Fiber bragg grating acceleration sensor parameter optimization method

Abstract

A fiber bragg grating acceleration sensor parameter optimization method comprises the steps that at least one key performance parameter of a sensor is selected as an optimization target, and the key performance parameters are sensitivity and resonant frequency; constructing a parameterized model of the sensor, wherein the parameterized model is used for representing key performance parameters as a function of a plurality of geometric parameters; taking the geometric parameter combination as an optimization variable, and setting a constraint condition according to an actual application scene and physical limitation of the sensor; and taking the optimization target as a moderate function, carrying out iterative optimization on optimization variables by using a genetic algorithm, and outputting an optimal geometric parameter combination. By constructing the parameterized model and optimizing the genetic algorithm, the problem that the sensitivity and the resonant frequency are difficult to consider in the traditional method is solved, and the high-frequency performance and the measurement precision of the sensor are remarkably improved.

Description

TECHNICAL FIELD

[0001] The application belongs to the field of measurement test technology, and particularly relates to a fiber grating acceleration sensor parameter optimization method. BACKGROUND

[0002] In the field of sensor technology, fiber grating acceleration sensors are widely used in complex scenes such as high-frequency vibration measurement and multi-component vector detection due to their high sensitivity and anti-electromagnetic interference characteristics. However, the existing sensor design has a contradiction between sensitivity and resonance frequency, making it difficult to meet the high-performance detection requirements. Sensitivity improvement often leads to a decrease in resonance frequency, limiting its application in high-frequency vibration measurement. At the same time, sensor performance is affected by multiple parameters such as geometric shape, material properties, etc. Traditional manual trial and error or experience design methods cannot balance performance and cannot meet the needs of high-frequency scenes or multi-component vector detection.

[0003] The two most important evaluation indicators of fiber grating acceleration sensors are resonance frequency and sensitivity , which determine the application scenarios and performance of the detector. Their values are determined by multiple variables such as detector structure, system stiffness, and vibrator mass. In general, and are mutually restrictive and inversely proportional. In previous detector research, the design is generally based on the detection requirements (low frequency and high sensitivity or wide frequency band, etc.), and the values of and cannot be accurately controlled, and it is impossible to determine whether the design is optimal. SUMMARY

[0004] The technical problem to be solved by the present application is to overcome the shortcomings of the prior art and provide a fiber grating acceleration sensor parameter optimization method that significantly improves sensor performance and reliability and stability.

[0005] The technical solution adopted to solve the above technical problems is: a fiber grating acceleration sensor parameter optimization method, comprising the following steps:

[0006] Step 1. Select at least one key performance parameter of the sensor as the optimization target, the key performance parameter being sensitivity, resonance frequency;

[0007] Step 2. Construct a parameterized model of the sensor, the parameterized model being used to characterize the key performance parameter as a function of multiple geometric parameters;

[0008] Step 3. Combine the geometric parameters as optimization variables and set the constraint conditions according to the actual application scenarios and physical limitations of the sensor;

[0009] Step 4. Iteratively optimize the optimization variables using a genetic algorithm with the optimization objective as a fitness function, and output the optimal combination of geometric parameters.

[0010] As a preferred technical solution, the specific steps of step 4 are:

[0011] Step 4.1. Randomly generate an initial value of a set of optimization variables, and map the value range of the optimization variables to the gene coding of the genetic algorithm;

[0012] Step 4.2. Evaluate the pros and cons of each individual according to the fitness function, and select the winning individual using the roulette method or the tournament method;

[0013] Step 4.3. Randomly select two individuals from the winning individuals for gene crossover to generate new individuals;

[0014] Step 4.4. Randomly mutate part of the genes of the new individuals generated in step 4.3 to increase population diversity;

[0015] Step 4.5. Repeat the selection, crossover and mutation operations until a preset number of iterations or fitness convergence is reached, and obtain the optimal combination of geometric parameters.

[0016] As a preferred technical solution, the sensor is a high-frequency hinge type single-component sensor, and the optimization objective is to maximize the resonant frequency f≥2000Hz and the sensitivity;

[0017] The parameterized model includes a sensitivity model and a resonant frequency model, and the sensitivity model S is:

[0018]

[0019]

[0020] In the formula, is the effective photoelastic coefficient of the optical fiber, is the center wavelength of the grating, is the distance between the fixed points of the optical fiber on both sides of the grating, is the tensile stiffness of the optical fiber, h is the height of the mass block, m is the mass of the mass block, R is the radius of the hinge groove, θ is the central angle of the hinge groove, b is the length of the mass block, and K is the hinge stiffness;

[0021] The resonant frequency model is:

[0022]

[0023]

[0024]

[0025] As a preferred technical solution, the moderate function is:

[0026]

[0027] In the formula, x is an optimization variable.

[0028] As a preferred technical solution, the sensor is a bending stretch type fiber grating acceleration sensor, the structure of the sensor is that an elliptical elastic sheet is arranged on a base, a fiber is arranged in the long axis direction of the elliptical elastic sheet, a grating is engraved on the fiber, a mass block is arranged on the top of the elliptical elastic sheet, the length of the mass block is the same as the long axis of the elliptical elastic sheet, and the width of the mass block is equal to the width of the elliptical elastic sheet.

[0029] The optimization target is to maximize the quality factor FM=S0·f0, S0 is sensitivity, and f0 is a resonance frequency.

[0030] The parameterized model includes a sensitivity model and a resonance frequency model, the sensitivity model is:

[0031]

[0032] In the formula, is the central wavelength of the grating, m is the mass of the mass block, L is the length of the long axis of the elliptical elastic sheet, is the thickness of the elliptical elastic sheet, E is the Young's modulus of the elliptical elastic sheet, and w is the width of the elliptical elastic sheet, is the density of the mass block, and e is the thickness of the mass block.

[0033] The resonance frequency model is:

[0034]

[0035] In the formula, is the equivalent stiffness of the acceleration sensor.

[0036] As a preferred technical solution, the moderate function is: ;

[0037] In the formula, x is an optimization variable.

[0038] The beneficial effects of the present application are as follows:

[0039] This invention solves the problem of balancing sensitivity and resonant frequency in traditional methods by constructing a parameterized model and optimizing it using a genetic algorithm, significantly improving the high-frequency performance and measurement accuracy of the sensor. Experimental verification shows that the error between the theoretical resonant frequency and the measured value is only 1.33%, and the sensitivity error is 14.57%, verifying the accuracy and reliability of the optimization method. This method eliminates the need for repeated manual trial and error, rapidly iteratively optimizing geometric parameters through a genetic algorithm, significantly improving design efficiency and reducing development costs. The optimized sensor exhibits good linear response characteristics and a wide operating frequency band (100Hz–1800Hz), meeting the requirements for high-frequency vibration measurement and multi-component vector detection.

[0040] This invention has strong universality and provides guidance for the parameter design of sensors, which can meet the sensing requirements of different frequencies and sensitivities. Attached Figure Description

[0041] Figure 1 This is a schematic diagram of the structure of the high-frequency hinge type single-component sensor in Embodiment 1 of the present invention.

[0042] Figure 2 This is a flowchart illustrating the fiber optic grating accelerometer parameter optimization method of Embodiment 1 of the present invention.

[0043] Figure 3 The following are the results displayed after running the Matlab program in Embodiment 1 of this invention: (a) Evolutionary generations and optimal individuals; (b) Fitness curve; (c) Verification result graph.

[0044] Figure 4 These are the amplitude-frequency characteristic curve and the acceleration characteristic curve at 600 Hz of the sensor in Embodiment 1 of the present invention.

[0045] Figure 5 This is a schematic diagram of the structure of the bending and stretching fiber optic grating accelerometer sensor according to Embodiment 2 of the present invention.

[0046] Figure 6 This is the curve showing the evolutionary generation, optimal individual, and fitness after running the Matlab program in Embodiment 2 of the present invention.

[0047] Figure 7 This is a simulation result diagram of Embodiment 2 of the present invention. Detailed Implementation

[0048] The present invention will be further described in detail below with reference to the accompanying drawings and embodiments, but the present invention is not limited to the following embodiments.

[0049] Example 1

[0050] This embodiment uses a high-frequency hinged single-component sensor as an example. The structure of the sensor is as follows:Figure 1 As shown, the geometric parameters of this sensor are: mass block width Length of mass block Mass block height Minimum hinge thickness Hinge groove radius Central angle Hinge height Hinge width .

[0051] exist Figure 2 The fiber Bragg grating accelerometer parameter optimization method of this embodiment includes the following steps:

[0052] Step 1. Select the sensor's resonant frequency f≥2000Hz and maximize its sensitivity as the optimization objective;

[0053] Step 2. Construct a parameterized model of the sensor. The parameterized model is used to characterize key performance parameters as functions of multiple geometric parameters. In this embodiment, the parameterized model includes a sensitivity model and a resonant frequency model, wherein the sensitivity model S is:

[0054]

[0055]

[0056] In the formula, For sensitivity, The effective elastic-optical coefficient of the optical fiber. The center wavelength of the grating This is the distance between the fiber optic fixing points on both sides of the grating. denoted as , where h is the tensile stiffness of the optical fiber, m is the height of the mass block, R is the radius of the hinge groove, θ is the central angle subtended by the hinge groove, b is the length of the mass block, and K is the hinge stiffness.

[0057] In this embodiment, the central angle θ subtended by the hinge groove is:

[0058]

[0059] Among them, the resonant frequency model for:

[0060]

[0061]

[0062]

[0063] Step 3. Use the combination of geometric parameters as optimization variables, and set constraints according to the actual application scenario and physical limitations of the sensor;

[0064] This embodiment will specify the length of the mass block. Mass block height Mass block width Hinge groove radius Minimum hinge thickness Hinge height Hinge width As an optimization variable, the combination of variables, for the needs of high-frequency exploration, includes the following constraints: , , , , °;

[0065] Step 4. Use the optimization objective as a fitness function. In this embodiment, the fitness function is... for:

[0066]

[0067] In the formula, x is the optimization variable. The resonant frequency, Sensitivity.

[0068] Step 5. Use a genetic algorithm to iteratively optimize the optimization variables and output the optimal combination of geometric parameters, specifically:

[0069] Step 5.1. Randomly generate a set of 500 individuals, each representing a set of optimization variables. This maps the range of values ​​for the optimization variables to the gene encoding of the genetic algorithm.

[0070] Step 5.2. Evaluate the merits of each individual based on the fitness function, and select the winning individual using roulette wheel selection;

[0071] Step 5.3. Randomly select two individuals from the winning individuals for gene crossover, with a crossover probability of 0.8, to generate new individuals;

[0072] Step 5.4. Randomly mutate some genes of the new individuals generated in Step 5.3 with a mutation probability of 0.3 to increase population diversity;

[0073] Step 5.5. Repeat steps 5.2 to 5.4 until the preset number of iterations (2000) is reached or the fitness converges, yielding the optimal combination of geometric parameters. =5.3 mm =15.3 mm =8.1 mm =7.6 mm =1.3mm =10.1 mm =7.4 mm, such as Figure 3 .

[0074] The optimal combination of geometric parameters corresponds to a resonant frequency f of 2072 Hz and a sensitivity of [missing information]. It was 15.378 pm / g.

[0075] The simulation results showed a resonant frequency of 2043.5 Hz, which is consistent with the simulation results obtained by the parameter optimization method of this invention.

[0076] To further verify the beneficial effects of the present invention, the inventors fabricated a physical sensor using the optimal geometric parameters obtained by the parameter optimization method of the present invention for testing. The specific steps are as follows:

[0077] First, the amplitude-frequency characteristics of the fiber optic sensor were tested: with a fixed acceleration of 2 g, measurements were taken from 100 Hz to 3000 Hz in 100 Hz steps. The drift of the FBG center wavelength was recorded, and the amplitude-frequency characteristic curve was plotted, as shown below. Figure 4 As shown in (a).

[0078] Depend on Figure 4 (a) It can be seen that:

[0079] (1) The flat operating frequency band of the sensor is 100 Hz to 1800 Hz, which proves that the sensor has a good response to high frequency vibration;

[0080] (2) The resonant frequency is around 2100 Hz, which is in good agreement with the previously calculated theoretical value of 2072 Hz and the simulated value of 2043.5 Hz. The error between the theoretical and experimental values ​​is [missing value]. .

[0081] Next, the output frequency of the vibration table was fixed at 600 Hz, and the acceleration value was set at 2 m / s². 2 The step size is from 2 m / s 2 Gradually increase to 10 m / s 2 The amplitude output by the sensor at different accelerations at 600 Hz is as follows: Figure 4 As shown in (b). By linearly fitting the relationship between the magnitude of acceleration and the amplitude of wavelength drift, the slope of the fitted line represents the sensor sensitivity, reaching 18.02 pm / g with a coefficient of determination of 0.999. This proves that the sensor has a linear response within the operating frequency band (600 Hz), and the error between the theoretical and experimental sensitivity values ​​is [missing value]. .

[0082] Therefore, the fiber optic grating accelerometer parameter optimization method of this invention successfully achieves the optimization goal of maximizing sensitivity while ensuring the resonant frequency is not lower than 2000 Hz. The optimal geometric parameter combination obtained through the genetic algorithm (b=5.3 mm, h=15.3 mm, c=8.1 mm, R=7.6 mm, t=1.3 mm, h1=10.1 mm, c1=7.4 mm) theoretically achieves a resonant frequency of 2072 Hz and a sensitivity of 15.378 pm / g. Simulation results (resonant frequency 2043.5 Hz) are basically consistent with the theoretical value with small errors, verifying the accuracy of the parameterized model. Physical testing further confirms the sensor's performance. In summary, this parameter optimization method effectively balances resonant frequency and sensitivity, and the designed sensor meets the requirements of high-frequency vibration measurement, possessing high engineering application value.

[0083] The optimized high-frequency accelerometer sensor of this invention is compared with the parameters of high-frequency FBG detectors published in recent years, as shown in Table 1. The sensor achieves a balance between good high-frequency response capability and high sensitivity, thus achieving a balance between the two core parameters.

[0084] Table 1. Comparison of parameters between the optimized high-frequency accelerometer of this invention and recently published high-frequency FBG detectors.

[0085] Example 2

[0086] This embodiment uses a bending-stretching fiber Bragg grating accelerometer as an example. The sensor's structure consists of an elliptical spring plate 3 mounted on a base 4, an optical fiber 2 positioned along the major axis of the elliptical spring plate 3, a grating etched onto the optical fiber, and a mass block 1 positioned on the top of the elliptical spring plate 3. The length of the mass block 1 is the same as the major axis of the elliptical spring plate 3, and its width is equal to the width of the elliptical spring plate 3. Figure 5 The sensor's geometric parameters are: the width of the elliptical spring piece 3... Thickness of elliptical spring piece 3 The length of the three major axes of the elliptical spring. ;

[0087] The parameter optimization method for the bending-stretching fiber Bragg grating accelerometer in this embodiment includes the following steps:

[0088] Step 1. Maximize the quality factor FM = S0·f0 as the optimization objective, where S0 is the sensitivity and f0 is the resonant frequency;

[0089] Step 2. Construct a parameterized model of the sensor. The parameterized model is used to characterize key performance parameters as functions of multiple geometric parameters. In this embodiment, the parameterized model includes a sensitivity model and a resonant frequency model. The sensitivity model... for:

[0090]

[0091] In the formula, E is the center wavelength of the grating, and E is the Young's modulus of the elliptical spring. The density of the mass block;

[0092] Resonant frequency model for:

[0093]

[0094] In the formula, It is the equivalent stiffness of the accelerometer.

[0095] Step 3. Use the combination of geometric parameters as optimization variables, and set constraints according to the actual application scenario and physical limitations of the sensor;

[0096] This embodiment will specify the thickness of the spring sheet. The combination of width w and mass block thickness e is used as optimization variables, with the following constraints: 20Hz≤f0≤100Hz, 0.5×10⁻10≤S0≤2×10⁻10;

[0097] Step 4. Use the optimization objective as a fitness function. In this embodiment, the fitness function is... for:

[0098]

[0099] In the formula, x is the optimization variable;

[0100] Step 5. Use a genetic algorithm to iteratively optimize the optimization variables and output the optimal combination of geometric parameters, specifically:

[0101] Step 5.1. Randomly generate a set of 500 individuals, each representing a set of optimization variables. This maps the range of values ​​for the optimization variables to the gene encoding of the genetic algorithm.

[0102] Step 5.2. Evaluate the merits of each individual based on the fitness function, and select the winning individual using a tournament method;

[0103] Step 5.3. Randomly select two individuals from the winning individuals for gene crossover, with a crossover probability of 0.8, to generate new individuals;

[0104] Step 5.4. Randomly mutate some genes of the new individuals generated in Step 5.3 with a mutation probability of 0.3 to increase population diversity;

[0105] Step 5.5. Repeat steps 5.2 to 5.4 until the preset number of iterations (2000) is reached or the fitness converges, yielding the optimal combination of geometric parameters. =0.1 mm =17.3 mm =5.1 mm, =40 mm, such as Figure 6 .

[0106] The optimal combination of geometric parameters corresponds to a resonant frequency f0 of 54.43 Hz and a sensitivity of... It was 1966.4 pm / g.

[0107] like Figure 7 The simulation results showed a resonant frequency of 52.893 Hz, which was verified by the parameter optimization method of this invention. The resonant frequency obtained was basically consistent with the simulation results.

Claims

1. A method for optimizing parameters of a fiber Bragg grating accelerometer, characterized in that, Includes the following steps: Step 1. Select at least one key performance parameter of the sensor as the optimization target, wherein the key performance parameter is sensitivity or resonant frequency; Step 2. Construct a parameterized model of the sensor, wherein the parameterized model is used to characterize the key performance parameters as functions of multiple geometric parameters; Step 3. Use the combination of geometric parameters as optimization variables, and set constraints according to the actual application scenario and physical limitations of the sensor; Step 4. Using the optimization objective as a fitness function, use a genetic algorithm to iteratively optimize the optimization variables and output the optimal combination of geometric parameters.

2. The method for optimizing fiber optic accelerometer parameters according to claim 1, characterized in that, The specific steps of step 4 are as follows: Step 4.

1. Randomly generate a set of initial values ​​for optimization variables and map the range of values ​​of the optimization variables to the gene encoding of the genetic algorithm; Step 4.

2. Evaluate the merits of each individual based on the fitness function, and select the winning individual using roulette wheel or tournament methods; Step 4.

3. Randomly select two individuals from the winning individuals to perform gene crossover and generate new individuals; Step 4.

4. Randomly mutate some genes of the new individuals generated in Step 4.3 to increase population diversity; Step 4.

5. Repeat the selection, crossover, and mutation operations until the preset number of iterations or fitness convergence is reached to obtain the optimal combination of geometric parameters.

3. The method for optimizing fiber optic accelerometer parameters according to claim 1 or 2, characterized in that: The sensor is a high-frequency hinged single-component sensor, and the optimization objective is to maximize the resonant frequency f ≥ 2000 Hz and the sensitivity. The parameterized model includes a sensitivity model and a resonant frequency model, wherein the sensitivity model S is: In the formula, The effective elastic-optical coefficient of the optical fiber. The center wavelength of the grating This is the distance between the fiber optic fixing points on both sides of the grating. denoted as , where h is the tensile stiffness of the optical fiber, m is the height of the mass block, R is the radius of the hinge groove, θ is the central angle subtended by the hinge groove, b is the length of the mass block, and K is the hinge stiffness. The resonant frequency model for: 。 4. The fiber optic accelerometer parameter optimization method according to claim 3, characterized in that, The appropriateness function is: In the formula, x is the optimization variable.

5. The method for optimizing fiber optic accelerometer parameters according to claim 1 or 2, characterized in that: The sensor is a bending and stretching fiber optic accelerometer. The sensor has the following structure: an elliptical spring is set on the base, an optical fiber is set along the long axis of the elliptical spring, a grating is engraved on the optical fiber, and a mass block is set on the top of the elliptical spring. The length of the mass block is the same as the long axis of the elliptical spring, and the width is equal to the width of the elliptical spring. The optimization objective is to maximize the quality factor FM = S0·f0, where S0 is the sensitivity and f0 is the resonant frequency. The parameterized model includes a sensitivity model and a resonant frequency model. The sensitivity model... for: In the formula, λ is the center wavelength of the grating, and m is the mass of the mass block. L is the length of the major axis of the elliptical spring. Let E be the thickness of the elliptical spring, E be the Young's modulus of the elliptical spring, and w be the width of the elliptical spring. Let be the density of the mass block, and e be the thickness of the mass block; The resonant frequency model for: In the formula, It is the equivalent stiffness of the accelerometer.

6. The method for optimizing fiber optic accelerometer parameters according to claim 5, characterized in that: The appropriateness function is: In the formula, x is the optimization variable.

Images