Loss measurement method and device based on frequency domain peak-valley difference and quadratic envelope compensation

Abstract

The application provides a loss measurement method and device based on frequency domain peak-valley difference and quadratic envelope compensation, acquires a time domain signal of fiber ring cavity ring-down, establishes an ideal exponential pulse attenuation model, performs frequency domain conversion and amplitude spectrum construction, analyzes and defines the peak-valley difference of frequency domain harmonic peak value and valley value characteristics, introduces a window function to truncate the pulse sequence, performs frequency domain response analysis, performs quadratic polynomial fitting of the frequency domain amplitude spectrum envelope, and performs spectrum compensation processing; without relying on complex microwave devices such as electro-optic modulators and network analyzers, the method can be directly integrated into an existing time domain FLRD measurement system, reduces the system hardware cost and implementation difficulty, and has a good engineering application prospect.

Description

Technical Field

[0001] This invention relates to electrical engineering and optical engineering technologies, and more particularly to a loss measurement method and apparatus based on frequency domain peak-valley difference and secondary envelope compensation. Background Technology

[0002] Fiber optic ring-down (FLRD) technology is a method for measuring loss based on the physical process of light propagating back and forth multiple times and gradually attenuating within a closed fiber optic ring cavity. Originating from cavity ring-down spectroscopy, this technique characterizes cavity loss by measuring the attenuation rate of the optical signal within the cavity, independent of the absolute intensity of the light source. Therefore, it has significant advantages in low-loss detection and high-sensitivity sensing. Compared to traditional optical power detection methods, FLRD technology excels in system stability, resistance to light source fluctuations, and measurement sensitivity, and has been widely applied in fiber optic device loss characterization, fiber optic sensors, and chemical and biological detection. Currently, the most widely used method in FLRD technology is still the time-domain pulse ringing-down method. This method acquires the attenuated pulse sequence after multiple round trips of the optical signal in the fiber optic loop cavity, and fits its envelope with an exponential function to extract the ringing-down time constant to characterize the cavity loss. However, in practical applications, this type of method still has certain shortcomings in demodulation accuracy and data processing. Specifically, the time-domain method usually relies on the fitting result of the overall shape of the ringing-down curve, and the exponential fitting process is sensitive to noise, baseline drift, and local anomalies, and is easily affected by single pulse fluctuations or detection noise, resulting in a significant decrease in the accuracy of ringing-down time extraction. When the signal-to-noise ratio is low or the ringing-down process is short, the selection of the fitting interval and the initial parameter settings have a significant impact on the final result, thereby reducing the repeatability and reliability of the measurement results.

[0003] Furthermore, time-domain methods require complex fitting calculations during data processing, resulting in cumbersome algorithmic flows that are not conducive to real-time measurement and engineering integration. Therefore, under complex environments or conditions with weak loss variation, traditional time-domain ringback methods still have room for improvement in terms of data processing robustness and loss characterization stability.

[0004] To overcome the aforementioned problems, some studies have attempted to convert FLRD signals from the time domain to the frequency domain for analysis, using spectral characteristics to invert cavity loss. Among these, the microwave photonics-based frequency-domain FLRD method incorporates an electro-optic modulator and network analyzer, treating the fiber optic ring cavity as a frequency-domain filtering structure, and extracting decay time information through amplitude-frequency or phase-frequency responses. This type of method reduces the dependence on high-speed time-domain sampling to some extent and improves the signal-to-noise ratio of demodulation. However, this approach typically requires additional microwave modulation devices, resulting in a complex system structure, high hardware costs, and difficulty in direct compatibility with existing time-domain FLRD systems, thus limiting its engineering application and widespread adoption.

[0005] Therefore, current technologies still lack a frequency domain method that can directly utilize existing time-domain FLRDs to acquire signals and achieve high-precision loss measurement through a simplified data processing flow without increasing the complexity of hardware. In particular, how to effectively utilize multi-harmonic information and suppress spectral distortion caused by finite sampling and window functions during frequency domain analysis remains to be further addressed. Summary of the Invention

[0006] The purpose of this invention is to address the shortcomings of the prior art by providing a loss measurement method and apparatus based on frequency domain peak-valley difference and secondary envelope compensation, which can significantly improve the robustness of peak-valley difference parameters to noise and spectral distortion, thereby achieving stable and accurate measurement of fiber optic ring cavity loss.

[0007] To achieve the above objectives, the present invention adopts the following technical solution: This invention provides a loss measurement method based on frequency domain peak-to-valley difference and secondary envelope compensation, comprising the following steps: S1. Obtain the time-domain signal of the fiber optic ring cavity oscillation and establish an ideal exponential pulse attenuation model; In fiber optic ring decay, after the optical signal is injected into the fiber optic ring by the laser, it propagates multiple times within the fiber optic ring and gradually attenuates during each round trip due to factors such as fiber transmission loss, coupler insertion loss, and medium scattering loss. If the time-domain signal received by the oscilloscope is represented as an exponentially weighted discrete pulse sequence, then the ideal exponential pulse decay model is established as follows: (1); in, The initial light intensity before entering the fiber optic ring cavity; The effective refractive index of the optical fiber; The speed of light in a vacuum; The total loss for each round trip consists of the transmission loss of the optical fiber, the insertion loss of the coupler, and the scattering loss of the medium. This refers to the total length of the optical fiber. The time interval between adjacent pulses is a fixed value; For time; It is a pulse function; S2, frequency domain transformation and amplitude spectrum construction; S3. Analysis and definition of peak-valley difference of frequency domain harmonic peak and valley characteristics; S4. Introduce window functions to truncate the pulse sequence and perform frequency domain response analysis; S5. Quadratic polynomial fitting of the frequency domain amplitude spectrum envelope; S6, Spectrum Compensation Processing.

[0008] Furthermore, S2 specifically refers to: Performing a Fourier transform on the above formula (1) yields the frequency domain expression as follows: (2); in, Indicates frequency; The corresponding amplitude spectrum expression is: (3).

[0009] Furthermore, S3 specifically refers to: When the frequency satisfies At that time, the amplitude spectrum reaches its maximum value. It is an integer; When the frequency satisfies When the amplitude spectrum reaches its minimum value, the corresponding peak and trough amplitudes are expressed as follows: (4); (5); Then the peak-to-valley difference of the frequency domain amplitude spectrum for: (6); when At that time, the peak-to-valley difference is expressed as: (7).

[0010] Furthermore, S4 specifically includes: S401. By introducing a window function to truncate the infinite sequence, the actual time-domain signal is simulated. The time-domain signal after introducing the window function is expressed as follows: (8); in, This is the original time-domain signal, i.e., the ideal exponential pulse decay model; For the window function, the expression is: (9); Let the attenuation factor The expression is: (10); in, For the length of the window, The number of pulses after truncation; the truncated signal is a finite sequence: (11); S402. According to the Fourier transform convolution theorem, the truncated spectrum... Convolution of the original spectrum and the window spectrum: (12); in, It is the Fourier transform of the window function; the Fourier transform of formula (9) is: (13); The phase is negligible in the amplitude spectrum; therefore, the amplitude spectrum is: (14); In low frequency hour, ; The oscillation at high frequencies decays to This results in the overall spectrum not being flat, but rather showing a decaying trend; S403. Apply a Fourier transform to the finite sequence to obtain: (15); The corresponding amplitude spectrum expression is: (16); when That is, the decay is sufficient, and the molecule is approximately 1; The peak appears after signal truncation The time, expressed as: (17); The peak appears after signal truncation. The time, expressed as: (18); S404. After signal truncation, the peak-to-valley difference expression is: (19).

[0011] Furthermore, S5 specifically includes: S501, the peak frequency is located at the frequency , Extracting peak values The envelope is fitted as a quadratic polynomial, where, The fitted envelope is represented by the fitted quadratic polynomial: (20); This is the coefficient vector, used to describe the envelope of the fitted quadratic polynomial; It is a constant term; It is the coefficient of the first-order term, representing the linear trend of the envelope; The coefficient of the quadratic term represents the curvature of the envelope; S502, Objective Function The summation of the squared differences between the peak value and the fitted value is expressed as: (twenty one); Establish the Vandermonde matrix for: (twenty two); S503. Differentiate equation (22) and set the gradient to zero, we get: (twenty three); in, The final envelope coefficient solution is: (twenty four).

[0012] Furthermore, S6 specifically includes: Using the obtained quadratic envelope estimation function, the truncated frequency domain amplitude spectrum is compensated, and the compensated amplitude spectrum is expressed as: (25); in, It is used to suppress the amplification of high-frequency noise during the compensation process. The value range is 0.001 to 0.01; After compensation, the frequency domain amplitude spectrum tends to flatten within the selected frequency range, and the peak-to-valley differences of multiple harmonics tend to be consistent. The approximate expression regresses to: (26); After compensation, the peak-to-valley difference tends to stabilize within the selected frequency range.

[0013] Furthermore, the loss measurement device based on frequency domain peak-valley difference and secondary envelope compensation, used to implement the loss measurement method based on frequency domain peak-valley difference and secondary envelope compensation, also includes: a function signal generator, a laser source, an optical fiber coupler, an optical fiber ring cavity, a sensor unit, a photodetector, an oscilloscope, and a data processing device. The function signal generator is used to output a periodic modulation signal to drive the laser source to generate pulsed light; The pulsed light output from the laser source is injected into the fiber ring cavity via a fiber coupler, and propagates back and forth multiple times within the ring cavity to form an exponentially decaying pulse sequence. The sensor unit is disposed in the optical fiber ring cavity and is used to introduce additional losses generated by the physical quantity to be measured. The photodetector converts the optical signal output from the fiber optic ring cavity into an electrical signal; The oscilloscope samples the electrical signal and outputs time-domain oscillation data; The data processing device is connected to the oscilloscope and is used to perform Fourier transform, frequency domain peak-valley and secondary envelope compensation algorithms. It also calculates the equivalent loss parameters of the fiber optic ring cavity based on the compensated frequency domain peak-valley difference. By substituting the frequency domain peak-valley difference obtained from actual measurement into the corresponding relationship, it realizes the quantitative calculation of the additional loss of the sensor unit, which is used to evaluate the sensitivity and measurement performance of the sensor unit.

[0014] The beneficial effects of this invention are as follows: This invention directly characterizes loss changes through frequency domain peak-to-valley difference, avoiding the process of fitting the exponential decay curve in the traditional time-domain FLRD method, significantly simplifying the signal processing flow and reducing the impact of fitting error on the measurement results. Secondly, this invention makes full use of the information of multiple harmonics in the frequency domain and improves the ability to suppress random noise through peak-valley difference averaging, thereby improving the stability and repeatability of the measurement results. Furthermore, by introducing a quadratic envelope compensation algorithm, the high-frequency amplitude attenuation problem caused by the finite number of pulses and the window function effect is effectively corrected, the flatness of the spectrum is restored, and the linearity and accuracy of the peak-valley difference as a loss-sensitive parameter are significantly improved. Furthermore, the method of the present invention does not rely on complex microwave devices such as electro-optic modulators and network analyzers, and can be directly integrated into existing time-domain FLRD measurement systems, reducing system hardware costs and implementation difficulty, and has good engineering application prospects. Attached Figure Description

[0015] Figure 1 The flowchart shows a loss measurement method based on frequency domain peak-valley difference and secondary envelope compensation. Figure 2 This is a schematic diagram of a loss measurement device based on frequency domain peak-valley difference and secondary envelope compensation. Figure 3 Frequency domain amplitude plots for different attenuation coefficients; Figure 4(a) shows the linear fit when the coefficient of determination is 0.9974; Figure 4(b) shows the linear fit when the coefficient of determination is 0.9972; Figure 4(c) shows the linear fit when the coefficient of determination is 0.9964; Figure 5 Frequency domain amplitude diagrams after compensation under different attenuation coefficients; Figure 6 The graph is a linear fit of the peak-to-valley difference and oscillation time after compensation. Detailed Implementation

[0016] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to the accompanying drawings. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the invention.

[0017] Please see Figure 1 The loss measurement method based on frequency domain peak-valley difference and secondary envelope compensation includes the following steps: S1. Obtain the time-domain signal of the fiber optic ring cavity oscillation and establish an ideal exponential pulse attenuation model; In fiber optic ring decay, after the optical signal is injected into the fiber optic ring by the laser, it propagates multiple times within the fiber optic ring and gradually attenuates during each round trip due to factors such as fiber transmission loss, coupler insertion loss, and medium scattering loss. For time-domain pulse decay signals, continuous-time exponential decay is used for description. Assume the initial light intensity before entering the fiber optic loop is... The relationship between light intensity and time can be expressed as: ; because, Since both are constants, the exponential decay coefficient only depends on the total loss. related.

[0018] The initial wave decay time is Defined as the light intensity decreasing to the initial light intensity. of The specific time is as follows: ; When additional loss is introduced into the fiber optic ring cavity, the ringing time of the fiber optic ring cavity will change. Let the ringing time after introducing the loss be... Then additional losses for: ; If the time-domain signal received by the oscilloscope is represented as an exponentially weighted discrete pulse sequence, then the ideal exponential pulse decay model is established as follows: (1); in, The initial light intensity before entering the fiber optic ring cavity; The effective refractive index of the optical fiber; The speed of light in a vacuum; The total loss for each round trip consists of the transmission loss of the optical fiber, the insertion loss of the coupler, and the scattering loss of the medium. This refers to the total length of the optical fiber. The time interval between adjacent pulses; For time; It is a pulse function; S2, frequency domain transformation and amplitude spectrum construction; S3. Analysis and definition of peak-valley difference of frequency domain harmonic peak and valley characteristics; S4. Introduce window functions to truncate the pulse sequence and perform frequency domain response analysis; S5. Quadratic polynomial fitting of the frequency domain amplitude spectrum envelope; S6, Spectrum Compensation Processing.

[0019] Specifically, S2 is: Performing a Fourier transform on the above formula (1) yields the frequency domain expression as follows: (2); in, Indicates frequency; The corresponding amplitude spectrum expression is: (3).

[0020] Specifically, S3 is: When the frequency satisfies At that time, the amplitude spectrum reaches its maximum value. It is an integer; When the frequency satisfies When the amplitude spectrum reaches its minimum value, the corresponding peak and trough amplitudes are expressed as follows: (4); (5); Then the peak-to-valley difference of the frequency domain amplitude spectrum for: (6); when At that time, the peak-to-valley difference is expressed as: (7).

[0021] Specifically, S4 is: In actual measurements, the time-domain decay signal contains only a finite number of pulses, which is equivalent to truncating an ideal infinite pulse sequence.

[0022] S401. By introducing a window function to truncate the infinite sequence, the actual time-domain signal is simulated. The time-domain signal after introducing the window function is expressed as follows: (8); in, This is the original time-domain signal, i.e., the ideal exponential pulse decay model; For the window function, the expression is: (9); Let the attenuation factor The expression is: (10); in, For the length of the window, The number of pulses after truncation; the truncated signal is a finite sequence: (11); S402. According to the Fourier transform convolution theorem, the truncated spectrum... Convolution of the original spectrum and the window spectrum: (12); in, It is the Fourier transform of the window function; the Fourier transform of formula (9) is: (13); The phase is negligible in the amplitude spectrum; therefore, the amplitude spectrum is: (14); The function introduces a decaying envelope.

[0023] In low frequency hour, ; The oscillation at high frequencies decays to This results in the overall spectrum not being flat, but rather showing a decaying trend; S403. Applying a Fourier transform to the finite sequence yields: (15); The molecule introduces a cutoff term. ; when Large enough and When the numerator term has a negligible effect on the amplitude spectrum, the effect of the numerator term on the amplitude spectrum can be ignored.

[0024] The corresponding amplitude spectrum expression is: (16); when That is, the decay is sufficient, and the molecule is approximately 1; Overall, the truncated spectrum exhibits a comb-like structure, but multiplying by the effective sinc envelope leads to high-frequency attenuation.

[0025] The peak appears after signal truncation. The time, expressed as: (17); The peak appears after signal truncation. The time, expressed as: (18); S404. After signal truncation, the peak-to-valley difference expression is: (19).

[0026] Truncation amplitude spectrum Envelope attenuation is introduced to extract and compensate for the flatness of the ideal infinite spectrum. Assuming the truncated amplitude spectrum has been calculated, the envelope approximates a sinc shape in the low-frequency harmonic range, and within a finite frequency range, it can be fitted with a quadratic polynomial to capture the attenuated signal.

[0027] Specifically, S5 is: S501, the peak frequency is located at the frequency Extracting peak values The envelope is fitted as a quadratic polynomial, where, The fitted envelope is represented by the fitted quadratic polynomial: (20); This is the coefficient vector, used to describe the envelope of the fitted quadratic polynomial; It is a constant term; It is the coefficient of the first-order term, representing the linear trend of the envelope; The coefficient of the quadratic term represents the curvature of the envelope; S502, Objective Function The summation of the squared differences between the peak value and the fitted value is expressed as: (twenty one); Establish the Vandermonde matrix for: (twenty two); S503. Differentiate equation (22) and set the gradient to zero, we get: (twenty three); in, The final envelope coefficient solution is: (twenty four).

[0028] Specifically, S6 is: Using the obtained quadratic envelope estimation function, the truncated frequency domain amplitude spectrum is compensated, and the compensated amplitude spectrum is expressed as: (25); in, It is used to suppress the amplification of high-frequency noise during the compensation process. The value range is 0.001 to 0.01; After compensation, the frequency domain amplitude spectrum tends to flatten within the selected frequency range, and the peak-to-valley differences of multiple harmonics tend to be consistent. The approximate expression regresses to: (26); After compensation, the peak-to-valley difference tends to stabilize within the selected frequency range.

[0029] By introducing a quadratic compensation algorithm, the influence of finite pulse number and window function effect on frequency domain characteristics is effectively reduced.

[0030] Please see Figure 2 A loss measurement device based on frequency domain peak-valley difference and secondary envelope compensation is used to implement the loss measurement method based on frequency domain peak-valley difference and secondary envelope compensation. It also includes: a function signal generator, a laser source, an optical fiber coupler, an optical fiber ring cavity, a sensor unit, a photodetector, an oscilloscope, and a data processing device. The function signal generator 11 is used to output a periodic modulation signal to drive the laser source to generate pulsed light; The pulsed light output from the laser source 22 is injected into the fiber ring cavity 55 via the fiber coupler 44, and propagates back and forth multiple times within the ring cavity to form an exponentially decaying pulse sequence. The sensor unit 66 is disposed in the optical fiber ring cavity and is used to introduce additional losses generated by the physical quantity to be measured. The photodetector 77 converts the optical signal output from the fiber optic ring cavity 55 into an electrical signal; The oscilloscope 88 samples the electrical signal and outputs time-domain oscillation data; The data processing device 99 is connected to the oscilloscope and is used to perform Fourier transform, frequency domain peak-valley and secondary envelope compensation algorithms, and calculate the equivalent loss parameters of the fiber ring cavity based on the compensated frequency domain peak-valley difference. By substituting the frequency domain peak-valley difference obtained from actual measurement into the corresponding relationship, the additional loss of the sensor unit is quantitatively calculated, which is used to evaluate the sensitivity and measurement performance of the sensor unit.

[0031] Isolator 33 prevents the optical signal from propagating bidirectionally within the loop and causing interference, ensuring unidirectional optical transmission to maintain a stable oscillating signal.

[0032] Simulation results: The effect of attenuation coefficient on spectral characteristics: To verify the effectiveness of the method of the present invention, an exponentially decaying pulse sequence was constructed to perform numerical simulation on the oscillation signal of the fiber optic ring cavity.

[0033] The initial light intensity was set to a fixed value, the pulse interval to a constant period, and the sampling frequency to meet the system sampling requirements. Different fiber optic ring cavity loss conditions were simulated by changing the attenuation coefficient. Under different attenuation coefficient conditions, the generated time-domain decaying signal was subjected to a Fourier transform to obtain the corresponding frequency-domain amplitude spectrum.

[0034] Simulation results show that as the attenuation coefficient increases, the amplitude of high-frequency harmonics in the frequency domain amplitude spectrum gradually decreases, the number of resolvable harmonics decreases, and the peak-to-valley difference number shows a significant decreasing trend with frequency, verifying the influence of the finite number of pulses on the spectral envelope.

[0035] To analyze the impact of the exponential decay coefficient on the spectral signal, an exponentially decaying pulse sequence model was constructed to simulate the ring cavity output signal in the FLRD system. A pulse format was used, with an initial light intensity of... Set to 4.0, total simulation time =10s, pulse interval is 1s (corresponding to a repetition frequency of 1Hz), sampling rate =5000Hz (dt=0.025s). Let the attenuation coefficient be... The values ​​were set to 0.2, 0.25, 0.3, 0.35, and 0.4 respectively to cover the typical low to medium loss range. The amplitude spectrum was obtained by directly performing FFT on the time-domain signal, and the simulation results are as follows. Figure 3 As shown in the figure. The results indicate that the larger the attenuation coefficient b, the more the peak amplitude gradually decreases with increasing frequency, the more significant the attenuation in the high-frequency part, and the fewer the number of peaks that can be distinguished.

[0036] Figure 4 shows the linear fit of the decay time of the average domain of different peak-valley differences. When only the first peak-valley difference is taken, the coefficient of determination of the fit is... The average peak-to-valley difference reached 0.9974, and the average peak-to-valley difference varied with the decay time. The linear increase is highly consistent with theoretical expectations, as shown in Figure 4(a); As the number of average points increases, when taking the first 10 peak-to-valley differences, the coefficient of determination for fitting increases. The accuracy dropped to 0.9972. Although the linear trend was still maintained, the accuracy showed a slight decrease, as shown in Figure 4(b). Further expanding to the first 20 peak-to-valley differences, the coefficient of determination for fitting was... The value will be reduced to 0.9964, as shown in Figure 4(c); The results show that as the peak-valley difference value increases, the linearity and accuracy of the peak-valley difference method in loss estimation gradually decrease. This gradual weakening of the correlation strength is mainly due to the inclusion of more high-frequency harmonic components. The amplitude reduction caused by the window function is more significant in the high-frequency region, exacerbating envelope deformation and noise effects, leading to increased fitting error and weakening the accuracy of the peak-valley difference in characterizing oscillation time.

[0037] After Fourier transform, the spectral envelope of a time-domain exponentially decaying signal often exhibits a frequency-dependent decay trend, which is mainly caused by spectral leakage and sidelobe effects resulting from the finite window function.

[0038] In practical FLRD systems, this attenuation amplifies noise interference and reduces the accuracy of loss characterization, especially in multiharmonic analysis, where the fitting coefficient of determination... It is prone to degradation. Therefore, it is possible to consider introducing a compensation algorithm to reverse envelope decay, ensuring a flatter spectrum, thereby improving the reliability of peak-to-valley difference as a loss indicator.

[0039] Quadratic envelope algorithm compensation: The primary approach employs a quadratic envelope compensation algorithm to compensate for the spectral signal. First, a peak-finding function is used to detect the main harmonic peaks in the amplitude spectrum, and a quadratic polynomial fitting is performed on the relationship between the peak frequency and the logarithmic peak amplitude to estimate the attenuation envelope curve. Then, the compensation factor is obtained by dividing the maximum peak amplitude by the inverse function of the fitted envelope, and this factor is multiplied by the original spectrum to achieve high-frequency amplitude enhancement.

[0040] like Figure 5 As shown, the frequency domain amplitude spectra of exponentially decaying pulse signals with different attenuation coefficients b=0.2, 0.25, 0.3, 0.35, and 0.4 are displayed. This compensation significantly reverses high-frequency attenuation, the peak distribution becomes more uniform, the number of high-frequency peaks increases, and the overall spectrum flattening is improved, highlighting the optimization of the spectral signal by the compensation. like Figure 6 As shown, the peak-to-valley difference and oscillation time coefficient of determination of fit The result is 0.9998, which is highly consistent with the theoretical value. The results show that the algorithm compensation achieves high-frequency signal enhancement, provides more accurate frequency domain characteristic signals for FLRD loss estimation, and expands the applicability of the method in noisy environments.

[0041] The embodiments described above are merely illustrative of implementation methods of the present invention, and while the descriptions are specific and detailed, they should not be construed as limiting the scope of the present invention. It should be noted that those skilled in the art can make various modifications and improvements without departing from the concept of the present invention, and these modifications and improvements all fall within the scope of protection of the present invention. Therefore, the scope of protection of this patent should be defined by the appended claims.

Claims

1. A loss measurement method based on frequency domain peak-valley difference and secondary envelope compensation, characterized in that, Includes the following steps: S1. Obtain the time-domain signal of the fiber optic ring cavity oscillation and establish an ideal exponential pulse attenuation model; In fiber optic ring decay, after the optical signal is injected into the fiber optic ring by the laser, it propagates multiple times within the fiber optic ring and gradually attenuates during each round trip due to factors such as fiber transmission loss, coupler insertion loss, and medium scattering loss. If the time-domain signal received by the oscilloscope is represented as an exponentially weighted discrete pulse sequence, then the ideal exponential pulse decay model is established as follows: （1）； in, The initial light intensity before entering the fiber optic ring cavity; The effective refractive index of the optical fiber; The speed of light in a vacuum; The total loss for each round trip consists of the transmission loss of the optical fiber, the insertion loss of the coupler, and the scattering loss of the medium. This refers to the total length of the optical fiber. The time interval between adjacent pulses is a fixed value; For time; It is a pulse function; S2, frequency domain transformation and amplitude spectrum construction; S3. Analysis and definition of peak-valley difference of frequency domain harmonic peak and valley characteristics; S4. Introduce window functions to truncate the pulse sequence and perform frequency domain response analysis; S5. Quadratic polynomial fitting of the frequency domain amplitude spectrum envelope; S6, Spectrum compensation processing.

2. The loss measurement method based on frequency domain peak-valley difference and secondary envelope compensation according to claim 1, characterized in that, Specifically, S2 is: Performing a Fourier transform on the above formula (1) yields the frequency domain expression as follows: （2）； in, Indicates frequency; The corresponding amplitude spectrum expression is: （3）。 3. The loss measurement method based on frequency domain peak-valley difference and secondary envelope compensation according to claim 2, characterized in that, Specifically, S3 is: When the frequency satisfies At that time, the amplitude spectrum reaches its maximum value. It is an integer; When the frequency satisfies When the amplitude spectrum reaches its minimum value, the corresponding peak and trough amplitudes are expressed as follows: （4）； （5）； Then the peak-to-valley difference of the frequency domain amplitude spectrum for: （6）； when At that time, the peak-to-valley difference is expressed as: （7）。 4. The loss measurement method based on frequency domain peak-valley difference and secondary envelope compensation according to claim 3, characterized in that, Specifically, S4 is: S401. By introducing a window function to truncate the infinite sequence, the actual time-domain signal is simulated. The time-domain signal after introducing the window function is expressed as follows: （8）； in, This is the original time-domain signal, i.e., the ideal exponential pulse decay model; For the window function, the expression is: （9）； Let the attenuation factor The expression is: （10）； in, For the length of the window, The number of pulses after truncation; the truncated signal is a finite sequence: （11）； S402. According to the Fourier transform convolution theorem, the truncated spectrum... Convolution of the original spectrum and the window spectrum: （12）； in, It is the Fourier transform of the window function; the Fourier transform of formula (9) is: （13）； The phase is negligible in the amplitude spectrum; therefore, the amplitude spectrum is: （14）； In low frequency hour, ; The oscillation at high frequencies decays to This results in the overall spectrum not being flat, but rather showing a decaying trend; S403. Apply a Fourier transform to the finite sequence to obtain: （15）； The corresponding amplitude spectrum expression is: （16）； when That is, the decay is sufficient, and the molecule is approximately 1; The peak appears after signal truncation The time, expressed as: （17）； The peak appears after signal truncation The time, expressed as: （18）； S404. After signal truncation, the peak-to-valley difference expression is: （19）。 5. The loss measurement method based on frequency domain peak-valley difference and secondary envelope compensation according to claim 4, characterized in that, Specifically, S5 is: S501, the peak frequency is located at the frequency , Extracting peak values The envelope is fitted as a quadratic polynomial, where, The fitted envelope is represented by the fitted quadratic polynomial: （20）； This is the coefficient vector, used to describe the envelope of the fitted quadratic polynomial; It is a constant term; It is the coefficient of the first-order term, representing the linear trend of the envelope; Here, represents the coefficient of the quadratic term, indicating the curvature of the envelope; S502, Objective Function The summation of the squared differences between the peak value and the fitted value is expressed as: （21）； Establish the Vandermonde matrix for: （22）； S503. Differentiate equation (22) and set the gradient to zero, we get: （23）； in, The final envelope coefficient solution is: （24）。 6. The loss measurement method based on frequency domain peak-valley difference and secondary envelope compensation according to claim 5, characterized in that, Specifically, S6 is: Using the obtained quadratic envelope estimation function, the truncated frequency domain amplitude spectrum is compensated, and the compensated amplitude spectrum is expressed as: （25）； in, It is used to suppress the amplification of high-frequency noise during the compensation process. The value range is 0.001 to 0.01; After compensation, the frequency domain amplitude spectrum tends to flatten within the selected frequency range, and the peak-to-valley differences of multiple harmonics tend to be consistent. The approximate expression regresses to: (26)； After compensation, the peak-to-valley difference tends to stabilize within the selected frequency range.

7. A loss measurement device based on frequency domain peak-valley difference and secondary envelope compensation, characterized in that, The loss measurement method based on frequency domain peak-valley difference and secondary envelope compensation as described in any one of claims 1 to 6 further includes: a function signal generator, a laser source, an optical fiber coupler, an optical fiber ring cavity, a sensor unit, a photodetector, an oscilloscope, and a data processing device. The function signal generator is used to output a periodic modulation signal to drive the laser source to generate pulsed light; The pulsed light output from the laser source is injected into the fiber ring cavity via a fiber coupler, and propagates back and forth multiple times within the ring cavity to form an exponentially decaying pulse sequence. The sensor unit is disposed in the optical fiber ring cavity and is used to introduce additional losses generated by the physical quantity to be measured. The photodetector converts the optical signal output from the fiber optic ring cavity into an electrical signal; The oscilloscope samples the electrical signal and outputs time-domain oscillation data; The data processing device is connected to the oscilloscope and is used to perform Fourier transform, frequency domain peak-valley and secondary envelope compensation algorithms. It also calculates the equivalent loss parameters of the fiber optic ring cavity based on the compensated frequency domain peak-valley difference. By substituting the frequency domain peak-valley difference obtained from actual measurement into the corresponding relationship, it realizes the quantitative calculation of the additional loss of the sensor unit, which is used to evaluate the sensitivity and measurement performance of the sensor unit.

Images