An acoustic-induced fiber grating and adaptive feedback-based hollow-core anti-resonant fiber mode control system and method

Abstract

This invention relates to a mode control system and method for hollow-core anti-resonant fiber based on an acousto-induced fiber grating and adaptive feedback. The system includes: an optical input module for providing incident light; a fiber alignment module for aligning and coupling a multimode transition fiber and a hollow-core anti-resonant fiber; an acousto-optic interaction module for applying mechanical vibration to the hollow-core anti-resonant fiber to form periodic micro-bending perturbations, constructing an acousto-induced fiber grating, and using the acousto-induced fiber grating to couple the incident light introduced into the hollow-core anti-resonant fiber into modes; and a feedback control module for acquiring the mode spot image output after mode coupling, performing mode recognition and analysis, obtaining updated driving signals to change the modulation state of the acousto-induced fiber grating and the output mode distribution, and achieving stable closed-loop control of the target mode proportion. This invention can realize dynamic excitation of higher-order modes in hollow-core anti-resonant fiber and controllable closed-loop optimization control of mode proportion.

Description

Technical Field

[0001] This invention relates to the fields of fiber photonics, hollow anti-resonant fiber, acousto-optic modulation and adaptive mode control, and in particular to a mode control system and method for hollow anti-resonant fiber based on acousto-induced fiber grating and adaptive feedback, which is used to realize the dynamic excitation of higher-order modes and closed-loop optimization control with controllable mode ratio in hollow anti-resonant fiber. Background Technology

[0002] Anti-resonant hollow-core fiber (AR-HCF) confines light within an air core for transmission via an anti-resonant reflection mechanism, offering advantages such as low transmission loss, high damage threshold, low dispersion, and low nonlinearity. It holds significant promise for applications in high-power laser transmission, optical communication, structured light field generation, and mode multiplexing transmission. However, due to the large core size and weak mode confinement capability of AR-HCF, multiple transverse modes often coexist during transmission. If higher-order modes cannot be stably and precisely excited and controlled, it will lead to a degraded output beam quality, reduced transmission efficiency, and system instability. Existing mode control methods often rely on fixed structural devices, static parameter design, or passive adjustment, typically suffering from insufficient dynamic adjustment capabilities, sensitivity to device errors and environmental drift, and difficulty in precisely controlling mode proportions. Acoustically Induced Fiber Grating (AIFG) technology achieves controllable coupling between the fundamental mode and higher-order modes by introducing periodic acoustic field perturbations into the optical fiber. It offers advantages such as reversibility, non-intrusiveness, and dynamic adjustability. However, most existing AIFG mode control schemes still employ open-loop driving methods. Driving frequency, voltage amplitude, and external perturbations can all affect the phase matching state and mode conversion efficiency, making it difficult to achieve stable control of the target mode's proportion. Summary of the Invention

[0003] The purpose of this invention is to provide a hollow-core anti-resonant fiber mode control system and method based on acousto-induced fiber gratings and adaptive feedback. This invention overcomes the shortcomings of existing hollow-core anti-resonant fiber mode control technologies, such as insufficient dynamic control capability, reliance on experience for parameter optimization, sensitivity to environmental disturbances, and difficulty in accurately controlling the mode ratio. This invention achieves controllable coupling between the fundamental mode and higher-order modes by exciting a tunable acousto-induced fiber grating in a hollow-core anti-resonant fiber. Furthermore, by combining mode recognition and SPGD adaptive algorithms, closed-loop optimization of the driving voltage and driving frequency is performed, thereby achieving stable excitation of higher-order modes and programmable control of the target mode ratio.

[0004] To achieve the above objectives, the present invention provides the following solution: A hollow-core anti-resonant fiber mode control system based on acousto-induced fiber grating and adaptive feedback includes: an optical input module, a fiber alignment module, an acousto-optic interaction module, and a feedback control module. The optical input module is used to provide incident light; The fiber alignment module is used to realize the alignment and coupling of the multimode transition fiber and the hollow anti-resonant fiber. After alignment, the incident light enters the hollow anti-resonant fiber through the multimode transition fiber. The acousto-optic module is used to apply mechanical vibration to the hollow anti-resonant fiber to form periodic micro-bending perturbations, construct an acousto-induced fiber grating, and perform mode coupling on the incident light introduced into the hollow anti-resonant fiber through the acousto-induced fiber grating. The feedback control module is used to acquire the pattern spot image output after mode coupling, perform pattern recognition and analysis, obtain the updated driving signal to change the modulation state of the acousto-induced fiber grating, and realize stable closed-loop control of the target mode proportion.

[0005] Optionally, the optical input module includes a narrowband laser, a circulator, and a multimode transition fiber, wherein the incident light output from the narrowband laser is guided into the multimode transition fiber via the circulator.

[0006] Optionally, the acousto-optic module includes: piezoelectric ceramic, energy-concentrating structure, and hollow anti-resonant optical fiber disposed on the substrate; The energy-focusing structure is fixed on a piezoelectric ceramic, which is mounted on a substrate. The energy-focusing structure is positioned at the junction of the uncoated portion and the coated portion of the hollow anti-resonant optical fiber and is fixed with UV adhesive.

[0007] Optionally, constructing an acousto-induced fiber grating includes: The radio frequency signal output by the signal generator is amplified by the signal amplifier and then applied to the piezoelectric ceramic. The piezoelectric ceramic generates mechanical vibration and is coupled to the hollow anti-resonant optical fiber through the energy-gathering structure. In the section where the coating layer is stripped, micro-bending disturbances are formed and periodically distributed along the axial direction, thus constructing the acousto-induced fiber grating.

[0008] Optionally, the feedback control module includes: The mode acquisition unit is used to acquire mode spot images at the output end of the hollow anti-resonant fiber and transmit them to the computer; The computer is used to identify and analyze the pattern spot image, obtain the target pattern ratio, calculate the updated driving signal through the SPGD algorithm, and control the signal generator to output the corresponding updated driving signal. The signal amplifier amplifies the updated driving signal. The piezoelectric ceramic generates mechanical vibration under the action of the amplified driving signal, and further changes the modulation state of the acousto-induced fiber grating to achieve stable closed-loop control of the target pattern ratio. The driving signal includes driving frequency and driving voltage.

[0009] Optionally, stable closed-loop control to achieve the target mode proportion includes: S1. Initialize control parameters; S2. The updated driving signal is calculated using the SPGD algorithm; S3. Change the modulation state of the acousto-induced fiber optic grating according to the updated driving signal, output the current mode spot image for recognition, obtain the current target mode ratio, and construct a performance evaluation function based on the error between the current target mode ratio and the actual mode ratio. S4. Determine whether the performance evaluation function meets the preset threshold or whether the target mode ratio meets the preset requirements. If it meets the requirements, proceed to S5; otherwise, return to S2. S5 outputs the converged drive frequency and drive voltage to achieve stable closed-loop control of the target mode proportion.

[0010] Optionally, the updated drive signals calculated using the SPGD algorithm include: Positive and negative perturbations are added to the current driving frequency to obtain the positive perturbation frequency and the reverse perturbation frequency; The output mode images corresponding to the forward and reverse perturbation frequencies are acquired, and the corresponding performance evaluation function values ​​are obtained through pattern recognition. The change in the first performance function is then calculated. The driving frequency is updated based on the updated gain coefficient of the frequency parameters, the frequency perturbation, and the change in the first performance function to obtain a new driving frequency. At the new driving frequency, positive and negative perturbations are added to the current driving voltage to obtain positive perturbation voltage and reverse perturbation voltage; The output mode images corresponding to the positive and reverse perturbation voltages are acquired, and the corresponding performance evaluation function values ​​are obtained through pattern recognition. The change in the second performance function is then calculated. The driving voltage is updated based on the updated gain coefficient of the voltage parameters, the voltage perturbation, and the change in the second performance function to obtain a new driving voltage.

[0011] This invention also provides a mode control method for hollow-core anti-resonant fiber based on acousto-induced fiber gratings and adaptive feedback, comprising: The incident light is guided into the multimode transition fiber through the optical input module, and then coupled into the hollow anti-resonant fiber through the fiber alignment module. Mechanical vibration is applied to the hollow anti-resonant fiber by an acousto-optic module to form periodic micro-bending perturbations in the fiber, thereby constructing an acousto-induced fiber grating. The incident light introduced into the hollow anti-resonant fiber is then coupled in mode through the acousto-induced fiber grating. The pattern spot image output after pattern coupling is acquired, and pattern recognition and analysis are performed. The updated driving signal is obtained to change the modulation state of the acousto-induced fiber optic grating. The pattern recognition and signal update process is repeated until the preset convergence condition is reached, so as to achieve stable closed-loop control of the target pattern proportion.

[0012] The beneficial effects of this invention are: 1. Mode conversion of hollow-core anti-resonant fiber is achieved by using an acoustic fiber grating, without the need for permanent structural rewriting of the fiber, and has the advantages of being dynamic, reversible and non-invasive; 2. By forming a closed-loop optimization control through pattern recognition and SPGD algorithm, the driving frequency and driving voltage can be adjusted in real time without the need for a precise analytical model, thereby reducing the impact of device manufacturing errors, environmental disturbances and parameter changes on the mode control effect.

[0013] 3. It can not only achieve high-purity excitation of the target's higher-order mode, but also programmable adjustment of the proportion of different preset target modes.

[0014] This invention combines mode conversion, mode ratio adjustment, and parameter self-optimization functions, and can be extended to applications such as hollow-core optical fiber communication, structured light field generation, mode multiplexing transmission, adaptive optical fiber devices, and intelligent photonic systems. Attached Figure Description

[0015] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0016] Figure 1 This is a schematic diagram of a hollow-core anti-resonant fiber mode control system based on acousto-induced fiber grating and adaptive feedback, according to an embodiment of the present invention. Figure 2 This is a schematic diagram of the optical input module according to an embodiment of the present invention; Figure 3 This is a schematic diagram of the fiber alignment module according to an embodiment of the present invention; Figure 4 This is a schematic diagram of the acousto-optic module according to an embodiment of the present invention; Figure 5 This is a schematic diagram of the feedback control module according to an embodiment of the present invention; Figure 6 This is a schematic diagram of the SPGD algorithm flow according to an embodiment of the present invention; Figure 7 The following is a convergence plot of the target value of LP11 under different initial control conditions in this embodiment of the invention, where (a) corresponds to the control condition of increasing initial voltage and increasing initial frequency, (b) corresponds to the control condition of increasing initial voltage and decreasing initial frequency, (c) corresponds to the control condition of decreasing initial voltage and increasing initial frequency, and (d) corresponds to the control condition of decreasing initial voltage and decreasing initial frequency. Figure 8 The diagram shows the convergence results of adaptive feedback control under different target mode proportions in this embodiment of the invention, where (a) corresponds to the convergence result under the frequency decreasing process, and (b) corresponds to the convergence result under the frequency increasing process. Among them, 1 is the optical input module, 2 is the fiber alignment module, 3 is the acousto-optic interaction module, 4 is the feedback control module, 1.1 is the narrowband laser, 1.2 is the circulator, 1.3 is the multimode transition fiber, 2.1 is the fusion splicer, 2.2 is the hollow-core anti-resonant fiber, 2.3 is the first port of the circulator, 2.4 is the second port of the circulator, 2.5 is the third port of the circulator, 2.6 is the fusion splice between the multimode fiber and the second port of the circulator, 3.1 is the acousto-induced fiber grating, 3.2 is the stripped coating section of the multimode fiber, 3.3 is the stripped coating section of the hollow-core anti-resonant fiber, 4.1 is the CCD mode acquisition unit, 4.2 is the computer, 4.3 is the signal generator, 4.4 is the signal amplifier, 4.5 is the aluminum taper, 4.6 is the piezoelectric ceramic, 4.7 is the substrate, 4.8 is the positive electrode of the piezoelectric ceramic, and 4.9 is the negative electrode of the piezoelectric ceramic. Detailed Implementation

[0017] 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, and 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.

[0018] To make the above-mentioned objects, features and advantages of the present invention more apparent and understandable, the present invention will be further described in detail below with reference to the accompanying drawings and specific embodiments.

[0019] This embodiment proposes a hollow-core anti-resonant fiber mode control system based on acousto-induced fiber grating and adaptive feedback, including: an optical input module, a fiber alignment module, an acousto-optic interaction module, and a feedback control module. The optical input module is used to provide incident light; The fiber alignment module is used to realize the alignment and coupling of the multimode transition fiber and the hollow anti-resonant fiber. After alignment, the incident light enters the hollow anti-resonant fiber through the multimode transition fiber. The acousto-optic module is used to apply mechanical vibration to the hollow anti-resonant fiber to form periodic micro-bending perturbations, construct an acousto-induced fiber grating, and perform mode coupling on the incident light introduced into the hollow anti-resonant fiber through the acousto-induced fiber grating. The feedback control module is used to acquire the pattern spot image output after mode coupling, perform pattern recognition and analysis, obtain the updated driving signal to change the modulation state of the acousto-induced fiber grating, and realize stable closed-loop control of the target mode proportion.

[0020] Furthermore, the optical input module includes a narrowband laser, a circulator, and a multimode transition fiber, wherein the incident light output from the narrowband laser is guided into the multimode transition fiber via the circulator.

[0021] Specifically, the optical input module provides incident light to the system, including a narrowband laser, a circulator, and a multimode transition fiber. The incident light output from the laser is guided into the multimode transition fiber via the circulator. The circulator is used to suppress the influence of reflected light on the light source, and the multimode transition fiber is used to improve the mode matching between the incident light field and the hollow anti-resonant fiber, thereby improving subsequent coupling efficiency.

[0022] Furthermore, the fiber alignment module is positioned between the multimode transition fiber and the hollow anti-resonant fiber. This module is used to achieve precise alignment and coupling between the multimode transition fiber and the hollow anti-resonant fiber through a fusion splicer, thereby reducing insertion loss and improving light guiding stability.

[0023] Furthermore, the acousto-optic module includes: piezoelectric ceramic, energy-concentrating structure, and hollow anti-resonant optical fiber disposed on the substrate; The energy-focusing structure is fixed on a piezoelectric ceramic, which is mounted on a substrate. The energy-focusing structure is positioned at the junction of the uncoated portion and the coated portion of the hollow anti-resonant optical fiber and is fixed with UV adhesive.

[0024] Furthermore, constructing an acousto-induced fiber grating includes: The radio frequency signal output by the signal generator is amplified by the signal amplifier and then applied to the piezoelectric ceramic. The piezoelectric ceramic generates mechanical vibration and is coupled to the hollow anti-resonant optical fiber through the energy-gathering structure. In the section where the coating layer is stripped, micro-bending disturbances are formed and periodically distributed along the axial direction, thus constructing the acousto-induced fiber grating.

[0025] Specifically, the acousto-optic module includes a hollow anti-resonant fiber, a piezoelectric ceramic, a focusing structure, and a substrate. After a driving signal is applied to the piezoelectric ceramic, the piezoelectric ceramic generates mechanical vibration. This mechanical vibration is coupled to the hollow anti-resonant fiber via the focusing structure, forming micro-bending perturbations that are periodically distributed along the axial direction in the fiber. This constructs a dynamic acousto-induced fiber grating, enabling coupling and conversion between the fundamental mode and the target higher-order mode.

[0026] Furthermore, the feedback control module includes: The mode acquisition unit is used to acquire mode spot images at the output end of the hollow anti-resonant fiber and transmit them to the computer; The computer is used to identify and analyze the pattern spot image, obtain the target pattern ratio, calculate the updated driving signal through the SPGD algorithm, and control the signal generator to output the corresponding updated driving signal. The signal amplifier amplifies the updated driving signal. The piezoelectric ceramic generates mechanical vibration under the action of the amplified driving signal, and further changes the modulation state of the acousto-induced fiber grating to achieve stable closed-loop control of the target pattern ratio. The driving signal includes driving frequency and driving voltage.

[0027] Furthermore, achieving stable closed-loop control of the target mode proportion includes: S1. Initialize control parameters; S2. The updated driving signal is calculated using the SPGD algorithm; S3. Change the modulation state of the acousto-induced fiber optic grating according to the updated driving signal, output the current mode spot image for recognition, obtain the current target mode ratio, and construct a performance evaluation function based on the error between the current target mode ratio and the actual mode ratio. S4. Determine whether the performance evaluation function meets the preset threshold or whether the target mode ratio meets the preset requirements. If it meets the requirements, proceed to S5; otherwise, return to S2. S5 outputs the converged drive frequency and drive voltage to achieve stable closed-loop control of the target mode proportion.

[0028] Specifically, the feedback control module includes a CCD mode acquisition unit, a computer, a signal generator, and a signal amplifier. The CCD mode acquisition unit is used to acquire mode spot images from the output end of the hollow anti-resonant fiber; the computer is used to identify, process, and analyze the acquired mode images to extract the target mode proportion; the signal generator is used to output a drive signal, and the signal amplifier is used to amplify the drive signal and apply it to the piezoelectric ceramic; based on the mode recognition results, the computer uses a stochastic parallel gradient descent algorithm to adjust the drive frequency and drive voltage to achieve closed-loop control of the target mode proportion.

[0029] Furthermore, the updated drive signals calculated using the SPGD algorithm include: Positive and negative perturbations are added to the current driving frequency to obtain the positive perturbation frequency and the reverse perturbation frequency; The output mode images corresponding to the forward and reverse perturbation frequencies are acquired, and the corresponding performance evaluation function values ​​are obtained through pattern recognition. The change in the first performance function is then calculated. The driving frequency is updated based on the updated gain coefficient of the frequency parameters, the frequency perturbation, and the change in the first performance function to obtain a new driving frequency. At the new driving frequency, positive and negative perturbations are added to the current driving voltage to obtain positive perturbation voltage and reverse perturbation voltage; The output mode images corresponding to the positive and reverse perturbation voltages are acquired, and the corresponding performance evaluation function values ​​are obtained through pattern recognition. The change in the second performance function is then calculated. The driving voltage is updated based on the updated gain coefficient of the voltage parameters, the voltage perturbation, and the change in the second performance function to obtain a new driving voltage.

[0030] This embodiment also provides a hollow-core anti-resonant fiber mode control method based on acousto-induced fiber gratings and adaptive feedback, for use in a control system, including: The incident light is guided into the multimode transition fiber through the optical input module, and then coupled into the hollow anti-resonant fiber through the fiber alignment module. Mechanical vibration is applied to the hollow anti-resonant fiber by an acousto-optic module to form periodic micro-bending perturbations in the fiber, thereby constructing an acousto-induced fiber grating. The incident light introduced into the hollow anti-resonant fiber is then coupled in mode through the acousto-induced fiber grating. The pattern spot image output after pattern coupling is acquired, and pattern recognition and analysis are performed. The updated driving signal is obtained to change the modulation state of the acousto-induced fiber optic grating. The pattern recognition and signal update process is repeated until the preset convergence condition is reached, so as to achieve stable closed-loop control of the target pattern proportion.

[0031] Example 1 like Figure 1 As shown, a hollow-core anti-resonant fiber mode control system based on acousto-induced fiber grating and adaptive feedback includes: an optical input module 1, a fiber alignment module 2, an acousto-optic interaction module 3, and a feedback control module 4. The optical input module 1 provides incident light to the system and includes a narrowband laser 1.1, a circulator 1.2, and a multimode transition fiber 1.3. The fiber alignment module 2, located between the multimode transition fiber 1.3 and the hollow-core anti-resonant fiber 2.2, includes a fusion splicer 2.1 for precise alignment and coupling of the multimode transition fiber 1.3 and the hollow-core anti-resonant fiber 2.2. The acousto-optic interaction module 3 includes an acousto-induced fiber grating 3.1. When a driving signal is applied to the acousto-induced fiber grating 3.1, periodic micro-bending perturbations are formed in the fiber, thereby constructing a dynamic acousto-induced fiber grating and achieving coupling conversion between the fundamental mode and the target higher-order mode. The feedback control module 4 includes a CCD mode acquisition unit 4.1, a computer 4.2, a signal generator 4.3, and a signal amplifier 4.4. The CCD mode acquisition unit 4.1 is used to acquire the mode light field image at the output end of the hollow anti-resonant fiber. The computer 4.2 is used to identify, process and analyze the acquired mode image, and adjust the driving parameters based on the identification results. The signal generator 4.3 is used to output the driving signal, and the signal amplifier 4.4 is used to amplify the driving signal to achieve closed-loop control of the target mode ratio.

[0032] Figure 2 This is a schematic diagram of the optical input module. First, the narrowband laser 1.1 outputs incident light with a wavelength of 1550 nm, which is input through the first port 2.3 of the circulator. The third port 2.5 of the circulator is left unconnected; its function is to suppress the influence of reflected light on the light source, thereby protecting the stable operation of the light source. The second port 2.4 of the circulator is connected to the multimode transition fiber 1.3, which has a specification of 62.5 / 125 μm. Its function is to improve the mode matching between the incident light field and the subsequent hollow-core anti-resonant fiber, preparing for the efficient introduction of light into the hollow-core anti-resonant fiber. The multimode transition fiber 1.3 and the hollow-core anti-resonant fiber 2.2 are connected to the fusion splice 2.6 of the second port of the circulator via the multimode fiber. A refractive index matching fluid is coated in this connection area, which improves optical coupling efficiency and transmission stability, and also helps to reduce the impact of end-face discontinuities on mode transmission.

[0033] Figure 3This is the fiber alignment module. During alignment, the stripped portion (3.2) of the multimode fiber is placed in the left clamping position of the fusion splicer (2.1), and the stripped portion (3.3) of the hollow-core anti-resonant fiber is placed in the right clamping position of the fusion splicer (2.1). The microscopic imaging and displacement adjustment functions of the fusion splicer are used to precisely align the end faces of the two fibers. The cutting angle of the two fibers is controlled within 1.8°, and the distance between the end faces of the two fibers is controlled at 20 μm to ensure connection stability and low insertion loss during coupling.

[0034] Figure 4 This is an acousto-optic module. In implementation, firstly, an aluminum taper 4.5 is fixed to a piezoelectric ceramic 4.6, the operating frequency of which is 100 kHz. Then, the piezoelectric ceramic 4.6 is mounted on a substrate 4.7, preferably a copper sheet. The tip of the aluminum taper 4.5 is positioned at the junction of the uncoated and coated portions of the hollow anti-resonant fiber and is fixed with UV adhesive. The effective length of the coated portion 3.3 of the hollow anti-resonant fiber is preferably 20 cm. After a driving signal is applied to the positive electrode 4.8 and the negative electrode 4.9 of the piezoelectric ceramic, the piezoelectric ceramic 4.6 generates mechanical vibration, which is coupled to the hollow anti-resonant fiber via the aluminum taper 4.5, thereby creating an acoustic disturbance in the coated portion 3.2 of the stripped multimode fiber, used to excite the acousto-optic fiber grating.

[0035] Figure 5 This is a feedback control module. During operation, the CCD mode acquisition unit 4.1 acquires the mode spot image at the output end of the hollow-core anti-resonant fiber and transmits the acquired mode image to the computer 4.2. The computer 4.2 identifies and analyzes the mode image and calculates an updated driving signal using the SPGD algorithm. The driving frequency is used to adjust the phase matching condition of the acousto-optic fiber grating, and the driving voltage is used to adjust the vibration amplitude of the piezoelectric ceramic and the modulation intensity of the acousto-optic fiber grating. The computer 4.2 controls the signal generator 4.3 to output the corresponding driving frequency and driving voltage. The driving signal is amplified by the signal amplifier 4.4 and then applied to the piezoelectric ceramic through the positive electrode 4.8 and the negative electrode 4.9. Under the action of the driving signal, the piezoelectric ceramic generates mechanical vibration, further changing the modulation state of the acousto-optic fiber grating, thereby changing the output mode distribution and achieving closed-loop feedback control of the output mode.

[0036] A mode control method for hollow-core anti-resonant fiber based on acousto-induced fiber gratings and adaptive feedback includes: Step 1: Establish a hollow-core anti-resonant fiber model and determine the phase matching condition: First, the hollow-core anti-resonant fiber used in the experiment was structurally modeled, and the core modes it supported were simulated and analyzed using the finite element method. (LP) 01 Modulus as input mode, with LP11 Using the fundamental mode as the target higher-order mode, the mode field distribution and effective refractive index of the corresponding mode within the working wavelength band are calculated. Based on the effective refractive index difference between the fundamental mode and the target higher-order mode, the beat length relationship between the two modes is established. Furthermore, considering the periodic micro-bending period formed by the acoustic wave in the optical fiber, the phase matching condition satisfied by the mode coupling of the acousto-optic fiber grating is determined. Simulations are used to obtain the correspondence between the driving frequency, the period of the acousto-optic fiber grating, and the resonant mode conversion wavelength, providing a theoretical basis for the parameter setting of the subsequent acousto-optic interaction module and the selection of mode conversion conditions.

[0037] Step 2: Provide incident light through the optical input module: like Figure 1 and Figure 2 As shown, incident light is provided to the system through an optical input module. The optical input module includes a narrowband laser, a circulator, and a multimode transition fiber. Preferably, the narrowband laser outputs incident light with a wavelength of 1550 nm. This incident light is input from port 1 of the circulator, while port 3 of the circulator is left unconnected to suppress the influence of reflected light on the light source and protect its stable operation. Port 2 of the circulator is connected to the multimode transition fiber, preferably a 62.5 / 125 μm fiber, to improve mode matching between the incident light field and the hollow anti-resonant fiber, preparing for the subsequent efficient introduction of light into the hollow anti-resonant fiber.

[0038] Step 3: Achieve light coupling and guidance through the fiber optic alignment module: like Figure 1 and Figure 3 As shown, the multimode transition fiber and the hollow-core antiresonant fiber are precisely aligned and coupled using a fiber alignment module. Specifically, the coated portion of the multimode transition fiber is placed in the left clamping position of the fusion splicer, and the coated portion of the hollow-core antiresonant fiber is placed in the right clamping position. The microscopic imaging and displacement adjustment functions of the fusion splicer are used to precisely align the end faces of the two fibers. Preferably, the cutting angle of the two fiber end faces is controlled within 1.8°, and the distance between the two fiber end faces is controlled at 20μm to ensure connection stability and low insertion loss during coupling. After alignment, the incident light output from the optical input module is coupled into the hollow-core antiresonant fiber through the multimode transition fiber.

[0039] Step 4: Constructing an acousto-induced fiber grating using an acousto-optic interaction module: like Figure 1 and Figure 4As shown, a dynamic acousto-optic fiber grating is constructed in a hollow anti-resonant fiber using an acousto-optic interaction module. The acousto-optic interaction module includes a piezoelectric ceramic, a focusing structure, a substrate, and a hollow anti-resonant fiber disposed thereon. Preferably, the piezoelectric ceramic is a piezoelectric ceramic with an operating frequency of approximately 100 kHz, the focusing structure is preferably an aluminum taper, and the substrate is preferably a copper sheet. In implementation, the aluminum taper is first fixed to the piezoelectric ceramic, and then the piezoelectric ceramic is mounted on the substrate; subsequently, the tip of the aluminum taper is positioned at the junction of the uncoated and coated portions of the hollow anti-resonant fiber and fixed with UV adhesive. Preferably, the working length of the coated portion of the hollow anti-resonant fiber is 20 cm. The radio frequency signal output from the signal generator is amplified by a signal amplifier and applied to the piezoelectric ceramic, causing the piezoelectric ceramic to vibrate mechanically and couple to the hollow anti-resonant fiber via the aluminum taper, thereby forming a periodically distributed micro-bending disturbance along the axial direction in the coated section, thus constructing the dynamic acousto-optic fiber grating.

[0040] Step 5: Implement the coupling transformation between the fundamental mode and the target higher-order mode: After the acoustic fiber grating is constructed, the incident fundamental mode light, introduced into the hollow anti-resonant fiber by the optical input module and fiber alignment module, undergoes mode coupling under the action of the acoustic fiber grating. When the driving frequency satisfies the phase matching condition, the period of the acoustic fiber grating is related to the LP. 01 Model and LP 11 The beat lengths between modes are matched, resulting in the strongest mode coupling and maximizing the proportion of the target higher-order modes. In a representative embodiment, the optimal driving frequency near the 1550nm band is approximately 44.01kHz, corresponding to the effective realization of LP. 01 The coupling conversion from the target mode to the LP11 mode is then performed. Subsequently, the driving voltage is further adjusted near the resonant frequency to examine the variation of the target mode proportion with the driving voltage, thereby determining the optimal mode conversion parameter range.

[0041] Step Six: Perform pattern acquisition and pattern recognition through the feedback control module: like Figure 1 and Figure 5 As shown, the output modes are acquired, identified, and processed through a feedback control module. This feedback control module includes a CCD, a computer, a signal generator, and a signal amplifier. During operation, the CCD acquires mode spot images from the output end of the hollow-core anti-resonant fiber and transmits the acquired mode images to the computer. The computer identifies and analyzes the mode images, extracts the proportion of the target mode, and thus obtains the composition of the current output modes.

[0042] Computers perform pattern recognition and analysis on images, extracting the proportion of target patterns, including: The acquired pattern spot images are processed by grayscale conversion, background noise removal, region cropping, and light intensity normalization to obtain the normalized output light field intensity distribution. Subsequently, a mode decomposition algorithm is used to calculate the weights of different mode components in the output light field, thereby extracting the target mode proportion (LP). 11 (Percentage of patterns).

[0043] Step 7: Construct the objective function and introduce the SPGD algorithm for closed-loop optimization: After obtaining the target mode proportion, an objective function for adaptive closed-loop control is constructed. Preferably, the objective function is defined as the absolute error between the target mode proportion and the actual mode proportion, i.e., minimizing the target mode proportion error is the optimization objective. Subsequently, the computer calls a stochastic parallel gradient descent algorithm to optimize the driving parameters, which include at least the driving frequency and driving voltage. In each iteration, the SPGD algorithm applies small random perturbations to the driving frequency and driving voltage, estimates the local gradient direction based on the change in the objective function before and after the perturbation, and then updates the control parameters. The updated driving frequency and driving voltage are sent to the signal generator and signal amplifier, and applied to the piezoelectric ceramic, thereby changing the modulation state and output mode distribution of the acousto-induced fiber grating.

[0044] like Figure 6 The SPGD algorithm flow shown includes: S01: Initialize control parameters. Set the target mode percentage. and initialize the drive frequency. and driving voltage At the same time, the frequency perturbation amount is set. Voltage perturbation Frequency update gain coefficient and voltage update gain coefficient The driving frequency and driving voltage constitute the control vector to be optimized.

[0045] S02: Add frequency perturbation at the current frequency. By adding positive and negative perturbations to the base, we obtain the positive perturbation frequency and the negative perturbation frequency: ; S03: Apply forward and reverse frequency perturbations respectively and perform pattern recognition to calculate the change in the performance function in the frequency direction. Maintain driving voltage. Unchanged, respectively and The system loads the corresponding output pattern images, acquires them, and obtains the corresponding performance evaluation function values ​​through pattern recognition. and .

[0046] The actual proportion of the target mode under positive perturbation conditions was obtained respectively. and the actual proportion of the target mode under reverse perturbation conditions Subsequently, the actual proportion of the target mode is compared with the preset target mode proportion. By comparing the results, we can obtain the corresponding performance evaluation function values: ; ; Based on the performance evaluation results under forward and reverse perturbations, the change in the performance function is calculated: ; S04: Update the drive frequency. Update the drive frequency according to the SPGD algorithm to obtain the new frequency parameters: ; in, This is the updated gain coefficient for the frequency parameters.

[0047] S05: Add voltage perturbation. At the updated frequency. Under the condition of keeping constant, for the current driving voltage By adding positive and negative perturbations, we obtain the positive perturbation voltage and the reverse perturbation voltage: ; S06: Apply forward and reverse voltage perturbations respectively and perform pattern recognition to calculate the change in the performance function in the voltage direction. Maintain the driving frequency. Unchanged, respectively and The system loads the corresponding output pattern images, acquires them, and obtains the corresponding performance evaluation function values ​​through pattern recognition. and Based on the performance evaluation results under forward and reverse perturbations, the change in the performance function is calculated: ; S07: Update the driving voltage. Update the driving voltage according to the SPGD algorithm to obtain new voltage parameters: in, This is the update gain coefficient for the voltage parameters.

[0048] S08: Load the updated driver parameters into the execution module and determine if the convergence condition is met. Update the driver frequency... and driving voltage The signal is sent to the signal generator and signal amplifier, driving the piezoelectric ceramic to operate, thereby changing the modulation state and output mode distribution of the acousto-induced fiber grating. The current target mode percentage is then obtained. A performance evaluation function is constructed based on the error between the target mode's proportion and the actual mode's proportion. ; Determine whether the updated performance evaluation function meets the preset threshold, or whether the target mode ratio reaches the preset requirement; if it does, proceed to step S09; if it does not, then... Then return to step S02 to continue iterating. This closed-loop optimization process achieves gradual convergence of the driving parameters by repeatedly executing the perturbation-identification-evaluation-update process.

[0049] S09: Output optimization results. Output the converged drive frequency and drive voltage, and maintain the system operating under the corresponding parameters to achieve stable closed-loop control of the target mode proportion. Experimental results show that this closed-loop method can achieve stable convergence under different initial conditions and make the target mode proportion approach the preset value.

[0050] Step 8: Achieve adaptive adjustment under the target mode proportion: After introducing the SPGD algorithm, a single-objective mode proportion closed-loop control experiment was first conducted. LP 11 The target mode ratio is set at 90%. During each iteration, the system calculates the objective function in real time based on the pattern recognition results and updates the driving frequency and driving voltage using the SPGD algorithm. As iterations proceed, the driving parameters gradually converge to the parameter region that satisfies the target mode ratio requirement, and the LP in the output mode is determined. 11 The proportion of the module gradually approaches the preset target value.

[0051] After completing single-objective closed-loop control, the proportions of multiple preset target modes are further adjusted. Preferably, different target mode proportions of 10%, 30%, 50%, and 70% are sequentially input into the control system, and adaptive feedback control is performed during the frequency decrease and frequency increase processes. For each target value, the system repeatedly executes the processes of mode acquisition, mode recognition, objective function calculation, and SPGD parameter update, so that the actual output mode proportion gradually approaches the corresponding target value.

[0052] Through the above implementation steps, this invention organically combines hollow-core anti-resonant fiber, acoustic fiber grating excitation structure, mode recognition, and SPGD adaptive feedback control. Without the need for permanent structural rewriting of the fiber, it realizes dynamic excitation of target high-order modes and programmable adjustment of mode ratio, and is suitable for application scenarios such as hollow-core fiber communication, mode multiplexing transmission, structured light field generation, and adaptive fiber optic devices.

[0053] Figure 7 For different initial control conditions LP 11 Convergence plot with a target value of 90%. Among them, Figure 7(a) Control conditions corresponding to an increase in both initial voltage and initial frequency; Figure 7 (b) Control conditions corresponding to an increase in initial voltage and a decrease in initial frequency; Figure 7 (c) Control conditions corresponding to a decrease in initial voltage and an increase in initial frequency; Figure 7 (d) Control conditions corresponding to decreasing initial voltage and decreasing initial frequency. Under the above four different initial control conditions, although the evolution paths of driving frequency and driving voltage are different, they can all converge to about 90% after pattern recognition and SPGD algorithm closed-loop optimization. This shows that the adaptive feedback control method proposed in this invention can achieve stable adjustment of the target mode under different initial conditions and has good repeatability, convergence and robustness.

[0054] Figure 8 The figure shows the convergence results of adaptive feedback control under different target mode proportions. Adaptive feedback control experiments were conducted under multiple preset target mode proportions, specifically during the frequency decreasing and increasing processes. The results are shown below. Figure 8 (a) and Figure 8 As shown in (b). Wherein, Figure 8 (a) Convergence results under the corresponding frequency decreasing process, Figure 8 (b) Convergence results under the corresponding frequency increase process. The target mode proportions in the two figures are set to 10%, 30%, 50%, and 70%, respectively. Experimental results show that under different target mode proportions, the system can stably converge to the vicinity of the corresponding target value through adaptive feedback control, and the steady-state error is small. This indicates that the mode control method described in this invention can achieve stable adjustment under multiple preset mode proportions.

[0055] The embodiments described above are merely preferred embodiments of the present invention and are not intended to limit the scope of the present invention. Various modifications and improvements made to the technical solutions of the present invention by those skilled in the art without departing from the spirit of the present invention should fall within the protection scope defined by the claims of the present invention.

Claims

1. A hollow-core anti-resonant fiber mode control system based on acousto-induced fiber optic gratings and adaptive feedback, characterized in that, include: Optical input module, fiber optic alignment module, acousto-optic module, and feedback control module; The optical input module is used to provide incident light; The fiber alignment module is used to realize the alignment and coupling of the multimode transition fiber and the hollow anti-resonant fiber. After alignment, the incident light enters the hollow anti-resonant fiber through the multimode transition fiber. The acousto-optic module is used to apply mechanical vibration to the hollow anti-resonant fiber to form periodic micro-bending perturbations, construct an acousto-induced fiber grating, and perform mode coupling on the incident light introduced into the hollow anti-resonant fiber through the acousto-induced fiber grating. The feedback control module is used to acquire the pattern spot image output after mode coupling, perform pattern recognition and analysis, obtain the updated driving signal to change the modulation state of the acousto-induced fiber grating, and realize stable closed-loop control of the target mode proportion.

2. The control system according to claim 1, characterized in that, The optical input module includes a narrowband laser, a circulator, and a multimode transition fiber, wherein the incident light output from the narrowband laser is guided into the multimode transition fiber via the circulator.

3. The control system according to claim 1, characterized in that, The acousto-optic module includes: piezoelectric ceramic, energy-concentrating structure, and hollow anti-resonant optical fiber disposed on the substrate; The energy-focusing structure is fixed on a piezoelectric ceramic, which is mounted on a substrate. The energy-focusing structure is positioned at the junction of the uncoated portion and the coated portion of the hollow anti-resonant optical fiber and is fixed with UV adhesive.

4. The control system according to claim 3, characterized in that, Constructing an acousto-induced fiber grating includes: The radio frequency signal output by the signal generator is amplified by the signal amplifier and then applied to the piezoelectric ceramic. The piezoelectric ceramic generates mechanical vibration and is coupled to the hollow anti-resonant optical fiber through the energy-gathering structure. In the section where the coating layer is stripped, micro-bending disturbances are formed and periodically distributed along the axial direction, thus constructing the acousto-induced fiber grating.

5. The control system according to claim 3, characterized in that, The feedback control module includes: The mode acquisition unit is used to acquire mode spot images at the output end of the hollow anti-resonant fiber and transmit them to the computer; The computer is used to identify and analyze the pattern spot image, obtain the target pattern ratio, calculate the updated driving signal through the SPGD algorithm, and control the signal generator to output the corresponding updated driving signal. The signal amplifier amplifies the updated driving signal. The piezoelectric ceramic generates mechanical vibration under the action of the amplified driving signal, and further changes the modulation state of the acousto-induced fiber grating to achieve stable closed-loop control of the target pattern ratio. The driving signal includes driving frequency and driving voltage.

6. The control system according to claim 5, characterized in that, Achieving stable closed-loop control of the target mode proportion includes: S1. Initialize control parameters; S2. The updated driving signal is calculated using the SPGD algorithm; S3. Change the modulation state of the acousto-induced fiber optic grating according to the updated driving signal, output the current mode spot image for recognition, obtain the current target mode ratio, and construct a performance evaluation function based on the error between the current target mode ratio and the actual mode ratio. S4. Determine whether the performance evaluation function meets the preset threshold or whether the target mode ratio meets the preset requirements. If it meets the requirements, proceed to S5; otherwise, return to S2. S5 outputs the converged drive frequency and drive voltage to achieve stable closed-loop control of the target mode proportion.

7. The control system according to claim 5, characterized in that, The updated driving signals calculated using the SPGD algorithm include: Positive and negative perturbations are added to the current driving frequency to obtain the positive perturbation frequency and the reverse perturbation frequency; The output mode images corresponding to the forward and reverse perturbation frequencies are acquired, and the corresponding performance evaluation function values ​​are obtained through pattern recognition. The change in the first performance function is then calculated. The driving frequency is updated based on the updated gain coefficient of the frequency parameters, the frequency perturbation, and the change in the first performance function to obtain a new driving frequency. At the new driving frequency, positive and negative perturbations are added to the current driving voltage to obtain positive perturbation voltage and reverse perturbation voltage; The output mode images corresponding to the positive and reverse perturbation voltages are acquired, and the corresponding performance evaluation function values ​​are obtained through pattern recognition. The change in the second performance function is then calculated. The driving voltage is updated based on the updated gain coefficient of the voltage parameters, the voltage perturbation, and the change in the second performance function to obtain a new driving voltage.

8. A hollow-core anti-resonant fiber mode control method based on acousto-induced fiber grating and adaptive feedback, used in the control system described in any one of claims 1-7, characterized in that, include: The incident light is guided into the multimode transition fiber through the optical input module, and then coupled into the hollow anti-resonant fiber through the fiber alignment module. Mechanical vibration is applied to the hollow anti-resonant fiber by an acousto-optic module to form periodic micro-bending perturbations in the fiber, thereby constructing an acousto-induced fiber grating. The incident light introduced into the hollow anti-resonant fiber is then coupled in mode through the acousto-induced fiber grating. The pattern spot image output after pattern coupling is acquired, and pattern recognition and analysis are performed. The updated driving signal is obtained to change the modulation state of the acousto-induced fiber optic grating. The pattern recognition and signal update process is repeated until the preset convergence condition is reached, so as to achieve stable closed-loop control of the target pattern proportion.

Images