Airborne combined measurement calculation method for closed-loop optical fiber

Abstract

The invention discloses an airborne combined measurement calculation method for closed-loop optical fibers, which comprises the following steps that: an orthogonally distributed three-axis closed-loop optical fiber gyroscope array acquires a Sagnac phase difference signal in real time, each optical fiber ring adopts an asymmetric pretightening force winding process, the pretightening forces of X / Y / Z axes are 12 + / -0.5 N, 8 + / -0.3 N and 15 + / -0.5 N respectively, the diameter ratio of the optical fiber rings is 1: 0.8: 1.2, environmental disturbance compensation is carried out, and the Sagnac phase difference signal is acquired in real time. Data of 48 PT1000 temperature sensors embedded in the axial direction of an optical fiber ring at the interval of 5 mm and full spectrum data of a three-axis vibration sensor at the frequency band of 0-2000 Hz are synchronously collected, and a coupling error model is constructed based on temperature gradient distribution and vibration spectrum characteristics. Therefore, through an asymmetric winding process, temperature-vibration coupling compensation, improved data fusion and dynamic closed-loop control, the problems of nonreciprocal errors caused by the symmetric winding process, insufficient environmental adaptability of single temperature compensation, limited nonlinear error processing capability of standard Kalman filtering and stability of a fixed parameter controller are solved; high-precision and high-reliability airborne combined measurement is realized.

Description

Technical Field

[0001] The present invention relates to the technical field of airborne combined measurement, and in particular to an airborne combined measurement calculation method for a closed-loop optical fiber. Background Art

[0002] Among existing airborne combined measurement technologies, traditional closed-loop fiber optic gyroscope systems typically employ a symmetrical winding process and a single temperature compensation method. Kalman filtering is used to achieve multi-source data fusion, and digital phase-locked technology is employed for closed-loop feedback control. In existing technologies, the fiber optic ring winding process is often symmetrically designed, temperature compensation relies primarily on a single temperature sensor, the data fusion algorithm typically employs a standard Kalman filter, and closed-loop control is based on a fixed-parameter PID controller. While this method meets measurement requirements in conventional flight environments to a certain extent, its performance in highly dynamic and complex environments still needs to be improved.

[0003] However, the symmetrical winding process cannot effectively suppress non-reciprocal errors, resulting in limited measurement accuracy. Secondly, the single temperature compensation method is difficult to cope with complex temperature gradient changes and has poor environmental adaptability. Thirdly, the standard Kalman filter has difficulty handling nonlinear errors in multi-source data fusion, resulting in insufficient fusion accuracy. Finally, the fixed-parameter PID controller is prone to control saturation in a high-dynamic environment, affecting system stability. In addition, the existing technology lacks refined compensation means for vibration interference and has insufficient fault self-healing capabilities, making it difficult to meet the needs of modern aircraft for high-precision and high-reliability navigation systems. Summary of the Invention

[0004] The present invention aims to solve one of the technical problems in the related art at least to a certain extent.

[0005] To this end, the purpose of the present invention is to propose an airborne combined measurement and calculation method for closed-loop optical fiber, which achieves high-precision measurement through multi-source data fusion and closed-loop feedback control, and solves the non-reciprocal error caused by the symmetrical winding process in the background technology, the insufficient environmental adaptability of the single temperature compensation, and the stability problem of the fixed parameter controller. Specifically, it includes: asymmetric preload winding to optimize the optical fiber ring structure, temperature-vibration coupling error model to improve environmental adaptability, improved volumetric Kalman filtering to enhance data fusion accuracy, and dynamic PID control to improve system stability.

[0006] To achieve the above object, the present invention proposes an airborne combined measurement and calculation method for a closed-loop optical fiber, comprising the following steps:

[0007] S1. Dynamic data acquisition: Sagnac phase difference signals are acquired in real time through an orthogonally distributed three-axis closed-loop fiber optic gyroscope array. Each fiber ring adopts an asymmetric preload winding process, with X / Y / Z axis preloads of 12±0.5N, 8±0.3N, and 15±0.5N, respectively. The fiber ring diameter ratio is 1:0.8:1.2.

[0008] S2. Environmental disturbance compensation: synchronously collect data from 48 PT1000 temperature sensors embedded at 5mm intervals along the axial direction of the fiber ring, as well as full spectrum data from the 0-2000Hz frequency band of the triaxial vibration sensor. A coupling error model is constructed based on the temperature gradient distribution and vibration spectrum characteristics.

[0009] S3. Multi-source data fusion: align the compensated gyro data with the outputs of the star sensor, MEMS accelerometer, and atmospheric data computer in time and space, and establish a 12-dimensional state vector containing position, velocity, attitude angle, gyro drift, and accelerometer bias. An improved cubature Kalman filter is used to achieve data fusion.

[0010] S4. Closed-loop feedback control, based on the six-degree-of-freedom kinematic equations of the aircraft, includes three translational degrees of freedom (X / Y / Z-axis acceleration) and three rotational degrees of freedom (angular velocity around the X / Y / Z axes). The Y-waveguide drive voltage and phase modulation depth are dynamically adjusted. The Sagnac interferometer signal is stabilized at the π / 2 operating point through digital phase-locked amplification technology, and feedforward compensation is introduced to eliminate phase noise caused by vibration.

[0011] The present invention discloses an airborne combined measurement and calculation method for a closed-loop optical fiber. The method stabilizes a Sagnac interference signal at a π / 2 operating point through a digital phase-locked amplification technology, introduces feedforward compensation to eliminate phase noise caused by vibration, adopts dynamic PID control (Kp=2.5±0.2, Ti=0.1s, Td=0.05s) and an anti-saturation control mode to ensure the stability of the system in a high-dynamic environment. The method solves the problems of non-reciprocal error caused by a symmetrical winding process, insufficient environmental adaptability of single temperature compensation, limited nonlinear error processing capability of a standard Kalman filter, and stability of a fixed parameter controller through an asymmetric winding process, thereby achieving high-precision and high-reliability airborne combined measurement.

[0012] Specifically, the closed-loop feedback control of step S4 includes: the main control loop uses a digital phase-locked amplifier to track the phase of the carrier signal in real time, with a locking accuracy of 0.01°; the auxiliary control loop generates an inverse phase disturbance signal based on the spectral characteristics of the vibration sensor, and the compensation bandwidth covers 10-1500Hz. The PID controller parameters for dynamically adjusting the Y-waveguide drive voltage are proportional coefficient Kp = 2.5±0.2, integral time Ti = 0.1s, and differential time Td = 0.05s. When the temperature change rate exceeds 5°C / min, it automatically switches to the anti-saturation control mode.

[0013] Specifically, the temperature compensation method in step S2 includes:

[0014] a. In the low temperature range of -40°C to 0°C, the first Vernier compensation coefficient matrix [V1] = [0.92, -0.15; 0.08, 1.03] is used, and the low temperature protection mode is activated to reduce the thermal stress of the fiber ring;

[0015] b. In the normal temperature range of 0°C to 40°C, the second Vernier compensation coefficient matrix [V2] = [1.05, 0.02; -0.03, 0.98] is used, and the standard compensation mode is activated to monitor the axial temperature gradient in real time;

[0016] c. In the high temperature range of 40°C to 80°C, the third Vernier compensation coefficient matrix [V3] = [0.85, 0.25; 0.12, 1.12] is used, and the high temperature heat dissipation mode is enabled. The fiber ring temperature rise rate is controlled to < 2°C / min through the active cooling device;

[0017] The action order of the compensation operators is dynamically adjusted according to the direction of the axial temperature gradient of the optical fiber ring. When the axial temperature difference is detected to be greater than 3°C, the gradient compensation mode is activated, and the parameters of the optical fiber ring heat conduction equation are identified online. The update cycle is 100ms and the identification error is less than 0.5°C.

[0018] Specifically, the vibration interference suppression method of step S2 includes: constructing a 6-degree-of-freedom vibration transfer function model, including 3 translational degrees of freedom (X / Y / Z axis acceleration) and 3 rotational degrees of freedom (angular velocity around the X / Y / Z axis), eliminating the phase error caused by the base angular vibration through inverse system compensation, using the db6 wavelet basis to perform 8-layer wavelet packet decomposition, retaining 3-5 layers of detail coefficients when reconstructing the signal, and automatically switching the anti-aliasing filter cutoff frequency to 2000Hz when detecting high-frequency vibration >500Hz, and the stopband attenuation >60dB.

[0019] Specifically, the multi-source data fusion in step S3 further includes: designing a sliding time window to achieve multi-sensor synchronization, with a time window length of 50ms and an overlap rate of 80%, building a three-dimensional heterogeneous data bus architecture, transmitting 10Gbps optical signals in the upper layer, transmitting 1Mbps control signals in the middle layer, and ensuring a ripple of less than 10mV in the lower power channel. A hardware scale factor calibration module is introduced in the extended state space model to correct the nonlinear response of the sensor in real time.

[0020] Specifically, it also includes fault self-healing processing: when the phase modulator fails, it switches to the dual-wavelength polarization diversity reception mode (1550nm and 1310nm), and adopts the attitude angle change rate smoothing algorithm when reconstructing the interference signal. By calculating the difference value of the attitude angle at adjacent moments in real time, the attitude angle jump error is compensated to less than 0.001°. When the backup channel is started, the mechanical gyroscope data is used to maintain the system stability time less than 10ms.

[0021] Specifically, step S1 dynamic data acquisition includes: alternating high-birefringence photonic crystal fiber segments and conventional polarization-maintaining fiber segments, forming a distributed microcavity strain sensor array with a spacing of 2 mm inside the optical fiber through femtosecond laser direct writing technology, and real-time correction of the Sagnac phase calculation value based on the Brillouin scattered light time domain reflection data, with a sampling rate of 1 kHz.

[0022] Specifically, step S4 closed-loop feedback control includes: executing a spiral maneuver flight trajectory with an angular velocity range of ±300° / s and an acceleration range of ±8g, online estimation of the first and second harmonic amplitudes of the scale factor nonlinearity, an update period of 200ms, and separation of deterministic error (residual <0.001%) and random error components through dual time correlation analysis.

[0023] Specifically, step S3 multi-source data fusion includes: using orthogonal phase-locked amplification technology to extract the first-order sideband component of the carrier signal, improving the signal-to-noise ratio by >20dB, performing sliding window singular value decomposition on the sensor data, achieving a fault detection sensitivity of 0.1mV, and realizing coordinate system conversion through the quaternion spherical linear interpolation (SLERP) method, with a calculation delay of <50μs.

[0024] Additional aspects and advantages of the present invention will be set forth in part in the description which follows and, in part, will be obvious from the description which follows, or may be learned through practice of the present invention. BRIEF DESCRIPTION OF THE DRAWINGS

[0025] The above and / or additional aspects and advantages of the present invention will become apparent and readily understood from the following description of the embodiments in conjunction with the accompanying drawings, in which:

[0026] Figure 1 Flowchart of the airborne combined measurement calculation method of the closed-loop optical fiber of the present invention;

[0027] Figure 2 This is a data diagram of Example 1 of the airborne combined measurement and calculation method for a closed-loop optical fiber of the present invention;

[0028] Figure 3 This is a data diagram of Example 2 of the airborne combined measurement calculation method for closed-loop optical fiber of the present invention. DETAILED DESCRIPTION

[0029] The embodiments of the present invention are described in detail below. Examples of the embodiments are shown in the accompanying drawings, wherein the same or similar reference numerals throughout represent the same or similar elements or elements having the same or similar functions. The embodiments described below with reference to the accompanying drawings are exemplary and are intended to be used to explain the present invention and are not to be construed as limiting the present invention. On the contrary, the embodiments of the present invention include all variations, modifications, and equivalents that fall within the spirit and scope of the appended claims.

[0030] The following describes an airborne combined measurement and calculation method for a closed-loop optical fiber according to an embodiment of the present invention with reference to the accompanying drawings.

[0031] like Figure 1-Figure 3 As shown, the airborne combined measurement and calculation method of a closed-loop optical fiber according to an embodiment of the present invention includes:

[0032] S1. Dynamic data acquisition: Real-time Sagnac phase difference signals are acquired through an orthogonally distributed three-axis closed-loop fiber gyroscope array. Each fiber ring is wound using an asymmetric preload force, with X / Y / Z axis preload forces of 12±0.5N, 8±0.3N, and 15±0.5N, respectively. The fiber ring diameter ratio is 1:0.8:1.2.

[0033] It should be noted that in the dynamic data acquisition stage described in this embodiment, the Sagnac phase difference signal is obtained in real time through an orthogonally distributed three-axis closed-loop fiber gyroscope array, wherein each fiber ring adopts an asymmetric preload winding process, and the X / Y / Z axis preload is 12±0.5N, 8±0.3N, and 15±0.5N, respectively. The fiber ring diameter ratio is 1:0.8:1.2, and high birefringence photonic crystal fiber segments (30% of the length) and conventional polarization-maintaining fiber segments are alternately arranged in the fiber ring to improve the ability to resist environmental interference. A distributed microcavity strain sensor array with a spacing of 2mm is formed inside the optical fiber through femtosecond laser direct writing technology to monitor the local strain distribution of the fiber ring in real time. Based on the Brillouin scattering light time domain reflection technology, the Sagnac phase calculation value is corrected in real time at a sampling rate of 1kHz to ensure the accuracy and reliability of dynamic data acquisition.

[0034] S2. To compensate for environmental disturbances, we simultaneously collect data from 48 PT1000 temperature sensors embedded at 5mm intervals along the fiber ring axis, as well as full-spectrum data from a triaxial vibration sensor covering the 0-2000Hz frequency range. We then construct a coupling error model based on the temperature gradient distribution and vibration spectrum characteristics.

[0035] It should be noted that in the environmental disturbance compensation stage described in this embodiment, data from 48 PT1000 temperature sensors embedded at 5mm intervals along the axial direction of the optical fiber ring and full spectrum data from the 0-2000Hz frequency band of the triaxial vibration sensor are synchronously collected. A coupling error model is constructed based on the temperature gradient distribution and vibration spectrum characteristics. The temperature compensation adopts a three-level interval division (-40℃~0℃, 0℃~40℃, and 40℃~80℃). Each interval corresponds to a different Vernier compensation coefficient matrix, and the compensation strategy is dynamically adjusted by online identification of the parameters of the optical fiber ring heat conduction equation (update period 100ms, identification error <0.5℃). Vibration interference suppression is achieved by constructing a 6-degree-of-freedom vibration transfer function model and using the db6 wavelet basis for 8-layer wavelet packet decomposition. 3-5 layers of detail coefficients are retained when reconstructing the signal. When high-frequency vibration >500Hz is detected, the anti-aliasing filter cutoff frequency is automatically switched to 2000Hz, and the stopband attenuation is >60dB, so as to achieve high-precision compensation for complex environmental disturbances.

[0036] S3. Multi-source data fusion: align the compensated gyro data with the outputs of the star sensor, MEMS accelerometer, and atmospheric data computer in time and space to establish a 12-dimensional state vector containing position, velocity, attitude angle, gyro drift, and accelerometer bias. An improved cubature Kalman filter is used to achieve data fusion.

[0037] It should be noted that in the multi-source data fusion stage described in this embodiment, the compensated gyro data is spatiotemporally aligned with the outputs of the star sensor, MEMS accelerometer, and atmospheric data computer to establish a 12-dimensional state vector containing position, velocity, attitude angle, gyro drift, and accelerometer bias. An improved cubature Kalman filter is used to achieve data fusion, in which a sliding time window (length 50ms, overlap rate 80%) is designed to achieve multi-sensor synchronization, and a scale factor nonlinear term δK = α·sin(ωt) + β·cos(2ωt) is introduced into the extended state space model. The fusion weight (λ = 0.02) of inertial navigation and satellite navigation is dynamically adjusted by the adaptive correlation factor β(t) = 1-exp(-λΔT). At the same time, orthogonal phase-locked amplification technology is used to extract the first-order sideband component of the carrier signal, and the signal-to-noise ratio is improved by > 20 dB. The coordinate system conversion is achieved by Lie group SE(3) spatial interpolation, and the calculation delay is < 50 μs, ensuring the accuracy and real-time performance of multi-source data fusion.

[0038] S4. Closed-loop feedback control dynamically adjusts the Y-waveguide drive voltage and phase modulation depth based on the vehicle's six-degree-of-freedom kinematic equations, including three translational and three rotational degrees of freedom. Digital phase-locked amplification technology is used to stabilize the Sagnac interferometer signal at the π / 2 operating point, and feedforward compensation is introduced to eliminate phase noise caused by vibration.

[0039] It should be noted that the closed-loop feedback control stage described in this embodiment dynamically adjusts the Y-waveguide drive voltage and phase modulation depth based on the six-degree-of-freedom kinematic equations of the aircraft. The Sagnac interferometer signal is stabilized at the π / 2 operating point through digital phase-locked amplification technology, and feedforward compensation is introduced to eliminate phase noise caused by vibration. The main control loop uses a digital phase-locked amplifier to track the carrier signal phase in real time, with a locking accuracy of 0.01°. The auxiliary control loop generates an inverse phase disturbance signal based on the spectral characteristics of the vibration sensor, with a compensation bandwidth covering 10-1500 Hz. The PID controller parameters for dynamically adjusting the Y-waveguide drive voltage are proportional coefficient Kp = 2.5 ± 0.2, integral time Ti = 0.1 s, and differential time Td = 0.05 s. It automatically switches to anti-saturation control mode when the temperature change rate exceeds 5°C / min. At the same time, the scale factor nonlinearity parameters α and β are estimated online by executing a spiral maneuver flight trajectory (angular velocity range ±300° / s, acceleration range ±8g), with an update period of 200ms, to ensure the stability and accuracy of the closed-loop feedback control.

[0040] Specifically, an orthogonally distributed three-axis closed-loop fiber optic gyroscope array is used to acquire Sagnac phase difference signals in real time. An asymmetric preload winding process (preload forces on the X / Y / Z axes are 12±0.5N, 8±0.3N, and 15±0.5N, respectively, with a diameter ratio of 1:0.8:1.2) is used to optimize the fiber ring structure, significantly reducing nonreciprocal errors. Data from 48 PT1000 temperature sensors embedded at 5mm intervals along the fiber ring axis and full-spectrum data from a three-axis vibration sensor in the 0-2000Hz frequency band are simultaneously collected to construct a temperature-vibration coupling error model. Environmental adaptability is improved through three-level temperature range division (-40℃~0℃, 0℃~40℃, and 40℃~80℃) and inverse compensation of the 6-DOF vibration transfer function. The compensated gyroscope data are spatiotemporally aligned with the outputs of the star sensor, MEMS accelerometer, and atmospheric data computer to establish a 12-dimensional state vector. Multi-source data fusion is achieved using an improved cubature Kalman filter. A sliding time window (50m s length, 80% overlap rate) and adaptive correlation factor β(t) = 1-exp(-λΔT) to dynamically adjust the fusion weight to solve the problem of nonlinear error. According to the six-degree-of-freedom kinematic equation of the aircraft, the Y waveguide driving voltage and phase modulation depth are dynamically adjusted. The Sagnac interference signal is stabilized at the π / 2 working point through digital phase-locked amplification technology, and feedforward compensation is introduced to eliminate the phase noise caused by vibration. Dynamic PID control (Kp = 2.5±0.2, Ti = 0.1s, Td = 0.05s) and anti-saturation control mode are adopted to ensure the stability of the system in a high dynamic environment. The present invention solves the non-reciprocal error caused by the symmetrical winding process in the background technology, the insufficient environmental adaptability of the single temperature compensation, the limited nonlinear error processing capability of the standard Kalman filter, and the stability problem of the fixed parameter controller through an asymmetric winding process, temperature-vibration coupling compensation, improved data fusion and dynamic closed-loop control, thereby realizing high-precision and high-reliability airborne combined measurement.

[0041] Furthermore, if Figure 1 As shown, the closed-loop feedback control of step S4 includes: the main control loop uses a digital lock-in amplifier to track the phase of the carrier signal in real time, with a locking accuracy of 0.01°; the auxiliary control loop generates an inverse phase disturbance signal based on the spectral characteristics of the vibration sensor, with a compensation bandwidth covering 10-1500 Hz; the PID controller parameters for dynamically adjusting the Y-waveguide drive voltage are: proportional coefficient Kp = 2.5 ± 0.2, integral time Ti = 0.1 s, differential time Td = 0.05 s; when the temperature change rate exceeds 5 ° C / min, it automatically switches to the anti-saturation control mode.

[0042] It should be noted that in the closed-loop feedback control stage described in this embodiment, the main control loop uses a digital phase-locked amplifier to track the phase of the carrier signal in real time, with a locking accuracy of 0.01°, and extracts the first-order sideband component of the carrier signal through orthogonal phase-locked amplification technology, with a signal-to-noise ratio improvement of >20dB. The auxiliary control loop generates an inverse phase disturbance signal based on the spectral characteristics of the vibration sensor, with a compensation bandwidth covering 10-1500Hz, and uses the db6 wavelet basis for 8-layer wavelet packet decomposition. When reconstructing the signal, 3-5 layers of detail coefficients are retained, and the anti-vibration function is automatically switched when high-frequency vibration >500Hz is detected. The aliasing filter cutoff frequency is set to 2000 Hz, with a stopband attenuation of >60 dB. The PID controller parameters for dynamically adjusting the Y-waveguide drive voltage are proportional coefficient Kp = 2.5 ± 0.2, integral time Ti = 0.1 s, and differential time Td = 0.05 s. When the temperature change rate exceeds 5°C / min, the system automatically switches to anti-saturation control mode. At the same time, the scale factor nonlinearity parameters α and β are estimated online by executing a spiral maneuver trajectory (angular velocity range ±300° / s, acceleration range ±8g), with an update period of 200 ms, to ensure the stability and accuracy of the closed-loop feedback control.

[0043] Furthermore, if Figure 1 As shown, the temperature compensation method in step S2 includes:

[0044] a. In the low temperature range of -40°C to 0°C, the first Vernier compensation coefficient matrix [V1] = [0.92, -0.15; 0.08, 1.03] is used, and the low temperature protection mode is activated to reduce the thermal stress of the fiber ring;

[0045] b. In the normal temperature range of 0°C to 40°C, the second Vernier compensation coefficient matrix [V2] = [1.05, 0.02; -0.03, 0.98] is used, and the standard compensation mode is activated to monitor the axial temperature gradient in real time;

[0046] c. In the high temperature range of 40°C to 80°C, the third Vernier compensation coefficient matrix [V3] = [0.85, 0.25; 0.12, 1.12] is used, and the high temperature heat dissipation mode is enabled. The fiber ring temperature rise rate is controlled to < 2°C / min through the active cooling device;

[0047] The compensation operator action sequence is dynamically adjusted according to the axial temperature gradient direction of the optical fiber ring. When the axial temperature difference is detected to be greater than 3°C, the gradient compensation mode is activated.

[0048] Online identification of fiber optic ring heat conduction equation parameters, with an update period of 100ms and an identification error of <0.5℃.

[0049] It should be noted that in the temperature compensation method described in this embodiment, the operating temperature is divided into three intervals: -40℃~0℃, 0℃~40℃, and 40℃~80℃. Each interval corresponds to a different Vernier compensation coefficient matrix, and the temperature gradient distribution is monitored in real time by 48 PT1000 temperature sensors embedded at a spacing of 5mm along the axial direction of the optical fiber ring. When the axial temperature difference is detected to be greater than 3℃, the gradient compensation mode is started, and the order of action of the compensation operator is dynamically adjusted. At the same time, the parameters of the optical fiber ring heat conduction equation are identified online with an update cycle of 100ms and an identification error of less than 0.5℃. The temperature-vibration coupling error model is constructed in combination with the vibration sensor data, and a three-level compensation strategy (interval compensation, gradient compensation, and coupling compensation) is adopted to achieve high-precision temperature compensation in complex environments.

[0050] Furthermore, if Figure 1 As shown, the vibration interference suppression method of step S2 includes:

[0051] A 6-DOF vibration transfer function model is constructed, including 3 translational degrees of freedom (X / Y / Z axis acceleration) and 3 rotational degrees of freedom (angular velocity around the X / Y / Z axes). The phase error caused by the base angular vibration is eliminated through inverse system compensation.

[0052] The db6 wavelet basis is used for 8-layer wavelet packet decomposition, and 3-5 layers of detail coefficients are retained when reconstructing the signal. When high-frequency vibration > 500Hz is detected, the anti-aliasing filter cutoff frequency is automatically switched to 2000Hz, and the stopband attenuation is > 60dB.

[0053] It should be noted that in the vibration interference suppression method described in this embodiment, a 6-degree-of-freedom vibration transfer function model is constructed, and the phase error caused by the base angular vibration is eliminated by inverse system compensation. Combined with the full spectrum data of the 0-2000Hz frequency band of the three-axis vibration sensor, the db6 wavelet basis is used to perform 8-layer wavelet packet decomposition, and 3-5 layers of detail coefficients are retained when reconstructing the signal. When high-frequency vibration >500Hz is detected, the anti-aliasing filter cutoff frequency is automatically switched to 2000Hz, and the stopband attenuation is >60dB. At the same time, a feedforward compensation mechanism is introduced to generate an inverse phase disturbance signal, and the compensation bandwidth covers 10-1500Hz to ensure the phase noise suppression effect in a high dynamic vibration environment.

[0054] Furthermore, if Figure 1 As shown, the multi-source data fusion in step S3 further includes:

[0055] Design a sliding time window to achieve multi-sensor synchronization, with a time window length of 50ms and an overlap rate of 80%;

[0056] Build a three-dimensional heterogeneous data bus architecture, with the upper layer transmitting 10Gbps optical signals, the middle layer transmitting 1Mbps control signals, and the lower layer power channel ripple less than 10mV;

[0057] A hardware scale factor calibration module is introduced into the extended state space model to correct the nonlinear response of the sensor in real time.

[0058] Furthermore, if Figure 1 As shown, the system also includes fault self-healing processing: when the phase modulator fails, it switches to the dual-wavelength polarization diversity reception mode (1550nm and 1310nm); when reconstructing the interference signal, an attitude angle change rate smoothing algorithm is used to compensate for the attitude angle jump error of less than 0.001° by calculating the difference between the attitude angles at adjacent moments in real time; when the backup channel is started, the mechanical gyroscope data is used to maintain the system stability time of less than 10ms.

[0059] It should be noted that in the fault self-healing process described in this embodiment, when the phase modulator fails, it switches to the dual-wavelength (1550nm and 1310nm) polarization diversity receiving mode, and uses femtosecond laser direct writing technology to form a distributed microcavity strain sensor array with a spacing of 2mm inside the optical fiber to monitor the signal status in real time. When reconstructing the Sagnac interference signal, the Lie group interpolation algorithm is introduced to compensate for the attitude angle jump error of less than 0.001°. When the backup channel is started, the mechanical gyroscope data is used to maintain the system stability time of less than 10ms, and combined with the deep belief network (DBN) fault prediction model to achieve early warning and rapid recovery of faults, ensuring the continuous and reliable operation of the system under abnormal conditions.

[0060] Furthermore, if Figure 1 As shown, step S1 dynamic data collection includes:

[0061] Alternately arrange high birefringence photonic crystal fiber segments and conventional polarization-maintaining fiber segments;

[0062] A distributed microcavity strain sensor array with a spacing of 2 mm is formed inside the optical fiber using femtosecond laser direct writing technology;

[0063] The Sagnac phase calculation value is corrected in real time based on the Brillouin scattered light time domain reflectometry data, with a sampling rate of 1kHz.

[0064] It should be noted that in the dynamic data acquisition described in this embodiment, high-birefringence photonic crystal fiber segments (30% of the length) and conventional polarization-maintaining fiber segments are arranged alternately, and the fiber ring structure is optimized through an asymmetric preload winding process (X / Y / Z axis preload is 12±0.5N, 8±0.3N, and 15±0.5N, respectively, and the diameter ratio is 1:0.8:1.2). At the same time, femtosecond laser direct writing technology is used to form a distributed microcavity strain sensor array with a spacing of 2mm inside the optical fiber to monitor the local strain distribution of the fiber ring in real time, and based on the Brillouin scattered light time domain reflection data, the Sagnac phase calculation value is corrected in real time at a sampling rate of 1kHz to ensure the accuracy and anti-interference ability of dynamic data acquisition.

[0065] Furthermore, if Figure 1 As shown, step S4 closed-loop feedback control includes: executing a spiral maneuver flight trajectory with an angular velocity range of ±300° / s and an acceleration range of ±8g, online estimating the first harmonic amplitude and second harmonic amplitude of the scale factor nonlinearity, with an update period of 200ms, and separating the deterministic error (residual <0.001%) and random error components through dual time correlation analysis. Step S3 multi-source data fusion includes: using orthogonal phase-locked amplification technology to extract the first-order sideband component of the carrier signal, with a signal-to-noise ratio improvement of >20dB, performing sliding window singular value decomposition on the sensor data, with a fault detection sensitivity of 0.1mV, and realizing coordinate system conversion through the quaternion spherical linear interpolation (SLERP) method, with a calculation delay of <50μs.

[0066] It should be noted that in the closed-loop feedback control described in this embodiment, a spiral maneuver flight trajectory (angular velocity range ±300° / s, acceleration range ±8g) is executed to stimulate the observability of all inertial channels, the first harmonic amplitude and the second harmonic amplitude of the scale factor nonlinearity are estimated online (update period 200ms), and the deterministic error (residual <0.001%) and the random error component are separated by dual time correlation analysis. At the same time, the orthogonal phase-locked amplification technology in multi-source data fusion is combined to extract the first-order sideband component of the carrier signal (signal-to-noise ratio improvement >20dB), the sliding window singular value decomposition (SVD, fault detection sensitivity of 0.1mV) is performed on the sensor data, and the coordinate system transformation is realized by Lie group SE(3) spatial interpolation (computation delay <50μs), ensuring the high-precision coordinated operation of closed-loop feedback control and data fusion.

[0067] Example 1:

[0068] Implementation of navigation measurements in highly dynamic flight environments

[0069] Application scenario: A certain type of UAV performs high-speed maneuvering missions (roll angular velocity ±280° / s, pitch acceleration ±7.5g)

[0070] Implementation steps and parameters:

[0071] Fiber ring preparation

[0072] High birefringence photonic crystal fiber (HB-PCF-1550, birefringence index 4.2×10 -4 ) and conventional polarization-maintaining fiber (PM1550) are arranged alternately, with photonic crystal fiber accounting for 30%.

[0073] Asymmetric preload winding of triaxial fiber ring: X-axis 12.3N (error ±0.3N), Y-axis 7.9N (error ±0.2N), Z-axis 15.1N (error ±0.4N), diameter ratio 1:0.8:1.2

[0074] Femtosecond laser direct writing (wavelength 800nm, pulse energy 0.5μJ) forms a microcavity array with a spacing of 2mm and a sensitivity of 0.8pm / με

[0075] Implementation of environmental compensation

[0076] PT1000 temperature sensors (48, 5mm spacing) collect data in real time, and activate gradient compensation when the temperature gradient is greater than 3°C

[0077] Vibration suppression uses db6 wavelet basis 8-layer decomposition, retaining the 4th layer detail coefficient (frequency band 125-250Hz), and the anti-aliasing filter switching threshold is set to 480Hz

[0078] System integration and testing

[0079] Three-dimensional heterogeneous data bus architecture: upper layer 10Gbps optical signal channel, middle layer 1Mbps control channel, lower layer power channel ripple <10mV

[0080] Thermal expansion coefficient gradient transition of metal-ceramic composite substrate: 4.2→9.6×10 -6 / ℃, inhibit thermal stress deformation

[0081] MEMS micromirror array realizes automatic optical path alignment, positioning accuracy of ±0.1μm, and response time of <1ms

[0082] Performance comparison (such as Figure 2 shown)

[0083] Example 2:

[0084] Reliability verification in extreme temperature and vibration environments

[0085] Application scenario: A helicopter performs a search and rescue mission in a desert environment (-35°C to 75°C) and encounters strong turbulent vibrations (0-1800Hz).

[0086] Key implementation details:

[0087] Temperature compensation optimization

[0088] Three-level compensation interval parameters:

[0089] Low temperature range (-40℃~0℃): Vernier matrix [V1] = [0.92, -0.15; 0.08, 1.03]

[0090] Normal temperature range (0℃~40℃): [V2]=[1.05,0.02;-0.03,0.98]

[0091] High temperature area (40℃~80℃): [V3]=[0.85,0.25;0.12,1.12]

[0092] Online identification of heat conduction equation parameters: thermal conductivity k update cycle 95ms, identification error 0.48℃.

[0093] Vibration-temperature coupling processing.

[0094] Inverse compensator design: residual phase error after zero-pole cancellation < 0.002°

[0095] Fault self-healing implementation

[0096] Dual wavelength switching: 1550nm (main) and 1310nm (backup) light source power are 20mW and 15mW respectively

[0097] Backup channel startup time is less than 10ms, using mechanical gyroscope data to maintain system stability

[0098] Measured performance data (such as Figure 3 shown).

[0099] In summary, the Sagnac phase difference signal is acquired in real time through an orthogonally distributed three-axis closed-loop fiber optic gyroscope array. The asymmetric preload winding process (the preload forces on the X / Y / Z axes are 12±0.5N, 8±0.3N, and 15±0.5N, respectively, and the diameter ratio is 1:0.8:1.2) is used to optimize the fiber ring structure, significantly reducing the non-reciprocal error. The data of 48 PT1000 temperature sensors embedded at a spacing of 5mm along the axial direction of the fiber ring and the full spectrum data of the 0-2000Hz frequency band of the three-axis vibration sensor are simultaneously collected to construct a temperature-vibration coupling error model. The environmental adaptability is improved through three-level temperature range division (-40℃~0℃, 0℃~40℃, 40℃~80℃) and inverse compensation of the 6-DOF vibration transfer function. The compensated gyroscope data are compared with the star sensor. The outputs of the sensor, MEMS accelerometer, and atmospheric data computer are aligned in time and space to establish a 12-dimensional state vector. An improved volumetric Kalman filter is used to achieve multi-source data fusion. The fusion weight is dynamically adjusted through a sliding time window (50ms length, 80% overlap) and an adaptive correlation factor to solve the problem of nonlinear error. According to the aircraft's six-degree-of-freedom kinematic equation, the Y-waveguide drive voltage and phase modulation depth are dynamically adjusted. The Sagnac interferometer signal is stabilized at the π / 2 operating point through digital phase-locked amplification technology, and feedforward compensation is introduced to eliminate phase noise caused by vibration. Dynamic PID control (Kp=2.5±0.2, Ti=0.1s, Td=0.05s) and anti-saturation control mode are used to ensure the stability of the system in a high dynamic environment. The present invention solves the non-reciprocal error caused by the symmetric winding process, the insufficient environmental adaptability of a single temperature compensation, the limited nonlinear error processing capability of the standard Kalman filter, and the stability problem of the fixed parameter controller through an asymmetric winding process, temperature-vibration coupling compensation, improved data fusion, and dynamic closed-loop control, thereby achieving high-precision and high-reliability airborne combined measurement.

[0100] Although the embodiments of the present invention have been shown and described above, it will be understood that the above embodiments are exemplary and are not to be construed as limitations on the present invention. A person skilled in the art may change, modify, replace and deform the above embodiments within the scope of the present invention.

Claims

1. An airborne combined measurement and calculation method for a closed-loop optical fiber, characterized in that: The following steps are involved: S1. Dynamic data acquisition: Sagnac phase difference signals are acquired in real time through an orthogonally distributed three-axis closed-loop fiber gyroscope array. The sampled data is time-stamped using a high-precision synchronous clock. Each fiber ring is wound using an asymmetric preload, with fixed preloads on the X / Y / Z axes. The fiber ring diameter ratio is 1:0.8:1.

2. S2. Environmental disturbance compensation: This system simultaneously collects data from 48 PT1000 temperature sensors embedded at 5mm intervals along the fiber ring axis, as well as full-spectrum data from a triaxial vibration sensor covering the 0-2000Hz frequency range. A coupling error model is constructed based on the temperature gradient distribution and vibration spectrum characteristics. An online identification and adaptive compensation algorithm is then used to update and compensate for measurement errors caused by temperature and vibration disturbances in real time. S3. Multi-source data fusion: align the compensated gyro data with the outputs of the star sensor, MEMS accelerometer, and atmospheric data computer in time and space, establishing a 12-dimensional state vector containing position, velocity, attitude angle, gyro drift, and accelerometer bias. A sliding time window is designed to achieve temporal alignment of multi-sensor data. An adaptive correlation factor is introduced into the extended Kalman filter to dynamically adjust the fusion weights. The attitude estimate of the integrated navigation system is corrected and smoothed in real time based on star sensor measurements, improving data fusion accuracy and attitude stability. S4. Closed-loop feedback control dynamically adjusts the Y-waveguide drive voltage and phase modulation depth based on the vehicle's six-degree-of-freedom kinematic equations, including three translational and three rotational degrees of freedom. Digital phase-locked amplification technology is used to stabilize the Sagnac interferometer signal at the operating point, and feedforward compensation is introduced to eliminate phase noise caused by vibration. An adaptive control algorithm or gain scheduling mechanism is used to dynamically optimize controller parameters, achieving real-time correction of gyro drift and attitude errors, thereby improving the system's responsiveness and attitude stability in highly dynamic environments.

2. The airborne combined measurement and calculation method for closed-loop optical fiber according to claim 1, characterized in that: The closed-loop feedback control of step S4 includes: The main control loop uses a digital lock-in amplifier to track the carrier signal phase in real time, with a locking accuracy of 0.01°; The auxiliary control loop generates an inverse phase disturbance signal based on the spectral characteristics of the vibration sensor, with a compensation bandwidth covering 10-1500Hz; The PID controller parameters for dynamically adjusting the Y-waveguide driving voltage are: proportional coefficient Kp = 2.5 ± 0.2, integral time Ti = 0.1 s, and differential time Td = 0.05 s; When the temperature change rate exceeds 5℃ / min, it automatically switches to anti-saturation control mode.

3. The airborne combined measurement and calculation method for closed-loop optical fiber according to claim 1, characterized in that: The temperature compensation method of step S2 includes: a. In the low temperature range of -40°C to 0°C, the first Vernier compensation coefficient matrix [V1] = [0.92, -0.15; 0.08, 1.03] is used, and the low temperature protection mode is activated to reduce the thermal stress of the fiber ring; b. In the normal temperature range of 0°C to 40°C, the second Vernier compensation coefficient matrix [V2] = [1.05, 0.02; -0.03, 0.98] is used, and the standard compensation mode is activated to monitor the axial temperature gradient in real time; c. In the high temperature range of 40°C to 80°C, the third Vernier compensation coefficient matrix [V3] = [0.85, 0.25; 0.12, 1.12] is used, and the high temperature heat dissipation mode is enabled. The fiber ring temperature rise rate is controlled to < 2°C / min through the active cooling device; The compensation operator action sequence is dynamically adjusted according to the axial temperature gradient direction of the optical fiber ring. When the axial temperature difference is detected to be greater than 3°C, the gradient compensation mode is activated. Online identification of fiber optic ring heat conduction equation parameters, with an update period of 100ms and an identification error of <0.5℃.

4. The airborne combined measurement and calculation method for closed-loop optical fiber according to claim 1, characterized in that: The vibration interference suppression method of step S2 includes: A 6-DOF vibration transfer function model is constructed, including 3 translational degrees of freedom and 3 rotational degrees of freedom, and the phase error caused by the base angular vibration is eliminated through inverse system compensation. The db6 wavelet basis is used for 8-layer wavelet packet decomposition, and 3-5 layers of detail coefficients are retained when reconstructing the signal. When high-frequency vibration > 500Hz is detected, the anti-aliasing filter cutoff frequency is automatically switched to 2000Hz, and the stopband attenuation is > 60dB.

5. The airborne combined measurement and calculation method for closed-loop optical fiber according to claim 1, characterized in that: The multi-source data fusion in step S3 further includes: Design a sliding time window to achieve multi-sensor synchronization, with a time window length of 50ms and an overlap rate of 80%; Build a three-dimensional heterogeneous data bus architecture, with the upper layer transmitting 10Gbps optical signals, the middle layer transmitting 1Mbps control signals, and the lower layer power channel ripple less than 10mV; A hardware scale factor calibration module is introduced into the extended state space model to correct the nonlinear response of the sensor in real time.

6. The airborne combined measurement and calculation method for closed-loop optical fiber according to claim 1, characterized in that: It also includes fault self-healing processing: When the phase modulator fails, it switches to dual-wavelength polarization diversity reception mode (1550nm and 1310nm); When reconstructing the interference signal, the attitude angle change rate smoothing algorithm is used to compensate for the attitude angle jump error of less than 0.001° by calculating the difference value of the attitude angle at adjacent moments in real time. When the backup channel is started, the mechanical gyroscope data is used to maintain the system stabilization time less than 10ms.

7. The airborne combined measurement and calculation method for closed-loop optical fiber according to claim 1, characterized in that: Step S1 dynamic data collection includes: Alternately arrange high birefringence photonic crystal fiber segments and conventional polarization-maintaining fiber segments; A distributed microcavity strain sensor array with a spacing of 2 mm is formed inside the optical fiber using femtosecond laser direct writing technology; The Sagnac phase calculation value is corrected in real time based on the Brillouin scattered light time domain reflectometry data, with a sampling rate of 1kHz.

8. The airborne combined measurement and calculation method for closed-loop optical fiber according to claim 1, characterized in that: Step S4 closed-loop feedback control includes: Execute spiral maneuver trajectory with angular velocity range of ±300° / s and acceleration range of ±8g; online estimate the first and second harmonic amplitudes of scale factor nonlinearity with an update period of 200ms; Deterministic error (residual < 0.001%) and random error components were separated by double time correlation analysis.

9. The airborne combined measurement and calculation method for closed-loop optical fiber according to claim 1, characterized in that: Step S3 multi-source data fusion includes: Orthogonal phase-locked amplification technology is used to extract the first-order sideband component of the carrier signal, improving the signal-to-noise ratio by more than 20dB. Sliding window singular value decomposition is performed on the sensor data, and the fault detection sensitivity reaches 0.1mV. Coordinate system conversion is achieved through the quaternion spherical linear interpolation method, and the calculation delay is less than 50μs.

Images