Multi-feature fusion sapphire fiber fabry-perot sensor demodulation method, system and device

Abstract

The present application relates to the technical field of optical fiber sensor demodulation, and particularly relates to a multi-feature fusion sapphire fiber Fabry-Perot sensor demodulation method, system and device. The method comprises the following steps: collecting an interference spectrum signal, constructing a two-dimensional feature representation form and a frequency domain feature representing interference characteristics; inputting the two-dimensional feature representation form and the frequency domain feature into a feature mapping model; extracting an interference fringe of a specific interference cavity according to the feature mapping model, calculating a rough interference order of the interference fringe of the specific interference cavity according to the spectrum signal; calculating an accurate interference order according to the rough interference order, and inputting an accurate optical path difference into the mapping model to establish a mapping relationship between the environment pressure value and the accurate optical path difference. The present application introduces different dimensional spectral feature information, realizes stable extraction of target cavity interference information, and overcomes the problems of insufficient demodulation accuracy and robustness of existing demodulation methods under the conditions of multi-cavity interference, non-ideal spectrum and environmental disturbance.

Description

Technical Field

[0001] This invention relates to the field of fiber optic sensor demodulation technology, and in particular to a method, system and apparatus for demodulating a multi-feature fusion sapphire fiber Fabry-Perot sensor. Background Technology

[0002] Fiber optic sensors are characterized by their small size, resistance to electromagnetic interference, flexible structure, high sensitivity, and fast response, making them suitable for measuring physical quantities such as temperature and pressure in harsh environments. Compared to traditional silica fiber, sapphire fiber has a high melting point of 2040℃ and good mechanical strength, thus showing promising application prospects in extreme environments such as aerospace. The performance of sapphire fiber MEMS Fabry-Perot sensors in practical measurement applications depends on targeted demodulation, and obtaining the specific interferometric cavity spectrum of the sensor is the most effective demodulation method.

[0003] Generally, existing fiber optic Fabry-Perot sensor demodulation mostly employs broadband light sources and spectral analyzers to acquire the complete interference spectrum. When multi-cavity length information is included, the interference fringes are approximated as multi-frequency superposition and modulated by the broadband light source, then demodulated using spectral domain algorithms. Most systems are limited by the spectrometer sampling rate and the cost of the optoelectronic system, preventing ultra-high-speed demodulation. Furthermore, equal wavenumber sampling is required, necessitating interpolation algorithms for equal-interval processing. Changes in interpolation or spectral envelope introduce errors, and in multi-cavity systems, the multiple peaks in the spectrum interfere with each other. If only Fourier spectrum analysis is used, the interference spectrum of the Fabry-Perot sensor's composite cavities is susceptible to changes in the light source spectral shape, interference contrast, and noise, leading to decreased demodulation stability based on peak values ​​or single-frequency domain features, thus affecting the cavity length demodulation accuracy.

[0004] Therefore, how to provide a high-frequency, high-stability, low-cost, and easily integrated fiber optic sensing demodulator has become a technical problem that urgently needs to be solved by those skilled in the art. Summary of the Invention

[0005] This invention aims to at least solve one of the technical problems existing in related technologies. To this end, this invention provides a demodulation method for a sapphire fiber Fabry-Perot sensor with multi-feature fusion, achieving stable extraction of interference information from the target cavity and overcoming the problems of insufficient demodulation accuracy and robustness of existing demodulation methods under conditions of multi-cavity interference, non-ideal spectra, and environmental disturbances.

[0006] This invention provides a demodulation method for a multi-feature fusion sapphire fiber Fabry-Perot sensor, comprising: S1: Acquire the interference spectrum signal reflected by the sapphire fiber optic Fabry-Perot sensor; S2: Reconstruct the interference spectral signal to build a two-dimensional feature representation that characterizes the interference properties; S3: Perform Fourier spectrum analysis on the interference spectrum signal, convert the interference spectrum signal from the spectral domain to the frequency domain, and apply frequency domain physical suppression constraints to the obtained spectrum signal to obtain frequency domain characteristics that reflect the characteristics of the interference cavity; S4: The two-dimensional feature representation is fused with the frequency domain features and input as a joint feature into the feature mapping model. The interference fringes of a specific interference cavity are extracted according to the feature mapping model. S5: Obtain the peak wavelength of the interference fringes, and calculate the rough interference order of the interference fringes of the specific interference cavity based on the spectrum signal; S6: Calculate the precise interference order based on the coarse interference order, and input the precise interference order and interference fringes into the nonlinear mapping model to predict the precise optical path difference; S7: Based on the precise optical path difference, establish a mapping relationship between the environmental pressure value and the precise optical path difference, and calculate the environmental pressure value according to the mapping relationship.

[0007] According to the multi-feature fusion sapphire fiber Fabry-Perot sensor demodulation method provided by the present invention, step S2 includes: S21: Randomly mask the interference spectral signal to obtain the masked spectral signal; S22: Perform reconstruction processing based on the masked spectral signal to obtain the reconstructed spectral signal; S23: Normalize the reconstructed spectral signal, convert the normalized reconstructed spectral signal from Cartesian coordinates to polar coordinates, and construct a two-dimensional feature representation based on the polar coordinates.

[0008] According to the demodulation method of a multi-feature fusion sapphire fiber Fabry-Perot sensor provided by the present invention, step S23 includes: S231: Normalize the reconstructed spectral signal: in, The reconstructed spectral signal after normalization. To reconstruct the spectral signal, It is a function with maximum value. It is a minimum value function; S232: Map the normalized reconstructed spectral signal from the Cartesian coordinate system to the polar coordinate system, and calculate the polar angle corresponding to each sampling point. in, This represents the first light intensity value of the reconstructed spectral signal after normalization. The polar angle corresponding to the first light intensity value; S233: Based on the polar coordinate representation, calculate the trigonometric function relationship between polar angles at different wavelengths, construct the Gram angle field, and obtain a two-dimensional feature representation. Matrix elements in two-dimensional feature representation form Defined as: in, This is the polar angle corresponding to the second light intensity value.

[0009] According to the demodulation method of a multi-feature fusion sapphire fiber Fabry-Perot sensor provided by the present invention, step S3 includes the following steps: S31: Based on the interference spectrum signal, the uniform wavelength interval is converted into a uniform wavenumber interval through reciprocal transformation and interpolation to obtain the wavenumber representation; S32: Perform a fast Fourier transform on the interference spectrum signal represented by the wavenumber to obtain the Fourier spectrum signal; S33: Divide the Fourier spectrum signal into several frequency domain sub-intervals according to the block division rules, and each frequency domain sub-interval corresponds to a set of continuous frequency domain sampling points; S34: Based on the energy distribution characteristics of the frequency domain sub-intervals, determine the target sub-interval set, retain the frequency domain information corresponding to the target sub-intervals, and obtain the frequency domain characteristics reflecting the specific interference cavity characteristics.

[0010] According to the multi-feature fusion sapphire fiber Fabry-Perot sensor demodulation method provided by the present invention, step S34 includes: S341: Based on the energy distribution characteristics of each frequency domain sub-interval in the aforementioned spectrum signal, determine the set of target sub-intervals: in, It is the spectral signal after Fast Fourier Transform. For signal ordinal number, For the first Sub-intervals, This represents the local energy value of a sub-interval in the frequency domain; Set dynamic energy threshold Filter to meet The subintervals are used as the target subinterval set. ; S342: Apply frequency domain physical suppression constraints to non-target sub-intervals and construct a physical bandpass mask function. : S343: Multiply the spectral signal by the bandpass mask function to obtain the frequency domain characteristics reflecting the specific interferometric cavity features. : .

[0011] According to the multi-feature fusion sapphire fiber Fabry-Perot sensor demodulation method provided by the present invention, step S4 includes: the two-dimensional feature representation and the frequency domain features are input in parallel to the feature mapping model based on the same interference spectral signal, and the feature extraction calculation expression is obtained as follows: in, The fused joint feature vector This indicates a feature channel splicing operation. The first feature transformation weight matrix, The second feature transformation weight matrix, A vector in two-dimensional feature representation form. To characterize the frequency domain eigenvectors representing the frequency domain energy and position information of a specific interferometric cavity, As a network bias term, the fused joint feature vector is fed into the decoding layer for feature reconstruction, and interference fringes of a specific interference cavity are output.

[0012] According to the demodulation method of a multi-feature fusion sapphire fiber Fabry-Perot sensor provided by the present invention, step S5 includes: S51: Determine the coarse peak wavelength using the local maximum algorithm. And select near the approximate peak wavelength A peak-finding window is constructed using consecutive sampling points; the centroid peak-finding algorithm is used to calculate the accurate peak wavelength. The calculation formula is: in, This is the width parameter of one side of the peak-finding window; The index of the sampling points within the peak-finding window, and ; For the first The wavelength values ​​corresponding to each sampling point; For the first The light intensity value corresponding to each sampling point; S52: Coarse optical path difference The calculation formula is: Where N is the total number of data points after cubic spline interpolation. It is the wavenumber sampling interval, and k is the peak position index in the spectrum of the spectral signal; S53: The formula for calculating the rough interference order is: in It is a rough interference level.

[0013] According to the multi-feature fusion sapphire fiber Fabry-Perot sensor demodulation method provided by the present invention, step S6 includes: S61: The precise interference order The result is a rounding of the coarse interference order, and the extracted interference fringes of the specific interference cavity are constructed into a one-dimensional sequence. The fringe hidden feature vector is extracted by inputting it into the first hidden layer of the nonlinear mapping model. The calculation formula is: in, This is the weight matrix of the first hidden layer. It is a non-linear activation function. This is the bias vector for the first hidden layer; S62: The stripe hidden feature vector With the precise interference order Concatenate the features along the feature dimension to construct a joint feature vector. The calculation formula is: S63: Transfer the joint feature vector After subsequent nonlinear mapping between the hidden layer and the output layer, the accurate optical path difference is calculated. The calculation formula is: in, This is the weight matrix of the second hidden layer. This is the output layer weight matrix. This is the bias vector for the second hidden layer. This is the output layer bias term.

[0014] This invention also provides a multi-feature fusion sapphire fiber Fabry-Perot sensor demodulation system, comprising the following modules: Two-dimensional feature acquisition module: used to acquire the interference spectrum signal reflected by the sapphire fiber optic Fabry-Perot sensor, reconstruct the interference spectrum signal, and construct a two-dimensional feature representation that characterizes the interference properties; Frequency domain feature acquisition module: used to perform Fourier spectrum analysis on the interference spectrum signal, convert the interference spectrum signal from the spectral domain to the frequency domain, perform frequency domain physical suppression constraint on the obtained spectrum signal, and obtain frequency domain features reflecting the characteristics of the interference cavity; The mapping model training module is used to fuse the two-dimensional feature representation with the frequency domain features and input them as joint features into the feature mapping model, and extract the interference fringes of a specific interference cavity based on the feature mapping model. The environmental pressure value calculation module is used to obtain the peak wavelength of the interference fringes, calculate the coarse interference order of the interference fringes of a specific interference cavity based on the spectral signal, calculate the precise interference order based on the coarse interference order, input the precise interference order and interference fringes into a nonlinear mapping model to predict the precise optical path difference, establish a mapping relationship between the environmental pressure value and the precise optical path difference based on the precise optical path difference, and calculate the environmental pressure value based on the mapping relationship.

[0015] The present invention also provides an electronic device, including a processor, a communication interface, a memory, and a communication bus, wherein the processor is used to execute a computer program to implement the steps of the multi-feature fusion sapphire fiber Fabry-Perot sensor demodulation method as described above.

[0016] This invention provides a multi-feature fusion demodulation method, system, and device for sapphire fiber Fabry-Perot sensors, effectively overcoming the shortcomings of existing deep learning demodulation methods, which mostly employ end-to-end black-box models, require massive amounts of training data, are prone to failure under complex operating conditions (such as broadband light source envelope drift and composite cavity crosstalk), and have poor generalization ability. This invention is not a simple application of existing Gram angle field and nonlinear mapping models, but rather a structural innovation targeting the specific physical mechanism of sapphire MEMS Fabry-Perot multi-cavity interferometry. On the one hand, by constructing a dual-branch architecture by fusing the physical interference morphology features and frequency domain features of the interference spectrum, the limitations of single frequency domain analysis are broken. In the presence of multi-cavity interference and composite reflection structures, the frequency domain features can provide frequency differentiation information corresponding to different cavities. Combined with spectral morphology features, selective extraction of interference information of the target cavity is achieved, effectively suppressing interference from non-target cavities, and making the demodulation results less sensitive to light source fluctuations, spectral envelope changes, and noise disturbances. On the other hand, in the nonlinear mapping stage, this invention creatively uses the "precise interference order" derived from traditional physics as a priori constraint and splices it into the hidden feature layer, forcing the network to focus on the real physical phase changes, effectively solving the technical bias of demodulation drift in pure data-driven models under extreme high temperature and high pressure environments.

[0017] Additional aspects and advantages of the invention will be set forth in part in the description which follows, and in part will be obvious from the description, or may be learned by practice of the invention. Attached Figure Description

[0018] To more clearly illustrate the technical solutions in this invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of this invention. For those skilled in the art, other drawings can be obtained from these drawings without creative effort.

[0019] Figure 1 This is a schematic flowchart of a multi-feature fusion demodulation method for a sapphire fiber Fabry-Perot sensor provided by the present invention.

[0020] Figure 2 This invention provides a specific interference cavity spectrum for a sapphire fiber MEMS Fabry-Perot sensor based on multi-feature fusion extraction.

[0021] Figure 3 This invention provides a sample-by-sample prediction accuracy of the optical path difference based on a multi-feature fusion intelligent demodulation method.

[0022] Figure 4 This invention provides an optical path difference residual distribution based on a multi-feature fusion intelligent demodulation method.

[0023] Figure 4 (a) is the optical path difference residual distribution of a training set based on a multi-feature fusion intelligent demodulation method provided in an embodiment of the present invention.

[0024] Figure 4 (b) is the optical path difference residual distribution of a verification set based on a multi-feature fusion intelligent demodulation method provided in an embodiment of the present invention.

[0025] Figure 4 (c) is the optical path difference residual distribution of a test set based on a multi-feature fusion intelligent demodulation method provided in an embodiment of the present invention.

[0026] Figure 5 This is a comparison chart of the demodulation results of the all-sapphire fiber MEMS Fabry-Perot pressure sensor using the traditional method and the multi-feature fusion method in an embodiment of the present invention.

[0027] Figure 6 This is a schematic diagram of the structure of a multi-feature fusion sapphire fiber Fabry-Perot sensor demodulation system provided by the present invention.

[0028] Figure 7 This is a schematic diagram of the structure of the electronic device provided by the present invention.

[0029] Figure label: 101. Two-dimensional feature acquisition module; 102. Frequency domain feature acquisition module; 103. Mapping model training module; 104. Environmental pressure value calculation module; 810. Processor; 820. Communication interface; 830. Memory; 840. Communication bus. Detailed Implementation

[0030] To make the objectives, technical solutions, and advantages of this invention clearer, the technical solutions of this invention will be clearly and completely described below. Obviously, the described embodiments are only some, not all, of the embodiments of this invention. Based on the embodiments of this invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this invention. The following embodiments are used to illustrate this invention but cannot be used to limit the scope of this invention.

[0031] In the description of this specification, the references to terms such as "one embodiment," "some embodiments," "example," "specific example," or "some examples," etc., refer to specific features, structures, materials, or characteristics described in connection with that embodiment or example, which are included in at least one embodiment or example of the present invention. In this specification, the illustrative expressions of the above terms do not necessarily refer to the same embodiment or example. Furthermore, the specific features, structures, materials, or characteristics described may be combined in any suitable manner in one or more embodiments or examples. Moreover, without contradiction, those skilled in the art can combine and integrate the different embodiments or examples described in this specification, as well as the features of different embodiments or examples.

[0032] The following is combined Figures 1 to 7 This invention is described.

[0033] Example like Figure 1 As shown, Figure 1 This is a flowchart illustrating a multi-feature fusion demodulation method for a sapphire fiber Fabry-Perot sensor provided by the present invention, including the following steps: S1: Acquire the interference spectrum signal reflected by the sapphire fiber optic Fabry-Perot sensor; S2: Reconstruct the interference spectral signal to build a two-dimensional feature representation that characterizes the interference properties; S3: Perform Fourier spectrum analysis on the interference spectrum signal, convert the interference spectrum signal from the spectral domain to the frequency domain, and apply frequency domain physical suppression constraints to the obtained spectrum signal to obtain frequency domain characteristics that reflect the characteristics of the interference cavity; S4: The two-dimensional feature representation is fused with the frequency domain features and input as a joint feature into the feature mapping model. The interference fringes of a specific interference cavity are extracted according to the feature mapping model. S5: Obtain the peak wavelength of the interference fringes, and calculate the rough interference order of the interference fringes of the specific interference cavity based on the spectrum signal; S6: Calculate the precise interference order based on the coarse interference order, and input the precise interference order and interference fringes into the nonlinear mapping model to predict the precise optical path difference; S7: Based on the precise optical path difference, establish a mapping relationship between the environmental pressure value and the precise optical path difference, and calculate the environmental pressure value according to the mapping relationship.

[0034] Specifically, step S2 includes: S21: Randomly mask the interference spectral signal to obtain the masked spectral signal; S22: Perform reconstruction processing based on the masked spectral signal to obtain the reconstructed spectral signal; S23: Normalize the reconstructed spectral signal, convert the normalized reconstructed spectral signal from Cartesian coordinates to polar coordinates, and construct a two-dimensional feature representation based on the polar coordinates.

[0035] Specifically, step S23 includes: S231: Normalize the reconstructed spectral signal: in, The reconstructed spectral signal after normalization. To reconstruct the spectral signal, It is a function with maximum value. It is a minimum value function; S232: Map the normalized reconstructed spectral signal from the Cartesian coordinate system to the polar coordinate system, and calculate the polar angle corresponding to each sampling point. in, This represents the first light intensity value of the reconstructed spectral signal after normalization. The polar angle corresponding to the first light intensity value; S233: Based on the polar coordinate representation, calculate the trigonometric function relationship between polar angles at different wavelengths, construct the Gram angle field, and obtain a two-dimensional feature representation. Matrix elements in two-dimensional feature representation form Defined as: in, This is the polar angle corresponding to the second light intensity value.

[0036] Specifically, step S3 includes the following steps: S31: Based on the interference spectrum signal, the uniform wavelength interval is converted into a uniform wavenumber interval through reciprocal transformation and interpolation to obtain the wavenumber representation; S32: Perform a fast Fourier transform on the interference spectrum signal represented by the wavenumber to obtain the Fourier spectrum signal; S33: Divide the Fourier spectrum signal into several frequency domain sub-intervals according to the block division rules, and each frequency domain sub-interval corresponds to a set of continuous frequency domain sampling points; S34: Based on the energy distribution characteristics of the frequency domain sub-intervals, determine the target sub-interval set, retain the frequency domain information corresponding to the target sub-intervals, and obtain the frequency domain characteristics reflecting the specific interference cavity characteristics.

[0037] Specifically, step S34 includes: S341: Based on the energy distribution characteristics of each frequency domain sub-interval in the aforementioned spectrum signal, determine the set of target sub-intervals: in, It is the spectral signal after Fast Fourier Transform. For signal ordinal number, For the first Sub-intervals, This represents the local energy value of a sub-interval in the frequency domain; Set dynamic energy threshold Filter to meet The subintervals are used as the target subinterval set. ; S342: Apply frequency domain physical suppression constraints to non-target sub-intervals and construct a physical bandpass mask function. : S343: Multiply the spectral signal by the bandpass mask function to obtain the frequency domain characteristics reflecting the specific interferometric cavity features. : .

[0038] The feature mapping model will use a two-dimensional feature matrix. As the first input to the feature mapping model, it is fed into the encoding branch based on the Visual Transformer (ViT). After processing through image patch embedding and wavelength position encoding, a two-dimensional feature vector representing the spatial morphology is extracted. In parallel, the frequency domain feature sequences of the suppression process are... As the second input, it is fed into the encoding branch of the multilayer perceptron (MLP) to extract the frequency domain feature vector representing the frequency domain energy and position information of a specific interferometric cavity. .

[0039] Specifically, the two-dimensional feature representation and the frequency domain features are input in parallel to the dual-branch feature mapping model based on the same interference spectrum signal, and feature extraction is performed separately. The feature extraction results of each branch are then fused to output the interference fringes corresponding to a specific interference cavity. The calculation expression for feature extraction is as follows: in, The fused joint feature vector This indicates a feature channel splicing operation. The first feature transformation weight matrix, A two-dimensional feature vector representing spatial morphology. The second feature transformation weight matrix, To characterize the frequency domain eigenvectors representing the frequency domain energy and position information of a specific interferometric cavity, This is the network bias term.

[0040] Finally, the fused joint feature vector is fed into the decoding layer of the feature mapping model for feature reconstruction, and the output is as follows: Figure 2 The interference fringes shown correspond to a specific interference cavity, thereby achieving stable demodulation of a specific target cavity in a complex crosstalk environment.

[0041] Specifically, step S5 includes: S51: Determine the coarse peak wavelength using the local maximum algorithm. And select near the approximate peak wavelength A peak-finding window is constructed using consecutive sampling points; the centroid peak-finding algorithm is used to calculate the accurate peak wavelength. The calculation formula is: in, This is the width parameter of one side of the peak-finding window; The index of the sampling points within the peak-finding window, and ; For the first The wavelength values ​​corresponding to each sampling point; For the first The light intensity value corresponding to each sampling point.

[0042] S52: Coarse optical path difference The calculation formula is: Where N is the total number of data points after cubic spline interpolation. Where k is the wavenumber sampling interval, and k is the peak position index in the spectrum of the spectral signal. S53: The formula for calculating the rough interference order is: in It is a rough interference level.

[0043] Specifically, step S6 includes: S61: The precise interference order The result is a rounding of the coarse interference order, and the extracted interference fringes of the specific interference cavity are constructed into a one-dimensional sequence. The fringe hidden feature vector is extracted by inputting it into the first hidden layer of the nonlinear mapping model. The calculation formula is: in, This is the weight matrix of the first hidden layer. It is a non-linear activation function. This is the bias vector for the first hidden layer; S62: The stripe hidden feature vector With the precise interference order Concatenate the features along the feature dimension to construct a joint feature vector. The calculation formula is: S63: Transfer the joint feature vector After subsequent nonlinear mapping between the hidden layer and the output layer, the accurate optical path difference is calculated. The calculation formula is: in, This is the weight matrix of the second hidden layer. This is the output layer weight matrix. This is the bias vector for the second hidden layer. This is the output layer bias term.

[0044] The neural network model that appears in step S6 is a nonlinear mapping model.

[0045] Step S7 involves establishing a linear function that maps environmental pressure values ​​to precise optical path differences, and then calculating the environmental pressure value based on this linear function.

[0046] Figure 3 The horizontal axis represents the random sampling points of the sensor pressure experiment data, and the vertical axis represents the optical path difference (OPD) in μm. It can be seen that the prediction accuracy of the optical path difference predicted by the multi-feature model is the same as that of the optical path difference measured in the pressure experiment.

[0047] like Figure 4 As shown, Figure 4This invention relates to a sample-by-sample error distribution of optical path difference demodulation based on a multi-feature fusion-based intelligent demodulation method. Figure 4 The horizontal axis represents the random sampling points of the sensor pressure experiment data, and the vertical axis represents the error value of the optical path difference (experimental measured value - predicted value), in μm. Subplots (a), (b), and (c) represent the training set, validation set, and test set, respectively.

[0048] Specifically, in this embodiment, the sample-by-sample prediction accuracy between the demodulated output and the true calibration value is such that, within the range of the test samples, the error between the accurate optical path difference prediction value and the true value of the present invention remains within a small range, and the error does not show obvious systematic drift with sample changes. This indicates that the demodulation method of the present invention, based on multi-feature parallel extraction, frequency domain physical suppression constraint and feature fusion, can achieve relatively stable optical path difference demodulation under composite cavity interference conditions. In the actual demodulation process, there is no need to rely on a complex optical system structure. This demodulation method can achieve an optical path difference demodulation accuracy of 0.014 μm, which greatly reduces the cost and complexity of the system, and has stable and superior demodulation performance.

[0049] like Figure 5 As shown, the demodulation results of the multi-feature fusion method of this invention are compared with those of the traditional method using only sensor spectrum. Pressure measurements from 0.5 kPa to 3500 kPa at 100℃ were performed using an all-sapphire fiber MEMS Fabry-Perot sensor, and the sensor spectrum was collected for demodulation. The results show that in low signal-to-noise ratio sensor measurements, due to issues such as blurred fringe contrast and multi-cavity crosstalk, the demodulation results of the traditional method exhibit continuous jumps. The multi-feature fusion method provided by this invention provides demodulation results with good linearity. Experiments have demonstrated that this method significantly reduces the impact of multi-cavity crosstalk of the all-sapphire fiber MEMS Fabry-Perot sensor on demodulation accuracy and avoids phase jump problems caused by high-frequency noise from non-target cavities and low-frequency background light noise.

[0050] Specifically, such as Figure 6 As shown, the present invention also provides a multi-feature fusion sapphire fiber Fabry-Perot sensor demodulation system, comprising the following modules: Two-dimensional feature acquisition module 101: used to acquire the interference spectrum signal reflected by the sapphire fiber Fabry-Perot sensor, reconstruct the interference spectrum signal, and construct a two-dimensional feature representation that characterizes the interference properties; Frequency domain feature acquisition module 102: used to perform Fourier spectrum analysis on the interference spectrum signal, convert the interference spectrum signal from the spectral domain to the frequency domain, perform frequency domain physical suppression constraint on the obtained spectrum signal, and obtain frequency domain features reflecting the characteristics of the interference cavity; Mapping model training module 103: used to fuse the two-dimensional feature representation with the frequency domain features, and input the joint features into the feature mapping model, and extract the interference fringes of a specific interference cavity according to the feature mapping model; Environmental pressure value calculation module 104: used to obtain the peak wavelength of the interference fringes, calculate the coarse interference order of the interference fringes of a specific interference cavity based on the spectrum signal; calculate the precise interference order based on the coarse interference order, input the precise interference order and interference fringes into a nonlinear mapping model to predict the precise optical path difference; establish a mapping relationship between the environmental pressure value and the precise optical path difference based on the precise optical path difference, and calculate the environmental pressure value based on the mapping relationship.

[0051] Figure 7 An example is a schematic diagram of the physical structure of an electronic device, such as... Figure 7 As shown, the electronic device may include: a processor 810, a communication interface 820, a memory 830, and a communication bus 840, wherein the processor 810, the communication interface 820, and the memory 830 communicate with each other via the communication bus 840. The processor 810 can call logical instructions in the memory 830 to execute a multi-feature fusion sapphire fiber Fabry-Perot sensor demodulation method, which includes: S1: Acquire the interference spectrum signal reflected by the sapphire fiber optic Fabry-Perot sensor; S2: Reconstruct the interference spectral signal to build a two-dimensional feature representation that characterizes the interference properties; S3: Perform Fourier spectrum analysis on the interference spectrum signal, convert the interference spectrum signal from the spectral domain to the frequency domain, and apply frequency domain physical suppression constraints to the obtained spectrum signal to obtain frequency domain characteristics that reflect the characteristics of the interference cavity; S4: The two-dimensional feature representation is fused with the frequency domain features and input as a joint feature into the feature mapping model. The interference fringes of a specific interference cavity are extracted according to the feature mapping model. S5: Obtain the peak wavelength of the interference fringes, and calculate the rough interference order of the interference fringes of the specific interference cavity based on the spectrum signal; S6: Calculate the precise interference order based on the coarse interference order, and input the precise interference order and interference fringes into the nonlinear mapping model to predict the precise optical path difference; S7: Based on the precise optical path difference, establish a mapping relationship between the environmental pressure value and the precise optical path difference, and calculate the environmental pressure value according to the mapping relationship.

[0052] Furthermore, the logical instructions in the aforementioned memory 830 can be implemented as software functional units and, when sold or used as independent products, can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, essentially, or the part that contributes to the prior art, or a part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of the present invention. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0053] The device embodiments described above are merely illustrative. The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the modules can be selected to achieve the purpose of this embodiment according to actual needs. Those skilled in the art can understand and implement this without any creative effort.

[0054] Through the above description of the embodiments, those skilled in the art can clearly understand that each embodiment can be implemented by means of software plus necessary general-purpose hardware platforms, and of course, it can also be implemented by hardware. Based on this understanding, the above technical solutions, in essence or the part that contributes to the prior art, can be embodied in the form of a software product. This computer software product can be stored in a computer-readable storage medium, such as ROM / RAM, magnetic disk, optical disk, etc., and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute the methods described in the various embodiments or some parts of the embodiments.

[0055] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, and not to limit them; although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features; and these modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.

[0056] It should be noted that the embodiments of this disclosure can be implemented using hardware, software, or a combination of both. The hardware portion can be implemented using dedicated logic; the software portion can be stored in memory and executed by a suitable instruction execution system, such as a microprocessor or dedicated-design hardware. Those skilled in the art will understand that the above-described devices and methods can be implemented using computer-executable instructions and / or included in processor control code, for example, such code provided on a programmable memory or a data carrier such as an optical or electronic signal carrier.

[0057] Furthermore, although the operation of the methods of this disclosure is described in a specific order in the accompanying drawings, this does not require or imply that these operations must be performed in that specific order, or that all the operations shown must be performed to achieve the desired result. Rather, the steps depicted in the flowcharts may be performed in a different order. Additionally or alternatively, certain steps may be omitted, multiple steps may be combined into one step, and / or one step may be broken down into multiple steps. It should also be noted that the features and functions of two or more devices according to this disclosure may be embodied in one device. Conversely, the features and functions of one device described above may be further divided and embodied by multiple devices.

[0058] While this disclosure has been described with reference to several specific embodiments, it should be understood that this disclosure is not limited to the specific embodiments disclosed. This disclosure is intended to cover various modifications and equivalent arrangements included within the spirit and scope of the appended claims.

Claims

1. A multi-feature fusion sapphire fiber Fabry-Perot sensor demodulation method, characterized in that, include: S1: Acquire the interference spectrum signal reflected by the sapphire fiber optic Fabry-Perot sensor; S2: Reconstruct the interference spectral signal to build a two-dimensional feature representation that characterizes the interference properties; S3: Perform Fourier spectrum analysis on the interference spectrum signal, convert the interference spectrum signal from the spectral domain to the frequency domain, and apply frequency domain physical suppression constraints to the obtained spectrum signal to obtain frequency domain characteristics that reflect the characteristics of the interference cavity; S4: The two-dimensional feature representation is fused with the frequency domain features and input as a joint feature into the feature mapping model. The interference fringes of a specific interference cavity are extracted according to the feature mapping model. S5: Obtain the peak wavelength of the interference fringes, and calculate the rough interference order of the interference fringes of the specific interference cavity based on the spectrum signal; S6: Calculate the precise interference order based on the coarse interference order, and input the precise interference order and interference fringes into the nonlinear mapping model to predict the precise optical path difference; S7: Based on the precise optical path difference, establish a mapping relationship between the environmental pressure value and the precise optical path difference, and calculate the environmental pressure value according to the mapping relationship.

2. The demodulation method for a multi-feature fusion sapphire fiber Fabry-Perot sensor according to claim 1, characterized in that, Step S2 includes: S21: Randomly mask the interference spectral signal to obtain the masked spectral signal; S22: Perform reconstruction processing based on the masked spectral signal to obtain the reconstructed spectral signal; S23: Normalize the reconstructed spectral signal, convert the normalized reconstructed spectral signal from Cartesian coordinates to polar coordinates, and construct a two-dimensional feature representation based on the polar coordinates.

3. The demodulation method for a multi-feature fusion sapphire fiber Fabry-Perot sensor according to claim 2, characterized in that, Step S23 includes: S231: Normalize the reconstructed spectral signal: in, The reconstructed spectral signal after normalization. To reconstruct the spectral signal, It is a function with maximum value. It is a minimum value function; S232: Map the normalized reconstructed spectral signal from the Cartesian coordinate system to the polar coordinate system, and calculate the polar angle corresponding to each sampling point. in, This represents the first light intensity value of the reconstructed spectral signal after normalization. The polar angle corresponding to the first light intensity value; S233: Based on the polar coordinate representation, calculate the trigonometric function relationship between polar angles at different wavelengths, construct the Gram angle field, and obtain a two-dimensional feature representation. Matrix elements in two-dimensional feature representation form Defined as: in, This is the polar angle corresponding to the second light intensity value.

4. The demodulation method for a multi-feature fusion sapphire fiber Fabry-Perot sensor according to claim 1, characterized in that, Step S3 includes the following steps: S31: Based on the interference spectrum signal, the uniform wavelength interval is converted into a uniform wavenumber interval through reciprocal transformation and interpolation to obtain the wavenumber representation; S32: Perform a fast Fourier transform on the interference spectrum signal represented by the wavenumber to obtain the Fourier spectrum signal; S33: Divide the Fourier spectrum signal into several frequency domain sub-intervals according to the block division rules, and each frequency domain sub-interval corresponds to a set of continuous frequency domain sampling points; S34: Based on the energy distribution characteristics of the frequency domain sub-intervals, determine the target sub-interval set, retain the frequency domain information corresponding to the target sub-intervals, and obtain the frequency domain characteristics reflecting the specific interference cavity characteristics.

5. The demodulation method for a multi-feature fusion sapphire fiber Fabry-Perot sensor according to claim 4, characterized in that, Step S34 includes: S341: Based on the energy distribution characteristics of each frequency domain sub-interval in the aforementioned spectrum signal, determine the set of target sub-intervals: in, It is the spectral signal after Fast Fourier Transform. For signal ordinal number, For the first Sub-intervals, This represents the local energy value of a sub-interval in the frequency domain; Set dynamic energy threshold Filter to meet The subintervals are used as the target subinterval set. ; S342: Apply frequency domain physical suppression constraints to non-target sub-intervals and construct a physical bandpass mask function. : S343: Multiply the spectral signal by the bandpass mask function to obtain the frequency domain characteristics reflecting the specific interferometric cavity features. : 。 6. The demodulation method for a multi-feature fusion sapphire fiber Fabry-Perot sensor according to claim 1, characterized in that, Step S4 includes: the two-dimensional feature representation and the frequency domain features are input in parallel to the feature mapping model based on the same interference spectral signal, resulting in the feature extraction calculation expression: in, The fused joint feature vector This indicates a feature channel splicing operation. The first feature transformation weight matrix, The second feature transformation weight matrix, A vector in two-dimensional feature representation form. To characterize the frequency domain eigenvectors representing the frequency domain energy and position information of a specific interferometric cavity, As a network bias term, the fused joint feature vector is fed into the decoding layer for feature reconstruction, and interference fringes of a specific interference cavity are output.

7. The demodulation method for a multi-feature fusion sapphire fiber Fabry-Perot sensor according to claim 6, characterized in that, Step S5 includes: S51: Determine the coarse peak wavelength using the local maximum algorithm. And select near the approximate peak wavelength A peak-finding window is constructed using consecutive sampling points; the centroid peak-finding algorithm is used to calculate the accurate peak wavelength. The calculation formula is: in, This is the width parameter of one side of the peak-finding window; The index of the sampling points within the peak-finding window, and ; For the first The wavelength values ​​corresponding to each sampling point; For the first The light intensity value corresponding to each sampling point; S52: Coarse optical path difference The calculation formula is: Where N is the total number of data points after cubic spline interpolation. It is the wavenumber sampling interval, and k is the peak position index in the spectrum of the spectral signal; S53: The formula for calculating the rough interference order is: in It is a rough interference level.

8. The demodulation method for a multi-feature fusion sapphire fiber Fabry-Perot sensor according to claim 7, characterized in that, Step S6 includes: S61: The precise interference order The result is a rounding of the coarse interference order, and the extracted interference fringes of the specific interference cavity are constructed into a one-dimensional sequence. The fringe hidden feature vector is extracted by inputting it into the first hidden layer of the nonlinear mapping model. The calculation formula is: in, This is the weight matrix of the first hidden layer. It is a non-linear activation function. This is the bias vector for the first hidden layer; S62: The stripe hidden feature vector With the precise interference order Concatenate the features along the feature dimension to construct a joint feature vector. The calculation formula is: S63: Transfer the joint feature vector After subsequent nonlinear mapping between the hidden layer and the output layer, the accurate optical path difference is calculated. The calculation formula is: in, This is the weight matrix of the second hidden layer. This is the output layer weight matrix. This is the bias vector for the second hidden layer. This is the output layer bias term.

9. A multi-feature fusion sapphire fiber Fabry-Perot sensor demodulation system, for performing a multi-feature fusion sapphire fiber Fabry-Perot sensor demodulation method as described in any one of claims 1 to 8, comprising the following modules: Two-dimensional feature acquisition module: used to acquire the interference spectrum signal reflected by the sapphire fiber optic Fabry-Perot sensor, reconstruct the interference spectrum signal, and construct a two-dimensional feature representation that characterizes the interference properties; Frequency domain feature acquisition module: used to perform Fourier spectrum analysis on the interference spectrum signal, convert the interference spectrum signal from the spectral domain to the frequency domain, perform frequency domain physical suppression constraint on the obtained spectrum signal, and obtain frequency domain features reflecting the characteristics of the interference cavity; The mapping model training module is used to fuse the two-dimensional feature representation with the frequency domain features and input them as joint features into the feature mapping model, and extract the interference fringes of a specific interference cavity based on the feature mapping model. Environmental pressure value calculation module: used to obtain the peak wavelength of the interference fringes and calculate the coarse interference order of the interference fringes of a specific interference cavity based on the spectrum signal; The precise interference order is calculated from the coarse interference order, and the precise interference order and interference fringes are input into the nonlinear mapping model to predict the precise optical path difference. Based on the precise optical path difference, a mapping relationship between the environmental pressure value and the precise optical path difference is established, and the environmental pressure value is calculated according to the mapping relationship.

10. An electronic device comprising a processor, a communication interface, a memory, and a communication bus, characterized in that, When the processor executes a computer program, it implements the steps of the multi-feature fusion sapphire fiber Fabry-Perot sensor demodulation method as described in any one of claims 1 to 8.

Images