An optical detection system and method

Abstract

This invention discloses an optical detection system and method, belonging to the field of optical non-contact measurement technology. The invention employs a low-coherence broadband light source, combining dispersive elements and an array detector to spatially separate its broadband spectrum to construct multiple parallel independent wavelength measurement channels. Simultaneously, it integrates an envelope extraction algorithm based on empirical mode decomposition, Hilbert transform, and window-length adaptive Savitzky-Golay filtering to perform real-time quality assessment and dynamic weighted fusion of the signals from each channel. This significantly improves the continuity and stability of vibration measurements under harsh speckle environments, achieving high-dimensional wavelength diversity at low cost. Through real-time quality assessment and weighted fusion using the adaptive envelope extraction algorithm, the final output vibration measurement values ​​show significantly improved signal-to-noise ratio and measurement accuracy.

Description

Technical Field

[0001] This invention relates to the field of optical non-contact measurement technology, and in particular to an optical detection system and method. Background Technology

[0002] Laser Doppler vibration measurement is a high-precision, non-contact optical measurement technique based on the optical Doppler effect and laser interference principle. It can accurately measure the vibration velocity or displacement of an object's surface. Due to its outstanding advantages such as high measurement accuracy, wide frequency response, and long measurement distance, this technology is widely used in various industrial and scientific research fields, including mechanical fault diagnosis, microelectromechanical system (MEMS) characteristic analysis, acoustic sensing, materials property research, and precision manufacturing.

[0003] Despite the significant advantages of laser Doppler vibration measurement technology, laser speckle becomes a major factor limiting its reliability and versatility in practical industrial applications, especially when measuring rough surfaces. In laser Doppler vibration meters, the random fluctuations of this speckle field directly lead to drastic fluctuations in the signal-to-noise ratio of the interference signal or signal loss, resulting in a large amount of burr noise in the demodulated vibration signal. In severe cases, this can make measurement impossible, limiting the technology's ability to perform stable and reliable testing of ordinary rough workpieces in industrial settings.

[0004] To suppress speckle effects and improve the stability of measurement signals, various diversity techniques have been proposed by those skilled in the art. Among them, "wavelength diversity" utilizes a dual-frequency laser or two independent single-frequency lasers combined to separate the Doppler signals corresponding to the two wavelengths by introducing a small frequency shift or utilizing the beat frequency characteristics of the two wavelengths, thus obtaining two independent vibration measurement channels and achieving wavelength diversity. However, the cost of both dedicated dual-frequency lasers and two independent single-frequency lasers is significantly higher than that of ordinary single-frequency lasers. Traditional single-frequency laser Doppler vibrometers, due to their extremely high spatiotemporal coherence, produce a strong speckle effect when measuring rough surfaces. The random fluctuations in the speckle field cause deep fading or even complete disappearance of the interference signal, resulting in measurement signal interruption and a sharp deterioration in the signal-to-noise ratio, severely limiting the reliability and universality of this technology in industrial rough surface detection.

[0005] Existing diversity technologies suffer from limited dimensionality and high costs. To suppress speckle, current wavelength diversity schemes typically employ dual-frequency lasers or dual-laser beam combining. These schemes provide only two extremely limited diversity channels, and when faced with complex surface motion, the probability of both channels fading simultaneously remains high, resulting in limited reliability improvements. Furthermore, these schemes rely on expensive light sources and complex wavelength division multiplexing optical paths, leading to high system costs and complex structures. Attempting to expand diversity dimensionality by increasing the number of lasers would increase costs, complexity, and the difficulty of optical path calibration, making them impractical. Therefore, existing wavelength diversity technologies are limited by their core contradictions of "low dimensionality, high cost, and difficulty in expansion," failing to achieve highly reliable diversity measurements that completely overcome speckle fading at an acceptable cost and complexity. This has become a key bottleneck hindering the widespread adoption and application of high-performance laser vibratory measurement technology in industry. To address this long-standing technical challenge, a novel solution capable of achieving high-dimensional, low-cost wavelength diversity is urgently needed. Summary of the Invention

[0006] To address the aforementioned problems, this invention provides an optical detection system. This system employs a low-coherence broadband light source, combining a dispersive element and an array detector to spatially separate its broadband spectrum, constructing multiple parallel independent wavelength measurement channels. Simultaneously, it integrates an envelope extraction algorithm based on empirical mode decomposition, Hilbert transform, and window-length adaptive Savitzky-Golay filtering. This algorithm performs real-time quality assessment and dynamic weighted fusion of the signals from each channel. This solves the technical problems of traditional laser vibrometers in rough surface measurements, such as random signal fading and unreliable measurements due to speckle effects, and the limitations of existing wavelength diversity techniques in achieving high-stability diversity measurements due to the limited number of channels and high cost. The system significantly improves the continuity and stability of vibration measurements under harsh speckle environments, achieving high-dimensional wavelength diversity at low cost. Real-time quality assessment and weighted fusion using an adaptive envelope extraction algorithm ultimately output vibration measurements with significantly improved signal-to-noise ratio and measurement accuracy. S ( t ).

[0007] To achieve the above objectives, the technical solution adopted by the present invention is: an optical detection method, comprising the following steps: S1: outputting a low-coherence broadband light beam through a broadband light source;

[0008] S2: Split the broadband beam into a reference beam and a measurement beam, and modulate the frequency of the reference beam;

[0009] S3: Guide the measurement light to the target surface to be measured, and receive the scattered light returned from the surface.

[0010] The returned scattered light interferes with the modulated reference light to produce interference light;

[0011] S4: The components of different wavelengths in the interference light are spatially separated by a dispersive element to form multiple parallel independent wavelength measurement channels, and the interference signals output from each measurement channel are synchronously acquired by an array detector. ;

[0012] S5: Parallel demodulation of the interference signals from each wavelength measurement channel. Preliminary vibration data for each channel were obtained. The interference signal :

[0013]

[0014] in, t For time; The signal amplitude envelope; For the first i The wavelengths corresponding to each channel;

[0015] S6: Calculate the final envelope of the interference signal in each channel using an adaptive envelope extraction algorithm. ;

[0016] S7: Normalize the final envelope to obtain the normalized weights of each channel. Combined with the aforementioned preliminary vibration data Perform weighted fusion to output the final vibration measurement value. :

[0017] .

[0018] Preferably, the bandwidth Δλ of the broadband beam and the roughness of the target surface being measured are... Ra Satisfies: Δλ = λ0 / 4 Ra ).

[0019] Preferably, the broadband light source is a superluminescent diode.

[0020] Preferably, the dispersive element is a diffraction grating.

[0021] The adaptive envelope extraction algorithm includes the following steps: Step 1: Synchronously acquire the interference signals of each channel in parallel. ;

[0022] Step 2: For the interference signal of each channel Empirical mode decomposition is performed to obtain multiple intrinsic mode functions. and residual components The signal is reconstructed using a preset number of preceding intrinsic mode functions and the residual components. ;

[0023] Step 3: Reconstruct the signal Perform Hilbert transform to construct analytic signals :

[0024]

[0025] in, j The imaginary unit; H Perform the Hilbert transform and calculate the analytic signal. The modulus value is used to obtain the initial envelope. ;

[0026] Step 4: Apply an adaptive Savitzky-Golay filter to the initial envelope. Smoothing optimization is performed; wherein, the window length of the adaptive Savitzky-Golay filter is... According to the initial envelope The rate of change is dynamically adjusted, and the adjustment method satisfies:

[0027]

[0028] in, Based on the window length, α These are adaptive coefficients;

[0029] Step 5: Calculate the final envelope :

[0030]

[0031] in, The window length for the adaptive Savitzky-Golay filter; This is the initial envelope; j The imaginary unit; t For time.

[0032] Preferably, the preset number of intrinsic mode functions and residual components is 3, and the signal is reconstructed using the first 3 intrinsic mode functions and the residual components. :

[0033]

[0034] in, For the first j One intrinsic mode function; m The number of intrinsic mode functions; This is the residual component.

[0035] Preferably, the basic window length The value range of the adaptive coefficient is 20 to 50.α The value range is from 5 to 20.

[0036] An optical detection system includes: a broadband light source for generating a low-coherence broadband beam; a beam splitting and modulation unit for splitting the broadband beam into a reference beam and a measurement beam, and frequency modulating the reference beam; an interference beam combining unit for guiding the measurement beam to a target surface and receiving scattered light returning from the surface, causing the returned scattered light to interfere with the modulated reference beam to generate interference light; and a dispersion detection unit including a dispersion element and an array detector, wherein the dispersion element is used to spatially expand the interference light according to wavelength components to form multiple parallel independent wavelength measurement channels; and the array detector is used to synchronously acquire interference signals of the spatially separated wavelength components. The signal processing module, connected to the array detector, is used to execute the above method and output the final vibration measurement value. .

[0037] Preferably, the beam splitting modulation unit includes an optical fiber coupler and an electro-optic modulator, and the interference beam combining unit includes a beam splitter.

[0038] Preferably, the signal processing module includes a programmable gate array or a digital signal processor.

[0039] By adopting the above technical solution, the present invention has the following beneficial effects.

[0040] (1) This invention uses a low-coherence broadband light source combined with dispersive elements and array detectors to construct a large number of parallel independent wavelength measurement channels. It also integrates an envelope extraction algorithm based on empirical mode decomposition and adaptive Savitzky-Golay filtering to perform real-time quality assessment and weighted fusion of the signals of each channel. This solves the technical problems of existing laser Doppler vibration measurement technology when measuring rough surfaces, such as random fading of interference signals due to speckle effect, measurement instability, poor reliability, and the limited number of channels, high cost, and difficulty in expansion of existing wavelength diversity schemes. This invention achieves the technical effects of significantly improving measurement stability and reliability, greatly reducing system cost and complexity, effectively improving the signal-to-noise ratio and accuracy of output signals, and realizing high-dimensional diversity that is easy to expand.

[0041] (2) This invention replaces multiple highly coherent single-frequency lasers with a single low-coherence broadband light source (such as an SLD), and uses dispersive elements (such as gratings) to spatially separate the continuous broadband spectrum, which is then synchronously acquired by an array detector, physically forming tens to hundreds of parallel wavelength channels. This constructs a parallel channel between the broadband light source and the dispersive-array detection, solving the technical problems of traditional wavelength diversity schemes (such as dual-frequency lasers) having very few channels (usually 2), high cost, and limited resistance to speckle fading. It achieves high-dimensional (far exceeding 2) wavelength diversity at low cost, fundamentally and significantly reducing the probability of all measurement channels simultaneously fading due to speckle. Furthermore, this invention establishes a matching relationship between broadband bandwidth and surface roughness: Δλ = λ0 / 4 ( Ra This ensures efficient utilization of broadband light source resources and achieves optimal speckle suppression and signal coverage under given roughness.

[0042] (3) This invention uses Empirical Mode Decomposition (EMD) to denoise and reconstruct the signal (e.g., the first 3 IMFs + residual components), and then uses Hilbert transform to construct an analytical signal to obtain the initial envelope. This scheme can more accurately extract the time-varying envelope reflecting the channel quality from the interference signal contaminated by speckle noise, avoiding the drastic random fluctuations in signal amplitude caused by speckle, thereby accurately and in real time evaluating the channel quality of each parallel channel. In addition, the Savitzky-Golay filter with adaptive window length can dynamically adjust according to the rate of change of the initial envelope. A longer window is used where the envelope changes gently to enhance the smoothing effect; the window is automatically shortened where the envelope changes rapidly to preserve details and reduce lag. This achieves adaptive optimization and smoothing of the envelope signal, achieving the best balance between suppressing noise and quickly tracking the real envelope change, thus obtaining a more accurate final envelope. Used for weight calculation to avoid over-smoothing or lag, which can lead to inaccurate weight calculation.

[0043] (4) The present invention demodulates the preliminary vibration data from each channel. The final envelope extracted and optimized by the above algorithm. Normalization is performed on the basis to obtain the normalized weights. Combined with the aforementioned preliminary vibration data Perform weighted fusion to output the final vibration measurement value. This scheme dynamically suppresses the contribution of fading channels, enhances the weight of high-quality channels, and achieves adaptive optimal synthesis of multi-channel signals, ultimately outputting a superior signal. The continuity, smoothness and signal-to-noise ratio are significantly improved.

[0044] (5) This invention employs a low-coherence broadband light source and combines a dispersive element and an array detector to spatially separate its broadband spectrum to construct dozens to hundreds of parallel independent wavelength measurement channels. Simultaneously, it integrates an envelope extraction algorithm based on empirical mode decomposition, Hilbert transform, and adaptive Savitzky-Golay filtering with window length to perform real-time quality assessment and dynamic weighted fusion of the signals from each channel. This solves the technical problems of traditional laser vibration measurement technology in rough surface measurements, such as random signal fading due to speckle effect and unreliable measurements, and the inability of existing wavelength diversity technology to achieve high-stability diversity measurement due to the extremely small number of channels and high cost. It significantly improves the continuity and stability of vibration measurements under harsh speckle environments, achieving high-dimensional wavelength diversity at low cost. Real-time quality assessment and weighted fusion are performed through an adaptive envelope extraction algorithm, ultimately outputting vibration measurement values ​​with significantly improved signal-to-noise ratio and measurement accuracy. . Attached Figure Description

[0045] The following provides a detailed discussion of the manufacture and application of preferred embodiments of the present invention. However, it should be understood that the present invention provides many applicable inventive concepts that can be embodied in various specific environments. The specific embodiments discussed are merely illustrative of specific ways of manufacturing and using the present invention and do not limit the scope of the invention. For those skilled in the art, other drawings can be obtained from these drawings without any creative effort.

[0046] Figure 1 This is a schematic diagram of the structure of the present invention.

[0047] Figure 2 This is a comparison chart showing the calculation of the initial weights for the interference signals of each channel in this invention.

[0048] Figure 3 This is a comparison chart of the normalized weights of each channel in this invention.

[0049] Figure 4 This is a comparison diagram of the demodulated speed signals of each channel and the weighted synthesized speed signals of this invention.

[0050] Among them, 1-broadband light source; 2-fiber optic coupler; 3-electro-optic modulator; 4-collimating lens; 5-beam splitter one; 6-beam splitter two; 7-dispersive element; 8-array detector; 9-signal processing module; 10-surface under test; S1-broadband beam; S2-reference light; S3-measurement light; S4-interference light; S5-spectral line. Detailed Implementation

[0051] The following provides a detailed discussion of the manufacture and application of preferred embodiments of the present invention. However, it should be understood that the present invention provides many applicable inventive concepts that can be embodied in various specific environments. The specific embodiments discussed are merely illustrative of specific ways of manufacturing and using the invention and do not limit the scope of the invention.

[0052] This invention proposes a wavelength diversity optical detection system based on a broadband light source. The core of this system lies in replacing a highly coherent single-frequency laser with a low-coherence broadband light source 1, and spatially expanding its continuous broadband spectrum through a dispersive element 7, which is simultaneously received by an array detector 8, thus naturally forming multiple independent wavelength measurement channels. The system mainly includes: a broadband light source 1, a beam splitting and modulation unit, an interference beam combining unit, a dispersion detection unit, and a signal processing unit 9. The beam splitting and modulation unit includes an optical fiber coupler 2 and an electro-optic modulator 3; the interference beam combining unit includes a collimator 4, a beam splitter 5, and a beam splitter 6; the dispersion detection unit includes a dispersive element 7 and an array detector 8.

[0053] The bandwidth of the broadband light source 1 is determined by the roughness of the measured surface 10 and the center wavelength: Δλ = λ0 / 4 ( Ra A broadband beam S1 emitted by a broadband light source 1 is split into a reference beam S2 and a measurement beam S3 by a 50 / 50 fiber coupler 2. The reference beam S2 is frequency-modulated by an electro-optic modulator 3 to introduce a heterodyne carrier frequency. After collimation, the two beams form parallel beams. The measurement beam S3 is guided to the target object, and its returned scattered / reflected light merges with the reference beam S2 at the fractionator 6 to produce interference, forming an interference beam S4. This interference beam S4 then passes through a dispersive element 7, which uses the dispersion effect to spatially separate its different wavelength components, and is synchronously acquired by an array detector 8, thereby obtaining a series of independent channel signals corresponding to different wavelengths. These signals are transmitted to a signal processing module 9, which performs parallel demodulation on the signals of each channel to obtain preliminary vibration data. Subsequently, these results are fused into a final, highly reliable vibration measurement value using an adaptive envelope extraction algorithm and weighted fusion. The weight of each channel is determined based on the fading of the interference signal (signal envelope). The more severe the signal fading, the lower the weight.

[0054] In this process, although the echo signal of a single wavelength channel may attenuate due to speckle effect, resulting in its intensity being lower than the effective detection threshold, the system can significantly reduce the probability of the overall measurement signal being completely overwhelmed by noise through weighted synthesis of a large number of channels. This mechanism based on broadband light source 1 and parallel demodulation achieves high-dimensional wavelength diversity and effectively suppresses the adverse effects of speckle on measurement stability.

[0055] The following is combined Figure 1-4This invention provides a schematic diagram of an optical detection system to address the technical problem of random signal fading and measurement instability caused by speckle effect in laser Doppler vibration measurement of rough surfaces. The present invention is illustrated in the figure below. Figure 1 As shown. The system of the present invention includes a broadband light source 1 for generating a low-coherence broadband beam S1. In this embodiment, the broadband light source 1 is a superluminescent diode with a center wavelength λ0 of 1550 nm and a bandwidth Δλ of 50 nm. The beam splitting and modulation unit includes a 50 / 50 fiber coupler 2 and an electro-optic modulator 3 for splitting the broadband beam S1 into a reference beam S2 and a measurement beam S3, and frequency modulating the reference beam S2 to introduce a heterodyne carrier frequency.

[0056] The interference beam combining unit includes a collimating lens 4, a beam splitter 5, and a beam splitter 6. It guides the measurement light S3 to the target surface 10 and receives the scattered light returning from that surface. The returned scattered light interferes with the modulated reference light S2 at beam splitter 6, generating interference light S4. In this embodiment, the surface 10 being measured is roughness. Ra The workpiece is made of aluminum with a diameter of 10 μm. The dispersion detection unit includes a dispersion element 7 (in this embodiment, a diffraction grating with 600 lines per millimeter) and an array detector 8 (a 1024-pixel linear array InGaAs detector). The dispersion element 7 is used to spatially expand the interference light S4 according to its wavelength components, forming multiple parallel independent wavelength measurement channels. At this time, each pixel of the detector receives an extremely narrow band (approximately 0.048 nm) of interference light S4, which is equivalent to 1024 laser vibrometers of different wavelengths working in parallel simultaneously. The array detector 8 is used to synchronously acquire the interference signals of the spatially separated wavelength components. In this embodiment, the light intensity signals from the 1024 channels are synchronously and rapidly converted into electrical signals and transmitted to the signal processing module 9. The signal processing module 9 is connected to the array detector 8 and is used to perform the following optical detection method and output the final vibration measurement value. .

[0057] The optical detection method is implemented according to the following steps: First, a broadband beam S1 is output through a broadband light source 1, the bandwidth of which Δλ is calculated according to the formula Δλ=λ0 / 4. Ra Determine to match the surface roughness of the tested surface by 10. RaThen, the beam is split into a reference beam S2 and a signal beam S3 by fiber coupler 2. The reference beam S2 is modulated by electro-optic modulator 3 and collimated by collimating lens 4 before being output to beam splitter 6. The measurement beam S3 is incident on the measured surface 10 through collimating lens 4 and beam splitter 5. After being reflected by the measured surface 10, the returned scattered light is output to beam splitter 6 through beam splitter 5 and interferes with the modulated reference beam S2 at beam splitter 6 to form interference beam S4. The interference beam S4 is dispersed by dispersive element 7 to form spatially expanded spectral lines S5 on array detector 8, and the interference signals of each channel are acquired synchronously. Its expression is:

[0058]

[0059] in, t For time; The signal amplitude envelope; For the first i The wavelengths corresponding to each channel; This provides preliminary vibration data. At a certain moment, due to the speckle effect, the signal in some wavelength channels may be significantly attenuated, while other channels remain at a relatively strong level. To suppress this effect, the system of this invention calculates the weights based on the envelope amplitude of the interference signal in each channel.

[0060] Subsequently, signal processing module 9 demodulates the preliminary vibration data corresponding to each channel in parallel. And execute an adaptive envelope extraction algorithm to calculate the final envelope of each channel. The adaptive envelope extraction algorithm specifically includes the following steps:

[0061] Step 1: Synchronously acquire interference signals from each channel in parallel Its expression is as described above.

[0062] Step 2: For the interference signal of each channel Empirical mode decomposition is performed to obtain multiple intrinsic mode functions. and residual components The signal is reconstructed using a preset number of preceding intrinsic mode functions and the residual components. .

[0063]

[0064] in, It is the first j One intrinsic mode function, m The number of intrinsic mode functions. These are residual components. Keep the first three. IMF The reconstructed signal consists of component and residual component components. :

[0065] .

[0066] Step 3: Reconstruct the signal Perform Hilbert transform to construct analytic signals :

[0067]

[0068] in, j The imaginary unit; H Perform the Hilbert transform and calculate the analytic signal. The modulus value is used to obtain the initial envelope. The initial envelope for:

[0069] .

[0070] Step 4: Apply an adaptive Savitzky-Golay filter to the initial envelope. Smoothing optimization is performed; wherein, the window length of the adaptive Savitzky-Golay filter is... According to the initial envelope The rate of change is dynamically adjusted, and the adjustment method satisfies:

[0071]

[0072] in, The base window length in this embodiment is... =31, α The adaptive coefficients are those used in this embodiment. α =10.

[0073] Step 5: Calculate the final envelope :

[0074]

[0075] in, The window length for the adaptive Savitzky-Golay filter; This is the initial envelope; j The imaginary unit; t The time is denoted by ; the polynomial order is 3.

[0076] In regions where the signal amplitude is low due to speckle fading, the weight of the corresponding channel is reduced, thus decreasing its contribution to the synthesis result. Subsequently, all weights are normalized to ensure that the sum of all weights is 1, i.e., the normalized channel weights. :

[0077]

[0078] in, These are the normalized channel weights; This forms the initial envelope. Finally, the vibration measurement results from each channel are weighted and synthesized according to their normalized weights to obtain the final vibration measurement values. :

[0079]

[0080] in, These are the normalized channel weights; This provides the preliminary vibration data for each channel.

[0081] In this embodiment, the 1024 pixels of the array detector 8 form 1024 independent wavelength channels. This embodiment takes four channels as an example, such as... Figure 2 As shown, the initial weights are calculated based on the interference signals of each channel, and the amplitude distribution of the interference signals and their envelopes for each channel is displayed. It can be seen from the figure that the envelope amplitude fluctuates due to speckle fading in each channel, reflecting the randomness of the speckle effect and the necessity of multi-channel diversity. Figure 3 Showing the Figure 2 The signal envelope is normalized and then weighted. The weights are proportional to the envelope amplitude, and the weights of fading channels are close to zero, thereby reducing the contribution of inferior channels and improving the overall signal-to-noise ratio.

[0082] Figure 4 The figure shows the demodulated speed signals of each channel and the weighted combined speed signal. From top to bottom, the figures represent channel 1, channel 2, channel 3, channel 4, and the weighted combined speed signal. As can be seen from the figure, the signal from a single channel exhibits noise and glitches due to speckle fading, but the weighted combined signal is smooth and continuous, with reduced glitches and amplitude, resulting in an improved signal-to-noise ratio. In this embodiment, at t=0.069s, the envelope amplitudes of each channel are as follows: A 1 = 0.0263; A 2 = 0.0593; A 3 = 0.0084; A 4 = 0.0237, indicating that the amplitude of channel 3 has weakened. The normalized weights obtained from the amplitude are... W 1 = 0.35; W 2=04; W 3 = 0.06; W 4 = 0.19. The demodulated speed signal for each channel is... S 1 = -0.036; S 2 = -0.036; S 3 = 0.0349;S 4 = -0.038. The signal fading in channel 3 due to speckle caused significant errors in the velocity calculation. The final weighted composite vibration measurement value... S ( t = -0.35×0.036 - 0.4×0.036 + 0.06×0.0349 - 0.19×0.038 = -0.032. Through weighting, the contribution of fading channels (such as channel 3) is suppressed, and the overall signal is more stable.

[0083] This figure visually demonstrates the advantages of this invention in suppressing speckle and improving measurement reliability. This embodiment combines a single broadband light source 1 with dispersive detection to achieve high-dimensional wavelength diversity at low cost. Furthermore, an adaptive weighted fusion algorithm dynamically suppresses the contribution of fading channels, effectively overcoming the speckle effect and ensuring high stability and reliability for vibration measurements on rough surfaces.

[0084] Existing technologies employ wavelength diversity, typically using dual-frequency lasers or multiple independent lasers, coupled with complex wavelength division multiplexing (WDM). This invention replaces multiple single-frequency lasers with a single broadband light source 1 (such as an SLD), eliminating the need for complex optical components and precise calibration steps required to combine multiple lasers. The system core only adds one dispersive element 7 and one array detector 8, resulting in lower optical path calibration difficulty and manufacturing costs, making it more suitable for industrial promotion and application. Compared to dual-frequency laser schemes that only provide two diversity channels, while multi-laser synthesis can increase the number of channels, it also increases cost and optical path complexity, and the practically achievable number of channels remains very limited.

[0085] This invention utilizes the continuous spectrum of a broadband light source 1, and through dispersion and array detection, increases the number of diversity channels from two to tens or hundreds, significantly reducing the probability of simultaneous fading across multiple channels and improving measurement reliability. Through signal processing module 9, this invention fully utilizes information from all channels, performs parallel demodulation on N channels, determines weights based on the fading of the interference signals, and performs weighted synthesis of the demodulation results, ensuring signal continuity and improving the signal-to-noise ratio and measurement accuracy of the final output signal.

[0086] Although the specification has provided a detailed description, it should be understood that various changes, substitutions, and modifications can be made without departing from the spirit and scope of the invention as defined by the appended claims. Furthermore, the specific embodiments described are not intended to limit the scope of the invention, and those skilled in the art will readily understand based on this invention that existing or future-developed processes, machines, manufactures, compositions of matter, means, methods, or steps can perform substantially the same functions or achieve substantially the same results as the embodiments of the invention. Therefore, the appended claims are intended to include such processes, machines, manufactures, compositions of matter, means, methods, or steps within their scope.

Claims

1. An optical detection method, characterized by, The method comprises the following steps: S1: output a low-coherence broadband light beam by a wide-spectrum light source, wherein the wide-spectrum light source is a super-radiation light-emitting diode; a bandwidth Δλ of the broadband light beam and a roughness of a target measured surface satisfy: Δλ = λ0 / 4 Ra Ra ), λ0 is a center wavelength of the wide-spectrum light source;​ S2: splitting a broadband light beam into a reference light and a measurement light, and frequency modulating the reference light; S3: directing the measurement light to a target measured surface, and receiving scattered light returned from the surface, causing the returned scattered light to interfere with the modulated reference light to generate interference light; S4: separating the components of different wavelengths in the interference light in space by the dispersive element to form a plurality of parallel independent wavelength measurement channels, and synchronously collecting the interference signals output by each measurement channel by an array detector ; S5: demodulating the interference signals of each of the wavelength measurement channels in parallel , obtaining preliminary vibration data corresponding to each channel , the interference signals : wherein, t is time; is signal amplitude envelope; is the wavelength corresponding to the i th channel; S6: Calculate the final envelope of each channel interference signal by adaptive envelope extraction algorithm ; S7: normalizing the final envelope to obtain the normalized weight of each channel , combining the preliminary vibration data Si ( t ) weighted fusion, output the final vibration measurement value : 。 2. The method of claim 1, wherein, The dispersion element is a diffraction grating.

3. The method of claim 1, wherein, The adaptive envelope extraction algorithm comprises the following steps: Step 1: Synchronously acquire interference signals of each channel in parallel ; Step 2: said interference signal for each channel performing empirical mode decomposition to obtain a plurality of intrinsic mode functions and a residual component and reconstructing a signal using a preset number of pre-intrinsic mode functions and said residual component ; Step 3: Hilbert transform to construct analytic signal Step 3: Hilbert transform to construct analytic signal : wherein j is the imaginary unit; H is the Hilbert transform and the modulus of the analytic signal is calculated, resulting in the initial envelope ; Step 4: performing smoothing optimization processing on the initial envelope by using an adaptive Savitzky-Golay filter; wherein a window length of the adaptive Savitzky-Golay filter is dynamically adjusted according to a change rate of the initial envelope Step 4: performing smoothing optimization processing on the initial envelope by using an adaptive Savitzky-Golay filter; wherein a window length of the adaptive Savitzky-Golay filter is dynamically adjusted according to a change rate of the initial envelope Step 4: performing smoothing optimization processing on the initial envelope by using an adaptive Savitzky-Golay filter; wherein a window length of the adaptive Savitzky-Golay filter is dynamically adjusted according to a change rate of the initial envelope Step 4: performing smoothing optimization processing on the initial envelope by using an adaptive Savitzky-Golay wherein is a base window length, α is an adaptation coefficient; Step 5: Calculate final envelope : wherein, is a window length of an adaptive Savitzky-Golay filter; is an initial envelope; j is an imaginary unit; t is time.

4. The method of claim 3, wherein, The preset number of the inherent modal functions and the residual components is 3, and the first three inherent modal functions are used to reconstruct the signal with the residual components : wherein is the number of the j th natural mode function; m is the number of natural mode functions; is the residual component.

5. The method of claim 3, wherein, The base window length The adaptive coefficient α The adaptive coefficient 6. An optical detection system, characterized in that It comprises: a wide-spectrum light source for generating a low-coherence broadband light beam; a light-splitting modulation unit for splitting the broadband light beam into a reference light and a measurement light, and frequency modulating the reference light; an interference combining unit for directing the measurement light to a target measured surface, and receiving scattered light returned from the surface, causing the returned scattered light to interfere with the modulated reference light to generate interference light; a dispersion detection unit comprising a dispersion element for spatially spreading the interference light by wavelength components to form a plurality of parallel independent wavelength measurement channels, and an array detector for synchronously collecting interference signals of the spatially separated wavelength components ; a signal processing module connected to the array detector for performing the method of any of claims 1-5, outputting a final vibration measurement value .

7. The system of claim 6, wherein, The light-splitting modulation unit comprises a fiber coupler and an electro-optical modulator, and the interference combining unit comprises a beam splitter.

8. The system of claim 6, wherein, The signal processing module comprises a programmable gate array or a digital signal processor.

Images