Miniature deep hole state detection ultrasonic signal acquisition system based on double FBG

Abstract

The application provides a micro deep hole state detection ultrasonic signal acquisition system based on double FBG, comprising a laser light source, a 1*2 optical fiber coupler, a first optical fiber circulator, a second optical fiber circulator, a sensing FBG, a reference FBG, an acoustic excitation device and a signal receiving and collecting module. The micro deep hole state detection ultrasonic signal acquisition system based on double FBG systematically solves the problems of high-fidelity in-situ collection of weak acoustic signals in a narrow space of a micro deep hole and multiplexing of sensors, guarantees high-precision detection, greatly simplifies the system structure, reduces the equipment cost and operation difficulty, and provides a high-value technical scheme for real-time and safe detection of the state of a micro deep hole.

Description

Technical Field

[0001] This invention relates to the field of micro-deep hole detection, and particularly to a system for acquiring ultrasonic signals for micro-deep hole state detection based on dual FBG and a system for detecting the state of micro-deep holes. Background Technology

[0002] Micro-deep holes (holes with a diameter less than 5 mm and an aspect ratio greater than 5) are widely used in nuclear power, energy and power, aerospace, semiconductor packaging, and precision medical equipment. The integrity of the inner surface of deep hole structures directly affects the overall reliability and service life of the device. Therefore, the detection technology of their inner wall morphology (including side holes, cracks, or steps) plays an important supporting role in improving device yield and process closed-loop.

[0003] The main current methods for detecting micro-deep hole structures include: I. Traditional optical observation methods. This observation method includes CCD imaging, laser triangulation, laser projection, and optical axis methods. These testing techniques are affected by the diffraction limit, which prevents the illumination beam from being effectively focused to the bottom region of the aperture in the far field. Furthermore, the light intensity decreases exponentially in the depth direction, making it impossible to support reliable topographic reconstruction.

[0004] II. Detection of Deep Hole Defect Morphology Based on Acousto-Optical-Magnetic Effects of Hole Walls. This technology utilizes acoustic, thermal, or magnetic signals to excite and receive signals on the outer wall. During testing, the effect of the hole wall on these physical quantities is used to achieve the detection purpose. This includes detection methods based on the Hall effect, solid-state ultrasonic technology, and acoustic vibration. These methods can achieve millimeter-level accuracy. However, the detection equipment is relatively large, mostly only suitable for testing outer surfaces, with low resolution for small holes on inner surfaces, insufficient transmission capacity of acoustic, thermal, or magnetic signals within the hole, and high detection costs and low efficiency.

[0005] III. Endoscopic Methods. This method utilizes image processing technology and relies on the miniaturization of imaging elements. Early assessments of internal wall morphology heavily depended on prior knowledge, thus limiting improvements in detection efficiency and accuracy. In recent years, with the development of machine vision, testing accuracy has significantly improved; however, in practical applications, the difficulty of acquiring images of internal surface defects and the lack of sufficient training samples still present challenges, hindering quantitative analysis and evaluation capabilities. Summary of the Invention

[0006] This invention provides a system for acquiring ultrasonic signals for micro deep hole state detection based on dual FBG, in order to solve one or more of the above-mentioned problems.

[0007] According to one aspect of the present invention, a system for acquiring ultrasonic signals for detecting the state of a miniature deep hole based on dual FBG (Fiber Bragg Grating, FBG) is provided, comprising: a laser source, a 1×2 fiber coupler, a first fiber circulator, a second fiber circulator, a sensing FBG, a reference FBG and an acoustic excitation device, and a signal receiving and acquisition module. The acoustic excitation device is used to generate ultrasonic waves on the first side inside a micro-deep hole; The laser emitted by the laser source passes through a 1×2 fiber coupler and is then incident on port one of the first fiber circulator and port one of the second fiber circulator, respectively. The laser incident on port one of the first fiber optic circulator is transmitted to the sensing FBG via port two, and the laser incident on port one of the second fiber optic circulator is transmitted to the reference FBG via port two. The sensing FBG is placed on the second side inside the micro deep hole to sense ultrasonic waves. Its reflected light signal is then transmitted to the signal receiving and acquisition module through port two to port three of the first fiber optic circulator. The reference FBG is placed on the second side inside the micro deep hole to sense ultrasonic waves. Its reflected light signal is then transmitted to the signal receiving and acquisition module through ports two to three of the second fiber optic circulator. The signal receiving and acquisition module is used to perform photoelectric conversion, differential amplification, and analog-to-digital conversion on the two optical signals mentioned above, and output digital signals.

[0008] In some embodiments, the invention also includes an isolator through which the laser emitted from the laser source enters a 1×2 fiber coupler.

[0009] In some embodiments, the micro-deep holes of the present invention are filled with hydrogel as an ultrasonic transmission medium, and both the sensing FBG and the reference FBG are placed inside the hydrogel.

[0010] In some embodiments, the sensing FBG of the present invention is located in the lower-middle part of the micro-deep hole, and the reference FBG is located inside the micro-deep hole near the opening.

[0011] In some embodiments, the acoustic excitation device of the present invention is an ultrasonic root canal irrigator.

[0012] In some embodiments, the signal receiving and acquisition module of the present invention includes a balanced detector and a data acquisition card; the optical signal passing through port three of the first fiber optic circulator enters the first input terminal of the balanced detector, and the optical signal passing through port three of the second fiber optic circulator enters the second input terminal of the balanced detector; the balanced detector is used to perform photoelectric conversion and differential amplification on the two optical signals and output a differential electrical signal to the data acquisition card.

[0013] This invention relates to a system for acquiring ultrasonic signals for micro deep-hole state detection based on dual-FBG fiber optic ultrasonic sensors. The system utilizes a needle-shaped ultrasonic excitation device to generate ultrasound within the deep hole, with hydrogel serving as the ultrasonic propagation medium. The FBGs are inserted into the desired locations within the deep hole to perform in-situ detection of the hole's inner wall morphology. Compared to existing technologies, its advantages are: 1. The FBG diameter and the needle-shaped ultrasonic excitation device are only in the sub-millimeter range, allowing for in-situ ultrasonic excitation and exploration at any location within a deep hole. In contrast, traditional solid-state ultrasonic methods place both the excitation and detection devices outside the deep hole. Therefore, in-situ detection is the main advantage of this invention compared to traditional optical and acousto-optic-magnetic effect detection methods.

[0014] 2. Hydrogels significantly improve the quality of ultrasonic transmission when used as an ultrasonic medium. Specifically, before gelation, the hydrogel is a liquid, and its fluidity allows it to penetrate into extremely fine side holes or cracks. After gelation, it becomes an excellent ultrasonic medium, enabling the ultrasonic excitation signal to propagate effectively within deep holes without overflowing due to the stress generated by the ultrasonic waves. After testing, the deep holes can be completely removed using a simple method, thus having no impact on the device.

[0015] This invention innovatively proposes an ultrasonic detection method for the state of micro-deep holes, combining a "miniature FBG sensing probe + dual-probe hardware differential noise reduction + biomimetic hydrogel acoustic coupling + compatibility with existing miniature ultrasonic excitation sources." Through a four-pronged collaborative design—sensor miniaturization, differential acquisition link, reversible propagation medium, and readily available excitation source—it systematically solves the challenges of high-fidelity in-situ acquisition and sensor multiplexing of weak acoustic signals within the confined space of micro-deep holes. While ensuring high detection accuracy, it significantly simplifies the system structure, reduces equipment costs and operational complexity, and provides a high-value technical solution for real-time and safe detection of the state of micro-deep holes.

[0016] According to another aspect of the present invention, a detection system for the state of a micro deep hole based on dual FBG is also provided, comprising the above-described system for acquiring ultrasonic signals for detecting the state of a micro deep hole based on dual FBG. Attached Figure Description

[0017] Figure 1 This is a schematic diagram of the structure of a micro deep hole state detection ultrasonic signal acquisition system based on dual FBG according to one embodiment of the present invention.

[0018] Figure 2 This is a schematic diagram showing the positional relationship between the dual FBG and the acoustic excitation device in the test structure (dental root canal) of a dual FBG-based micro deep hole state detection ultrasonic signal acquisition system according to one embodiment of the present invention.

[0019] Figure 3 This is a cross-sectional schematic diagram of the test structure (dental root canal) under five different working conditions; among them, Figure 3 (a) is a cross-sectional view of a tooth accessory root canal; (b) is a cross-sectional view of a tooth root lateral branch root canal; (c) is a cross-sectional view of a tooth crown lateral branch root canal; (d) is a cross-sectional view of a tooth step; and (e) is a cross-sectional view of a normal tooth.

[0020] Figure 4(a) shows the ultrasonic signal acquisition system for micro deep hole state detection based on dual FBG according to one embodiment of the present invention. Figure 3 Time-domain signal and spectrum of the operating condition (a).

[0021] Figure 4(b) shows the ultrasonic signal acquisition system for micro deep hole state detection based on dual FBG according to one embodiment of the present invention. Figure 3 Time-domain signal and spectrum of the medium (b) working condition.

[0022] Figure 4(c) shows the ultrasonic signal acquisition system for micro deep hole state detection based on dual FBG according to one embodiment of the present invention. Figure 3 Time-domain signal and spectrum diagram of the (c) working condition.

[0023] Figure 4(d) shows the ultrasonic signal acquisition system for micro deep hole state detection based on dual FBG according to one embodiment of the present invention. Figure 3 Time-domain signal and spectrum of the medium (d) working condition.

[0024] Figure 4(e) shows the ultrasonic signal acquisition system for micro deep hole state detection based on dual FBG according to one embodiment of the present invention. Figure 3 Time-domain signal and spectrum of the operating condition (e).

[0025] Figure 5 The above are superimposed comparison diagrams of the signal spectra of the five test structures (dental root canals) shown in Figures 4(a)-4(e).

[0026] Figure 6 This invention provides a process for identifying ultrasonic signals for micro-deep hole state detection based on dual FBG, according to one embodiment of the present invention.

[0027] Figure 7 This invention provides an enhancement process for ultrasonic signals used in micro-deep hole state detection based on dual FBG, according to one embodiment of the present invention.

[0028] Figure 8 This is an evaluation diagram of the enhancement effect of ultrasonic signals for micro deep hole state detection based on dual FBG according to one embodiment of the present invention.

[0029] Figure 9 The selection results are used to compare the classification model selection and cross-validation performance of each stage in the two-stage classification process.

[0030] Figure 10 This is a performance comparison chart of normal and abnormal binary classification in the first stage of classification.

[0031] Figure 11 This is the normalized confusion matrix diagram for the first stage.

[0032] Figure 12 This is a detailed performance comparison chart for the four abnormal operating conditions in the second stage.

[0033] Figure 13 This is a normalized confusion matrix diagram of the four abnormal operating conditions in the second stage.

[0034] Figure 14 The normalized confusion matrix diagram for two-stage integrated classification. Detailed Implementation

[0035] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0036] It should be noted that, unless otherwise specified, the embodiments and features described in this application can be combined with each other.

[0037] Finally, it should be noted that in this document, relational terms such as "first" and "second" are used only to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising" or "including" include not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus. Unless otherwise specified, an element defined by the phrase "comprising..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes said element.

[0038] The present invention will now be described in further detail with reference to the accompanying drawings.

[0039] Figure 1 The structure of a micro-deep hole state detection ultrasonic signal acquisition system based on dual FBG (fiber Bragg grating) according to an embodiment of the present invention is schematically shown.

[0040] In this system, dual FBGs are used as sensing probes to detect the state of micro-deep holes. Different types of micro-deep hole structures are identified by the changes in the propagation characteristics of ultrasonic waves within the micro-deep hole system.

[0041] refer to Figure 1 As shown, the acquisition system for obtaining micro-deep hole state detection signals based on dual Bragg fiber gratings includes: a narrowband tunable laser source 10, an isolator 20, a 1×2 fiber coupler 30, a first fiber circulator 41, a second fiber circulator 42, a sensing FBG 51, a reference FBG 52, and an acoustic excitation device 81, as well as a signal receiving and acquisition module (including a balanced detector 82 and a data acquisition card 83) and a computer 84.

[0042] The light emitted from the narrowband tunable laser source passes through a 1×2 fiber coupler 30 and is then incident on port one of the first fiber circulator and port one of the second fiber circulator, respectively.

[0043] The laser light incident on port one of the first fiber optic circulator 41 is transmitted to the sensor FBG51 via port two, and the laser light incident on port one of the second fiber optic circulator is transmitted to the reference FBG52 via port two.

[0044] The FBG51 sensor is placed on the second side inside the micro-deep hole to sense ultrasonic waves. Its reflected light signal is transmitted to the signal receiving and acquisition module through ports two to three of the first fiber optic circulator 41.

[0045] The reference FBG52 is placed on the second side inside the micro-deep hole to sense ultrasonic waves. Its reflected light signal is then transmitted to the signal receiving and acquisition module through port two to port three of the second fiber optic circulator 42.

[0046] The signal receiving and acquisition module is used to perform photoelectric conversion, differential amplification, photoelectric conversion and analog-to-digital conversion on the two optical signals mentioned above, and output digital signals.

[0047] Isolator 20 is connected in series between the output of the laser source and the input of the 1×2 fiber coupler 30. Its function is to prevent reflected light generated in the subsequent optical path from returning to the laser cavity, thus avoiding output power fluctuations and frequency instability.

[0048] The 1×2 fiber coupler 30 has a splitting ratio of 50:50, which equally divides the input optical power into two output paths: a sensing channel and a reference channel. This characteristic is crucial for ensuring the effectiveness of subsequent differential cancellation. If there is a deviation in the splitting ratio between the two paths, the initial amplitude of the common interference in the two paths will be unequal, reducing the suppression effect of differential cancellation.

[0049] The laser source uses a narrowband tunable laser, which outputs a continuous narrowband laser with a wavelength of 1550 nm, and can be precisely locked to the linear operating range of the FBG reflection spectrum. The narrowband laser enters a 1×2 fiber coupler 30 after passing through isolator 20, and is divided into two paths: a sensing channel and a reference channel.

[0050] The optical path employs a strictly symmetrical structure. This symmetrical optical path layout ensures that both optical signals experience the same fiber optic link length and optical element type. Assume the laser output optical power is... The power at both output terminals of the coupler is Under ideal conditions where link loss differences are not considered, the interference components in the two signals caused by laser intensity noise, fiber optic link thermal drift, and ambient temperature changes are highly consistent in amplitude and phase, providing a physical basis for differential cancellation of subsequent balanced detection.

[0051] In the sensing channel, the first fiber optic circulator 41 is a three-port device: port one is the input, port two is a bidirectional port (also connected to the sensing FBG 51), and port three is the reflected signal output. It is necessary to ensure that the incident light at port one does not directly leak to port three and interfere with the measurement of the reflected signal. The sensing FBG 51 senses the dynamic strain caused by the ultrasonic field within the micro-deep hole, causing a dynamic shift in the center wavelength of its reflected spectrum. The reflected light carries strain modulation information. This reflected light returns through ports two and three of the circulator and is connected to the first input of the balanced detector 82.

[0052] The reference channel adopts a structure that is completely symmetrical with the sensing channel. The optical signal is connected to the reference FBG52 via the second fiber optic circulator 42 (the same model and parameters as the first fiber optic circulator). The reflected light from the reference FBG52 returns through ports two to three of the second fiber optic circulator 42 and is connected to the second input terminal of the balanced detector 82.

[0053] Both the sensing FBG51 and the reference FBG52 are fabricated from bare fiber gratings customized in the same batch, with a center wavelength of 1550nm, a grating length of 1mm, and a fiber radial dimension of 125μm. This batch-customization ensures the consistency of the spectral characteristics (center wavelength, reflection spectrum shape, 3dB bandwidth) of the two FBGs, a prerequisite for effective differential cancellation of the two signals in the balanced detector 82. Addressing the fundamental physical limitation of traditional sensors being too large to penetrate deep into micro-holes, this application employs a 125μm radial dimension FBG sensing probe, which can be non-destructively implanted into the core region of the micro-hole acoustic field. By sensing ultrasonic strain signals, this application overcomes the long-standing limitation of deployment in the confined space of micro-holes, achieving non-destructive in-situ detection within micro-holes, offering advantages such as simple structure and no damage to the micro-hole structure.

[0054] The balanced detector 82 is the core component for hardware-level noise suppression in this system. It contains two highly matched photodiodes and a transimpedance differential amplifier circuit, which can respond to the fundamental frequency and higher harmonic components of a 45kHz ultrasonic signal without distortion.

[0055] The balanced detector 82 differentially amplifies the two optical signals and outputs a voltage signal, which is then converted into a digital signal by the data acquisition card 83 and transmitted to the computer 84. The data acquisition card 83 has 16-bit data precision and a sampling frequency of 1MHz.

[0056] Because tooth root canals have a naturally small and complex micro-deep pore structure, this embodiment uses an excised double-root canal tooth as the test structure. The original pulp in the pulp cavity has been removed, and hydrogel is injected into the pulp cavity as an ultrasonic transmission medium before the test.

[0057] The acoustic excitation device 81 is used to generate an ultrasonic field inside the first lateral root canal of the tooth. The sound waves are transmitted to the contralateral root canal through the hydrogel and the dentin wall between the root canals. The dual FBG probes implanted in the second lateral root canal sense the ultrasonic strain, causing dynamic drift of their respective center wavelengths. The sensing and demodulation optical path module converts the wavelength drift into a light intensity change signal. The signal receiving and acquisition module performs photoelectric conversion, differential amplification, and analog-to-digital conversion on the two optical signals, and outputs a digital signal to the computer 84 to complete the signal acquisition.

[0058] To address the constraints of the confined space in the root canal, an FBG fiber with a radial dimension of approximately 125 μm is used, much smaller than the minimum diameter of the main root canal (approximately 0.5 mm), allowing for non-destructive insertion. To ensure that the excitation device does not damage the test structure during testing, a low-power ultrasonic root canal irrigator 70 is used as the acoustic excitation device 81. For the identification of minute structures such as accessory root canals, lateral canals, and steps, hardware-level noise suppression is achieved through a dual FBG depth differential layout and a balanced detection architecture.

[0059] The response of a FBG to ultrasound is essentially due to strain. The Bragg wavelength of an FBG... satisfy:

[0060] in, The effective refractive index of the optical fiber. The period of the grating. When ultrasound propagates along the optical fiber axis, it causes periodic axial strain in the fiber, which in turn modulates the Bragg wavelength through two physical mechanisms: one is a geometric effect, where the strain directly changes the physical period of the grating. Secondly, there is the elasto-optic effect, where strain affects the refractive index of the fiber core material. Changes occur. Under the influence of ultrasound, the center wavelength shift of the FBG changes. With axial strain The relationship between them can be expressed as:

[0061] in, The effective elastic coefficient.

[0062] This application employs an edge-filter demodulation method based on wavelength locking: the narrowband laser wavelength is locked within the linear operating region with the largest slope within a 3dB bandwidth of the reflection spectra of sensor FBG51 and reference FBG52, and tuned to the compromise linear operating point of their reflection spectra, ensuring that both channels operate within the linear response range. When ultrasonic strain causes a slight shift in the center wavelength of the FBG, the reflected light power at the locked wavelength changes linearly, thus converting wavelength modulation into intensity modulation. The wavelength shift can be demodulated by real-time detection of the light intensity, and the ultrasonic strain can be inverted. This method is simple in structure, fast in response, and requires no interferometer, making it suitable for real-time detection of dynamic strain.

[0063] Balanced Detection Design: The balanced detector 82 serves as the core of hardware-level noise suppression. It receives the sensing and reference optical signals via two highly matched photodiodes, converting them into photocurrents. A transimpedance differential amplifier then outputs a voltage proportional to the difference between the two photocurrents. Based on a symmetrical optical path design, interference from laser intensity fluctuations, thermal drift, and mechanical vibrations is similar in amplitude and phase in both channels. The sensing channel's optical power contains both useful and interference signals, while the reference channel primarily contains interference signals. After differential processing, the interference components are directly canceled out. This significantly suppresses common noise at the acquisition source before analog-to-digital conversion, effectively improving the output signal-to-noise ratio.

[0064] Figure 2 The diagram schematically illustrates the position of the dual FBG in the test structure (dental root canal) and the point of application of the acoustic excitation device in the acquisition system of the dual FBG-based micro deep hole condition detection ultrasonic signal.

[0065] refer to Figure 2As shown, the test object (test structure 100) is an excised double-rooted human tooth with fixed support roots. The root canal filling medium is a biosafe sodium alginate hydrogel. This sodium alginate hydrogel has a sound velocity of approximately 1561±8 m / s, an attenuation coefficient of approximately 0.54±0.18 dB / cm / MHz, and an acoustic impedance within the range of corresponding parameters for human soft tissue. This material effectively reduces sound wave reflection energy loss at the interface between the root canal wall and the filling medium, establishing a continuous sound transmission path. Its viscoelastic properties effectively prevent filling medium overflow caused by ultrasonic stress. Simultaneously, the hydrogel can be removed from the root canal through irrigation. Addressing the challenge of signal drift and uneven coupling caused by the flowability of traditional fluid media in complex root canal anatomy, this application proposes an embedded coupling scheme based on in-situ cured sodium alginate hydrogel. This material has good biocompatibility, and its acoustic parameters are highly similar to those of human dental pulp tissue, effectively constructing a continuous, low-loss sound transmission path. Leveraging its "flow-first, solidify-later" physical properties, this gel can fully saturate the minute gaps within the root canal during the initial injection phase. Placing the FBG sensor probe within this space allows the gel to solidify in situ, forming a stable viscoelastic envelope. This design helps suppress disturbances to the test signal caused by fluid flow, enabling precise and stable sensor positioning within the root canal and eliminating residual air at the interface, significantly improving acoustic coupling efficiency and signal consistency. Furthermore, utilizing its soluble nature after solidification, it can be thoroughly removed with gentle irrigation after testing, ensuring the non-destructive recovery and reuse of the FBG sensor probe while offering advantages in ease of operation and low cost.

[0066] The ultrasonic excitation device uses an ultrasonic root canal sizing device 70, with a nominal center frequency of 45kHz. This device is a routinely used irrigation and activation instrument in root canal treatment, and its working tip size and structure are designed to fit the confined space of the root canal. Therefore, directly using this existing device as the excitation source eliminates the need for additional development or configuration of a dedicated excitation device, resulting in low system complexity and good compatibility. During use, the ultrasonic working tip extends into one side of the root canal, exciting the ultrasonic sound field by contacting the hydrogel medium filling the root canal, establishing a sound wave path that propagates across the root canal. When the sound waves encounter acoustic impedance mismatch interfaces such as accessory canals, lateral canals, or steps during propagation, reflection, scattering, and mode conversion occur, carrying information about the internal structural state of the root canal. The working tip of the ultrasonic root canal sizing device 70 extends into the middle of one side of the root canal, and the sound waves are transmitted to the other side of the root canal via the intercanal dentin wall. A dual FBG probe implanted in the hydrogel medium of the other root canal senses ultrasonic strain, causing a dynamic drift in its center wavelength. The amount of FBG center wavelength drift can be demodulated by monitoring changes in the reflected light intensity. In this invention, the signal noise originates from the vibration of the system under test and the excitation system. The vibration is amplified at certain frequencies, so the noise is a typical example of colored noise.

[0067] Both the FBG51 sensor and the reference FBG52 are deployed in the same root canal on the opposite side of the hydrogel filling, but their implantation depths are intentionally differentiated: the grating area of ​​the FBG51 sensor is implanted in the middle of the root canal, approximately 10 mm from the root canal orifice, at the core path of ultrasonic wave propagation, used to sense dynamic changes in the ultrasonic sound field within the root canal; the grating area of ​​the reference FBG52 is located on the surface of the hydrogel near the root canal orifice in the same root canal, approximately 2 mm from the orifice, primarily sensing common interferences such as environmental vibration, temperature drift, and light source noise, while its response to changes in the sound field structure deep within the root canal is weaker. To address the problem of low signal-to-noise ratio caused by the inability to separate common environmental noise and useful signals at the source in single-sensor acquisition schemes, this application employs a depth-differential layout, implanting the FBG51 sensor and the reference FBG52 into the same root canal. By utilizing the physical barrier of the rapid attenuation of ultrasonic signals with depth, the environmental interference noise in the two signals is kept highly correlated, while the target ultrasonic signal has a significant gradient difference. Combined with the differential operation of the balanced detector 82, the common noise is directly canceled at the hardware level, which improves the signal-to-noise ratio and extraction fidelity of the weak acoustic signal from the source. It has the advantages of significant noise reduction effect and low signal distortion.

[0068] In this configuration, the amplitude of the ultrasonic differential signal sensed by sensor FBG51 is significantly greater than that of the reference FBG52, while the common interference sensed by the two FBGs is highly correlated. After differential operation by the balanced detector 82, the common interference cancels each other out, the differential signal is effectively preserved, and the system signal-to-noise ratio is significantly improved. Furthermore, the spatial differential information in the depth direction provides a basis for signal source tracing: when local structural anomalies in the root canal (such as steps or lateral openings) cause changes in the acoustic field at the FBG, this change has a weak response at the FBG in the shallow layer of the root canal, and the differential output can effectively reflect the local characteristics of this change; if the signals at the two FBGs change proportionally due to fluctuations in external excitation power, the differential output remains basically stable. The dual-probe depth differential configuration not only achieves noise suppression but also provides spatial resolution to distinguish the signal source.

[0069] The complete system calibration process is as follows: 1. Adjust the narrowband laser wavelength to 1550.00 nm and monitor it with a spectrometer to ensure that the reflection spectra of both the sensing FBG51 and the reference FBG52 are operating in the linear region of the rising edge of the reflection spectrum, and that the reflected light power of the two FBGs at this wavelength is close to 50% of their respective peak reflection power; 2. Insert the sensor FBG51 and reference FBG52 into the root canal at the specified depths according to the depth difference layout described above, without bending or hard contact with the canal wall. 3. The laser is preheated for 30 minutes. After the output power stabilizes, the system enters the acquisition state.

[0070] Figure 3The schematic diagram shows the cross-sectional views of the test structure (dental root canal) under five different working conditions.

[0071] The preparation of five typical working conditions is as follows. Each healthy tooth is prepared independently according to one working condition to avoid cumulative damage caused by repeated processing of the same tooth.

[0072] refer to Figure 3 As shown, where: (a) For accessory canal preparation: The first healthy tooth is harvested, and a small accessory canal channel is prepared in the middle of the main root canal wall using a micro-file. This structure simulates the common small branching channels originating from the pulp chamber floor or root canal wall. After preparation, the root canal is filled with sodium alginate hydrogel.

[0073] (b) Lateral branching canal construction – A second healthy tooth was harvested, and a lateral branching channel was prepared in the apical third of the root canal using micro-instruments. This structure simulates common lateral branching canals or apical bifurcation structures located in the apical region. After preparation, the root canal was filled with sodium alginate hydrogel.

[0074] (c) For the coronal lateral canal case—a third healthy tooth was taken, and the preparation method was similar to that for the root lateral canal case, except that the lateral canal opening was located in the coronal 1 / 3 region of the root canal, in order to study the influence of the lateral canal position parameters on sound field propagation and signal characteristics. After preparation, the root canal was filled with sodium alginate hydrogel.

[0075] (d) Stepped defect – A fourth healthy tooth was harvested, and a geometrically controlled stepped defect was prefabricated at a predetermined depth in the root canal wall using a micro-file, with a step depth of approximately 0.16 mm. This structure simulates an iatrogenic step caused by improper instrumentation during root canal preparation. After preparation, the root canal was filled with sodium alginate hydrogel.

[0076] (e) Normal tooth condition – The fifth healthy tooth was selected, with the root canal system maintaining its original anatomical shape and the interior uniformly filled with sodium alginate hydrogel, without introducing any additional anatomical variations or artificial defects, as the baseline control group.

[0077] Sodium alginate hydrogel can flow into double root canals, lateral root canals, and microfissures before forming a stable gel. It forms a continuous and stable viscoelastic acoustic coupler between the root canal wall and the FBG sensing probe. This hydrogel coupler fixes the position and orientation of the FBG sensing probe, suppressing media flow caused by ultrasonic excitation, overflow of the filling medium, and signal drift caused by probe displacement, thus maintaining the stability of the acoustic wave propagation boundary during detection. After multiple reflections between the inner walls of double root canals, lateral root canals, and irregular root canals, ultrasonic waves generate ultrasonic information with distinct characteristics. This ultrasonic information acts on the embedded FBG sensing probe and is converted into a corresponding grating strain response, enabling the FBG to stably sense multiple reflection signals from the root canal wall and obtain repeatable arrival time, amplitude attenuation, and spectral variation characteristics. Sodium alginate hydrogel can simultaneously achieve deep in-situ gel formation, stable FBG embedding, repeated sensing of multiple reflection signals, and post-detection removal, thus providing stable and reliable raw data for subsequent extraction of multiple root canal wall reflection characteristics and signal processing.

[0078] Figures 4(a)-4(e) schematically show the original time-domain signals and spectrograms of five types of root canal conditions acquired by the dual-FBG-based micro deep hole condition detection ultrasonic signal acquisition system according to an embodiment of the present invention.

[0079] The experiments were conducted sequentially, following the order of accessory root canals, lateral root canals, coronal lateral root canals, ledges, and normal canals. For each case, after completing hydrogel filling, FBG probe implantation, and the aforementioned optical path calibration, the ultrasonic root canal irrigator 70 was activated at its nominal center frequency of 45 kHz in continuous operation mode. Data acquisition was initiated after the acoustic field within the root canal stabilized.

[0080] The differential voltage signal output by the balanced detector 82 is converted into a digital signal by the data acquisition card 83 and then transmitted to the computer 84. The sampling frequency of the data acquisition card 83 is set to 1MHz, and the data precision is 16 bits. According to the Nyquist sampling theorem, the fundamental frequency and its multiple harmonic components of the 45kHz ultrasonic excitation signal can be effectively extracted. The single sampling duration is set to 0.01s, and its time domain coverage is sufficient to completely capture the multiple reflections and attenuation processes experienced by the ultrasonic wave as it travels from the excitation side through the dentin wall to the sensing side.

[0081] One hundred samples were collected for each operating condition, resulting in a total experimental dataset of 500 samples across the five operating conditions. Data acquisition for all operating conditions was conducted under the same experimental environment, with an ambient temperature of approximately 25°C. Environmental conditions and device status were monitored in real time during the acquisition process to ensure consistency of experimental conditions across all operating conditions.

[0082] The differential voltage signal output by the balanced detector 82 is input to the data acquisition card 83, and the original time-domain voltage waveform is recorded in real time by LabVIEW software on the computer 84. During the acquisition process, the signal waveform quality can be observed intuitively, and the acquisition status can be monitored.

[0083] Referring to Figures 4(a)-4(e), the time-domain amplitude, waveform morphology, and frequency-domain energy distribution of the FBG response signal under different root canal conditions are shown.

[0084] Among them, Figure 4(a) is Figure 3 Figure 4(a) shows the time-domain signal and spectrum of the accessory root canal in the tooth; Figure 4(b) shows... Figure 3 Figure 4(b) shows the time-domain signal and spectrum of the lateral canal at the root of the tooth; Figure 4(c) shows... Figure 3 Figure 4(c) shows the time-domain signal and spectrum of the lateral canal of the tooth crown under working conditions; Figure 4(d) shows... Figure 3 Figure 4(d) shows the time-domain signal and spectrum of the tooth step condition; Figure 4(e) shows... Figure 3 (e) Time-domain signal and spectrum of normal dental working condition.

[0085] The location indicated by the dashed box in the figure is the location of the main response frequency, and the response frequency range is 45kHz ± 10% of the nominal frequency of the excitation device.

[0086] This application employs a dual-probe differential layout, with both probes simultaneously acquiring signals. After differential processing, common interferences cancel each other out, effectively suppressing environmental interference. The signals acquired using the dual FBG differential layout exhibit good stability, indicating that FBG ultrasonic sensing technology is feasible for acquiring response signals under different root canal conditions.

[0087] Figure 5 The diagram schematically illustrates the signal spectrum superposition and comparison of five types of root canal conditions acquired by the dual-FBG-based micro deep hole condition detection ultrasonic signal acquisition system according to an embodiment of the present invention.

[0088] refer to Figure 5 As shown in the figure, the position marked by the dashed box is the location of the main response frequency, and the response frequency range is the nominal frequency of the excitation device, 45kHz±10%.

[0089] Figure 6 The illustration schematically shows the enhanced identification process of ultrasonic signals for micro deep hole state detection based on dual FBG according to an embodiment of the present invention.

[0090] refer to Figure 6 As shown, the process includes the following steps: Step S10: Reading and preprocessing of ultrasonic signals from micro-deep holes; Step S20: Enhancement of the random resonance signal; Step S30: Fixed sequence length conversion; Step S40: Construct three-domain fusion features; Step S50: Two-stage soft voting identification.

[0091] The following sections will provide a detailed explanation of each step.

[0092] Regarding step S10, the reading and preprocessing of the micro-deep hole response signal.

[0093] The root canal response ultrasonic signal acquired by the above-mentioned dual FBG-based micro deep hole state detection ultrasonic signal acquisition system is read. In this project, after the ultrasonic signal is excited by the cusp oscillation tip in the root canal, the signal obtained by the FBG includes the vibration of the alveolar tissue, the vibration of the cusp oscillation tip, and the noise of the demodulation part of the FBG ultrasonic sensor itself. Therefore, it has the characteristics of low signal-to-noise ratio. For this reason, this application uses an improved stochastic resonance model to improve the signal-to-noise ratio of the sensing signal.

[0094] First, the read micro-deep-hole ultrasonic response signal is mean-removed and standardized. To adjust the input signal to the optimal response range of the bistable barrier, the input peak value is calculated. Amplitude scaling is applied during processing, with a threshold value of 0.5 used as an empirical value. After the above processing, Gaussian white noise is added at a noise level equal to the signal standard deviation of 0.0323, resulting in a random resonance input signal. .

[0095] Secondly, using Fourier analysis, with the nominal excitation frequency of 45kHz of the excitation device as the center frequency, and based on the 10% error frequency response range specified in the instruction manual of the ultrasonic root canal irrigator used in this implementation, a signal strength search was performed within a narrow band to obtain the actual sensing signal's main frequency. 44kHz.

[0096] Finally, based on the actual sensing signal frequency The effective signal range is defined as the narrow band centered at 44kHz [43kHz, 45kHz]. The signal-to-noise ratio (SNR) of the random resonant input signal is calculated to obtain an initial SNR of -3.617dB.

[0097] Figure 7 The flowchart of a method for enhancing ultrasonic signals for micro-deep hole state detection based on dual FBG according to an embodiment of the present invention is illustrated.

[0098] refer to Figure 7 As shown, after the ultrasonic signal reading and preprocessing in step S10 (micro-deep hole response), the random resonance signal enhancement process in step S20 includes the following steps: Step S21, standard stochastic resonance structure parameter optimization: The input signal and random noise are used as excitations to enhance the system based on the standard stochastic resonance model to obtain the standard stochastic resonance enhanced signal. The system structure parameters of the standard stochastic resonance model are obtained by traversal combined with nested search optimization methods.

[0099] Step S22: Construct an improved stochastic resonance model: Add an overdamped term, a time delay term, and a fractional derivative parameter term to the standard stochastic resonance model to form an improved stochastic resonance model; input the standard stochastic resonance enhancement signal and random noise as excitations to the enhancement system based on the improved stochastic resonance model to obtain the improved stochastic resonance enhancement signal; the core parameters added in the improved stochastic resonance model are obtained through a combination of traversal and nested search optimization methods.

[0100] Step S23, signal enhancement based on nested search and screening of random resonance parameters using optimization method: In the ergonomic combined nested search optimization method, the evaluation index of optimization is signal-to-noise ratio (SNR); the method for obtaining the initial SNR and the SNR of the enhanced signal is as follows: using Fourier analysis, with the nominal excitation frequency of the acoustic excitation device as the center frequency, the signal intensity is searched within the narrow band of the acoustic excitation instrument error to obtain the main frequency of the actual sensing signal, and the narrow band range centered on the main frequency of the actual sensing signal is taken as the effective signal interval, and the SNR is calculated.

[0101] Step S24: Enhance the evaluation of the treatment effect.

[0102] In step S21, the standard random resonance structure parameters are first optimized.

[0103] The mathematical model of standard stochastic resonance is:

[0104] in, The signal is enhanced by standard random resonance. For linear gain terms, These are nonlinear constraint terms; , These are structural parameter items, parameters Parameters used to adjust the potential well position and the system's response to weak signals. Used to limit the output amplitude and adjust the barrier height; Indicates time, It is a random noise term.

[0105] Here, the Euler method is used to solve the differential equation. , The values ​​of the two parameters in the search space are:

[0106] The optimal values ​​for the stochastic resonance-enhanced signal-to-noise ratio (SNR) gain and the magnitude of the post-processed main frequency shift are evaluated to obtain the optimal parameter combination for the best post-processing effect. .

[0107] In step S22, the optimal parameters obtained in step S21 are... Combine and substitute into the improved stochastic resonance model constructed in step S22:

[0108] in, This is a signal enhanced by an improved stochastic resonance model. This improved stochastic resonance model adds three adjustment terms compared to the standard stochastic resonance model. Their functions and parameters are as follows: Overdamped term Its function is to reduce the spontaneous oscillations caused by underdamped characteristics, which lead to false resonance peaks masking the true characteristics, and to ensure the stability and convergence accuracy of the numerical solution process. Its characteristic parameters are: ; Delay Item Its function is to modulate the system by introducing historical information to induce stochastic resonance, and its characteristic parameters are: and , Using an empirical value of 1, For the delay time, Obtained by traversal optimization; Fractional derivative parameter terms Its definition is:

[0109] in, For the order of the fractional differential, Represents the gamma function. (Caputo) indicates the fractional derivative type; its function is as follows: because the viscoelastic sodium alginate hydrogel used in acquiring micropore root canal response signals exhibits parametric excitation under high-frequency acoustic excitation, fractional parametric excitation models provide a higher fit to the actual system in viscoelastic material parametric excitation systems, and the order of the fractional derivative is also higher. When the acoustic excitation is working continuously, the fractional derivative contributes to viscoelasticity; in addition, it can also modulate the memory characteristics of the system to better match the characteristics of the FBG sensing signal as multiple reflections from the root canal wall.

[0110] The above equations are solved using numerical methods, where, , The two parameters are the values ​​obtained from the optimization in step S21. , and The information was obtained through optimization using a traversal method, and then enhanced to include... .

[0111] In step S23, the fractional derivative order of the key parameters of the improved stochastic resonance model in step S22 is calculated. Overdamping coefficient and delay time The process involves screening, using signal-to-noise ratio (SNR) as the evaluation metric, and traversing the preset parameter space to select the parameter combination that maximizes the SNR at the target frequency. The fractional derivative term is output by the prediction-correction iteration. First, the prediction state is obtained based on the current state, time delay state, input signal, and fractional derivative. Then, the enhanced signal is obtained by correction based on the prediction state.

[0112] Here, the Euler method is used to solve the differential equations. In this model, the parameters... , and Using the traversal method The optimal parameter combination is obtained by optimizing the SNR gain and the post-processed main frequency offset through random resonance enhancement. , and After processing with optimal parameters, the signal is The signal-to-noise ratio is improved to SNR=3.751dB, the peak output frequency is 44000.00 Hz, the gain after the improved random resonance enhancement process is 7.368dB, and the frequency offset is 0Hz.

[0113] Figure 8 The illustration schematically demonstrates the enhancement effect of ultrasonic signals for micro-deep hole state detection based on dual FBG according to one embodiment of the present invention.

[0114] refer to Figure 8 As shown, the spectral peaks of the signal near the target frequency are not prominent enough before processing, making it susceptible to interference from background noise and adjacent frequency components. After the improved stochastic resonance enhancement, the spectral peaks near the target dominant frequency are significantly raised, and the dominant frequency characteristics are more concentrated and clear. The signal-to-noise ratio in the figure increases from -3.62dB to 3.75dB, with an enhancement gain of approximately 7.37dB, while the frequency shift is 0Hz. This indicates that the enhancement method improves the target frequency energy and signal-to-noise ratio without changing the dominant frequency position of the original ultrasonic response signal.

[0115] Step S30: Fixed sequence length conversion.

[0116] Since the response sequence lengths of different samples may vary, in order to eliminate the impact of inconsistent sampling lengths between different samples on subsequent feature construction and the input dimension of the classification model, the system adopts a time-series sequence equal-length normalization processing method to enhance the response sequence. Convert to a uniform response sequence of preset length N=900 .

[0117] Step S40: Construct three-domain fusion features.

[0118] To improve the separability between different root canal states and the stability of subsequent identification, this embodiment reads the signal after conversion with a fixed sequence length. Extract time-domain statistical features, frequency-domain features, and wavelet multi-scale features.

[0119] Among them, the time-domain statistical features are used to reflect the signal energy, wave intensity, and characterize the overall envelope of the ultrasonic response and the morphological changes in different propagation stages. These include standard deviation, maximum value, minimum value, peak-to-peak value, mean square energy, skewness, kurtosis, waveform factor, impulse factor, peak factor, energy, and 22-dimensional morphological characteristics of the output waveform obtained by resampling, forming a total of 33-dimensional time-domain statistical feature vector. .

[0120] Frequency domain features are used to reflect the dominant frequency, spectral energy distribution, and harmonic structure, including: dominant frequency, spectral centroid, spectral entropy, spectral mean square value, and amplitudes of the first to sixth harmonics, forming a total of 10-dimensional frequency domain features. .

[0121] Wavelet multi-scale features reflect information such as the low-frequency profile, approximate information, high-frequency abrupt changes, details, and noise of a signal. In this embodiment, Haar wavelet transform is used to extract the signal. Three-layer wavelet approximate mean and variance, and first, second, and third layer wavelet detail mean and variance, totaling 8-dimensional features. .

[0122] The time-domain features, frequency-domain features, and wavelet-domain features are concatenated to form the fused feature quantity.

[0123] Step S50, two-stage soft voting identification.

[0124] The signal is fused into three-domain features after feature extraction. This serves as the input to the two-stage identification model. In this embodiment, 200 sets of discarded samples are removed, and 60 sets of samples are selected for each working condition during the two-stage classification, for a total of 300 samples. 20% of these samples are used for the test set, and 80% are used for the training set. The first stage is used to distinguish between normal and abnormal samples, and the second stage is used to identify subclasses of the four types of abnormal samples.

[0125] The first stage transforms the five-class classification problem into a binary classification problem of normal and abnormal. The candidate models for the first stage classification include Random Forest, K-Nearest Neighbor, and Support Vector Machine, with scores of 0.8655, 0.8225, and 0.7689, respectively. All three were selected for soft voting ensemble.

[0126] The second-stage abnormality category set includes accessory root canals, lateral root canals of the root, coronal lateral root canals, and steps. Candidate models for the second stage include: Extreme Gradient Boosting (XGBoost), Lightweight Gradient Boosting (LightGBM), Random Forest, Support Vector Machine, and K-Nearest Neighbors. The top two models, ranked according to cross-validation scores, are selected for soft voting ensemble. The selected models are Extreme Gradient Boosting and Lightweight Gradient Boosting, yielding the classification results for abnormal conditions (see [link to documentation]). Figure 9 (As shown).

[0127] Figure 9 The illustration schematically shows the soft voting results of different classification models and the selection of classification models at each stage in the two-stage soft voting identification process of the ultrasonic signal identification for micro deep hole state detection based on dual FBG according to an embodiment of the present invention, during the first and second stage identification processes.

[0128] refer to Figure 9 As shown, a confidence judgment mechanism was added in the second-stage classification. After threshold cross-validation, when d1=0.450 and d2=0.400 (where d1 and d2 are the confidence judgment thresholds for the first and second stages, respectively), the coverage of the two-stage ensemble model was 1.0000, the accuracy of non-rejected samples was 0.8833, and the accuracy of rejecting samples calculated as errors was 0.8833. Under the current threshold setting, there were no rejecting samples in any of the five test samples, resulting in classification results for four types of abnormal working conditions.

[0129] Figure 10 The diagram illustrates the performance comparison between normal and abnormal binary classification in the first stage of classification.

[0130] refer to Figure 10 As shown in the figure, during the first-stage classification of normal and abnormal root canals, the precision of the abnormal class is 0.980, the recall is 1.000, and the F1 score (the harmonic mean of precision and recall) is 0.990; while the precision of the normal class is 1.000, the recall is 0.917, and the F1 score is 0.957. This demonstrates that the first-stage model can effectively distinguish between normal and abnormal root canal conditions, especially exhibiting a strong detection capability for abnormal samples, making it suitable as a preliminary screening step for subsequent anomaly subdivision and identification.

[0131] Figure 11 The normalized confusion matrix for the first stage is shown schematically.

[0132] refer to Figure 11 As shown, the results of the binary classification of normal and abnormal samples are described. The figure depicts the normalized confusion matrix of the first-stage binary classification. It shows that the proportion of abnormal samples correctly identified as abnormal is 0.980%, and the proportion of normal samples correctly identified as normal is 0.917%. Only a small number of misclassifications exist, with the proportion of abnormal samples misclassified as normal being 0.020% and the proportion of normal samples misclassified as abnormal being 0.083%. These results indicate that the first-stage classification model has good ability to distinguish between normal and abnormal samples, with a low misclassification rate.

[0133] Figure 12 The diagram illustrates the detailed performance comparison of the four abnormal operating conditions in the second stage.

[0134] refer to Figure 12 As shown, the identification performance of the four abnormal conditions during the second stage of detailed classification of abnormal samples is evident. The lateral root canal category performed best, with a precision of 0.923, a recall of 1.000, and an F1 score of 0.960. Accessory root canals and coronal lateral root canals both achieved an F1 score of 0.880. The step category had a relatively low recall of 0.667 and an F1 score of 0.762. These results indicate that the second-stage model can further distinguish different abnormal root canal structures.

[0135] Figure 13 The normalized confusion matrix of the four abnormal operating conditions in the second stage is shown schematically.

[0136] refer to Figure 13 As shown, the normalized confusion matrix for the four types of abnormal conditions in the second stage is presented. The matrix shows that the lateral root canal at the root level has the best identification performance, with a correct identification rate of 1.00; the correct identification rates for accessory root canals and coronal lateral root canals are both 0.92; the correct identification rate for step-type canals is 0.67, and there are cases where they are misidentified as accessory root canals and coronal lateral root canals, with misidentification rates of 0.17 and 0.17, respectively. These results indicate that the second-stage model has a good ability to distinguish most abnormal structures.

[0137] Figure 14 The normalized confusion matrix of the two-stage integrated classification is illustrated schematically.

[0138] refer to Figure 14As shown in the figure, this figure illustrates the final normalized confusion matrix of the two-stage ensemble classification model for five root canal conditions. The coverage rate is 100%, the correct identification rate for lateral root canals is 1.00, the correct identification rates for accessory root canals, coronal lateral root canals, and normal canals are all 0.92, and the correct identification rate for step-like canals is 0.67. The figure shows that the errors are mainly concentrated between step-like canals and accessory root canals, coronal lateral root canals, and a small number of normal samples being classified as abnormal. These results demonstrate that the two-stage ensemble classification method can successfully identify both normal and multiple abnormal conditions and exhibits high classification stability for most root canal conditions.

[0139] The above-described method for enhancing and identifying ultrasonic signals for detecting the state of micro deep holes based on dual FBG, and the detection system for the state of micro deep holes of the present invention can also be applied to the detection of the inner wall morphology (including side holes, cracks, or steps) of materials in engineering fields such as nuclear power, energy and power, aerospace, semiconductor packaging, and precision medical equipment.

[0140] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of this application, and are not intended to limit them. Although this application 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. Such 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 this application.

Claims

1. A system for acquiring ultrasonic signals for micro deep hole state detection based on dual FBG, characterized in that, include: The system includes a laser source, a 1×2 fiber optic coupler, a first fiber optic circulator, a second fiber optic circulator, a sensing FBG, a reference FBG, an acoustic excitation device, and a signal receiving and acquisition module. The acoustic excitation device is used to generate ultrasonic waves on the first side inside the micro-deep hole, and the ultrasonic waves are transmitted to the inner wall of the micro-deep hole. The laser emitted by the laser source passes through a 1×2 fiber coupler and is then incident on port one of the first fiber circulator and port one of the second fiber circulator, respectively. The laser incident on port one of the first fiber optic circulator is transmitted to the sensing FBG via port two, and the laser incident on port one of the second fiber optic circulator is transmitted to the reference FBG via port two. The sensing FBG is placed on the second side inside the micro-deep hole to sense ultrasonic waves. Its reflected light signal is then transmitted to the signal receiving and acquisition module through port two to port three of the first fiber optic circulator. The reference FBG is placed on the second side inside the micro-deep hole to sense ultrasonic waves. Its reflected light signal is then transmitted to the signal receiving and acquisition module through ports two to three of the second fiber optic circulator. The signal receiving and acquisition module is used to perform photoelectric conversion, differential amplification, and analog-to-digital conversion on the two optical signals mentioned above, and output digital signals.

2. The acquisition system according to claim 1, characterized in that, It also includes an isolator, through which the laser emitted by the laser source enters a 1×2 fiber coupler.

3. The acquisition system according to claim 1, characterized in that, The micro-deep hole is filled with hydrogel as an ultrasonic transmission medium, and both the sensing FBG and the reference FBG are placed inside the hydrogel.

4. The acquisition system according to claim 3, characterized in that, The sensing FBG is located in the lower middle part of the micro-deep hole, and the reference FBG is located inside the micro-deep hole near the opening.

5. The acquisition system according to claim 1, characterized in that, The acoustic excitation device is an ultrasonic root canal irrigator.

6. The acquisition system according to claim 1, characterized in that, The signal receiving and acquisition module includes a balanced detector and a data acquisition card; The optical signal passing through port three of the first fiber optic circulator enters the first input terminal of the balanced detector, and the optical signal passing through port three of the second fiber optic circulator enters the second input terminal of the balanced detector. The balanced detector is used to perform photoelectric conversion and differential amplification on two optical signals and then output a differential electrical signal to the data acquisition card.

7. A micro deep hole state detection system based on dual FBG, characterized in that, The system includes the acquisition system for ultrasonic signals for micro deep hole state detection based on dual FBG as described in any one of claims 1-6.

Images