband sound pressure level characteristic curve, and set the spectrum analysis parameter rules; 2) Based on the aforementioned time-domain data, perform the following two analyses in parallel: Line spectrum calculation: After designing an FIR bandpass filter, the filtered signal is windowed in frames and FFT transformation is performed. The power spectral density of each frame is calculated and then smoothed to output the final line spectrum of the radiated noise of the unmanned surface vessel at each frequency point, so as to achieve accurate capture of the narrowband noise characteristics of the unmanned surface vessel. 1 / 3 Oct Spectrum Calculation: Determine the center frequency, upper and lower limit frequencies, and corresponding FFT frequency indexes of each 1 / 3 Oct frequency band. Calculate the total power in each frequency band and perform exponential averaging. Output the average power data of each 1 / 3 Oct frequency band to quickly grasp the broadband noise distribution trend of the unmanned surface vessel. 3) Based on the average power data of the 1 / 3 Oct frequency band, the total power is obtained by summation. Combined with the reference sound pressure, the total sound pressure level of the unmanned surface vessel is calculated. At the same time, the line spectrum and the 1 / 3 Oct spectrum are plotted and analyzed according to the analysis requirements. Finally, the total sound pressure level, the total sound source level and the optional spectrum are output to achieve accurate quantification of the overall noise level of the unmanned surface vessel.

[0011] Furthermore, the rules for setting the spectrum analysis parameters are as follows: the Hanning window function is used for line spectrum analysis, and the smoothing process adopts exponential averaging or linear averaging; the 1 / 3 Oct spectrum analysis adopts exponential averaging. Furthermore, an FIR bandpass filter is designed, specifically based on the discrete-time impulse response formula of an ideal linear bandpass filter, combined with the lower cutoff frequency of the filter. Upper limit frequency With sampling frequency Calculate the ideal impulse response, extract the impulse response points corresponding to odd-order M and apply the Hanning window function to obtain the initial FIR bandpass filter coefficients. Then, correct these initial coefficients to obtain the final filter coefficients.

[0012] Beneficial effects 1. The measurement system is convenient, efficient, and provides reliable results: The unmanned surface vessel (USV) navigation noise measurement system of this invention consists of simplified components such as a computer, a data acquisition module, and a hydrophone. The structure is simple and easy to assemble. By using lead weights to ensure the hydrophone is vertical and rubber bands to cushion the movement of the measurement vessel, external interference is effectively reduced, ensuring accurate test results. At the same time, deployment and retrieval operations are simple, significantly saving manpower, material resources, and time costs for field tests, and solving the problems of dispersed equipment and complex operation in traditional measurement systems.

[0013] 2. Data processing methods adapted to the core characteristics of unmanned surface vessels: In view of the characteristics of unmanned surface vessels, such as small size, simple power equipment and rapid change of the slope of the navigation noise characteristic curve, we innovatively select short time domain data of 1-2 seconds around the maximum value of broadband sound pressure level for analysis. This not only avoids the problem of blurred noise time-varying characteristics caused by long-term data, but also accurately matches the noise propagation law as a point sound source, solving the pain point of poor adaptability of existing standards.

[0014] 3. Significantly improved accuracy in noise parameter calculation: By optimizing spectrum analysis parameters (using Hanning window function and adaptive exponential / linear averaging for line spectrum analysis, and exponential averaging for 1 / 3 Oct spectrum analysis), combined with targeted processes such as FIR bandpass filtering, frame windowing, and power spectral density smoothing, data processing deviations are effectively reduced. This allows for accurate reflection of the sound pressure level, source level, and spectral characteristics of unmanned surface vessel (USV) navigation noise, solving the problem of large deviations between existing standard data processing results and actual noise levels.

[0015] 4. Balancing safety and practicality in test operation: The measurement system adopts a single-point suspended deployment design, eliminating outward-extending cables and ropes. This avoids the risk of equipment getting tangled in the chassis of the unmanned surface vessel or towing equipment during testing, protecting the unmanned surface vessel under test and the measurement system from damage. At the same time, the collected parameters can be flexibly adjusted according to the target analysis frequency, adapting to the testing needs under different working conditions, making it more practical. Attached Figure Description

[0016] Figure 1 Underwater acoustic system structure diagram Figure 2 Schematic diagram of ship maneuvering methods Figure 3 Unmanned surface vessel (USV) navigation noise testing and data analysis process Figure 4 Schematic diagram of unmanned surface vessel (USV) navigation noise measurement deployment. Detailed Implementation

[0017] (1) Unmanned surface vessel navigation noise measurement system Traditional underwater radiated noise measurement system (see schematic diagram) Figure 1 The bottom-mounted measurement method has the following main problems: 1) The neutral support used in the system is heavy and requires multiple people to deploy and retrieve, which is time-consuming and inflexible; 2) There are many connecting cables and ropes, which are easy to get tangled in the chassis of the unmanned surface vessel and its towing equipment, causing damage to the unmanned surface vessel and the measurement system.

[0018] The single-point suspended hydrophone testing system used in this invention (see schematic diagram) Figure 4The equipment is simple, easy to deploy and retrieve, and can significantly save on labor and time costs; the hydrophone and counterweight are also vertically placed on the side of the ship, without any outward-extending cables and ropes, which improves safety.

[0019] The following will describe in detail the unmanned surface vessel (USV) navigation noise measurement system described in this invention: a) Deployment of the test system This scheme uses a single-point suspended hydrophone to measure the navigation noise of the unmanned surface vessel (USV). A deployment diagram of the USV navigation noise measurement system is shown below. Figure 4 The unmanned surface vessel (USV) navigation noise measurement system consists of a computer, a data acquisition module, a hydrophone, a lead weight, a nylon rope, and rubber bands. In use, the computer, data acquisition module, and hydrophone are connected, and the nylon rope and hydrophone cable are secured together. A lead weight is suspended from the end of the nylon rope to ensure the hydrophone is vertical; one end of the rubber band is connected to the nylon rope, and the other end is secured to the measurement vessel to reduce the impact of the vessel's movement on the hydrophone measurement results.

[0020] b) Data Acquisition Parameter Settings After the test system is connected, the acquisition parameters should be set. The sampling frequency should be set to be at least three times higher than the upper limit of the target analysis frequency. When setting up line spectrum acquisition, in order to observe the target radiated noise characteristics in real time, if the upper limit of the target analysis frequency is lower than 5kHz, the number of analysis lines should be no less than 1600 lines, and can be further refined. If the upper limit of the target analysis frequency is higher than 5kHz, the number of analysis lines should be no less than 6400 lines, and can be further refined. When setting up 1 / 3 Octave spectrum, the real-time measurement results of the broadband radiated noise level should be set and displayed to facilitate the initial recording of test results and to facilitate the adjustment of the transverse distance between the hydrophone and the target.

[0021] c) Data acquisition and storage After completing the signal acquisition settings, follow these steps to collect and store the measurement data: 1) Test and record the background noise; the measurement time should be no less than 3 minutes; 2) After the target is deployed, run it 1-2 times and observe the total source level of the radiated noise in real time. Based on the test results, plan the test lateral distance; 3) The unmanned surface vessel (USV) should proceed according to... Figure 2 Maneuver as shown, and the data recording time for each voyage should not be less than 20 seconds. Calculate whether the signal-to-noise ratio is greater than 6dB based on the background noise test results and the real-time flight noise test results to determine the validity of the measurement data. Each operating condition should include at least two single-trip valid voyages. 4) After completing the test of all operating conditions, the background noise should be tested and recorded again, and the measurement time should not be less than 3 minutes.

[0022] (2) A method for processing navigation noise data applicable to small-sized unmanned surface vessels and other vessels with simple power equipment. Traditional sound field testing standards are primarily applicable to large ships with numerous mechanical noise sources. These ships typically have low speeds and relatively flat radiated noise transmission characteristic curves. Therefore, to obtain stable spectral characteristics, a relatively long time-domain data period is selected when processing navigation noise data. Unmanned surface vessels (USVs), however, are characterized by their small size, simple propulsion systems, and relatively high speeds, resulting in rapidly changing transmission characteristic curve slopes. Using longer time-domain data periods can obscure the time-varying characteristics and lead to lower radiated noise source level test results, failing to accurately reflect the true level of USV navigation noise sources.

[0023] This invention proposes a method for processing unmanned surface vessel (USV) navigation noise data that is suitable for small-sized and simple-powered vessels. It uses shorter time-domain data, preserves the time-varying characteristics of USV navigation noise, and can accurately reflect the source level and spectral characteristics of USV navigation noise.

[0024] The following will describe in detail the data processing method of the present invention: a) Analysis of raw sound field data To select the analysis duration for the unmanned surface vessel (USV) navigation noise test data, in order to match the characteristics of USV navigation noise, signals with a bandwidth of 1-2 seconds to the left and right of the maximum value of the characteristic curve should be selected as the analysis object. When analyzing the line spectrum using a signal analyzer, the Hanning window function should be selected, and an exponential or linear averaging method should be used. When analyzing with a 1 / 3 Oct digital spectrum analyzer, an exponential averaging method should be used. The specific calculation method is as follows: 1) Calculate the radiated noise line spectrum For a selected time-domain data segment of finite length 1-2 seconds, the sampling frequency is... The data length is ; The discrete-time impulse response of an ideal linear bandpass filter is: (1) in: For integers, the theoretical range is... ; This is the lower cutoff frequency of the filter. This is the upper limit frequency of the filter;

[0025] center point The value at this location needs to be calculated separately:

[0026] To obtain an M-order (M is an odd number) finite-length FIR filter, truncation is performed. The number of points is determined, and a Hanning window is applied. The Hanning window function is defined as follows: (2) The discrete-time impulse response of an ideal linear bandpass filter is multiplied by a window function to obtain the coefficients of an FIR bandpass filter of length M. : (3) Correcting filter coefficients: (4) Signal The bandpass filtered signal is obtained through this FIR filter. : (5) The length of the filtered signal is .

[0027] Spectral analysis parameters FFT length is If the frame length is L and the number of overlapping samples between adjacent frames is R, then the total number of frames is calculated as follows: .

[0028] For the i-th frame of data ( ):

[0029] Add the Hanning window function: (6) Window normalization factor: (7) Data after windowing: (8) FFT transform: (9) The power spectral density is: (10) ① Exponential average The formula for calculating the exponentially weighted moving average is: (11) in: As a smoothing factor,

[0030] The formula for calculating the exponential average spectrum is: (12) in: .

[0031] ② Linear average The formula for calculating the power spectral density after linear averaging is: (13) 2) Calculate the radiated noise 1 / 3 Oct spectrum For each 1 / 3 Oct frequency band b, the center frequency is At that time, lower limit frequency upper limit frequency ,bandwidth .

[0032] The corresponding FFT frequency index is: (14) Calculate the power within the frequency band: (15) Perform an exponential average to obtain the band average power, where the exponential average power of the b-th band in the i-th frame is... It is expressed as follows: (16) in: .

[0033] b) Calculation of Results Calculation of total sound source level based on the 1 / 3 Oct spectrum: (17) in: For the power of the b-th 1 / 3 Oct, = .

[0034] The total sound pressure level is: (18) in: .

[0035] Based on the calculation results and analysis requirements, line spectra and 1 / 3 Oct spectra can be plotted and analyzed.

Claims

1. A system for measuring the navigation noise of an unmanned surface vessel, characterized in that, It consists of a computer, a data acquisition module, a hydrophone, a lead weight, a nylon rope, and a rubber band. In use, the computer, data acquisition module, and hydrophone are connected together, and the nylon rope and hydrophone cable are fixed together. The lead weight is suspended from the end of the nylon rope to ensure that the hydrophone is vertical. One end of the rubber band is connected to the nylon rope, and the other end is tied to the survey vessel to reduce the impact of the survey vessel's swaying on the hydrophone measurement results.

2. The system according to claim 1, characterized in that, The system's acquisition parameters are set as follows: the sampling frequency is more than three times higher than the upper limit of the target analysis frequency; when acquiring line spectra, if the upper limit of the target analysis frequency is below 5kHz, the number of analysis lines is not less than 1600 lines, and if the upper limit of the target analysis frequency is above 5kHz, the number of analysis lines is not less than 6400 lines; when setting 1 / 3 Octave spectrum, the broadband radiated noise level is set and displayed in real time for preliminary recording of test results and adjustment of the transverse distance between the hydrophone and the unmanned surface vessel under test.

3. A method for processing unmanned surface vessel (USV) navigation noise data, characterized in that, 1) Select effective unmanned surface vessel (USV) navigation noise time-domain data within 1-2 seconds to the left and right of the maximum value of the broadband sound pressure level characteristic curve, and set the spectrum analysis parameter rules; 2) Based on the aforementioned time-domain data, perform the following two analyses in parallel: Line spectrum calculation: After designing an FIR bandpass filter, the filtered signal is windowed in frames and FFT transformation is performed. The power spectral density of each frame is calculated and then smoothed to output the final line spectrum of the radiated noise of the unmanned surface vessel at each frequency point, so as to achieve accurate capture of the narrowband noise characteristics of the unmanned surface vessel. 1 / 3 Oct Spectrum Calculation: Determine the center frequency, upper and lower limit frequencies, and corresponding FFT frequency indexes of each 1 / 3 Oct frequency band. Calculate the total power in each frequency band and perform exponential averaging. Output the average power data of each 1 / 3 Oct frequency band to quickly grasp the broadband noise distribution trend of the unmanned surface vessel. 3) Based on the average power data of the 1 / 3 Oct frequency band, the total power is obtained by summation. Combined with the reference sound pressure, the total sound pressure level of the unmanned surface vessel is calculated. At the same time, the line spectrum and the 1 / 3 Oct spectrum are plotted and analyzed according to the analysis requirements. Finally, the total sound pressure level, the total sound source level and the optional spectrum are output to achieve accurate quantification of the overall noise level of the unmanned surface vessel.

4. The method according to claim 1, characterized in that, The rules for setting spectrum analysis parameters are as follows: the Hanning window function is used for line spectrum analysis, and the smoothing process adopts exponential averaging or linear averaging; the 1 / 3 Oct spectrum analysis adopts exponential averaging.

5. The method according to any one of claims 3-4, characterized in that, Design an FIR bandpass filter, specifically based on the discrete-time impulse response formula of an ideal linear bandpass filter, combined with the filter's lower cutoff frequency. Upper limit frequency With sampling frequency Calculate the ideal impulse response, extract the impulse response points corresponding to odd-order M and apply the Hanning window function to obtain the initial FIR bandpass filter coefficients. Then, correct these initial coefficients to obtain the final filter coefficients.

Images