Optical accelerometer control method based on nonlinear optical coupling

Abstract

The invention relates to an optical accelerometer control method based on nonlinear optical coupling, and the method comprises the steps: building a nonlinear optical coupling module, carrying out the laser initial calibration, and obtaining system parameters; and establishing a blue detuning light field to drive the mechanical oscillator to enter a pre-oscillation-starting state. During operation, the output optical power spectrum of the resonant cavity is acquired in real time, and the laser wavelength and the coupling spacing are dynamically adjusted by using a PID feedback algorithm, so that dynamic optimization adaptation of optical coupling is realized. The system monitors laser input power in real time and intelligently switches working modes according to the laser input power, parameter optical mechanical oscillation is started in a low-power interval to improve sensitivity, and the risk of mode interlocking is avoided by reducing power or adjusting coupling in a high-power interval. And finally, establishing a linear relation between the amplitude of a low-frequency modulation sideband signal generated by a mechanical oscillator in an output optical power spectrum and the external acceleration by resolving the low-frequency modulation sideband signal, so as to accurately calculate the magnitude and the direction of the acceleration. Dynamic adaptation, wide-power-range noise suppression and high-precision segmented solution are effectively realized, and the measurement accuracy is remarkably improved.

Description

Technical Field

[0001] This invention belongs to the field of optomechanical sensing technology, and particularly relates to a control method for an optical accelerometer based on nonlinear optical coupling. Background Technology

[0002] Accelerometers, as high-precision sensors based on the principle of optomechanical coupling for acceleration measurement, have significant application value in aerospace navigation, precision instrument calibration, and geological exploration due to their advantages such as resistance to electromagnetic interference, high sensitivity, and wide dynamic range. With the development of micro-nano fabrication and optoelectronic technologies, silicon-based photonic crystal resonators, possessing high optical quality factor (Q value), strong light field confinement capability, and compatibility with micromechanical structures, have become the core sensing unit of optical accelerometers. Related technology research and development focuses on "optical-mechanical-electronic" coordinated control to improve measurement accuracy and stability. Currently, the technical solutions for silicon-based optical accelerometers mainly revolve around three major stages: laser excitation, optomechanical coupling, and signal processing. Existing research has made some progress: for example, optical field confinement is achieved by designing an air hole array structure for the photonic crystal resonator (e.g., fixing the lattice constant and air hole radius); an optical transmission link is constructed by connecting a trapezoidal silicon waveguide with the photonic crystal waveguide; and a mechanical oscillator is formed by supporting a large mass block with a double-folded cantilever beam to respond to displacement caused by external acceleration. In terms of laser excitation control, a tunable laser outputs a specific wavelength of laser light, and a polarization controller is used to adjust the polarization state to ensure efficient coupling of the laser to the resonator. In terms of signal processing, the change in the output optical power of the resonator is collected to calculate the magnitude of the acceleration. However, existing technologies still have the following key shortcomings in practical applications, making it difficult to meet the requirements for high-precision and high-stability measurements: I. Static limitations of optical coupling structures and control processes: In existing technologies, the coupling parameters of silicon-based photonic crystal resonators, trapezoidal silicon waveguides, and bent photonic crystal waveguides are mostly fixed designs. Coupling efficiency is ensured only through initial assembly, without considering the impact of mechanical oscillator displacement caused by external acceleration on the optical coupling state. When the mechanical oscillator is driven by acceleration and displaces, it causes a change in the cavity length of the photonic crystal resonator, which in turn causes a shift in the optical resonant frequency. The fixed coupling structure cannot dynamically adapt to this shift, leading to a decrease in optical coupling efficiency and even optical field decoupling, which seriously affects measurement accuracy. At the same time, laser wavelength adjustment often adopts a single fixed mode, without building a dynamic feedback mechanism based on real-time changes in the optical resonant frequency, making it difficult to maintain a stable optomechanical coupling state. II. The noise suppression strategies are too singular and lack adaptability: The measurement accuracy of optical accelerometers is affected by multiple sources of noise, including thermal noise, shot noise, and free carrier oscillation noise. Existing noise suppression strategies have significant limitations: Firstly, in low-power laser excitation scenarios, thermal noise is reduced solely by increasing the mass of the mechanical oscillator or optimizing the cantilever beam structure. This fails to utilize parametric optomechanical oscillation modes to enhance optomechanical coupling strength and actively suppress thermal noise, resulting in thermal noise accounting for over 60% of the total noise, thus limiting sensitivity improvement. Secondly, in high-power laser excitation scenarios, the free carrier oscillation mode is active and prone to interlocking with the parametric optomechanical oscillation mode, manifesting as frequency division and superimposed clutter in the spectrum. Existing technologies passively avoid this by reducing laser power, failing to combine this with dynamic adjustment of the waveguide coupling angle to actively avoid mode interlocking, and neglecting the superposition effect of temperature drift on noise. When the temperature deviates from the optimal operating range of 25℃±0.5℃, the optical resonant frequency shifts further, leading to increased noise fluctuations and failing to meet the noise stability requirements of high-precision measurements. III. Frequency band coverage and accuracy deficiencies in acceleration calculation: Current acceleration calculation techniques rely on sensitivity calibration within a single frequency band, currently only calibrating the average sensitivity in the 1kHz or 2-27.6kHz bands, without constructing sensitivity-frequency relationships for the mechanical oscillator response characteristics across different frequency bands. In practical applications, the frequency distribution of external acceleration is wide, and the sensitivity of the mechanical oscillator varies significantly across different frequency bands; for example, the sensitivity at 6.4kHz is more than six times that at 2kHz. Single sensitivity calibration leads to calculation errors exceeding 8% in the mid-to-high frequency band (10-27.6kHz). Furthermore, the lack of error correction mechanisms for parameters such as coupling spacing deviation and temperature drift during the calculation process further amplifies measurement errors, making it difficult to meet the demands of high-precision applications. Therefore, this invention aims to provide a control method for an optical accelerometer based on nonlinear optical coupling, which is an optical accelerometer control method with dynamic optical coupling control, full-power range adaptive noise suppression, and high-precision piecewise calculation capabilities. Summary of the Invention

[0003] The purpose of this invention is to provide a control method for an optical accelerometer based on nonlinear optical coupling, so as to solve the technical problems existing in the prior art.

[0004] To solve the above-mentioned technical problems, the present invention adopts the following technical solution: The control method for an optical accelerometer based on nonlinear optical coupling includes the following steps: S1: Construct a nonlinear optical coupling module, including a silicon-based photonic crystal resonator, a trapezoidal silicon waveguide and a bent photonic crystal waveguide, and integrate a mechanical oscillator structure; S2: Construct a laser excitation module, including a tunable laser, a three-ring polarization controller, and a vacuum fiber optic flange. Perform initial calibration on the polarization and power of the laser excitation module to obtain initial calibration parameters. S3: Blue detuning state establishment: Based on the initial calibration parameters, the output wavelength of the tunable laser is adjusted to the blue detuning range of the optical resonant frequency, and the mechanical oscillator is driven into the pre-oscillation state. S4: Real-time acquisition of the output optical power spectrum of the silicon-based photonic crystal resonator. When external acceleration causes the mechanical oscillator to shift and the optical resonant frequency to deviate, the laser wavelength is adjusted through a PID feedback algorithm, and the coupling distance between the trapezoidal silicon waveguide and the bent photonic crystal waveguide is also adjusted. S5: Real-time acquisition of laser input power. When the laser input power is lower than the low power threshold, the parametric optomechanical oscillation mode is enabled. When the laser input power is higher than the high power threshold, the resonant frequency shift of the free carrier oscillation mode is monitored, and mode interlocking is avoided by reducing the laser power or adjusting the waveguide coupling angle. S6: Acquire the low-frequency modulation signal of the mechanical oscillator in the output optical power spectrum of the silicon-based photonic crystal resonator. Based on the preset sensitivity-frequency relationship, the sensitivity is ≥100mV / g in the 2-10kHz band and ≥660mV / g at 6.4kHz. Establish a linear relationship between the modulation signal amplitude and acceleration. Based on the acquired modulation signal amplitude, calculate the magnitude of the external acceleration in reverse by combining the linear relationship. At the same time, determine the acceleration direction by the in-plane vibration direction of the mechanical oscillator (perpendicular to the air groove of the optical resonator) and output the acceleration measurement result.

[0005] Preferably, in step S1, the silicon-based photonic crystal resonator adopts an air hole array periodic structure, and a defect region is formed by progressive displacement perturbation of the air holes in the central region to achieve optical field confinement. The width of the trapezoidal silicon waveguide input port is adapted to the spot diameter of the tapered fiber lens, and the width of the output port is matched with the defect width of the standard line of the photonic crystal waveguide. The gradient length is 10 times the wavelength of the input light. The bent photonic crystal waveguide forms a transmission path by removing horizontal and 60° oblique air holes, so that the input light is coupled to the silicon-based photonic crystal resonator at a 60° oblique incidence angle. Integrated mechanical oscillator structure: 10 double-folded cantilever beams support a 500nm thick mass block.

[0006] Preferably, the specific process for initial calibration of the polarization and power of the laser excitation module in step S2 is as follows: S21: A laser transmission link is built based on the laser excitation module: The tunable laser outputs laser in a specified band, and the polarization state is adjusted by a three-ring polarization controller with a specified fiber turns ratio to cover the full polarization range of the Poincaré sphere. The laser is then transmitted to the input end of the trapezoidal silicon waveguide through a vacuum fiber flange. S22: Adjust the laser power to 3dBm and traverse the polarization state through the three-ring polarization controller; S23: Monitor the optical transmission spectrum of the silicon-based photonic crystal resonator in real time. When a symmetrical Lorentz-type concave peak appears, lock the polarization state parameters and laser power at this time to complete the initial calibration.

[0007] Preferably, the specific process of step S3 is as follows: S31: Based on step S23, retrieve the locked specified parameters, including optical resonant frequency, optimal polarization state parameters, and initial laser power; S32: Frequency difference conversion: Calculate the laser frequency adjustment range based on the definition of blue detuning; S33: Adjust the output wavelength from the initial calibrated resonant wavelength to a shorter wavelength by a specified step; S34: Preset photon number calculation formula: The relationship between the intracavity photon number N and the laser power P and cavity lifetime τ is as follows: N=P×τ / h ×ω l , h Let be Planck's constant. oh l The laser frequency; S35: If the number of photons in the cavity is less than 386 under the initial laser power, the laser power is increased in increments of 0.1dBm. Each time the power is increased, the number of photons is calculated using the above photon count formula. When the number of photons reaches 386, the laser power is locked. S36: Pre-start oscillation characteristic judgment: Collect the vibration signal of the mechanical oscillator. If a weak sine wave signal appears near the mechanical resonant frequency of 70.3kHz and the signal frequency is stable, it is determined that the mechanical oscillator has been driven to the pre-start oscillation state by the intracavity optical field. Preferably, the specific process of adjusting the laser wavelength using a PID feedback algorithm in step S4, while simultaneously adjusting the coupling distance between the trapezoidal silicon waveguide and the bent photonic crystal waveguide, is as follows: S41: Extract specified core reference values ​​based on the established blue detuning state, including laser wavelength reference, coupling distance reference, and output optical power reference; S42: Real-time acquisition of deviation signal: Optical signal acquisition and conversion: The output light of the silicon-based photonic crystal resonator is received in real time, converted into an electrical signal, then into a digital signal, and the digital signal is filtered to extract the real-time value and fluctuation characteristics of the output optical power. At the same time, the offset of the optical resonant frequency is monitored through spectrum analysis. Calculate the deviation value: Laser wavelength deviation Δλ: Using the reference wavelength as the target value, the monitored resonant frequency offset is converted into wavelength deviation according to the frequency-wavelength conversion relationship λ=c / ω; Coupling spacing deviation Δ d : Using the reference coupling spacing as the target value d 0 According to the transmittance-coupling spacing correlation formula T =T 0 −k·d T real-time transmittance T 0 Standard transmittance, k The transmittance attenuation coefficient is... d The real-time coupling spacing is calculated using real-time transmittance. d This leads to the spacing deviation Δ d= | d−d 0 | ; S43: Set parameters for laser wavelength and coupling distance separately and calculate them independently, and set up dual-parameter linkage logic: Priority setting: Laser wavelength adjustment takes precedence over coupling distance adjustment. When the wavelength deviation is >0.01nm, wavelength adjustment is performed first. After the wavelength deviation is ≤0.01nm, coupling distance adjustment is performed. S44: Actuator response adjustment achieves parameter correction: Laser wavelength adjustment execution: The position of the laser cavity mirror is finely adjusted through the internal piezoelectric ceramic driving mechanism to correct the output wavelength from the current value to the target value; Coupling spacing adjustment: The displacement stage drives the trapezoidal silicon waveguide to move along the direction perpendicular to the waveguide transmission through a stepper motor, thereby correcting the coupling spacing.

[0008] Preferably, the specific process of step S5 is as follows: S5: Real-time acquisition of laser input power. When the laser input power is lower than the low power threshold, the parametric optomechanical oscillation mode is enabled. When the laser input power is higher than the high power threshold, the resonant frequency shift of the free carrier oscillation mode is monitored, and mode interlocking is avoided by reducing the laser power or adjusting the waveguide coupling angle. S51: The tunable laser has a built-in power monitoring unit that collects the laser input power and output optical power three times per second and performs smoothing filtering. Oscillation mode identification: In low-power scenarios, i.e., laser input power < 5dBm, the parametric optomechanical oscillation mode is enabled; in high-power scenarios, i.e., laser input power > 8dBm, it is determined that there is a risk of mode interlocking. S52: Enabling parametric optomechanical oscillation mode to suppress thermal noise in low-power scenarios: Mode triggering and parameter configuration: Configure mode parameters based on the current laser input power, lock the driving signal frequency of the mechanical oscillator to its mechanical resonant frequency; synchronously adjust the optical field distribution of the silicon-based photonic crystal resonator, and concentrate the optical field energy in the cavity to the defect area by fine-tuning the three-ring polarization controller; S53: Avoiding interlocking of free carrier oscillation modes under high power: Prioritize laser power adjustment: If a risk of mode interlock is detected, reduce the laser input power in 0.2dBm steps; Each time the power is reduced, observe the change in clutter using a spectrum analyzer: if the clutter amplitude decreases by ≥50%, continue to fine-tune the power to the 8.0-8.5dBm range; if there is no significant change in clutter after reducing the power by 0.5dBm, stop power adjustment and switch to waveguide coupling angle adjustment.

[0009] Preferably, step S5 further includes a process for noise suppression through temperature compensation, as detailed below: S54: Set the target operating temperature to 25±0.5℃, the temperature-wavelength correction coefficient to 0.01nm / ℃, the heater power adjustment range to 1-5W, and the noise control target: after temperature compensation, the total noise fluctuation ≤±5%; S55: Temperature Data Acquisition and Filtering The current operating temperature is collected 10 times per second. The collected temperature values ​​are then filtered by moving average to eliminate instantaneous temperature fluctuations and obtain a stable current temperature value T1. Temperature deviation and correction calculation: Calculate the temperature deviation ΔT. If |ΔT|>0.5℃, trigger the temperature compensation process. Calculate the laser wavelength correction Δλ: According to the correction coefficient, Δλ = ΔT × 0.01 nm / ℃; Predict the direction of heater power adjustment: If T1 < 24.5℃, increase the heater power; if T1 > 25.5℃, decrease the heater power. S56: Set a two-dimensional adjustment strategy of heater power regulation + laser wavelength correction: Heater power adjustment: Heater power is adjusted based on the absolute value of the temperature deviation ΔT: when |ΔT| = 0.6-1.0℃, the heater power is adjusted to 2-3W; when |ΔT| > 1.0℃, the heater power is adjusted to 4-5W. Continuously monitor temperature changes until the current temperature T1 returns to the range of 25±0.5℃. At this point, lock the heater power and enter the laser wavelength correction stage. S57: Laser wavelength correction: The calculated wavelength correction Δλ is converted into a control command and sent to the tunable laser. After receiving a command, the tunable laser will correct its output wavelength from the current value to the target value.

[0010] Preferably, the specific process of step S6 is as follows: S61: Set core reference parameters, including target signal frequency band, segmented sensitivity standard, signal amplitude-acceleration correspondence and direction determination basis; S62: Receives optical signals, converts them into corresponding electrical signals, and amplifies the amplitude of the electrical signals to 0-5V through a signal amplification module. The spectrum analyzer performs spectrum analysis on the amplified electrical signals to generate a power spectrum diagram of the 2-27.6kHz frequency band and extracts the low-frequency modulation signals in this frequency band. S63: Digital filtering of low-frequency modulated signals: Butterworth low-pass filter is used to remove high-frequency noise, 50Hz notch filter is used to eliminate power frequency interference, and the filtered signal is smoothed to obtain a smooth frequency-amplitude curve. Analyze the smoothed curve to identify the frequency band with the largest signal amplitude; if the signal covers multiple sub-frequency bands, extract the maximum amplitude of each sub-frequency band, and then calculate the corresponding frequency band sensitivity separately, finally taking the average value as the acceleration result; S64: Precise amplitude extraction: For the determined main frequency band, read the maximum amplitude of the signal within that frequency band; If it is a multi-band signal, extract the maximum effective amplitude of the 2-5kHz and 5-10kHz sub-bands respectively, and record the center frequency of each sub-band; Sensitivity-frequency relationship matching: Recalls pre-stored sensitivity-frequency relationships and finds the sensitivity value for the corresponding frequency band based on the signal's main / sub-band frequencies. If the main frequency band is 6.4kHz, directly match the sensitivity at 6.4kHz in the curve; If the sub-band is 3.5kHz, the sensitivity at 3.5kHz in the matching curve; If the signal frequency is between the two calibration values, the corresponding sensitivity is calculated using linear interpolation. S65: Acceleration magnitude calculation: Frequency band signals: according to the formula a = Calculated by effective amplitude V / corresponding sensitivity; For multi-band signals: calculate the acceleration value corresponding to each sub-band separately, and then calculate the average value; Acceleration direction determination: Receive the vibration direction monitoring signal of the mechanical oscillator, which reflects the vibration direction of the mechanical oscillator; By combining the structural orientation of the silicon-based photonic crystal resonator, the vibration direction is converted into the acceleration direction.

[0011] The beneficial effects of this invention include: 1. By adjusting the laser wavelength in real time through a PID feedback algorithm and simultaneously linking a vacuum electric displacement stage to correct the coupling distance between the trapezoidal silicon waveguide and the bent photonic crystal waveguide, the problem of resonant frequency shift caused by the inability of fixed coupling parameters to adapt to acceleration in existing technologies is solved. This ensures a stable bilateral transmittance of ≥80% for the "silicon waveguide-photonic crystal waveguide-resonant cavity" system, avoiding measurement distortion caused by decreased coupling efficiency. The dynamic response time in the optical coupling state is ≤100ms, matching the response frequency band requirements of 2-27.6kHz acceleration.

[0012] 2. Based on the initial calibration parameters, the laser wavelength is adjusted to the blue detuning range. The number of photons is precisely controlled to be ≥386 using the intracavity photon count calculation formula, driving the mechanical oscillator to stably enter the pre-oscillation state. Compared with the fuzzy control of blue detuning in existing technologies, which only sets an undefined range of blue detuning, this solution can ensure the optomechanical coupling rate, providing a stable energy foundation for subsequent acceleration response and avoiding coupling failure due to insufficient photon count.

[0013] 3. For low-power scenarios with laser power < 5dBm, the parametric optomechanical oscillation mode is enabled to improve the mechanical quality factor and thermal noise suppression rate, solving the problem of excessive thermal noise ratio at low power in existing technologies; for high-power scenarios with power > 8dBm, the power is reduced in 0.2dBm steps or the waveguide coupling angle is adjusted by ±5° to reduce the mode interlocking rate and avoid interlocking interference between free carrier oscillation and parametric optomechanical oscillation.

[0014] 4. Real-time temperature monitoring: When the temperature deviates from 25±0.5℃, the heater power is adjusted by 1-5W, and the laser wavelength is simultaneously corrected by 0.01nm / ℃ to reduce overall noise fluctuation. This dual-dimensional compensation strategy fills the gap in existing technologies that ignore the effect of temperature on noise superposition, ensuring the measurement stability of the sensor under varying ambient temperature conditions.

[0015] 5. The 2-27.6kHz frequency band is subdivided into sub-bands such as 2-5kHz and 5-10kHz, with preset independent sensitivity frequency relationships for each band. Compared to existing technologies with single sensitivity calibration, the calculation error in the mid-to-high frequency band (10-27.6kHz) is reduced. Simultaneously, linear interpolation is used to calculate sensitivity at the transition frequency (8.5kHz) between bands, ensuring the continuity and accuracy of the calculation across the entire frequency band. Attached Figure Description

[0016] Figure 1 This is a flowchart illustrating the control method of the optical accelerometer based on nonlinear optical coupling according to the present invention.

[0017] Figure 2 The curves showing the relationship between mechanical sensitivity and frequency under different mechanical Q values ​​according to the present invention are shown.

[0018] Figure 3 The curves showing the relationship between mechanical sensitivity and frequency under different mechanical frequencies according to the present invention are shown. Detailed Implementation

[0019] The following is in conjunction with the appendix Figure 1~Figure 3 The present invention will be further described in detail below: Example 1 See appendix Figure 1 As shown, the control method for a light accelerometer based on nonlinear optical coupling includes the following steps: S1: Construct a nonlinear optical coupling module, which includes a silicon-based photonic crystal resonator, a trapezoidal silicon waveguide and a bent photonic crystal waveguide, and integrates a mechanical oscillator structure.

[0020] The silicon-based photonic crystal resonator adopts an air-hole array periodic structure with a lattice constant of 505 nm and an air-hole radius of 187.5 nm. By progressively displacing the air holes in the central region to create a defect region, optical field confinement is achieved. The displacement scales of the progressive displacement perturbation are 15 nm, 10 nm, and 5 nm, so that the optical resonant frequency is stabilized at 193.38 THz, corresponding to a wavelength of 1549.8 nm, ensuring that the working wavelength band of 1500-1600 nm falls within the photonic bandgap of 135.13-213.6 THz.

[0021] The trapezoidal silicon waveguide has an input port width of 2μm, adapted to a tapered fiber lens with a spot diameter of 2±0.5μm. The output port width is set to 874.66nm, matching the defect width of the photonic crystal waveguide standard line. The gradient length is 10 times the input wavelength (15.5μm) to ensure a transmittance ≥99%. The bent photonic crystal waveguide forms a transmission path by removing horizontal and 60° oblique air holes, allowing the input light to couple to the silicon-based photonic crystal resonator at a 60° oblique incidence angle, ensuring a bilateral transmittance ≥80.5% for the "silicon waveguide-photonic crystal waveguide-optical resonator" structure.

[0022] Integrated mechanical oscillator structure: 10 double-folded cantilever beams support a 500nm thick mass block with dimensions of 1.065mm×1mm. The middle long arm of the double-folded cantilever beam has dimensions of 120μm×19μm, and the two side long arms have dimensions of 115μm×7μm, so that the resonant frequency of the mechanical oscillator is stabilized at 70.3kHz, forming an optomechanical coupling with the optical resonant cavity, with a coupling rate ≥31.4GHz / nm.

[0023] S2: Construct the laser excitation module, which includes a tunable laser, a three-ring polarization controller, and a vacuum fiber optic flange. Perform initial calibration on the polarization and power of the laser excitation module to obtain initial calibration parameters. The laser link is: tunable laser → three-ring polarization controller → vacuum fiber optic flange → trapezoidal silicon waveguide.

[0024] S3: Blue detuning state establishment: Based on the initial calibration parameters, adjust the output wavelength of the tunable laser to the blue detuning range of the optical resonant frequency, i.e., the blue detuning amount Δ= oh l - oh opt =5-10GHz, so that the number of photons in the cavity is ≥386, corresponding to an energy of 45aJ, driving the mechanical oscillator into the pre-oscillation state.

[0025] S4: Real-time acquisition of the output optical power spectrum of the silicon-based photonic crystal resonator. When external acceleration causes the mechanical oscillator to shift and the optical resonant frequency to deviate, the laser wavelength is adjusted through a PID feedback algorithm, while the coupling distance between the trapezoidal silicon waveguide and the bent photonic crystal waveguide is also adjusted.

[0026] S5: Real-time acquisition of laser input power. When the laser input power is lower than the low power threshold of 5dBm, the parametric optomechanical oscillation mode is enabled to suppress thermal noise by enhancing the optomechanical coupling strength. When the laser input power is higher than the high power threshold of 8dBm, the resonant frequency shift of the free carrier oscillation mode is monitored. Mode interlocking is avoided by reducing the laser power or adjusting the waveguide coupling angle. The adjustment range of the waveguide coupling angle is ±5°.

[0027] S6: Acquire the low-frequency modulation signal of the mechanical oscillator in the output optical power spectrum of the silicon-based photonic crystal resonator. Based on the preset sensitivity-frequency relationship, the sensitivity is ≥100mV / g in the 2-10kHz band and ≥660mV / g at 6.4kHz. Establish a linear relationship between the modulation signal amplitude and acceleration. Based on the acquired modulation signal amplitude, calculate the magnitude of the external acceleration in reverse by combining the linear relationship. At the same time, determine the acceleration direction by the in-plane vibration direction of the mechanical oscillator (perpendicular to the air groove of the optical resonator) and output the acceleration measurement result.

[0028] In this embodiment, the specific process of initial calibration of the polarization and power of the laser excitation module in step S2 is as follows: S21: A laser transmission link is built based on the laser excitation module: A TSL550C tunable laser is used to output laser light in the 1480-1630nm band. The polarization state is adjusted by a three-ring polarization controller with a fiber turns ratio of 2:4:2 to cover the full polarization range of the Poincaré sphere. The laser light is then transmitted to the input end of the trapezoidal silicon waveguide through a four-channel bare fiber through-vacuum fiber flange.

[0029] S22: Adjust the laser power to 3dBm and traverse the polarization state through the three-ring polarization controller; S23: Monitor the optical transmission spectrum of the silicon-based photonic crystal resonator in real time. When a symmetrical Lorentz-type concave peak appears (optical quality factor ≥14575), lock the polarization state parameters and laser power at this time to complete the initial calibration.

[0030] Example 2 Based on Example 1, the specific process of step S3 is as follows: S31: Retrieve the specified parameters for locking based on step S23, including: Optical resonant frequency oh opt The inherent resonant frequency of the silicon-based photonic crystal resonator is determined by cavity structure parameters including a lattice constant of 505 nm, an air hole radius of 187.5 nm, and a central defect region displacement of 15 nm / 10 nm / 5 nm. The calibration value is 193.38 THz, corresponding to a wavelength of 1549.8 nm.

[0031] Optimal polarization state parameters: The angle combination of the three-ring polarization controller with fiber turns ratio of 2:4:2 is set to fast axis rotation angle θ=45° and polarization state ellipticity correlation angle φ=90° to ensure the highest laser coupling efficiency. At this time, the Lorentz concave peak of the optical transmission spectrum is most significant, and the quality factor Q value is ≥14575.

[0032] Initial laser power: The calibrated 3dBm output power, which is the reference power before noise suppression, to avoid power fluctuations affecting detuning.

[0033] S32: Frequency difference conversion: According to the definition of blue detuning, Δ= oh l - oh opt Δ is the blue detuning quantity, ω l The laser frequency, oh opt It is the optical resonant frequency. oh opt =193.38THz, calculate the laser frequency adjustment range: When Δ=5GHz, oh l =193.38THz + 5GHz = 193385MHz; When Δ=10GHz oh l = 193.38THz + 10GHz = 193390MHz; Then through the wavelength-frequency relationship λ=c / ω l, c The speed of light is 3×10 8 m / s, which translates to a laser wavelength adjustment range of 1549.72nm-1549.76nm, ensuring that the wavelength falls within the 1500-1600nm working band corresponding to the photonic crystal band gap of 135.13-213.6THz.

[0034] S33: In 0.01nm increments, the output wavelength is adjusted from the initial calibrated resonant wavelength of 1549.8nm to a shorter wavelength. Blue detuning corresponds to a laser wavelength shorter than the resonant wavelength.

[0035] The output photocurrent of the silicon-based photonic crystal resonator is monitored synchronously using a PDB440C balanced photodetector. When the photocurrent drops to 60%-70% of the initial resonant state, which is the typical light intensity characteristic of the blue detuned state, the adjustment is paused, and the laser wavelength and frequency at this time are recorded.

[0036] The monitored laser frequency oh l Substitute Δ= oh l - oh opt Calculate the actual detuning amount. If it is not within the 5-10GHz range, repeat the above adjustment steps, fine-tuning by 0.005nm each time, until Δ stabilizes in the target range with an error ≤0.2GHz. This avoids insufficient optomechanical coupling strength (<31.4GHz / nm) due to excessive detuning amount, or a sudden drop in the number of photons in the cavity due to excessive detuning amount.

[0037] S34: Preset photon number calculation logic: Based on the energy requirements of optomechanical coupling, the relationship between the intracavity photon number N and the laser power P and cavity lifetime τ is as follows: N=P×τ / h×ω l , h τ is Planck's constant, where the cavity lifetime τ of a silicon-based photonic crystal resonator is determined by the mechanical Q value (≥14575) and the optical resonant frequency. oh opt The calculation yields τ = Q / oh opt ≈14575 / (2π×193380×10 6 )≈1.2×10 -8 s.

[0038] S35: If the number of photons in the cavity is less than 386 at the initial power of 3dBm, the laser power is increased in increments of 0.1dBm. Each time the power is increased, the number of photons is calculated using the formula above.

[0039] When the photon number reaches 386, the corresponding energy is: E=N×h× oh l≈386×6.626×10 -34 ×2π×193380×10 6 ≈45aJ; With the laser power locked, the energy of the optical field inside the cavity is sufficient to effectively couple with the mechanical oscillator, satisfying the energy threshold of optomechanical coupling rate ≥ 31.4 GHz / nm.

[0040] The laser power feedback loop is activated to monitor the output optical power in real time. If the power fluctuation exceeds ±2%, the drive current of the laser is automatically adjusted to ensure that the number of photons in the cavity is stable within the range of 386-400, so as to avoid the instability of the mechanical oscillator pre-oscillation state caused by the fluctuation of the number of photons.

[0041] S36: Pre-oscillation characteristic judgment: The vibration signal of the mechanical oscillator is collected by the KEYSIGHT N9010B EXA spectrum analyzer. If a weak sine wave signal with an amplitude of 10-20mV appears near the mechanical resonant frequency of 70.3kHz, which is the typical signal strength of pre-oscillation, and the signal frequency is stable with fluctuation ≤0.1kHz, it indicates that the mechanical oscillator has been driven to the pre-oscillation state by the intracavity optical field.

[0042] Simultaneously observe the output optical power spectrum of the silicon-based photonic crystal resonator. If a sideband signal of 70.3 kHz appears, which is the characteristic sideband of optomechanical coupling, it further verifies that the pre-oscillation state is effective. The sideband amplitude is positively correlated with the number of photons in the cavity, and the sideband amplitude corresponding to 45 aJ energy is ≥5 mV.

[0043] After confirming the pre-start-up state, the wavelength of the tunable laser is locked within the blue detuning range, the power is maintained at a photon number ≥ 386, and the angle parameters of the three-ring polarization controller are stored in the register of the FPGA controller as reference parameters for subsequent dynamic feedback adjustment.

[0044] Record the vacuum level, ambient temperature (25±0.5℃), and other environmental parameters at this time to avoid subsequent environmental changes affecting the stability of the blue detuning state. See Table 1 below for parameter recording during the blue detuning state establishment process: The specific process of adjusting the laser wavelength using the PID feedback algorithm in step S4, while simultaneously adjusting the coupling distance between the trapezoidal silicon waveguide and the bent photonic crystal waveguide, is as follows: S41: Extract specified core baseline values ​​based on the established blue detuning state, including: Laser wavelength reference: The wavelength locked within the blue detuning range, such as 1549.74nm, corresponding to Δ=7.5GHz; Coupling spacing reference: The initial coupling spacing between the trapezoidal silicon waveguide and the bent photonic crystal waveguide is 874.66 nm, ensuring bilateral transmittance ≥80%; Output optical power reference: The output optical power at blue detuning stability is 2.8 dBm, corresponding to 386 photons in the cavity; S42: Real-time acquisition of deviation signal: Optical signal acquisition and conversion: The PDB440C photodetector receives the output light from the silicon-based photonic crystal resonator in real time, converts the optical signal into an electrical signal (current range 0-10mA), and then converts the analog electrical signal into a digital signal, which is transmitted to the FPGA controller. The FPGA controller filters the digital signal using a 50Hz notch filter to remove power frequency interference, extracts the real-time value and fluctuation characteristics of the output optical power, and monitors the shift of the optical resonant frequency through spectrum analysis. For example, due to acceleration causing changes in cavity length, the resonant frequency shifts from 193.38THz to 193.378THz.

[0045] Calculate the deviation value: Laser wavelength deviation Δλ: Using the reference wavelength as the target value, the monitored resonant frequency offset is converted into wavelength deviation according to the frequency-wavelength conversion relationship λ=c / ω. For example, a frequency offset of 2GHz corresponds to a wavelength deviation of approximately 0.016nm, i.e., Δλ=+0.016nm.

[0046] Coupling spacing deviation Δ d : Using the reference coupling spacing as the target value d 0 According to the transmittance-coupling spacing correlation formula T =T 0 −k•d , T For real-time transmittance, T 0 Standard transmittance, k The transmittance attenuation coefficient is... d The real-time coupling spacing is calculated using real-time transmittance. d This leads to the spacing deviation Δ d =| d - d 0 |

[0047] S43: The FPGA controller has a built-in PID algorithm adapted to the optomechanical system, which sets parameters for laser wavelength and coupling distance and calculates them independently, and sets up dual-parameter linkage logic: Priority setting: Since the laser wavelength directly affects the blue detuning state and the spacing affects the transmittance, laser wavelength adjustment takes precedence over coupling spacing adjustment. When the wavelength deviation is >0.01nm, wavelength adjustment is performed first, and coupling spacing adjustment is performed only after the wavelength deviation is ≤0.01nm.

[0048] Synchronous verification: After each adjustment, the FPGA controller needs to monitor both wavelength deviation and spacing deviation simultaneously. Only when both are ≤5% of the target deviation, i.e. wavelength ≤0.0008nm and spacing ≤0.5nm, will the current adjustment cycle be stopped and the next round of signal acquisition begin.

[0049] S44: Actuator response adjustment (to achieve parameter correction): Laser wavelength adjustment execution: The FPGA controller sends wavelength adjustment commands to the wavelength control module of the TSL550C tunable laser via RS485 communication protocol. After receiving the command, the laser fine-tunes the position of the laser cavity mirror through its internal piezoelectric ceramic drive mechanism, correcting the output wavelength from the current value of 1549.756nm to the target value of 1549.7506nm. The adjustment process takes ≤100ms, ensuring that the dynamic response speed matches the acceleration change, with an acceleration response frequency band of 2-27.6kHz.

[0050] Coupling spacing adjustment execution: The FPGA controller sends the spacing adjustment command to the drive module of the vacuum electric displacement stage. The displacement stage drives the trapezoidal silicon waveguide to move in a direction perpendicular to the waveguide transmission direction through a stepper motor. For example, it moves 2.103nm closer to the bent photonic crystal waveguide to correct the coupling spacing. After adjustment, the actual displacement is fed back through the grating ruler built into the displacement stage with an accuracy of ±5nm, ensuring that the adjustment error is ≤10% of the command value.

[0051] Validation of adjustment effects and closed-loop iteration: Real-time verification metrics: Blue detuning verification: The difference Δ between the laser frequency and the optical resonant frequency is monitored by a spectrum analyzer to ensure that Δ remains stable in the 5-10GHz range with an error ≤0.2GHz.

[0052] Transmittance verification: Transmittance is calculated by converting the output optical power to ensure that the transmittance is ≥80%. For example, the transmittance recovers from 78% to 81.2% after adjustment.

[0053] Photon count verification: Calculate the number of photons in the cavity based on the laser power and cavity lifetime, and ensure that the number of photons is ≥386. If the calculated value is 392, it meets the optomechanical coupling threshold.

[0054] Closed-loop iterative logic: If all the above indicators meet the standards, the FPGA controller maintains the current adjustment parameters and enters the next round of signal acquisition. If any indicator fails to meet the standards, such as Δ=10.5GHz exceeding the range, the deviation value is recalculated, and the PID calculation → adjustment → effect verification process is repeated until all indicators return to the target range, thus realizing dynamic closed-loop control.

[0055] Example 3 Based on Example 1 or Example 2, the specific process of step S5 is as follows: Start-up power monitoring module: The TSL550C tunable laser has a built-in power monitoring unit and a PDB440C balanced photodetector to assist in verifying the output optical power, ensuring smooth real-time monitoring of input / output power; Initialization adjustment actuators: Parameter optomechanical oscillation mode trigger module (integrated into FPGA controller), laser power adjustment unit (drive current adjustment range 0-500mA), vacuum electric displacement stage (with angle adjustment function for fine adjustment of waveguide coupling angle).

[0056] S51: Real-time power acquisition: The laser's built-in power monitoring unit acquires the laser input power value 3 times per second and transmits it to the FPGA controller; at the same time, the PDB440C photodetector acquires the output optical power. The input power and output power are linearly correlated, and the proportionality coefficient is determined by the previous calibration and can be 0.85. The input power is cross-validated to avoid single monitoring errors.

[0057] The FPGA controller performs smoothing filtering on the acquired power value (using a 5-point moving average) to eliminate instantaneous fluctuation interference, such as random fluctuations of ±0.2dBm, and determines the current stable power value, such as 4.8dBm or 8.5dBm.

[0058] Oscillation mode identification: Low power scenario (<5dBm): Monitor the vibration spectrum of the mechanical oscillator. If thermal noise causes the signal-to-noise ratio of the vibration signal to be <20dB, it is determined that the parametric optomechanical oscillation mode needs to be enabled. High power scenario (>8dBm): Monitor the spectrum. If the free carrier oscillation peak and the parametric optomechanical oscillation peak overlap / divide, such as the appearance of 68kHz or 72kHz clutter peaks next to the 70.3kHz mechanical resonance peak, it is determined that there is a risk of mode interlocking.

[0059] S52: Enables parametric optomechanical oscillation mode at low power to suppress thermal noise. Mode triggering and parameter configuration: The FPGA controller sends a trigger command to the parameter optomechanical oscillation module and configures the mode parameters based on the current laser input power of 4.2dBm: the drive signal frequency of the mechanical oscillator is locked to its mechanical resonant frequency of 70.3±0.5kHz (the inherent resonant frequency of the mechanical oscillator). 4.2dBm corresponds to an amplitude of 15mV, ensuring that the oscillation intensity matches the power.

[0060] Synchronous adjustment of the optical field distribution of silicon-based photonic crystal resonator: By finely adjusting the three-ring polarization controller with an angle step of 1°, the optical field energy in the cavity is concentrated in the defect area, and the uniformity of the optical field distribution is improved to more than 90%, thereby enhancing the interaction between light and mechanical oscillators.

[0061] Enhancement and Verification of Optomechanical Coupling Strength: Once the mode is enabled, monitor the optomechanical coupling strength. If the coupling strength is <31.4GHz / nm, gradually increase the amplitude of the driving signal by 2mV each time until the coupling strength meets the standard.

[0062] Thermal noise suppression verification: Compare the vibration signal signal-to-noise ratio before and after the mode is enabled to ensure that the signal-to-noise ratio increases from <20dB to ≥30dB and the thermal noise power spectral density decreases from ≥5pW / Hz¹ / ² to ≤3.5pW / Hz¹ / ². At the same time, monitor the stability of the output optical power to ensure that the fluctuation is ≤±2% to avoid power anomalies caused by the mode being enabled.

[0063] S53: Avoiding interlocking of free carrier oscillation modes under high power: Prioritize laser power adjustment: If a mode interlock risk is detected, the FPGA controller first sends a command to the laser power adjustment unit to reduce the laser input power in 0.2dBm steps, such as from 8.5dBm to below 8.0dBm. Each power reduction is monitored using a spectrum analyzer to observe clutter changes: if the clutter amplitude decreases by ≥50%, the interlock risk is reduced, and the power is further fine-tuned to the 8.0-8.5dBm range; if there is no significant change in clutter after reducing the power by 0.5dBm, such as an amplitude decrease of <20%, power adjustment is stopped, and the system switches to waveguide coupling angle adjustment.

[0064] Waveguide coupling angle fine-tuning and effect verification: Activate the angle adjustment function of the vacuum electric displacement stage, and fine-tune the coupling angle between the trapezoidal silicon waveguide and the bent photonic crystal waveguide in 0.5° increments. The adjustment range is ±5°, with an initial angle of 60°. Wait 3 seconds after each adjustment to ensure the optical field is stable.

[0065] Mode interlock verification: Monitor whether the spurious peaks disappear or the amplitude is ≤10% of the original value using a spectrum analyzer. At the same time, verify that the bilateral transmittance of the silicon waveguide-photonic crystal waveguide-resonant cavity is ≥80% to avoid the decrease in coupling efficiency caused by angle adjustment. If the standard is not met when the angle is adjusted to ±3°, continue to increase the adjustment range to a maximum of ±5° until the interlock risk is eliminated.

[0066] Continuous power and mode tracking: Regardless of whether it is a low-power or high-power scenario, the FPGA controller continuously monitors the laser input power, oscillation mode and key indicators, including coupling strength, signal-to-noise ratio and clutter status, every 500ms. If the power fluctuation exceeds ±0.3dBm, such as a sudden increase from 4.8dBm to 5.2dBm, the corresponding processing mode is automatically switched, from parametric optomechanical oscillation mode to normal mode, or from normal mode to interlock monitoring state.

[0067] Parameter memory and fast response: The effective parameters after each adjustment, such as the amplitude of the drive signal at low power and the angle value at high power, are stored in the FPGA register. When the same power and mode scenario occurs later, the historical parameters are directly called, and the response time is shortened from ≥100ms to ≤30ms, improving the efficiency of dynamic adaptation.

[0068] Since temperature drift can cause a shift in the optical resonant frequency, the process of setting up noise suppression through temperature compensation is as follows: S54: The target operating temperature is set to 25±0.5℃, which is the optimal operating temperature for the silicon-based photonic crystal resonator and mechanical oscillator. At this temperature, the optical resonant frequency is stable at 193.38THz, and the mechanical resonant frequency is stable at 70.3kHz. The temperature-wavelength correction factor is 0.01nm / ℃, which is the measured temperature drift compensation value. That is, for every 1℃ deviation of the temperature from the target value, the laser wavelength needs to be corrected by 0.01nm to match the change in optical resonant frequency. The heater power adjustment range is 1-5W, which is the rated power range of the thin-film heater built into the LTCC substrate, ensuring a temperature adjustment accuracy of ±0.1℃ and a noise control target: after temperature compensation, the total noise fluctuation is ≤±5%, avoiding the superposition of thermal noise caused by temperature drift, which affects the accuracy of acceleration calculation.

[0069] Start the temperature monitoring module: The PT1000 temperature sensor built into the LTCC substrate is connected to the FPGA controller via the I2C communication protocol to realize real-time temperature data acquisition.

[0070] Initialize the actuator adjustment: The thin film heater drive module receives power control commands from the FPGA and the wavelength adjustment module of the TSL550C tunable laser, with a minimum step of 0.001nm and a correction accuracy of 0.01nm / ℃, ensuring that the actuator is in standby mode.

[0071] S55: Temperature Data Acquisition and Filtering The PT1000 temperature sensor collects the current operating temperature 10 times per second, such as 24.3℃ and 25.7℃, and transmits it to the FPGA controller. The FPGA controller performs a moving average filter on the collected temperature values ​​to eliminate instantaneous temperature fluctuations and obtain a stable current temperature value T1.

[0072] Temperature deviation and correction calculation: Calculate the temperature deviation ΔT: ΔT = T1 - 25℃. If |ΔT| > 0.5℃, it exceeds the allowable range and triggers the temperature compensation process. Calculate the laser wavelength correction Δλ: According to the correction coefficient in the document, Δλ = ΔT × 0.01nm / ℃. For example, when ΔT = -0.8℃, Δλ = -0.008nm; when ΔT = +0.6℃, Δλ = +0.006nm.

[0073] Predict the direction of heater power adjustment: If T1 < 24.5℃ (low temperature deviation), the heater power needs to be increased; if T1 > 25.5℃ (high temperature deviation), the heater power needs to be reduced.

[0074] S56: Set a two-dimensional adjustment strategy of heater power regulation + laser wavelength correction: Heater power adjustment: The FPGA controller outputs power control commands based on the absolute value of the temperature deviation ΔT: When |ΔT|=0.6-1.0℃ (moderate deviation): adjust the heater power to 2-3W. For example, when T1=24.2℃, set the power to 2.5W; when T1=25.8℃, set the power to 1.5W.

[0075] When |ΔT|>1.0℃ (severe deviation): adjust the heater power to 4-5W. For example, when T1=23.9℃, set the power to 4.2W; when T1=26.1℃, set the power to 1.2W.

[0076] The power adjustment adopts a "stepped approximation": each time the power is adjusted by 0.2W, the temperature is collected again after waiting for 2 seconds (temperature stabilization time) to avoid temperature overshoot caused by sudden power changes.

[0077] Continuously monitor temperature changes until the current temperature T1 returns to the range of 25±0.5℃. For example, if it rises from 24.2℃ to 24.7℃, lock the heater power to 2.5W and enter the laser wavelength correction stage.

[0078] S57: Laser wavelength correction: The FPGA controller converts the calculated wavelength correction Δλ (e.g., -0.008nm) into control commands and sends them to the TSL550C tunable laser via RS485 protocol. Upon receiving the commands, the laser fine-tunes the laser cavity mirror position using its internal piezoelectric ceramic drive mechanism, correcting the output wavelength from the current value (e.g., 1549.74nm) to the target value (1549.74nm + Δλ). For example, if ΔT = -0.8℃, the corrected wavelength is 1549.732nm, ensuring the laser wavelength matches the optical resonant frequency after temperature drift and maintaining a blue detuned state (Δ = 5~10GHz).

[0079] Wavelength correction accuracy verification: Monitor the optical resonant frequency to confirm that the detuning between the corrected laser frequency and the resonant frequency is still in the 5-10 GHz range. If the deviation exceeds the range, fine-tune the wavelength by 0.001 nm each time until the target is met.

[0080] Noise data acquisition: Acquire the noise spectrum of the output optical power after temperature compensation; Extract key noise indicators: thermal noise power spectral density and shot noise power spectral density, and calculate the real-time fluctuation value of total noise (compared with the noise baseline value before compensation).

[0081] Determination of compensation effect: If the total noise fluctuation is ≤±5%, such as decreasing from ±8% before compensation to ±4.2%, it indicates that the temperature compensation is effective. Maintain the current heater power and laser wavelength parameters and enter the continuous monitoring stage. If the total noise fluctuation is > ±5%, or if it is still ±7.5%, recheck the temperature regulation accuracy (whether it returns to 25±0.5℃) and wavelength correction (whether it accurately matches ΔT). Repeat the power adjustment → wavelength correction → noise monitoring process until the noise fluctuation meets the standard.

[0082] Continuous temperature and noise tracking: Temperature and noise data are continuously collected in 100ms cycles. If the temperature deviates from 25±0.5℃ again, such as T1=25.6℃ due to changes in ambient temperature, a new round of compensation process is automatically triggered. After each compensation, the effective parameters (heater power, wavelength correction) are stored in the FPGA register. When the same temperature deviation occurs in the future, the historical parameters are directly called, and the response time is shortened from ≥500ms to ≤100ms, thus improving the compensation efficiency.

[0083] Exception handling: If the temperature sensor detects T1 < 23℃ or T1 > 27℃ (out of safe range), the FPGA controller immediately triggers an alarm signal, adjusts the heater power to 5W (low temperature) or 1W (high temperature), and suspends acceleration measurement to avoid damage to the silicon-based photonic crystal structure from extreme temperatures. The structural safe temperature range is 23-27℃.

[0084] The specific process of step S6 is as follows: S61: Set core baseline parameters: Target signal frequency band: 2-27.6kHz, the range of low-frequency vibration signals after the mechanical oscillator is driven by acceleration, avoiding the mechanical resonant frequency of 70.3kHz and high-frequency noise interference.

[0085] Sensitivity-frequency relationship: See [link to sensitivity-frequency relationship] Figure 2 and Figure 3 It has been pre-stored in the industrial computer.

[0086] Signal amplitude-acceleration correspondence: acceleration value a =Signal amplitude V / corresponding frequency band sensitivity, e.g., when the amplitude is 660mV at 6.4kHz, a =660mV / 660mV / g=1g.

[0087] Direction determination criteria: The vibration direction of the mechanical oscillator is consistent with the acceleration direction, and the vibration direction is perpendicular to the air hole arrangement direction of the silicon-based photonic crystal resonator. This is the inherent vibration characteristic of the mechanical oscillator structure design.

[0088] S62: Optical signal reception and electrical signal conversion: The optical power output of a silicon-based photonic crystal resonator is modulated by the displacement of the mechanical oscillator driven by external acceleration. For example, when acceleration is applied, the optical power fluctuates by ±0.5dBm from the reference value of 2.8dBm.

[0089] The PDB440C photodetector receives the modulated optical signal and converts it into a corresponding electrical signal. The current signal range is 0-10mA, and the signal amplification module amplifies the amplitude of the electrical signal to 0-5V to match the input range of the spectrum analyzer, thus avoiding signal attenuation and distortion.

[0090] Spectrum Analyzer Signal Acquisition: KEYSIGHT N9010B spectrum analyzer acquisition parameters: center frequency 15kHz, i.e., the midpoint of the 2-27.6kHz frequency band, bandwidth 30kHz to cover the target frequency band, resolution bandwidth 10Hz to ensure frequency band subdivision accuracy, acquisition time 1 second to obtain a stable signal spectrum.

[0091] The spectrum analyzer performs spectral analysis on the amplified electrical signal, generates a power spectrum diagram in the 2-27.6kHz frequency band, extracts the low-frequency modulation signal in this band, which is represented by continuous amplitude fluctuation spectrum lines, distinguishing it from discrete noise spikes, and transmits the frequency-amplitude correspondence of the signal data to the industrial computer in real time.

[0092] S63: Signal Filtering and Denoising: The industrial computer performs digital filtering on the received signal data: a Butterworth low-pass filter with a cutoff frequency of 30kHz is used to remove high-frequency noise, and a 50Hz notch filter is used to eliminate power frequency interference, ensuring that the signal-to-noise ratio of the filtered signal is ≥30dB. The filtered signal is then smoothed (using a 10-point moving average) to eliminate instantaneous fluctuations and obtain a smooth frequency-amplitude curve.

[0093] Signal main frequency band determination: Analyze the smoothed frequency-amplitude curve to identify the frequency band with the largest signal amplitude, i.e., the main vibration frequency band. For example, under acceleration, the signal amplitude is the largest at 6.4kHz, which is determined to be the main frequency band.

[0094] If the signal covers multiple sub-bands, such as 2-5kHz and 5-10kHz, with significant amplitudes, then the maximum amplitude of each sub-band is extracted, and the corresponding band sensitivity is calculated separately. Finally, the average value is taken as the acceleration result.

[0095] S64: Precise Amplitude Extraction For the determined main frequency band, such as 6.4kHz, the maximum amplitude of the signal within this frequency band is read as 660mV. The background noise amplitude of 10mV needs to be subtracted, and the actual effective amplitude is 650mV. If it is a multi-band signal, extract the maximum effective amplitude of the sub-bands such as 2-5kHz and 5-10kHz respectively, such as 150mV at 2-5kHz and 200mV at 5-10kHz, and record the center frequency of each sub-band, such as 3.5kHz and 7.5kHz.

[0096] Sensitivity-frequency relationship matching: The industrial computer retrieves pre-stored sensitivity-frequency relationships and looks up the sensitivity value for the corresponding frequency band based on the frequency of the main / sub-band of the signal. If the main frequency band is 6.4kHz, directly match the sensitivity at 6.4kHz in the curve (≥660mV / g, such as the actual calibration value of 660mV / g).

[0097] If the sub-band is 3.5kHz (belonging to the 2-10kHz range), the sensitivity at 3.5kHz in the matching curve is ≥100mV / g, such as the actual calibration value of 120mV / g.

[0098] If the signal frequency is between two calibration values, such as 8.5kHz, the corresponding sensitivity is calculated using linear interpolation. For example, if the sensitivity is 110mV / g at 8kHz and 105mV / g at 9kHz, then the sensitivity at 8.5kHz is (110+105) / 2=107.5mV / g. S65: Acceleration magnitude calculation: Frequency band signals: according to the formula a = Calculated by effective amplitude V / corresponding sensitivity. For example, if the effective amplitude at 6.4kHz is 650mV, S = 660mV / g, then a = 650mV / 660mV / g ≈ 0.98g.

[0099] Multi-band signals: Calculate the acceleration value corresponding to each sub-band, such as at 3.5kHz. a 1 = 150mV / 120mV / g = 1.25g, at 7.5kHz a 2 = 200mV / 105mV / g ≈ 1.90g, then calculate the average value a. avg =(1.25+1.90) / 2≈1.58g, reducing single-band error.

[0100] Acceleration direction determination: The industrial computer receives the vibration direction monitoring signal of the mechanical oscillator (acquired by a micro-strain sensor integrated on the mechanical oscillator, an additional monitoring unit in the mechanical structure design). This signal reflects the vibration direction of the mechanical oscillator (such as the +X direction, -X direction, consistent with the acceleration direction).

[0101] Based on the structural orientation of the silicon-based photonic crystal resonator, the X-axis is preset to be perpendicular to the air hole arrangement direction. The vibration direction is converted into the acceleration direction. For example, if the vibration direction is +X, then the acceleration direction is +X, and the magnitude result is recorded together.

Claims

1. A control method for an optical accelerometer based on nonlinear optical coupling, characterized in that, Includes the following steps: S1: Construct a nonlinear optical coupling module, including a silicon-based photonic crystal resonator, a trapezoidal silicon waveguide and a bent photonic crystal waveguide, and integrate a mechanical oscillator structure; S2: Construct a laser excitation module, including a tunable laser, a three-ring polarization controller, and a vacuum fiber optic flange. Perform initial calibration on the polarization and power of the laser excitation module to obtain initial calibration parameters. S3: Blue detuning state establishment: Based on the initial calibration parameters, the output wavelength of the tunable laser is adjusted to the blue detuning range of the optical resonant frequency, and the mechanical oscillator is driven into the pre-oscillation state. S4: Real-time acquisition of the output optical power spectrum of the silicon-based photonic crystal resonator. When external acceleration causes the mechanical oscillator to shift and the optical resonant frequency to deviate, the laser wavelength is adjusted through a PID feedback algorithm, and the coupling distance between the trapezoidal silicon waveguide and the bent photonic crystal waveguide is also adjusted. S5: Real-time acquisition of laser input power. When the laser input power is lower than the low power threshold, the parametric optomechanical oscillation mode is enabled. When the laser input power is higher than the high power threshold, the resonant frequency shift of the free carrier oscillation mode is monitored, and mode interlocking is avoided by reducing the laser power or adjusting the waveguide coupling angle. S6: Acquire the low-frequency modulation signal of the mechanical oscillator in the output optical power spectrum of the silicon-based photonic crystal resonator, establish a linear relationship between the amplitude of the modulation signal and the acceleration, and output the acceleration measurement result based on the acquired amplitude of the modulation signal.

2. The control method for a light accelerometer based on nonlinear optical coupling according to claim 1, characterized in that, In step S1, the silicon-based photonic crystal resonator adopts an air hole array periodic structure. By progressively displacing and perturbing the air holes in the central region, a defect region is formed to achieve optical field confinement. The width of the trapezoidal silicon waveguide input port is adapted to the spot diameter of the tapered fiber lens, and the width of the output port is matched with the defect width of the standard line of the photonic crystal waveguide. The gradient length is 10 times the wavelength of the input light. The bent photonic crystal waveguide forms a transmission path by removing horizontal and 60° oblique air holes, so that the input light is coupled to the silicon-based photonic crystal resonator at a 60° oblique incidence angle. Integrated mechanical oscillator structure: 10 double-folded cantilever beams support a 500nm thick mass block.

3. The control method for a light accelerometer based on nonlinear optical coupling according to claim 1, characterized in that, The specific process for initial calibration of the polarization and power of the laser excitation module in step S2 is as follows: S21: A laser transmission link is built based on the laser excitation module: The tunable laser outputs laser in a specified band, and the polarization state is adjusted by a three-ring polarization controller with a specified fiber turns ratio to cover the full polarization range of the Poincaré sphere. The laser is then transmitted to the input end of the trapezoidal silicon waveguide through a vacuum fiber flange. S22: Adjust the laser power to 3dBm and traverse the polarization state through the three-ring polarization controller; S23: Monitor the optical transmission spectrum of the silicon-based photonic crystal resonator in real time. When a symmetrical Lorentz-type concave peak appears, lock the polarization state parameters and laser power at this time to complete the initial calibration.

4. The control method for a light accelerometer based on nonlinear optical coupling according to claim 2, characterized in that, The specific process of step S3 is as follows: S31: Based on step S23, retrieve the locked specified parameters, including optical resonant frequency, optimal polarization state parameters, and initial laser power; S32: Frequency difference conversion: Calculate the laser frequency adjustment range based on the definition of blue detuning; S33: Adjust the output wavelength from the initially calibrated resonant wavelength to a shorter wavelength by a specified step; S34: Preset photon number calculation formula: The relationship between the intracavity photon number N and the laser power P and cavity lifetime τ is as follows: N=P×τ / h× ω l , h Let be Planck's constant. ω l The laser frequency; S35: If the number of photons in the cavity is less than 386 under the initial laser power, the laser power is increased in increments of 0.1dBm. Each time the power is increased, the number of photons is calculated using the above photon count formula. When the number of photons reaches 386, the laser power is locked. S36: Pre-start oscillation characteristic judgment: Collect the vibration signal of the mechanical oscillator. If a weak sine wave signal appears near the mechanical resonant frequency of 70.3kHz and the signal frequency is stable, it is determined that the mechanical oscillator has been driven to the pre-start oscillation state by the intracavity optical field.

5. The control method for a light accelerometer based on nonlinear optical coupling according to claim 4, characterized in that, The specific process of adjusting the laser wavelength using the PID feedback algorithm in step S4, while simultaneously adjusting the coupling distance between the trapezoidal silicon waveguide and the bent photonic crystal waveguide, is as follows: S41: Extract specified core reference values ​​based on the established blue detuning state, including laser wavelength reference, coupling distance reference, and output optical power reference; S42: Real-time acquisition of deviation signal: Optical signal acquisition and conversion: The output light of the silicon-based photonic crystal resonator is received in real time, converted into an electrical signal, then into a digital signal, and the digital signal is filtered to extract the real-time value and fluctuation characteristics of the output optical power. At the same time, the offset of the optical resonant frequency is monitored through spectrum analysis. Calculate the deviation value: Laser wavelength deviation Δλ: Using the reference wavelength as the target value, the monitored resonant frequency offset is converted into wavelength deviation according to the frequency-wavelength conversion relationship λ=c / ω; Coupling spacing deviation Δ d : Using the reference coupling spacing as the target value d 0 According to the transmittance-coupling spacing correlation formula T=T 0 − k·d T represents the real-time transmittance. T 0 Standard transmittance, k The transmittance attenuation coefficient is... d The real-time coupling spacing is calculated using real-time transmittance. d This leads to the spacing deviation Δ d= | d−d 0 | ; S43: Set parameters for laser wavelength and coupling distance separately and calculate them independently, and set up dual-parameter linkage logic: Priority setting: Laser wavelength adjustment takes precedence over coupling distance adjustment. When the wavelength deviation is >0.01nm, wavelength adjustment is performed first. After the wavelength deviation is ≤0.01nm, coupling distance adjustment is performed. S44: Actuator response adjustment achieves parameter correction: Laser wavelength adjustment execution: The position of the laser cavity mirror is finely adjusted through the internal piezoelectric ceramic driving mechanism to correct the output wavelength from the current value to the target value; Coupling spacing adjustment: The displacement stage drives the trapezoidal silicon waveguide to move along the direction perpendicular to the waveguide transmission through a stepper motor, thereby correcting the coupling spacing.

6. The control method for a light accelerometer based on nonlinear optical coupling according to claim 1, characterized in that, The specific process of step S5 is as follows: S5: Real-time acquisition of laser input power. When the laser input power is lower than the low power threshold, the parametric optomechanical oscillation mode is enabled. When the laser input power is higher than the high power threshold, the resonant frequency shift of the free carrier oscillation mode is monitored, and mode interlocking is avoided by reducing the laser power or adjusting the waveguide coupling angle. S51: The tunable laser has a built-in power monitoring unit that collects the laser input power and output optical power three times per second and performs smoothing filtering. Oscillation mode identification: In low-power scenarios, i.e., laser input power < 5dBm, the parametric optomechanical oscillation mode is enabled; in high-power scenarios, i.e., laser input power > 8dBm, it is determined that there is a risk of mode interlocking. S52: Enabling parametric optomechanical oscillation mode to suppress thermal noise in low-power scenarios: Mode triggering and parameter configuration: Configure mode parameters based on the current laser input power, lock the driving signal frequency of the mechanical oscillator to its mechanical resonant frequency; synchronously adjust the optical field distribution of the silicon-based photonic crystal resonator, and concentrate the optical field energy in the cavity to the defect area by fine-tuning the three-ring polarization controller; S53: Avoiding interlocking of free carrier oscillation modes under high power: Prioritize laser power adjustment: If a risk of mode interlock is detected, reduce the laser input power in 0.2dBm steps; Each time the power is reduced, observe the change in clutter using a spectrum analyzer: if the clutter amplitude decreases by ≥50%, continue to fine-tune the power to the 8.0-8.5dBm range; if there is no significant change in clutter after reducing the power by 0.5dBm, stop power adjustment and switch to waveguide coupling angle adjustment.

7. The control method for a light accelerometer based on nonlinear optical coupling according to claim 6, characterized in that, Step S5 also includes a process for noise suppression through temperature compensation, as detailed below: S54: Set the target operating temperature to 25±0.5℃, the temperature-wavelength correction coefficient to 0.01nm / ℃, the heater power adjustment range to 1-5W, and the noise control target: after temperature compensation, the total noise fluctuation ≤±5%; S55: Temperature Data Acquisition and Filtering The current operating temperature is collected 10 times per second. The collected temperature values ​​are then filtered by moving average to eliminate instantaneous temperature fluctuations and obtain a stable current temperature value T1. Temperature deviation and correction calculation: Calculate the temperature deviation ΔT. If |ΔT|>0.5℃, trigger the temperature compensation process. Calculate the laser wavelength correction Δλ: According to the correction coefficient, Δλ = ΔT × 0.01 nm / ℃; Predict the direction of heater power adjustment: If T1 < 24.5℃, increase the heater power; if T1 > 25.5℃, decrease the heater power. S56: Set a two-dimensional adjustment strategy of heater power regulation + laser wavelength correction: Heater power adjustment: Heater power is adjusted based on the absolute value of the temperature deviation ΔT: when |ΔT| = 0.6-1.0℃, the heater power is adjusted to 2-3W; when |ΔT| > 1.0℃, the heater power is adjusted to 4-5W. Continuously monitor temperature changes until the current temperature T1 returns to the range of 25±0.5℃. At this point, lock the heater power and enter the laser wavelength correction stage. S57: Laser wavelength correction: The calculated wavelength correction Δλ is converted into a control command and sent to the tunable laser. After receiving a command, the tunable laser will correct its output wavelength from the current value to the target value.

8. The control method for a light accelerometer based on nonlinear optical coupling according to claim 1, characterized in that, The specific process of step S6 is as follows: S61: Set core reference parameters, including target signal frequency band, sensitivity frequency relationship, signal amplitude-acceleration correspondence, and direction determination criteria; S62: Receives optical signals, converts them into corresponding electrical signals, and amplifies the amplitude of the electrical signals to 0-5V through a signal amplification module. The spectrum analyzer performs spectrum analysis on the amplified electrical signals to generate a power spectrum diagram of the 2-27.6kHz frequency band and extracts the low-frequency modulation signals in this frequency band. S63: Digital filtering of low-frequency modulated signals: Butterworth low-pass filter is used to remove high-frequency noise, 50Hz notch filter is used to eliminate power frequency interference, and the filtered signal is smoothed to obtain a smooth frequency-amplitude curve. Analyze the smoothed curve to identify the frequency band with the largest signal amplitude; if the signal covers multiple sub-frequency bands, extract the maximum amplitude of each sub-frequency band, and then calculate the corresponding frequency band sensitivity separately, finally taking the average value as the acceleration result; S64: Precise amplitude extraction: For the determined main frequency band, read the maximum amplitude of the signal within that frequency band; If it is a multi-band signal, extract the maximum effective amplitude of the 2-5kHz and 5-10kHz sub-bands respectively, and record the center frequency of each sub-band; Sensitivity-frequency relationship matching: Recalls pre-stored sensitivity-frequency relationships and finds the sensitivity value for the corresponding frequency band based on the signal's main / sub-band frequencies. If the main frequency band is 6.4kHz, directly match the sensitivity at 6.4kHz in the curve; If the sub-band is 3.5kHz, the sensitivity at 3.5kHz in the matching curve; If the signal frequency is between the two calibration values, the corresponding sensitivity is calculated using linear interpolation. S65: Acceleration magnitude calculation: Frequency band signals: according to the formula a = Calculated by effective amplitude V / corresponding sensitivity; For multi-band signals: calculate the acceleration value corresponding to each sub-band separately, and then calculate the average value; Acceleration direction determination: Receive the vibration direction monitoring signal of the mechanical oscillator, which reflects the vibration direction of the mechanical oscillator; By combining the structural orientation of the silicon-based photonic crystal resonator, the vibration direction is converted into the acceleration direction.

Images