A method and system for calculating sensitivity of a fiber optic microphone

Abstract

The present application relates to a kind of optical fiber microphone sensitivity calculation method and system, its method includes the following steps: step 1, test environment building and sound pressure and light phase signal acquisition;Step 2, data preprocessing and effective channel screening;Step 3, signal cutting and optimal sequence extraction;Step 4, sensitivity and sensitivity level calculation;Step 5, output and visualization.The beneficial effects of the present application are: significantly improve the accuracy, statistical stability and anti-interference ability of optical fiber microphone sensitivity measurement.In the test scene level, the scene designed by the present application eliminates the physical interference of the device main body to the sound field to the greatest extent.In the signal processing level, the present application uses multiple signal processing methods to effectively avoid the influence of transient mutation noise and outliers from the statistical point of view, which can better reflect the real performance of the device compared to single measurement and mean method.The present application realizes efficient, accurate and strong anti-noise optical fiber sensitivity characteristic calibration.

Description

Technical Field

[0001] This invention relates to the field of fiber optic sensing technology, specifically to a method and system for calculating the sensitivity of a fiber optic microphone. Background Technology

[0002] Fiber optic sensing technology uses optical fibers as the sensing medium to reflect changes in the external environment by monitoring changes in parameters such as light intensity, phase, frequency, and wavelength. It can be used to measure various physical quantities such as sound vibration, temperature, and strain. With its strong resistance to electromagnetic interference and ability to achieve distributed and ultra-long-distance monitoring, this technology has been widely applied in fields such as geophysical exploration, industrial condition monitoring, and perimeter security.

[0003] The function of an optical fiber microphone is to enhance the sensitivity of optical fibers to sound waves, thereby enabling the detection of weak sound signals. Sensitivity is a key performance indicator for optical fiber microphones. The following are patents related to measuring the sensitivity of optical fiber microphones:

[0004] (1) An optical microphone testing platform and testing system, application publication number: CN118612645A;

[0005] (2) Fiber optic sensitivity testing method and apparatus, application publication number: CN117433746A.

[0006] However, the aforementioned existing technologies still have significant limitations:

[0007] Although patent (1) proposes a test platform that is easy to build, its sensitivity definition is inconsistent with the sensitivity definition of fiber optic microphones commonly used in academia. The test parameter is the light wavelength rather than the light phase, which is a different category, and it lacks an effective algorithm to eliminate environmental disturbances.

[0008] Patent (2) is mainly used for underwater acoustic testing, while fiber optic microphones also have a large number of applications in the air environment. Their sensitivity calculation method is relatively simple, but they have weak ability to suppress environmental interference and the results are not stable enough.

[0009] Overall, most existing technologies focus on the design of test platforms or hardware structures, while lacking in-depth research on the sensitivity calculation methods themselves. This results in actual measurement results being easily affected by environmental noise, with limited stability and accuracy. Summary of the Invention

[0010] In summary, to address the technical problems of existing data processing algorithms lacking standardization and being susceptible to environmental interference, the present invention aims to provide a method and system for calculating the sensitivity of fiber optic microphones. This method effectively reduces environmental interference, making the measured sensitivity more accurate and reliable, thereby improving the scientific rigor and consistency of fiber optic microphone performance evaluation.

[0011] The technical solution of this invention to solve the above-mentioned technical problems is as follows: A method for calculating the sensitivity of an optical fiber microphone, comprising the following steps:

[0012] Step 1: Set up a standard noise reduction test environment. Play single-frequency sound waves at gradient frequencies using an acoustic excitation source. Collect the average sound pressure level data and optical phase signal matrix at each frequency point to obtain the average sound pressure level data and full-band optical phase signal matrix at each frequency point, providing raw input data for subsequent calculations.

[0013] Step 2: Based on the full-band optical phase signal matrix obtained in Step 1, the effective channel phase sequence with the largest signal fluctuation at each frequency point is screened out through low-frequency filtering and noise reduction and channel standard deviation analysis, thereby obtaining the effective channel phase sequence at each frequency point and removing invalid channel interference.

[0014] Step 3: Based on the effective channel phase sequence of each frequency point obtained in Step 2, extract signal features and cut effective segments through short-time Fourier transform, and finally select the most representative optimal phase sequence to obtain the optimal phase sequence and avoid interference from invalid signals.

[0015] Step 4: Based on the average sound pressure level at each frequency point obtained in Step 1 and the optimal phase sequence obtained in Step 3, the core index calculation is completed by calculating the sensitivity and sensitivity level of the fiber optic microphone at each frequency point and obtaining the full-band sensitivity and sensitivity level data.

[0016] Step 5: Based on the full-band sensitivity and sensitivity level data obtained in Step 4, output specific values ​​and plot the sensitivity response curve to intuitively present the sensitivity characteristics of the fiber optic microphone.

[0017] Based on the above technical solution, the present invention can be further improved as follows:

[0018] Furthermore, step 1 specifically involves:

[0019] Step 1.1: Fix the fiber optic microphone to be tested on a bracket with a height of ≥1.2m, and place a speaker at the same horizontal height at a distance of ≥1m directly in front of it, with the sound wave emission direction of the speaker facing the fiber optic microphone;

[0020] Step 1.2: Remove the fiber optic microphone and place the sound level meter in its original position;

[0021] Step 1.3, control the speaker at a frequency f A single-frequency sound wave is played periodically, and the sound pressure level reading is recorded during each 2-second effective playback period. The average sound pressure level at that frequency is then calculated. L f ;

[0022] Step 1.4, according to the intervalC The speaker's playback frequency increases by Hz. f Repeat step 1.3 until the highest frequency within the required test frequency range is reached, and record the average sound pressure level at each frequency point across the entire frequency band. L 500 , L 750 , L 1000 , ..., L 9750 , L 10000 ;

[0023] Step 1.5: Remove the sound level meter and reposition the fiber optic microphone in its original location;

[0024] Step 1.6, control the speaker at a frequency f Periodically play single-frequency sound waves, by sampling frequency f k The optical phase demodulator acquires the optical phase signal matrix Φ f The data is transmitted to the data processing terminal.

[0025] Step 1.7, by interval C The speaker's playback frequency increases by Hz. f Repeat step 1.6 until the highest frequency within the required test frequency range is reached, and record the optical phase signal matrix sequence Φ at each frequency point across the entire frequency band. 500 , Φ 750 , Φ 1000 , …, Φ 9750 , Φ 10000 .

[0026] Furthermore, all equipment in step 1.1 is placed in an anechoic chamber or sound-absorbing box.

[0027] Furthermore, the microphone and main unit of the sound level meter described in step 1.2 are arranged separately, and the two are connected by an extension line.

[0028] Furthermore, the playback method in step 1.3 is as follows: after playing 2 seconds of audio, pause for 1 second and loop the playback. T Second-rate;

[0029] The playback method in step 1.6 is as follows: play 2 seconds of audio, pause for 1 second, and repeat continuously. t s; the optical phase signal matrix Φ f For size D × N The matrix, where N For the number of channels, D The number of sampling points per channel, and satisfying the following conditions: D / f k = t .

[0030] Furthermore, step 2 specifically involves:

[0031] Step 2.1, the data processing terminal reads the optical phase signal matrix sequence Φ 500 , Φ 750 , Φ 1000 , …, Φ 9750 , Φ 10000 The first item Φ 500 Perform a Fast Fourier Transform on each column of the matrix to select columns with frequencies below the low-frequency cutoff. f d The components are set to 0, and then the denoised optical phase signal matrix Φ* is obtained by inverse fast Fourier transform. 500 The f d The lowest frequency below the range of frequencies to be measured;

[0032] Step 2.2, calculate the noise-reduced optical phase signal matrix Φ* 500 Phase standard deviation per channel s Select the channel with the largest standard deviation, and delete the first and last 5% of data points in the phase sequence of that channel to obtain the effective channel phase sequence. f 500 ;

[0033] Step 2.3, following the methods in steps 2.1 and 2.2, process the optical phase signal matrix sequence Φ. 500 , Φ 750 , Φ 1000 , …, Φ 9750 , Φ 10000 Φ 500 The external frequency matrices are processed sequentially to obtain the complete effective channel phase sequence. f 500 , f 750 , f 1000 , ..., f 9750 , f 10000 .

[0034] Furthermore, step 3 specifically involves:

[0035] Step 3.1, for the effective channel phase sequence f 500 Perform a short-time Fourier transform, record the maximum amplitude of the spectrum for each calculation segment, and form a sequence of maximum amplitudes of the spectrum U:{ U 1, U2, ..., U i-1 , U i}, and simultaneously record the starting point position of each calculation segment in the original signal, forming the calculation segment starting point sequence Z:{ Z 1, Z 2, ..., Z i-1 , Z i},in, i Calculate the number of segments for the short-time Fourier transform;

[0036] Step 3.2, find the maximum value in the sequence U. U max The calculation segment in step 3.1 is judged, and the segment with the largest amplitude of the spectrum is greater than 100%. U max 30% of the calculated segments are determined to be generalized effective signal segments. Each consecutive generalized effective signal segment is merged to obtain multiple transition signal segments. 15% of the data before and after each transition signal segment is deleted to obtain multiple effective signal segments.

[0037] Step 3.3: Calculate the maximum amplitude of the spectrum after Fast Fourier Transform for each effective signal segment, and select the signal segment with the amplitude at the median level as the optimal phase sequence. i 500 ;

[0038] Step 3.4: Following the methods in steps 3.1 to 3.3, process the effective channel phase sequence. f 500 , f 750 , f 1000 , ..., f 9750 , f 10000 Except f 500 The sequences of each frequency point outside the range are processed sequentially to finally obtain the complete optimal phase sequence. i 500 , i 750 , i 1000 , ..., i 9750 , i 10000 .

[0039] Furthermore, step 4 specifically involves:

[0040] Step 4.1, Calculate the optimal phase sequence i500 The Fast Fourier Transform is used to obtain the maximum amplitude of the spectrum. A 500 ;

[0041] Step 4.2: Calculate the sound pressure level obtained in step 1.4. L 500 Corresponding sound pressure p 500 ;

[0042] Step 4.3, calculate the sensitivity at 500 Hz. S 500 ;

[0043] Step 4.4, calculate the sensitivity level at 500Hz. M 500 ;

[0044] Step 4.5: Following the methods in steps 4.1 to 4.4, process the optimal phase sequence. i 500 , i 750 , i 1000 , ..., i 9750 , i 10000 Except i 500 The frequency sequences outside the range are processed sequentially to obtain the complete sensitivity. S 500 , S 750 , S 1000 , ..., S 9750 , S 10000 With sensitivity level M 500 , M 750 , M 1000 , ..., M 9750 , M 10000 .

[0045] Furthermore, step 5 specifically involves:

[0046] Step 5.1: Output the sensitivity of all frequency points calculated in step 4.5. S 500 , S 750 , S 1000 , ...,S 9750 , S 10000 With the sensitivity level M 500 , M 750 , M 1000 , ..., M 9750 , M 10000 ;

[0047] Step 5.2, plot the sensitivity response curve of the fiber optic microphone: using frequency... f The x-axis represents the number of points, and the y-axis represents the number of points, respectively, based on sensitivity. S and sensitivity level M Using the vertical axis as the ordinate, the test frequencies ( f , S f )and( f , M f Plot the points and connect them to generate a visual sensitivity response curve for the fiber optic microphone.

[0048] The present invention also provides a fiber optic microphone sensitivity calculation system for applying the above-described fiber optic microphone sensitivity calculation method, comprising:

[0049] The sound pressure and optical phase signal acquisition module provides raw input data for sensitivity calculation, establishes a standardized test environment, and acquires sound pressure and optical phase signals to ensure the accuracy and consistency of the data source.

[0050] The data preprocessing and channel filtering module removes interference noise from the original signal and filters out effective signal channels to provide a high-quality phase sequence for subsequent calculations.

[0051] The signal segmentation and optimal sequence extraction module extracts the optimal phase segment that best represents the acoustic excitation characteristics from the effective channel phase sequence, eliminates invalid signals and abnormal data, and improves the statistical rationality of the calculation.

[0052] The sensitivity and sensitivity level calculation module, based on the optimal phase sequence and measured sound pressure data, calculates the sensitivity and sensitivity level of the fiber optic microphone at each frequency using a standardized formula, thereby quantifying the core indicators.

[0053] The output and visualization module presents the final calculation results, providing users with an intuitive and easy-to-interpret basis for sensitivity evaluation.

[0054] The beneficial effects of this invention are: significantly improved accuracy, statistical stability, and anti-interference capability of fiber optic microphone sensitivity measurement. At the testing scenario level, the scenario designed in this invention minimizes physical interference from the device itself to the sound field. At the signal processing level, this invention employs multiple signal processing methods to effectively avoid the influence of transient noise and outliers from a statistical perspective, reflecting the true performance of the device more accurately than single measurements and the mean method. This invention achieves efficient, accurate, and noise-resistant fiber optic sensitivity characteristic calibration. Attached Figure Description

[0055] Figure 1 This is a flowchart illustrating the method described in this invention;

[0056] Figure 2 This is a block diagram of the system composition described in this invention. Detailed Implementation

[0057] The principles and features of the present invention are described below with reference to the accompanying drawings. The examples given are only for explaining the present invention and are not intended to limit the scope of the present invention.

[0058] like Figure 1 As shown, a method for calculating the sensitivity of an optical fiber microphone includes the following steps:

[0059] Step 1: Set up a standard noise reduction test environment. Play single-frequency sound waves at gradient frequencies using an acoustic excitation source, and collect the average sound pressure level data and optical phase signal matrix at each frequency point. Obtain the full-band optical phase signal matrix from the optical phase signal matrices at each frequency point, thus obtaining the average sound pressure level data and the full-band optical phase signal matrix for subsequent calculations. This step is the data source for sensitivity calculations, providing the original input data for the calculations, specifically including:

[0060] Step 1.1: Fix the fiber optic microphone under test on a bracket with a height ≥1.2m. Place a speaker at the same horizontal height ≥1m directly in front of the microphone, with the speaker's sound wave emission direction facing the microphone. The speaker, as the acoustic excitation source, is also placed on a bracket, with its sound wave emission direction facing the fiber optic microphone under test. All the above equipment should be placed in an anechoic chamber or sound-absorbing box, preferably an anechoic chamber environment.

[0061] Step 1.2: Remove the microphone under test and place the sound level meter in its original position. The sound level meter must be located in the same sound field as the subsequent microphone to accurately measure the actual sound pressure level produced by the speaker at that point, avoiding data distortion due to placement errors. The sound level meter consists of a microphone (probe) and a main unit (display and processing unit). Typically, the microphone is directly mounted on the main unit. To prevent the large size of the main unit from affecting the sound field, the microphone and main unit can be arranged separately and connected by an extension cable. Using an extension cable to connect the microphone and main unit of the sound level meter reduces the main unit's obstruction and reflection interference in the sound field, ensuring sound field uniformity and avoiding significant interference with the sound field distribution at the probe, thus achieving more accurate sound measurements.

[0062] Step 1.3, control the speaker at a frequency f The system plays a single-frequency sound wave periodically, with a 1-second pause after every 2 seconds of audio playback, and then repeats the playback in a loop. T During each 2-second effective audio playback period, the sound pressure level (in dB) reading of the sound level meter was recorded. L f, 1 , L f, 2 , ..., L f,T This reading is allowed to fluctuate by approximately ±0.2 dB. Calculated from this sequence... f Average sound pressure level at frequency L f . f The initial value is the lowest frequency in the range to be measured.

[0063] In one embodiment, f The initial value is 500 Hz. T The value is set to 10. Single-frequency sound waves can avoid sound pressure interference caused by the superposition of multiple frequencies and accurately match the frequency correspondence of subsequent optical phase signals; the "play 2s + pause 1s" cycle mode can ensure stable acquisition of sound pressure level (2s is enough for the sound pressure signal to reach a steady state) and eliminate the residual influence of the sound waves from the previous playback by pausing; the average value is taken after 10 cycles, which can reduce the interference of random environmental noise (such as background noise in an anechoic chamber) on sound pressure data and improve data stability.

[0064] Step 1.4, according to the interval C Hz increments in the frequency of the loudspeaker f Repeat step 1.3 until the highest frequency within the required test frequency range is reached, and record the average sound pressure level at each frequency point across the entire frequency band. L 500 , L 750 , L 1000 , ..., L9750 , L 10000 .

[0065] In one embodiment, the step size C is 250 Hz, and the highest frequency is 10 kHz. 500Hz~10kHz is the core operating frequency band of the fiber optic microphone, covering most practical application scenarios (such as voice communication and industrial inspection); the 250Hz frequency interval can ensure the integrity of the frequency band coverage, avoid the low testing efficiency caused by too close an interval, and at the same time meet the resolution requirements for subsequent sensitivity response curve plotting, ensuring that the curve is smooth and can reflect the characteristics of key frequency points.

[0066] Step 1.5: Remove the sound level meter and reposition the fiber optic microphone under test in its original location. The sound level meter and microphone cannot be in the same test position simultaneously (as they will block and interfere with each other's sound field). Therefore, the order of "measuring sound pressure first, then phase" must be adopted, and the placement of the microphone and sound level meter must be completely consistent to ensure that the optical phase signal and sound pressure level data accurately correspond to the same sound field conditions during subsequent calculations, avoiding systematic errors introduced by positional deviations.

[0067] Step 1.6, control the speaker at a frequency f The system plays a single-frequency sound wave periodically, with a 2-second audio playback followed by a 1-second pause, repeating continuously. t s. By sampling frequency f k The optical phase demodulator acquires the optical phase signal matrix Φ f The optical phase signal matrix Φ is transmitted to the data processing terminal. f For size D × N The matrix, N For the number of channels, D The number of sampling points per channel, and satisfying the following conditions: D / f k = t . N The settings need to cover all points on the fiber optic microphone or 20 points near the maximum point.

[0068] In one embodiment, t It is 120 seconds. f k 100 kHz fThe initial frequency is 500 Hz. Using the same sound wave playback mode as the sound pressure level test ensures consistent sound excitation conditions and precise frequency and intensity matching between the optical phase signal and the sound pressure level data. Continuous acquisition for 120 seconds covers 10 cycles (3 seconds each, totaling 30 seconds), reserving ample data redundancy for subsequent filtering of valid signals. The 100kHz sampling frequency is significantly higher than the highest test frequency of 10kHz (satisfying the Nyquist sampling theorem, the sampling frequency must be ≥2 times the highest signal frequency), avoiding phase signal aliasing distortion. Multi-channel (N-channel) acquisition covers key points in the microphone's sensitive area, capturing the phase signal that best reflects the sound excitation response.

[0069] Step 1.7, by interval C Hz increments in the frequency of the loudspeaker f Repeat step 1.6 until the highest frequency within the required test frequency range is reached, and record the optical phase signal matrix sequence Φ at each frequency point across the entire frequency band. 500 , Φ 750 , Φ 1000 , …, Φ 9750 , Φ 10000 .

[0070] In one embodiment, similar to step 1.4, the step size C is 250 Hz, and the highest frequency is 10 kHz. Maintaining consistency with the frequency interval of the sound pressure level test ensures that each frequency point has a corresponding "sound pressure level-optical phase" data pair, providing complete input for subsequent sensitivity calculations at each frequency point. Full-band coverage comprehensively reflects the sensitivity characteristics of the fiber optic microphone at different frequencies, avoiding incomplete characteristic evaluation due to missing frequency points.

[0071] Step 2: Based on the full-band optical phase signal matrix obtained in Step 1, the effective channel phase sequences with the largest signal fluctuations at each frequency point are selected through low-frequency filtering and noise reduction, and channel standard deviation analysis. This yields the effective channel phase sequences for each frequency point, removing invalid channel interference. This step aims to remove low-frequency noise interference and select effective channel phase sequences for sensitivity calculation, specifically including:

[0072] Step 2.1, the data processing terminal reads the optical phase signal matrix sequence Φ 500 , Φ 750 , Φ 1000 , …, Φ 9750 , Φ 10000 The first item Φ 500 Perform a Fast Fourier Transform (FFT) on each column of the matrix to select columns with frequencies below the low-frequency cutoff. f d The components are set to 0, and then the denoised optical phase signal matrix Φ* is obtained by inverse fast Fourier transform (IFFT). 500 .f d It should be lower than the lowest frequency in the frequency range to be measured.

[0073] In one embodiment, take f d = 490 Hz. The optical phase signal may be mixed with low-frequency environmental noise (such as equipment vibration and power supply interference, mostly below 490 Hz). This noise is unrelated to the test frequency band (500 Hz~10 kHz) signal and needs to be removed. FFT can convert the time-domain phase signal into a frequency-domain signal, which is convenient for accurately screening the target frequency band components. After setting the low-frequency components to zero, the time-domain signal is restored by IFFT, which can reduce noise without damaging the effective test signal.

[0074] Step 2.2, calculate the denoised optical phase signal matrix Φ* 500 Phase standard deviation per channel s This is used to characterize the fluctuation level of the phase sequence in each channel. The standard deviation of all channels is calculated. s 1, s 2, ..., s N Then, the channel with the largest standard deviation is selected, and the first and last 5% of data points of the phase sequence of that channel are deleted to reduce edge effects, thus obtaining the effective channel phase sequence. f 500 In multi-channel phase signals, some channels may have weak signals and small fluctuations (small standard deviation) due to distance from the microphone's sensitive area or fiber optic loss, which cannot reflect the acoustic excitation response. It is necessary to select the channel with the largest standard deviation (the phase fluctuation caused by acoustic excitation is the most obvious and the most effective information). The data at the beginning and end of the phase sequence may be affected by the transition interference (edge ​​effect) of the signal start / end. Removing 5% of the beginning and end can eliminate transition noise and retain the steady-state effective signal.

[0075] Step 2.3, following the methods in Steps 2.1 and 2.2, process the optical phase signal matrix sequence Φ. 500 , Φ 750 , Φ 1000 , …, Φ 9750 , Φ 10000 Φ 500 The frequency point matrices outside are processed sequentially. Specifically, Φ in sub-steps S21 and S22 is processed sequentially. 500 Replace with Φ in sequence 750 , Φ 1000 , …, Φ 9750 , Φ 10000 The denoised optical phase signal matrix Φ* at the corresponding frequency points is obtained respectively. 750 ,Φ* 1000 , …, Φ* 9750 ,Φ* 10000 and its effective channel phase sequence f 750 , f 1000 , ..., f 9750 , f 10000 Finally, a complete effective channel phase sequence is obtained. f 500 , f 750 , f 1000 , ..., f 9750 , f 10000 .

[0076] The effects of acoustic excitation at different frequencies on fiber optic microphones vary. The noise distribution and channel response characteristics of the phase signals at each frequency point differ. Therefore, noise reduction and channel selection need to be performed separately for each frequency point to ensure that the optimal effective phase sequence can be obtained for each frequency point. This provides accurate input for subsequent frequency division sensitivity calculations and avoids errors caused by cross-frequency processing.

[0077] Step 3: Based on the effective channel phase sequences obtained in Step 2, signal features are extracted and effective segments are cut using short-time Fourier transform. Finally, the most representative optimal phase sequence is selected, thus avoiding interference from invalid signals. This step involves cutting effective information segments from the effective channel phase sequences at each frequency and selecting the optimal phase sequence from among many segments. Specifically, this includes:

[0078] Step 3.1, for the effective channel phase sequence f 500 Perform a Short-Time Fourier Transform (STFT), record the maximum amplitude of the spectrum for each calculation segment, and form a sequence of maximum amplitudes of the spectrum U:{ U 1, U 2, ..., U i-1 , U i}, and simultaneously record the starting point position of each calculation segment in the original signal, forming the calculation segment starting point sequence Z:{ Z 1, Z 2, ..., Z i-1 , Z i},in, i Calculate the number of segments for STFT.

[0079] f 500As a long-term phase sequence, containing signals from multiple acoustic excitation cycles and noise during pauses, the STFT can divide the long-term signal into multiple short-term calculation segments (windowed processing) and convert each segment into the frequency domain. The effective acoustic excitation signal segment (the effective signal amplitude is much larger than the noise) can be quickly located by the maximum amplitude of the spectrum, providing a basis for subsequent signal segmentation. The starting point sequence Z can accurately mark the position of each calculation segment, facilitating subsequent screening and merging.

[0080] Step 3.2, find the maximum value in sequence U. U max The calculation segment in step 3.1 is judged, and the segment with the largest amplitude of the spectrum is greater than 100%. U max 30% of the calculated segments are identified as generalized effective signal segments. Each consecutive generalized effective signal segment is merged to obtain multiple transition signal segments. The data before and after each transition signal segment is removed to obtain multiple effective signal segments. Each effective signal segment can characterize the phase information generated by the corresponding 2-second audio excitation in step 1.

[0081] U max The effective signal segment corresponding to the strongest acoustic excitation has an amplitude greater than U max The 30% threshold can balance the integrity of the effective signal and the noise removal effect (below 30% is mostly noise or signal attenuation during pauses); the continuous generalized effective segment is the signal of one complete sound excitation (2s playback). After merging, 15% of the data before and after are deleted, which can remove the interference of the rising edge at the start of the signal and the falling edge at the end (non-steady-state signal), and retain the intermediate steady-state effective signal segment.

[0082] Step 3.3: Calculate the maximum amplitude of the spectrum after Fast Fourier Transform for each effective signal segment, and select the signal segment with the amplitude at the median level as the optimal phase sequence. i 500 This sequence is statistically significant and reasonable, and can represent the characteristics of sound source excitation at the corresponding frequency.

[0083] In the effective signal segments of multiple acoustic excitation cycles, there may be individual segments with abnormally high or low amplitudes due to instantaneous interference. Selecting the signal segment at the median level can avoid the influence of extreme values ​​and ensure the statistical representativeness of the optimal sequence. The phase characteristics of this signal segment can truly reflect the stable response of the microphone to acoustic excitation at this frequency point, providing the most reliable feature sequence for sensitivity calculation.

[0084] Step 3.4: Following the methods in steps 3.1 to 3.3, process the effective channel phase sequence. f 500 , f 750 , f1000 , ..., f 9750 , f 10000 Except f 500 The frequency sequences outside of these steps are processed sequentially. Specifically, steps 3.1 to 3.3 are processed sequentially. f 500 Replace with in sequence f 750 , f 1000 , ..., f 9750 , f 10000 The optimal phase sequence for each corresponding frequency point is obtained. i 750 , i 1000 , ..., i 9750 , i 10000 Finally, the complete optimal phase sequence is obtained. i 500 , i 750 , i 1000 , ..., i 9750 , i 10000 .

[0085] The phase signal characteristics corresponding to the acoustic excitation frequency and intensity at different frequencies are different, and there are differences in the distribution of extreme values ​​and the amplitude characteristics of the effective signal segment. It is necessary to extract the optimal phase sequence for each frequency point separately to ensure that the sensitivity calculation of each frequency point is based on the most stable and representative phase characteristics of that frequency point, and to avoid deviations caused by cross-frequency multiplexing processing results.

[0086] Step 4: Based on the average sound pressure level at each frequency point obtained in Step 1 and the optimal phase sequence obtained in Step 3, the core performance indicators are calculated by calculating the sensitivity and sensitivity level of the fiber optic microphone at each frequency point. This step, based on the optimal phase sequence and measured sound pressure data, calculates the sensitivity and sensitivity level at each frequency point, specifically including:

[0087] Step 4.1, Calculate the optimal phase sequence i 500 The Fast Fourier Transform (FFT) is used to obtain the maximum amplitude of the spectrum. A 500 .

[0088] i 500Given a time-domain phase sequence, FFT can convert it into a frequency-domain spectrum, with the maximum amplitude of the spectrum being... A 500 The peak value of the phase response induced by the acoustic excitation at this frequency (500Hz) is used to quantify the phase sensitivity of the microphone to the acoustic signal at that frequency, and is the core characteristic parameter for calculating sensitivity.

[0089] Step 4.2: Calculate the sound pressure level obtained in step 1.4. L 500 Corresponding sound pressure p 500 The calculation formula is as follows:

[0090] ;

[0091] In the formula, p N For reference sound pressure, take 20 uPa.

[0092] Average sound pressure level measured by a sound level meter L f Relative decibel values ​​(based on reference sound pressure) p N =20uPa (the sound pressure level at which the human ear can hear), the decibel value needs to be converted to an absolute sound pressure level using this formula. p 500 The formula derivation is based on the definition of sound pressure level, and the reverse calculation can accurately restore the actual sound pressure level at the test point, providing sound pressure input for sensitivity calculation.

[0093] Step 4.3, calculate the sensitivity at 500 Hz. S 500 The calculation formula is as follows:

[0094] ;

[0095] In the formula, S 500 The unit is rad / Pa.

[0096] The core definition of fiber optic microphone sensitivity is "the amount of phase change produced under a unit sound pressure level". A 500 The peak value (rad) of the phase response induced by the sound pressure at that frequency. p 500 The ratio of rad / Pa to the actual sound pressure level (Pa) directly quantifies the microphone's sensitivity to the sound signal at that frequency. The unit rad / Pa can intuitively reflect the correspondence between sound pressure and phase change.

[0097] Step 4.4, calculate the sensitivity level at 500Hz. M 500 The calculation formula is as follows:

[0098] ;

[0099] In the formula, M 500 The unit is dB re rad / μPa. M N The sensitivity reference value is set to 1 rad / μPa.

[0100] Sensitivity S These are absolute values; different microphones... S The differences can be several orders of magnitude; converting them to decibel values ​​(sensitivity levels) allows for a more intuitive comparison of sensitivity differences. The formula is based on the definition of decibels, with reference values... M N =1rad / μPa is the industry standard, and the conversion form conforms to the decibel quantification of acoustic parameters, making it easy to benchmark against industry standards and data of similar products.

[0101] Step 4.5: Following the methods in steps 4.1 to 4.4, process the optimal phase sequence. i 500 , i 750 , i 1000 , ..., i 9750 , i 10000 Except i 500 The frequency sequences outside the range are processed sequentially. Specifically, the sequences are processed sequentially. i 500 Replace with in sequence i 750 , i 1000 , ..., i 9750 , i 10000 The sensitivity at the corresponding frequency points was obtained respectively. S 750 , S 1000 , ..., S 9750 , S 10000 With sensitivity level M 750 , M 1000 , ..., M 9750 , M 10000 Finally, the complete sensitivity was obtained. S 500 ,S 750 , S 1000 , ..., S 9750 , S 10000 With sensitivity level M 500 , M 750 , M 1000 , ..., M 9750 , M 10000 .

[0102] The sensitivity of fiber optic microphones is frequency-dependent, with varying degrees of sensitivity at different frequencies (i.e., frequency response characteristics). It is necessary to calculate the sensitivity and sensitivity level at each frequency point in order to comprehensively characterize the microphone's performance within its core operating frequency band and avoid the one-sided performance evaluation caused by calculation at a single frequency point.

[0103] Step 5: Based on the full-band sensitivity and sensitivity level data obtained in Step 4, output the specific values ​​and plot the sensitivity response curve to visually represent the sensitivity characteristics of the fiber optic microphone. Specifically, this includes:

[0104] Step 5.1: Output the sensitivity of all frequency points calculated in step 4.5. S 500 , S 750 , S 1000 , ..., S 9750 , S 10000 With sensitivity level M 500 , M 750 , M 1000 , ..., M 9750 , M 10000 .

[0105] The output of specific values ​​can provide accurate quantitative indicators to meet the needs of product testing, performance calibration, quality control and other scenarios; full-band data can fully present the frequency response details of the microphone, providing data support for subsequent product optimization (such as fiber optic layout adjustment, demodulation algorithm improvement), and at the same time, it is easy to compare with the design specifications to verify whether the requirements are met.

[0106] Step 5.2, plot the sensitivity response curve of the fiber optic microphone: using frequency... fThe x-axis represents the number of points, and the y-axis represents the number of points, respectively, based on sensitivity. S and sensitivity level M Using the vertical axis as the ordinate, the test frequencies ( f , S f )and( f , M f Plot the points separately and connect them to generate a visual sensitivity response curve for the fiber optic microphone.

[0107] Compared to pure numerical values, visualized curves can more intuitively present the trend of microphone sensitivity changing with frequency (such as which frequency has the highest sensitivity and which frequency has significant attenuation), making it easier to quickly determine the flatness of the frequency response (a high-quality microphone should maintain a stable curve within the core frequency band); the dual vertical axis curve can simultaneously display absolute sensitivity and relative sensitivity levels, taking into account both the need for accurate quantification and intuitive comparison, and adapting to the analysis needs of different scenarios.

[0108] like Figure 2 As shown, the present invention also provides a fiber optic microphone sensitivity calculation system for applying the above-described fiber optic microphone sensitivity calculation method, comprising:

[0109] (1) Acquisition module for sound pressure and optical phase signals

[0110] It provides raw input data for sensitivity calculations, establishes a standardized testing environment, and collects sound pressure and optical phase signals to ensure the accuracy and consistency of the data source. Specifically, this includes:

[0111] Establish a suitable testing environment: Place the fiber optic microphone and speaker under test on a bracket ≥1.2m high, maintaining the same horizontal height and a distance ≥1m between them, and place the entire setup in an anechoic chamber or sound-absorbing box to minimize environmental noise interference. Collect sound pressure level data: Record sound pressure level readings at different frequencies (increasing at fixed intervals from the lowest to the highest value in the test frequency range) using a sound level meter, and calculate the average sound pressure level at each frequency as a sound pressure reference standard for sensitivity calculation. Collect optical phase signals: After removing the sound level meter, collect the optical phase signals of the fiber optic microphone at the corresponding frequencies using an optical phase demodulator, forming an optical phase signal matrix and transmitting it to the data processing terminal, ensuring signal coverage of key points of the fiber optic microphone.

[0112] (2) Data preprocessing and channel filtering module

[0113] It removes interference noise from the original signal, filters out the effective signal channels, and provides a high-quality phase sequence for subsequent calculations. Specifically, it includes:

[0114] Signal Denoising: Fast Fourier Transform (FFT) is performed on the optical phase signal matrix at each frequency to filter out noise components below the low-frequency cutoff frequency (less than the lowest value in the measured frequency range). Then, Inverse Fast Fourier Transform (IFFT) is used to obtain the denoised optical phase signal matrix. Effective Channel Selection: The standard deviation of the phase sequence of each channel in the denoised matrix (characterizing the degree of signal fluctuation) is calculated. The channel with the largest standard deviation (the most significant signal change, best reflecting the acoustic excitation response) is selected, and the first and last 5% of data points are deleted to reduce edge effects, forming the effective channel phase sequence. Batch Processing Adaptation: Denoising and channel selection operations are sequentially performed on the optical phase signal matrix across the entire frequency band (all test frequencies) to generate a complete set of effective channel phase sequences.

[0115] (3) Signal cutting and optimal sequence extraction module

[0116] It extracts the optimal phase segment that best characterizes the acoustic excitation features from the effective channel phase sequence, eliminating invalid signals and anomalous data, and improving the statistical rationality of the calculation. Specifically, it includes:

[0117] Feature Extraction and Signal Segmentation: Short-Time Fourier Transform (STFT) is performed on the effective channel phase sequences to identify the calculation segment corresponding to the maximum spectral amplitude. Consecutive effective signal segments are merged, and the first and last 15% of useless data are removed to obtain multiple effective signal segments that can characterize the 2-second audio excitation. Optimal Sequence Selection: The maximum spectral amplitude after FFT for each effective signal segment is calculated, and the signal segment with an amplitude at the median level is selected as the optimal phase sequence (balancing statistical representativeness and stability, avoiding the influence of extreme values). Full-Band Processing: Segmentation and selection operations are sequentially performed on the effective channel phase sequences corresponding to all test frequencies to generate a set of optimal phase sequences for the entire frequency band.

[0118] (4) Sensitivity and sensitivity level calculation module

[0119] Based on optimal phase sequences and measured sound pressure levels, it calculates the sensitivity and sensitivity level of fiber optic microphones at various frequencies using standardized formulas, thus quantifying core performance indicators. Specifically, this includes:

[0120] Amplitude Acquisition: Perform FFT on the optimal phase sequence to extract the maximum amplitude of the spectrum, which serves as the core quantitative indicator of the phase response. Sensitivity Calculation: Combine the average sound pressure level (SPL) at the corresponding frequency (converted from SPL level) to calculate the sensitivity (unit: rad / Pa) using a formula, reflecting the phase response intensity of the fiber optic microphone to SPL. Sensitivity Level Calculation: Convert the sensitivity to sensitivity levels (unit: dB re rad / μPa) for easy comparison and industry standard adaptation. Full-Band Batch Calculation: Sequentially perform amplitude extraction, sensitivity, and sensitivity level calculations on the optimal phase sequences of all test frequencies to obtain the core indicator data for the entire frequency band.

[0121] (5) Output and Visualization Module

[0122] It presents the final calculation results, providing users with an intuitive and easily interpretable basis for sensitivity evaluation. Specifically, this includes:

[0123] Data Output: Exports full-band sensitivity (rad / Pa) and sensitivity level (dB re rad / μPa) quantized data to meet precise analysis needs. Visualization: Plots sensitivity response curves with frequency on the x-axis and sensitivity and sensitivity level on the y-axis respectively. Connecting the plotted points visually presents the sensitivity characteristics of the fiber optic microphone at different frequencies, facilitating rapid performance assessment.

[0124] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A method for calculating the sensitivity of an optical fiber microphone, characterized in that, Includes the following steps: Step 1: Set up a standard noise reduction test environment. Play single-frequency sound waves at gradient frequencies using an acoustic excitation source. Collect the average sound pressure level data and optical phase signal matrix at each frequency point to obtain the average sound pressure level data and full-band optical phase signal matrix at each frequency point, providing raw input data for subsequent calculations. Step 2: Based on the full-band optical phase signal matrix obtained in Step 1, the effective channel phase sequence with the largest signal fluctuation at each frequency point is screened out through low-frequency filtering and noise reduction and channel standard deviation analysis, thereby obtaining the effective channel phase sequence at each frequency point and removing invalid channel interference. Step 3: Based on the effective channel phase sequence of each frequency point obtained in Step 2, the signal features are extracted and effective segments are cut through short-time Fourier transform. Finally, the most representative optimal phase sequence is selected to obtain the optimal phase sequence and avoid interference from invalid signals. Step 4: Based on the average sound pressure level at each frequency point obtained in Step 1 and the optimal phase sequence obtained in Step 3, the core index calculation is completed by calculating the sensitivity and sensitivity level of the fiber optic microphone at each frequency point and obtaining the full-band sensitivity and sensitivity level data. Step 5: Based on the full-band sensitivity and sensitivity level data obtained in Step 4, output specific values ​​and plot the sensitivity response curve to intuitively present the sensitivity characteristics of the fiber optic microphone.

2. The fiber optic microphone sensitivity calculation method according to claim 1, characterized in that, Step 1 is as follows: Step 1.1: Fix the fiber optic microphone to be tested on a bracket with a height of ≥1.2m, and place a speaker at the same horizontal height at a distance of ≥1m directly in front of it, with the sound wave emission direction of the speaker facing the fiber optic microphone; Step 1.2: Remove the fiber optic microphone and place the sound level meter in its original position; Step 1.3, control the speaker at a frequency f A single-frequency sound wave is played periodically, and the sound pressure level reading is recorded during each 2-second effective playback period. The average sound pressure level at that frequency is then calculated. L f ; Step 1.4, according to the interval C The speaker's playback frequency increases by Hz. f Repeat step 1.3 until the highest frequency within the required test frequency range is reached, and record the average sound pressure level at each frequency point across the entire frequency band. L 500 , L 750 , L 1000 , ..., L 9750 , L 10000 ; Step 1.5: Remove the sound level meter and reposition the fiber optic microphone in its original location; Step 1.6, control the speaker at a frequency f Periodically play single-frequency sound waves, by sampling frequency f k The optical phase demodulator acquires the optical phase signal matrix Φ f The data is transmitted to the data processing terminal. Step 1.7, by interval C The speaker's playback frequency increases by Hz. f Repeat step 1.6 until the highest frequency within the required test frequency range is reached, and record the optical phase signal matrix sequence Φ at each frequency point across the entire frequency band. 500 Φ 750 Φ 1000 , …, Φ 9750 Φ 10000 .

3. The fiber optic microphone sensitivity calculation method according to claim 2, characterized in that, All equipment in step 1.1 is placed in an anechoic chamber or sound-absorbing box.

4. The fiber optic microphone sensitivity calculation method according to claim 2, characterized in that, In step 1.2, the microphone and main unit of the sound level meter are arranged separately and connected by an extension cable.

5. The fiber optic microphone sensitivity calculation method according to claim 2, characterized in that, The playback method in step 1.3 is as follows: play the audio for 2 seconds, pause for 1 second, and then loop. T Second-rate; The playback method in step 1.6 is as follows: play 2 seconds of audio, pause for 1 second, and repeat continuously. t s; the optical phase signal matrix Φ f For size D × N The matrix, where N For the number of channels, D The number of sampling points per channel, and satisfying the following conditions: D / f k = t .

6. The fiber optic microphone sensitivity calculation method according to claim 2, characterized in that, Step 2 is as follows: Step 2.1, the data processing terminal reads the optical phase signal matrix sequence Φ 500 Φ 750 Φ 1000 , …, Φ 9750 Φ 10000 The first item Φ 500 Perform a Fast Fourier Transform on each column of the matrix to select columns with frequencies below the low-frequency cutoff. f d The components are set to 0, and then the denoised optical phase signal matrix Φ* is obtained by inverse fast Fourier transform. 500 The f d The lowest frequency below the range of frequencies to be measured; Step 2.2, calculate the noise-reduced optical phase signal matrix Φ* 500 Phase standard deviation per channel σ Select the channel with the largest standard deviation, and delete the first and last 5% of data points in the phase sequence of that channel to obtain the effective channel phase sequence. φ 500 ; Step 2.3, following the methods in steps 2.1 and 2.2, process the optical phase signal matrix sequence Φ. 500 Φ 750 Φ 1000 , …, Φ 9750 Φ 10000 Φ 500 The external frequency matrices are processed sequentially to obtain the complete effective channel phase sequence. φ 500 , φ 750 , φ 1000 , ..., φ 9750 , φ 10000 .

7. The fiber optic microphone sensitivity calculation method according to claim 6, characterized in that, Step 3 specifically involves: Step 3.1, for the effective channel phase sequence φ 500 Perform a short-time Fourier transform, record the maximum amplitude of the spectrum for each calculation segment, and form a sequence of maximum amplitudes of the spectrum U:{ U 1, U 2, ..., U i-1 , U i }, and simultaneously record the starting point position of each calculation segment in the original signal, forming the calculation segment starting point sequence Z:{ Z 1, Z 2, ..., Z i-1 , Z i }, in, i Calculate the number of segments for the short-time Fourier transform; Step 3.2, find the maximum value in the sequence U. U max The calculation segment in step 3.1 is judged, and the segment with the largest amplitude of the spectrum is greater than 100%. U max 30% of the calculated segments are determined to be generalized effective signal segments. Each consecutive generalized effective signal segment is merged to obtain multiple transition signal segments. 15% of the data before and after each transition signal segment is deleted to obtain multiple effective signal segments. Step 3.3: Calculate the maximum amplitude of the spectrum after Fast Fourier Transform for each effective signal segment, and select the signal segment with the amplitude at the median level as the optimal phase sequence. θ 500 ; Step 3.4: Following the methods in steps 3.1 to 3.3, process the effective channel phase sequence. φ 500 , φ 750 , φ 1000 , ..., φ 9750 , φ 10000 Except φ 500 The sequences of each frequency point outside the range are processed sequentially to finally obtain the complete optimal phase sequence. θ 500 , θ 750 , θ 1000 , ..., θ 9750 , θ 10000 .

8. The fiber optic microphone sensitivity calculation method according to claim 7, characterized in that, Step 4 is as follows: Step 4.1, Calculate the optimal phase sequence θ 500 The Fast Fourier Transform is used to obtain the maximum amplitude of the spectrum. A 500 ; Step 4.2: Calculate the sound pressure level obtained in step 1.

4. L 500 Corresponding sound pressure p 500 ; Step 4.3, calculate the sensitivity at 500 Hz. S 500 ; Step 4.4, calculate the sensitivity level at 500Hz. M 500 ; Step 4.5: Following the methods in steps 4.1 to 4.4, process the optimal phase sequence. θ 500 , θ 750 , θ 1000 , ..., θ 9750 , θ 10000 Except θ 500 The frequency sequences outside the range are processed sequentially to obtain the complete sensitivity. S 500 , S 750 , S 1000 , ..., S 9750 , S 10000 With sensitivity level M 500 , M 750 , M 1000 , ..., M 9750 , M 10000 .

9. The fiber optic microphone sensitivity calculation method according to claim 8, characterized in that, Step 5 specifically involves: Step 5.1: Output the sensitivity of all frequency points calculated in step 4.

5. S 500 , S 750 , S 1000 , ..., S 9750 , S 10000 With the sensitivity level M 500 , M 750 , M 1000 , ..., M 9750 , M 10000 ; Step 5.2, plot the sensitivity response curve of the fiber optic microphone: using frequency f The x-axis represents the number of points, and the y-axis represents the number of points, respectively, based on sensitivity. S and sensitivity level M Using the vertical axis as the ordinate, the test frequencies ( f , S f )and( f , M f Plot the points and connect them to generate a visual sensitivity response curve for the fiber optic microphone.

10. A fiber optic microphone sensitivity calculation system, used to apply the fiber optic microphone sensitivity calculation method as described in any one of claims 1-9, characterized in that, include: The sound pressure and optical phase signal acquisition module provides raw input data for sensitivity calculation, establishes a standardized test environment, and acquires sound pressure and optical phase signals to ensure the accuracy and consistency of the data source. The data preprocessing and channel filtering module removes interference noise from the original signal and filters out effective signal channels to provide a high-quality phase sequence for subsequent calculations. The signal segmentation and optimal sequence extraction module extracts the optimal phase segment that best represents the acoustic excitation characteristics from the effective channel phase sequence, eliminates invalid signals and abnormal data, and improves the statistical rationality of the calculation. The sensitivity and sensitivity level calculation module, based on the optimal phase sequence and measured sound pressure data, calculates the sensitivity and sensitivity level of the fiber optic microphone at each frequency using a standardized formula, thereby quantifying the core indicators. The output and visualization module presents the final calculation results, providing users with an intuitive and easy-to-interpret basis for sensitivity evaluation.

Images