A resonant fiber-optic gyroscope fast operation simulation method

Abstract

This invention provides a fast computational simulation method for resonant fiber optic gyroscopes, comprising a resonant fiber optic gyroscope signal processing and closed-loop control model. The method includes: initially setting simulation parameters, selecting modulation waveforms, simulating the closed-loop control model, wherein multi-turn interference coupling of the simulated resonant cavity employs multi-core parallel computation; synchronous sampling, demodulation, and filtering; and state machine judgment of state transitions to determine whether to continue frequency sweeping or frequency locking control of the light source, thereby updating the output data. The systematic and rapid computational simulation method provided by this invention can shorten simulation time, improve simulation efficiency, and completely simulate the resonant fiber optic gyroscope signal processing closed-loop system, facilitating observation of the output results of each waveform modulation and demodulation, and providing theoretical support for hardware debugging.

Description

Technical Field

[0001] This invention belongs to the field of sensor technology, and relates to inertial sensor technology based on the Sagnac effect, and in particular to a fast computation simulation method for resonant fiber optic gyroscopes. Background Technology

[0002] A resonator-fiber-optic gyroscope (RFOG) measures the gyroscope's angular velocity by detecting the difference in resonant frequencies of two opposing beams within a fiber ring resonator (FRR). In the RFOG signal processing system, the laser frequency needs to be locked to the resonant frequency of one of the beams, and the difference in frequency between the two beams is proportional to the rotational speed of the fiber ring resonator. Fluctuations in various related parameters affect the overall performance of the RFOG and the gyroscope's detection sensitivity. Therefore, modeling the resonator-fiber-optic gyroscope signal processing system allows for simulation verification of the relevant theories.

[0003] Currently, most signal processing systems for resonant fiber optic gyroscopes (RFOGs) are implemented in FPGAs. However, problems only become apparent during the debugging phase. RFOG systems are highly complex, time-consuming, and involve cumbersome procedures, requiring pre-debugging system modeling and simulation to provide theoretical support for hardware debugging. In recent years, numerous simulation methods and models related to RFOGs have emerged, but a fast and efficient simulation method is still lacking. Therefore, a fast and effective system simulation model is needed to simulate and verify the signal processing of RFOG systems. However, the simulation computation is enormous and time-consuming, requiring further optimization of the computational methods to meet the requirements. Summary of the Invention

[0004] To address the issues of incomplete and time-consuming simulation of RFOG signal processing systems, this invention provides a systematic and fast parallel computing method for simulating resonant fiber optic gyroscopes.

[0005] To achieve the above objectives, the present invention provides the following technical solution:

[0006] A fast computation simulation method for resonant fiber optic gyroscopes includes resonant fiber optic gyroscope signal processing and a closed-loop control model, comprising the following steps: initially setting simulation-related parameters, selecting modulation waveforms, simulating the closed-loop control model, wherein multi-turn interference coupling of the simulated resonant cavity adopts multi-core parallel computation; synchronous sampling, demodulation, and filtering, and performing state machine judgment to determine state transition, thereby deciding whether to continue frequency sweeping or frequency locking control of the light source, and updating the output data.

[0007] Furthermore, the specific steps include the following:

[0008] Step 1: Set the simulation parameters;

[0009] Step 2: Allocate storage space for each output value array and select the code for the input waveform according to the required modulation waveform;

[0010] Step 3: Start system simulation. The two modulated light beams are coupled into the fiber optic resonant cavity, where they interfere to form clockwise and counterclockwise beams that propagate multiple times around the cavity. The beams are then converted into voltage signals by a photodetector. The detector's output signal is input to an A / D converter module, where the time signal is converted into an AD sampling time format and then into a digital signal. Synchronous demodulation and filtering are applied to the sampled output signal, followed by low-pass filtering. This achieves the function of phase-locked amplification of the photodetector's output signal.

[0011] Step 4: The PID feedback module calculates the error feedback quantity based on the demodulation output and the frequency sweep control quantity. If the error feedback quantity is not 0, the closed-loop control continues until the laser center frequency is locked at the clockwise CW resonant frequency point, and the final error signal is 0. At this time, the counterclockwise CCW digital signal is used as the gyroscope open-loop output signal after synchronous demodulation and low-pass filtering by the demodulation filter 1 module. Whether to perform frequency sweep control or frequency locking control on the light source needs to be determined by the state machine to determine the state transition.

[0012] Step 5: Update the output data, pass the signal values ​​obtained from the simulation calculation to each output value array, calculate each performance parameter, including: free spectral width FSR, full width at half maximum (FWHM), fineness F, and resonance depth; analyze and compare the data results obtained from the simulation.

[0013] Furthermore, step 1 specifically includes:

[0014] (1) Set the global runtime and step runtime of the simulation, define the number of coupling cycles inside the resonant cavity (looptimes), the effective refractive index (n), and the control signal V fed back to the laser; simulate the FPGA clock pulse (stepTFPGA), the number of bits of the DAC (dacN), the maximum input range of the DAC (dacVFs), the sweep frequency (scanFreq), the sweep interval (scanScalr), the sweep threshold (scanRange), the number of bits of the ADC (adcN), and simulate the maximum input voltage of the AD (adcVFs);

[0015] (2) The least significant bit (dacLSBV) of the D / A conversion can be calculated based on the number of bits (dacN) of the DAC and the maximum input range (dacVFs) of the DAC, and is expressed as: Similarly, the least significant bit adcLSBV of the A / D sample is expressed as:

[0016] (3) Set the relevant parameters of the laser, including: the wavelength λ of the output light wave, and the center frequency f of the laser output light. laser Output optical power I0, laser output light amplitude E0, effective refractive index n, free spectral width FSR, and number of longitudinal mode q near the wavelength;

[0017] (4) Set the relevant parameters of the fiber optic resonator, including: cavity length L, coupling coefficient k of the coupler. c The loss coefficient α of the coupler c The loss factor α of light waves propagating one revolution within the FRR L ;

[0018] (5) Calculate the relevant parameters of the resonant cavity, including: the transmission coefficient of the light wave propagating one revolution in the FRR: Transmission coefficient of the through end: The transmission coefficient of the light wave after it enters the FRR through the coupler cross-end, travels one revolution, and returns to the straight-through end: Resonance depth of the resonance spectral line: The transit time of a light wave for one revolution is τ = n * L / c, where c is the speed of light, n is the refractive index of the light wave, and L is the cavity length.

[0019] (6) Set modulation and demodulation related parameters: Select the corresponding modulation and demodulation waveforms, and set the clockwise and counterclockwise modulation frequencies and modulation amplitudes;

[0020] (7) Set the parameters related to simulated optical noise, including: backscattering coefficient R b Shot noise of photodetectors, laser frequency noise, relative intensity noise, including white noise, pink noise, 1 / f noise, and thermal fluctuations;

[0021] (8) Set the relevant parameters of the low-pass filter, including: passband cutoff frequency fp, cutoff frequency fc, sampling frequency fs, maximum passband ripple attenuation, stopband attenuation, and filter order; based on the relevant parameters of the low-pass filter, obtain the numerator coefficient group IIRb and the denominator coefficient group IIRa.

[0022] (9) Set the relevant parameters for PID feedback control, including: proportional gain k p Integral amplification factor k I Differential amplification factor k d .

[0023] Furthermore, a direct type II low-pass filter is used for low-pass filtering.

[0024] Furthermore, the output value array in step 2 specifically includes:

[0025] The outputs of the two photodetectors include the backscatter output, the control signal fed back to the laser, the center frequency of the laser output light, the amplitude-frequency characteristics and frequency response of the low-pass filter, the analog running time, the AD input signal, the synchronous demodulation output, and the two-channel filter output.

[0026] The rules for selecting the corresponding code for the input waveform according to the required modulation waveform are as follows: no modulation is 0, sine modulation is 1, triangular wave modulation is 2, sawtooth wave modulation is 3, square wave modulation is 4, and trapezoidal wave modulation is 5.

[0027] Furthermore, the simulation in step 3 includes simulation of the optical part and the electrical signal processing part. The optical part includes the coupling transmission process of the resonant cavity. The transmission process is divided into two types: one is the output light field superposition form, and the other is the transfer function form. The electrical part includes the real-time adjustment of the laser center frequency, that is, the sum of the initial value and the frequency sweep feedback signal. When PI control is introduced, it is adjusted to the sum of the initial value and the feedback control signal.

[0028] Furthermore, the optical section includes two interference coupling mathematical models: if it is an output light field superposition form, the output superposition light field is calculated based on the Jones matrix and the input waveform; if it is a transfer function form, the output light field magnitude is calculated based on the amplitude transfer function and the phase transfer function. First, the clockwise output frequency, counterclockwise output frequency, laser output center frequency, resonant cavity light field output, photodetector output, synchronous demodulation output, low-pass filter output, and state machine state judgment are calculated in parallel. After one round of calculation, the corresponding signal value is output. At the same time, during the continuous calculation process, the values ​​of each register are updated in parallel, and the generated variable values ​​are stored in the corresponding registers, including laser frequency sweep or frequency locking control, backscatter output update, control quantity update, A / D output update, demodulation output update, and time update modules.

[0029] Furthermore, in step 4, whether to perform frequency sweeping or frequency locking control on the light source depends on the state transition process determined by the state machine:

[0030] The state machine consists of four states: state0, state1, state2, and state3. Combinational logic is used to determine state transition conditions and describe the state transition rules. State0 is the initial state, and its output is to sweep the laser frequency. The state transition is determined based on the first valley entry criterion: if the valley entry criterion is met, it transitions to state1; otherwise, it returns to state0. State1 outputs to continue sweeping the laser frequency, and the state transition is determined based on the second valley entry criterion: if the valley entry criterion is met, it transitions to state2; otherwise, it returns to state1. State2 has a delay period before frequency locking to ensure accuracy. It remains in state2 until the set delay value is reached, after which it transitions to state3. State3 outputs frequency locking control, calculating the error feedback quantity V based on the proportional gain, integral gain, and differential gain, and feeding it back to the laser. When the control signal V exceeds the maximum input range, V becomes zero.

[0031] Furthermore, step 5 also includes the following processes: releasing the remaining storage space, analyzing and comparing the simulation results obtained under different modulation and demodulation frequencies, performing time-domain waveform analysis and spectrum analysis on the output data, and drawing graphs.

[0032] Compared with the prior art, the present invention has the following advantages and beneficial effects:

[0033] The systematic and rapid computational simulation method provided by this invention can shorten simulation time, improve simulation efficiency, and completely simulate the closed-loop system of resonant fiber optic gyroscope signal processing, facilitating the observation of the output results of each waveform modulation and demodulation. Therefore, this invention can provide theoretical support for hardware debugging. Parallel computing can save time and costs, reduce the resources and time consumed in the simulation process, and simplify complex simulation procedures. The simulation program of this invention implements parallel design, which facilitates clear thinking and results, and can further improve simulation efficiency. Attached Figure Description

[0034] Figure 1 This is a schematic diagram of a resonant fiber optic gyroscope signal detection system.

[0035] Figure 2 Here is a flowchart of the simulation algorithm for a resonant fiber optic gyroscope.

[0036] Figure 3 A schematic diagram of parallel computing simulation for a resonant fiber optic gyroscope;

[0037] Figure 4 This is a diagram showing the resonance curve of a resonant fiber optic gyroscope.

[0038] Figure 5 This is a demodulation output curve of a resonant fiber optic gyroscope.

[0039] Figure 6 Output curve of a frequency-locked control gyroscope for a resonant fiber optic gyroscope;

[0040] Figure 7 A comparison chart of the simulation time performance of a resonant fiber optic gyroscope with frequency locking. Detailed Implementation

[0041] The technical solutions provided by the present invention will be described in detail below with reference to specific embodiments. It should be understood that the following specific embodiments are only used to illustrate the present invention and are not intended to limit the scope of the present invention.

[0042] This method uses MATLAB software to establish a signal processing and closed-loop control model for a resonant fiber optic gyroscope. The closed-loop control model is simulated, employing multi-core parallel computation to simulate multi-turn interference coupling in the resonant cavity. Synchronous sampling, demodulation, and filtering are performed, and a state machine is used to determine state transitions, thereby deciding whether to continue frequency sweeping or frequency locking control of the light source, updating the output data, shortening computation time, and improving simulation efficiency. The parallel computation simulation method for a resonant fiber optic gyroscope provided in this embodiment of the invention specifically includes the following steps:

[0043] Step 1: Establish a model of the resonant fiber optic gyroscope signal detection system:

[0044] Figure 1 This is a schematic diagram of a resonant fiber optic gyroscope signal detection system. The laser output light is split into two beams with the same intensity and frequency. One beam (clockwise CW) is used for frequency locking, and the other beam (counterclockwise CCW) is used as the gyroscope signal output.

[0045] The phase modulator modulates two light waves. Taking a sine wave as an example, modulation wave 1 is a sine wave with amplitude V1 and modulation frequency f1, V1sin(2πf1t); modulation wave 2 is a sine wave with amplitude V2 and modulation frequency f2, V2sin(2πf2t). The two modulated light waves are coupled into the fiber resonant cavity and interfere with each other in the cavity to form clockwise and counterclockwise beams that propagate around the cavity multiple times.

[0046] Next, the CCW output and CW output undergo photoelectric conversion through photodetector 1 and photodetector 2 respectively, converting the optical signal into a voltage signal. After signal amplification, the outputs of photodetector 1 and photodetector 2 are obtained. The output signals of the detectors are input to the A / D converter module, which converts them into digital quantities. After synchronous demodulation by demodulation and filtering module 2, the CW digital signal determines whether sawtooth wave frequency sweep control or PID feedback control is used. The module calculates the error feedback based on the demodulated output. If the error feedback is not zero, closed-loop control continues until the laser frequency is locked at the CW resonant frequency point, and the final error signal is zero. At this point, the CCW digital signal, after synchronous demodulation and low-pass filtering by demodulation and filtering module 1, is amplified and used as the open-loop output signal of the gyroscope.

[0047] Step 2: Based on the above resonant fiber optic gyroscope signal detection system model, according to... Figure 2 The flowchart of the simulation algorithm for a resonant fiber optic gyroscope, and the specific implementation steps are as follows:

[0048] Step 1: Set simulation parameters. Specifically, the simulation duration for each simulation step can be arbitrarily set by the user. Initially, set the global runtime and step runtime of the simulation, define the number of coupling cycles inside the resonant cavity (looptimes), the effective refractive index (n), and the control signal V fed back to the laser; simulate the FPGA clock pulse (stepTFPGA), the DAC bit depth (dacN), the DAC maximum input range (dacVFs), the sweep frequency (scanFreq), the sweep interval (scanScalr), the sweep threshold (scanRange), the ADC bit depth (adcN), and the simulated maximum AD input voltage (adcVFs).

[0049] Step 2: Further, based on the number of bits DAC (dacN) and the maximum input range DAC (dacVFs) from Step 1, the least significant bit (dacLSBV) of the D / A conversion can be calculated, which can be expressed as: Similarly, the least significant bit adcLSBV of A / D sampling can be expressed as:

[0050] Step 3: Set the relevant parameters of the laser, including: the wavelength λ of the output light wave is 1550nm, and the center frequency f of the laser output light. laser =c / λ=193THz, output optical power I0 is 10mW, laser output light amplitude E0 is 86.80J, effective refractive index n=1.455, free spectral width FSR=c / (n*L), where L is the cavity length of the resonant cavity, and the number of longitudinal modes near the wavelength q=floor(f laser / FSR), floor() points to the floor pointer and rounds down to the nearest integer less than or equal to the element;

[0051] Step 4: Set the relevant parameters of the fiber optic resonant cavity, including: cavity length L is 13m, and the coupling coefficient k of the coupler. c The loss coefficient α of the coupler is 0.05. c The loss coefficient α of light waves propagating one revolution within the FRR is 0.11. L It is 0.0065;

[0052] Step 5: Based on the parameters in Step 4, calculate the relevant parameters of the resonant cavity, including the transmission coefficient of the light wave propagating one revolution within the FRR. Transmission coefficient of the through end: The transmission coefficient of the light wave after it enters the FRR through the coupler cross-end, travels one revolution, and returns to the straight-through end: Resonance depth of the resonance spectral line: The transit time of a light wave for one revolution is τ = n * L / c, where c is the speed of light, n is the effective refractive index of the light wave, and L is the cavity length.

[0053] Step 6: Set the modulation and demodulation parameters. Select the corresponding case for the modulation and demodulation waveforms, and set the modulation frequency f. CW and f CCW The modulation amplitude is V CW and V CCW Set the parameters related to simulated optical noise: backscattering coefficient R b The noise level is 0.01. This includes shot noise, laser frequency noise, relative intensity noise (including white noise, pink noise, 1 / f noise), and thermal fluctuations in the photodetector. Low-pass filter parameters are set as follows: passband cutoff frequency fp = 50Hz, cutoff frequency fc = 5kHz, sampling frequency fs = 1 / stepTFPGA, maximum passband ripple attenuation, stopband attenuation, and filter order. A direct type II low-pass filter is used, which saves buffer space and speeds up simulation compared to a direct type I low-pass filter. Under the same performance conditions, the order is much lower than that of an FIR filter. Furthermore, based on the low-pass filter parameters, the numerator coefficient group IIRb and the denominator coefficient group IIRa can be obtained: [IIRb,IIRa] = butter(filterOrder,fc / (fs / 2)), where butter() refers to the Butterworth filter used for designing the low-pass filter; filterOrder refers to the filter order. PID feedback control parameters are set as follows: proportional gain k. p Integral amplification factor k I Differential amplification factor k d ;

[0054] Step 7: Allocate storage space for each output value array, including: the two light intensities Icw and Iccw output from the photodetector, the backscatter outputs BWIcw and BWIccw, the control signal LaserV fed back to the laser, the center frequency of the laser output light LaserFreq, the amplitude-frequency characteristics and frequency response of the low-pass filter, the analog running time time, the AD input signals adCWVal and adCCWVal, the demodulation outputs DemodIcwVal and DemodIccwVal, and the two filter outputs CWFilter and CCWFilter.

[0055] Step 8: Input the waveform according to the required modulation waveform and select the code: 0 for no modulation, 1 for sine modulation, 2 for triangular wave modulation, 3 for sawtooth wave modulation, 4 for square wave modulation, and 5 for trapezoidal wave modulation. Then start the system simulation.

[0056] Step 9: The system then performs simulations of the optical and electrical signal processing parts. The optical part includes the coupling and transmission process of the resonant cavity, which is divided into two types: one is the output light field superposition form, and the other is the transfer function form. The electrical part includes the real-time adjustment of the laser center frequency, i.e., the sum of the initial value and the frequency sweep feedback signal. When PI control is introduced, the adjustment is the sum of the initial value and the feedback control signal. Simulate the FPGA's A / D sampling. adCWIn and adCCWIn respectively acquire the signals pdCWout and pdCCWout output by the photodetector. The two A / D outputs are: adCWIn = floor(pdCWout / adcLSBV)*adcLSBV, adCCWIn = floor(pdCCWout / adcLSBV)*adcLSBV. After the CW digital signal is synchronously demodulated by the demodulation and filtering module 2, the output is: demodIcw = demod_wave_CW(2*pi*f CW *tFPGA)*adCWIn;

[0057] Then, a low-pass filter is performed using a fourth-order direct type II filter, and the algorithm is as follows:

[0058] CWIIRX(1) = demodIcw;

[0059] CWs0=demodIcw-a1*CWs1-a2*CWs2-a3*CWs3-a4*CWs4;

[0060] CWYout=(CWs0*b0+CWs1*b1+CWs2*b2+CWs3*b3+CWs4*b4);

[0061] CWIIRY(1) = CWYout;

[0062] CWs4=CWs3; CWs3=CWs2; CWs2=CWs1; CWs1=CWs0;

[0063] After signal amplification, the output is: outFiltCW = CWYout * 50; where CWYout is the CW filter output, CWs0, CWs1, CWs2, CWs3, and CWs4 are temporary variables, and CWIIRX and CWIIRY are the numerator and denominator coefficients, respectively. The PID feedback module calculates the error feedback quantity based on the demodulated output and the frequency sweep control quantity. If the error feedback quantity is not zero, the closed-loop control continues until the laser frequency is locked at the CW resonant frequency point, and the final error signal is zero.

[0064] Furthermore, in step 9, the sawtooth wave sweep frequency module is used to scan the resonant frequency, find the resonant frequency, and then use a PI controller to lock the resonant frequency of the resonant cavity. Its principle is to scan the voltage; when the output of the demodulation and filtering module drops from its peak value to zero, a trough-judging range is set, and the sweep frequency interval after D / A conversion determines the sweep frequency accuracy.

[0065] Then, the AD signal is synchronously demodulated and low-pass filtered. The valley range is then determined based on the filtered output value. If the valley criterion is met, PI control is performed, and the PI error output is fed back to the laser to lock the laser center frequency at the CW resonant frequency point. If it is not within the valley range, the laser output frequency is swept until the next demodulation zero point is found.

[0066] Further, in step 9, whether to perform frequency sweeping or frequency locking control on the light source depends on the state transition of the state machine. The state machine's state transition judgment module includes four states: state0, state1, state2, and state3. Combinatorial logic is used to determine the state transition conditions and describe the state transition rules. The initial state is state0. If the state is state0 or state1, then frequency sweeping is performed on the light source; if the state is state3, then frequency locking control is performed on the light source. state=0: Initial state, the output is to sweep the laser frequency, and the state transition is determined according to valley entry criterion 1. If valley entry criterion 1 is met, it transitions to state1; otherwise, it returns to state0. state=1: The output is to sweep the laser frequency, and the state transition is determined according to valley entry criterion 2. If valley entry criterion 2 is met, it transitions to state2; otherwise, it returns to state1. state=2: Delay, a delay is performed before frequency locking to ensure the accuracy of frequency locking. State2 is maintained until the set delay value is reached, and then it transitions to state3 after the set delay value. state=3: The output is frequency locking control, based on the proportional amplification factor k. p Integral gain kI Differential amplification factor k d The error feedback quantity V is calculated and fed back to the laser. When the control signal V exceeds the maximum input range dacVFs, V becomes zero. In state 3, the angular velocity is input, and the relationship between the frequency difference and the angular velocity is obtained by observing the open-loop output. The PI control algorithm is as follows:

[0067] PID.Intergate=PID.Intergate+(outFiltCW);

[0068] out = k p *(outFiltCW)+k I *PID.Intergate+k d *(outFiltCW-PID.lastVal);

[0069] PID.lastVal = outFiltCW;

[0070] V = PID.Vinit + floor(out / dacLSBV) * dacLSBV; where PID.Intergate is the sum of errors under integral action, PID.Vinit represents the initial error, and k is the derivative amplification factor in PI control. d Set to zero;

[0071] Step 10: Update the output data, transferring the simulation-calculated data to the storage space of each output value; calculate performance parameters such as FSR, full width at half maximum (FWHM), fineness F, and resonance depth; release the remaining storage space, analyze and compare the simulation results under different modulation and demodulation frequencies, perform time-domain waveform analysis and spectrum analysis on the output data, and draw graphs, including the resonance curve of the photodetector output, the demodulated filter output, the resonance curve of the backscattered light wave, the power spectrum of the demodulated output, and the signal control process.

[0072] The schematic diagram of the parallel computing simulation of the resonant fiber optic gyroscope of the present invention is shown below. Figure 3 As shown, T is defined k-1 For the previous sampling period, T k For the current sampling period, T k+1The next sampling period is set. Simulated resonant cavity optical wave interference coupling is the core of the parallel computing module. If the output light field is superimposed, the superimposed output light field is calculated based on the Jones matrix and the input waveform; if it is a transfer function, the magnitude of the output light field is calculated based on the amplitude transfer function and phase transfer function. In the parallel computing module, each computational unit works simultaneously, greatly improving computational efficiency. First, the CW resonant frequency, CCW resonant frequency, laser output center frequency, resonant cavity light field output, photodetector output, synchronous demodulation output, low-pass filter output, and state machine state judgment are calculated in parallel. After one round of computation, the corresponding signal value is output. Simultaneously, during continuous computation, the values ​​of various registers are updated in parallel, and the generated variable values ​​are stored in the corresponding registers, including laser frequency sweep or frequency locking control, backscatter output update, control quantity update, A / D output update, demodulation output update, and time update modules. The parallel computing process is executed once in each sampling period, with an overall simulation time of approximately 16 seconds, demonstrating high computational efficiency.

[0073] The resonant curve of the resonant fiber optic gyroscope of the present invention is shown in the figure below. Figure 4 As shown, the modulation waveform is selected as sinusoidal modulation, the modulation frequency is 400kHz, and the modulation amplitude is set to 2.405. The resonance curve output by the photodetector is the resonance output under high-frequency modulation. The demodulation output curve of the resonant fiber optic gyroscope of this invention is shown in the figure below. Figure 5 As shown, in frequency sweep mode, the CW demodulated output signal is used as the open-loop output of the gyroscope. The output curve of the frequency-locked control gyroscope for the resonant fiber optic gyroscope is shown in the figure below. Figure 6 As shown, in frequency-locked control mode, and with angular velocity introduced, it can be seen that the demodulated output includes angular velocity information, and frequency locking is achieved. The simulation time performance comparison graph of the resonant fiber optic gyroscope frequency locking is shown below. Figure 7 As shown, the simulation running time required for frequency locking in the original and parallel light field superposition forms is mainly compared. The number of cyclic couplings in the light field superposition form is set to 100, 150, 200, 250, and 300, respectively. Figure 7 This indicates that when the number of loop couplings is the same, parallel computing can reduce the simulation frequency locking time. As the number of loops increases, the required frequency locking time also increases, and the superiority of parallel computing becomes more obvious.

[0074] The technical means disclosed in this invention are not limited to those disclosed in the above embodiments, but also include technical solutions composed of any combination of the above technical features. It should be noted that those skilled in the art can make various improvements and modifications without departing from the principles of this invention, and these improvements and modifications are also considered within the scope of protection of this invention.

Claims

1. A fast computation simulation method for a resonant fiber optic gyroscope, characterized in that, The model includes resonant fiber optic gyroscope signal processing and closed-loop control, comprising the following steps: initial setting of simulation parameters, selection of modulation waveform, simulation of closed-loop control model, wherein multi-turn interference coupling of simulated resonant cavity adopts multi-core parallel operation; synchronous sampling, demodulation and filtering, and state machine judgment of state transition, thereby determining whether to continue frequency sweeping or frequency locking control of the light source, and updating output data; Specifically, the steps include the following: Step 1: Set the simulation parameters; Step 2: Allocate storage space for each output value array and select the code for the input waveform according to the required modulation waveform; Step 3: Start system simulation. The two modulated light waves are coupled into the fiber optic resonant cavity and interfere with each other inside the cavity, forming clockwise and counterclockwise beams that propagate around the cavity multiple times. The photoelectric conversion is then completed by a photodetector, converting the optical signal into a voltage signal. The output signal of the detector is input to the A / D converter module, where the time is converted into AD sampling time form and then into a digital quantity. Synchronous demodulation and filtering are performed on the sampled output signal, followed by low-pass filtering. To achieve phase-locked amplification of the output signal of the photodetector; Step 4: The PID feedback module calculates the error feedback quantity based on the demodulation output and the frequency sweep control quantity. If the error feedback quantity is not 0, the closed-loop control continues until the laser center frequency is locked at the clockwise CW resonant frequency point, and the final error signal is 0. At this time, the counterclockwise CCW digital signal is used as the gyroscope open-loop output signal after synchronous demodulation and low-pass filtering by the demodulation filter 1 module. Whether to perform frequency sweep control or frequency locking control on the light source needs to be determined by the state machine to determine the state transition. Step 5: Update the output data, pass the signal values ​​obtained from the simulation calculation to each output value array, and calculate each performance parameter, including: Free Spectral Width (FSR), Full Width at Half Maximum (FWHM), Fineness (F), and Resonance Depth. Analyze and compare the data results obtained from the simulation.

2. The fast computation simulation method for resonant fiber optic gyroscopes according to claim 1, characterized in that, Step 1 specifically includes: (1) Set the global runtime and step runtime of the simulation, define the number of coupling cycles inside the resonant cavity (looptimes), the effective refractive index (n), and the control signal V fed back to the laser; simulate the FPGA clock pulse (stepTFPGA), the number of bits of the DAC (dacN), the maximum input range of the DAC (dacVFs), the sweep frequency (scanFreq), the sweep interval (scanScalr), the sweep threshold (scanRange), the number of bits of the ADC (adcN), and simulate the maximum input voltage of the AD (adcVFs); (2) Calculate the least significant bit (dacLSBV) of the D / A conversion based on the number of bits (dacN) of the DAC and the maximum input range (dacVFs) of the DAC, expressed as: Similarly, the least significant bit adcLSBV of the A / D sample is expressed as: ; (3) Set the relevant parameters of the laser, including: the wavelength λ of the output light wave, and the center frequency f of the laser output light. laser Output optical power I0, laser output light amplitude E0, effective refractive index n, free spectral width FSR, and number of longitudinal mode q near the wavelength; (4) Set the relevant parameters of the fiber optic resonator, including: cavity length L, coupling coefficient k of the coupler. c The loss coefficient α of the coupler c The loss factor α of light waves propagating one revolution within the FRR L ; (5) Calculate the relevant parameters of the resonant cavity, including: the transmission coefficient of the light wave propagating one revolution in the FRR: Q= Transmission coefficient at the straight-through end: The transmission coefficient of the light wave after it enters the FRR through the coupler cross-end, travels one revolution, and returns to the straight-through end: The resonance depth of the resonance spectral line: The transit time of a light wave for one revolution is τ = n * L / c, where c is the speed of light, n is the refractive index of the light wave, and L is the cavity length. (6) Set modulation and demodulation related parameters: Select the corresponding modulation and demodulation waveforms, and set the clockwise and counterclockwise modulation frequencies and modulation amplitudes; (7) Set the parameters related to simulated optical noise, including: backscattering coefficient R b Shot noise of photodetectors, laser frequency noise, relative intensity noise, including white noise, pink noise, 1 / f noise, and thermal fluctuations; (8) Set the relevant parameters of the low-pass filter, including: passband cutoff frequency fp, cutoff frequency fc, sampling frequency fs, maximum passband ripple attenuation, stopband attenuation, and filter order; based on the relevant parameters of the low-pass filter, obtain the numerator coefficient group IIRb and the denominator coefficient group IIRa. (9) Set the relevant parameters for PID feedback control, including: proportional gain k p Integral amplification factor k I Differential amplification factor k d .

3. The fast computation simulation method for resonant fiber optic gyroscopes according to claim 2, characterized in that, The low-pass filter uses a direct type II low-pass filter.

4. The fast computation simulation method for resonant fiber optic gyroscopes according to claim 1, characterized in that, The output value array in step 2 specifically includes: The outputs of the two photodetectors include the backscatter output, the control signal fed back to the laser, the center frequency of the laser output light, the amplitude-frequency characteristics and frequency response of the low-pass filter, the analog running time, the AD input signal, the synchronous demodulation output, and the two-channel filter output. The rules for selecting the corresponding code for the input waveform according to the required modulation waveform are as follows: no modulation is 0, sine modulation is 1, triangular wave modulation is 2, sawtooth wave modulation is 3, square wave modulation is 4, and trapezoidal wave modulation is 5.

5. The fast computation simulation method for resonant fiber optic gyroscopes according to claim 1, characterized in that, The simulation in step 3 includes the simulation of the optical part and the electrical signal processing part. The optical part includes the coupling transmission process of the resonant cavity. The transmission process is divided into two types: one is the output light field superposition form, and the other is the transfer function form. The electrical part includes the real-time adjustment of the laser center frequency, that is, the sum of the initial value and the frequency sweep feedback signal. When PI control is introduced, it is adjusted to the sum of the initial value and the feedback control signal.

6. The fast computation simulation method for resonant fiber optic gyroscopes according to claim 5, characterized in that, The optical part includes two interference coupling mathematical models: if it is a superposition of output light fields, the superimposed output light field is calculated based on the Jones matrix and the input waveform; If the output is in transfer function form, the magnitude of the output light field is calculated based on the amplitude transfer function and the phase transfer function. First, the clockwise output frequency, counterclockwise output frequency, laser output center frequency, resonant cavity light field output, photodetector output, synchronous demodulation output, low-pass filter output, and state machine state judgment are calculated in parallel. After one round of calculation, the corresponding signal value is output. Simultaneously, during continuous calculation, the values ​​of various registers are updated in parallel, and the generated variable values ​​are stored in the corresponding registers. This includes laser frequency sweep or frequency locking control, backscatter output update, control quantity update, A / D output update, demodulation output update, and time update modules.

7. The fast computation simulation method for resonant fiber optic gyroscopes according to claim 1, characterized in that, In step 4, whether to perform frequency sweeping or frequency locking control on the light source depends on the state machine's determination of the state transition process. The state machine consists of four states: state0, state1, state2, and state3. Combinational logic is used to determine state transition conditions and describe the state transition rules. State0 is the initial state, and its output is to sweep the laser frequency. The state transition is determined based on the first valley entry criterion: if the valley entry criterion is met, it transitions to state1; otherwise, it returns to state0. State1 outputs to continue sweeping the laser frequency, and the state transition is determined based on the second valley entry criterion: if the valley entry criterion is met, it transitions to state2; otherwise, it returns to state1. State2 has a delay period before frequency locking to ensure accuracy. It remains in state2 until the set delay value is reached, after which it transitions to state3. State3 outputs frequency locking control, calculating the error feedback quantity V based on the proportional gain, integral gain, and differential gain, and feeding it back to the laser. When the control signal V exceeds the maximum input range, V becomes zero.

8. The fast computation simulation method for resonant fiber optic gyroscopes according to claim 1, characterized in that, Step 5 also The process includes the following steps: releasing the remaining storage space, analyzing and comparing the simulation results obtained under different modulation and demodulation frequencies, performing time-domain waveform analysis and spectrum analysis on the output data, and plotting graphs.

Images