Improved vibration compensation method suitable for atom interference gravimeter

Abstract

The invention discloses an improved vibration compensation method suitable for an atom interference gravimeter. The method comprises the following steps: acquiring an atom interference signal and a vibration voltage signal output by a seismometer; initializing population parameters of the improved whale optimization algorithm, wherein the population parameters comprise a search space of a population size, a maximum number of iterations, a gain coefficient K and a delay coefficient tau; the method comprises the following steps: generating an initial whale population by adopting a method of combining Logistic chaotic mapping and Latin hypercube sampling LHS; calculating a phase shift caused by vibration noise and an atomic interference phase delta phi, and calculating the fitness of the whale individual by taking a root-mean-square error of compensated interference fringes as a target function; iteratively updating the population position through an improved whale optimization algorithm, outputting an optimal gain coefficient and an optimal delay coefficient, and obtaining an atomic interference signal after vibration compensation; according to the atomic interference gravimeter vibration compensation method, the global search capability, the local development precision and the stability of a vibration compensation algorithm are improved.

Description

Technical Field

[0001] The present invention relates to the technical field of gravity measurement, and in particular to an improved vibration compensation method applicable to atomic interferometric gravimeters. Background Art

[0002] Absolute gravity measurement is one of the important means in the fields of geophysical exploration, resource development, earth information monitoring, aided navigation, and national defense. At present, high-precision absolute gravity measurement mainly relies on two types of instruments: optical absolute gravimeters and atomic interferometric absolute gravimeters. Among them, the atomic interferometric absolute gravimeter utilizes the motion characteristics of cold atomic clouds freely falling in a vacuum, and excites atomic interference phenomena through laser pulses, thereby measuring the absolute value of the gravitational acceleration. Due to its high sensitivity and long-term stability, it has become the main development direction of the next generation of high-precision gravity measurement instruments.

[0003] In practical applications, the measurement accuracy of atomic interferometric gravimeters is easily interfered by ground vibration noise, which mainly enters the system through Raman mirrors and affects the phase of interference fringes. High-frequency vibrations above 1 Hz are generated by human activities, mechanical equipment, buildings, etc. Medium-frequency vibrations between 0.1 and 1 Hz originate from natural phenomena such as ground pulsations and earthquakes. Low-frequency vibrations below 0.1 Hz, such as atmospheric pressure fluctuations, although changing slowly, cannot be ignored in long-term measurements. These vibrations cause the movement of the mirror, resulting in phase noise of the laser, interfering with the precise measurement of atomic interference fringes, and seriously reducing the accuracy and stability of gravity measurement.

[0004] In response to ground vibration noise, scholars at home and abroad have proposed various vibration compensation techniques: The Paris Observatory in France combines passive vibration isolation systems with vibration compensation, and uses IIR filters to process the output signals of seismometers to correct the phase noise caused by vibrations. The University of Hanover in Germany uses digital high-pass and low-pass filters to process seismometer signals to improve the accuracy of mirror vibration signals. Zhejiang University proposes to correct the transfer function of seismometers through inversion software to improve the accuracy of vibration signals; National University of Defense Technology considers the vertical vibration of the mirror and horizontal coupling noise, and uses accelerometers and methods of accurately calibrating transfer functions for compensation.

[0005] Although various vibration compensation methods at home and abroad have achieved certain compensation effects, there are still technical bottlenecks such as local optimal traps, insufficient boundary constraint processing, and low search efficiency in complex dynamic environments. Summary of the Invention

[0006] Based on the technical problems existing in the background art, the present invention proposes an improved vibration compensation method applicable to atomic interferometric gravimeters, which improves the global search ability, local development accuracy, and stability of the vibration compensation algorithm.

[0007] An improved vibration compensation method applicable to an atomic interferometer gravimeter proposed by the present invention includes:

[0008] Obtain the atomic interference signal and the vibration voltage signal output by the seismometer;

[0009] Initialize the population parameters of the improved whale optimization algorithm, including the population size, the maximum number of iterations, the search space of the gain coefficient K and the delay coefficient τ;

[0010] Generate the initial whale population by combining the Logistic chaotic mapping and the Latin hypercube sampling LHS;

[0011] Calculate the phase shift caused by the vibration noise and the atomic interference phase ΔΦ, use the root mean square error of the compensated interference fringes as the objective function, and calculate the fitness of the whale individuals;

[0012] Iteratively update the population position through the improved whale optimization algorithm, output the optimal gain coefficient and the optimal delay coefficient, obtain the atomic interference signal after vibration compensation, and realize the vibration compensation of the atomic absolute gravimeter.

[0013] Further, in the method of generating the initial whale population by combining the Logistic chaotic mapping and the Latin hypercube sampling LHS, specifically:

[0014] Divide the population N particles into two parts, the population size N logistic generated by the Logistic chaotic mapping = β·N particles , and the population size N lhs generated by the Latin hypercube sampling LHS = (1-β)·N particles , where β is the proportionality coefficient;

[0015] Initialize N logistic and N lhs respectively, and combine the two generated sequences to obtain the initial whale population.

[0016] Further, the sample points generated by the Logistic chaotic mapping:

[0017]

[0018] where a and b are the minimum and maximum values of the parameters respectively, r is the chaotic parameter of the Logistic mapping, usually taking the value of 4, x i ∈(0,1) is the Logistic chaotic sequence, is the parameter value after mapping x i ;

[0019] Divide the parameter range [a, b] evenly into N intervals, and the sample points generated by Latin Hypercube Sampling (LHS):

[0020]

[0021] where is the interval width, m is the index of the current interval width, rand represents a random number in the range [0, 1), is the parameter value after mapping the width of the j-th interval.

[0022] Furthermore, in the process of iteratively updating the population position by the improved whale optimization algorithm, specifically:[[]]

[0023] Use a non-linear Sigmoid control factor and an adaptive factor to dynamically adjust the search parameters of the vector, and balance global search and local optimization;

[0024] According to the probability parameter p and the search vector A, select the strategies of surrounding the prey, spiral attack or random search to update the population position;

[0025] Randomly apply Gaussian distribution mutation to the updated whale positions to increase the population diversity in the next iteration;

[0026] Use the mirror reflection boundary treatment method to perform reflection mapping on the variables that exceed the boundary and return to the search space;

[0027] Iterate into the step of calculating the fitness of whale individuals until the maximum number of iterations is reached.

[0028] Furthermore, in the process of using a non-linear Sigmoid control factor and an adaptive factor to dynamically adjust the search parameters of the vector, the expressions of the non-linear Sigmoid control factor and the adaptive factor are as follows:

[0029]

[0030] where a Sigmoid is the non-linear Sigmoid control factor, a Adaptive is the adaptive factor, k is the steepness of the Sigmoid, iter is the current number of iterations, and iter Max is the maximum number of iterations.

[0031] Furthermore, in the process of selecting the strategies of surrounding the prey, spiral attack or random search to update the population position according to the probability parameter p, specifically:[[]]

[0032] When p < 0.5 and |A| < 1, use the strategy of surrounding the prey to update the population position;

[0033] When p < 0.5 and |A| ≥ 1, use the random search strategy to update the population position;

[0034] When p ≥ 0.5, the spiral attack is adopted to update the population position.

[0035] Furthermore, in the random Gaussian distribution mutation applied to the updated whale position, specifically: when rand[0,1] < 0.2, the Gaussian distribution mutation is applied to the updated whale position, and the position update with disturbance is as follows:

[0036]

[0037] where, is the whale position after introducing disturbance, is the updated whale position, and iter is the current iteration number.

[0038] Furthermore, in the mirror reflection boundary processing method for reflecting and mapping variables beyond the boundary, the mirror reflection boundary processing method is specifically as follows:

[0039] When x is less than the lower bound x min , x = x min +(x min -x);

[0040] When x is greater than or equal to the upper bound x max , x = x max -(x max -x);

[0041] where, x is the current whale position, x min , x max are the lower limit value and the upper limit value of the search space respectively.

[0042] A computer device, comprising a memory, a processor, and a computer program stored on the memory and executable on the processor, wherein when the processor executes the computer program, the vibration compensation method of the atomic interferometer gravity meter as described above is implemented.

[0043] A computer-readable storage medium, on which a number of classification programs are stored, and the number of classification programs are used to be called by a processor and execute the vibration compensation method of the atomic interferometer gravity meter as described above.

[0044] The advantages of the vibration compensation method for atomic interferometric gravimeters provided by the present invention are as follows: It is used to estimate the simplified model of the transfer function of the Raman reflector and is used for the correction of the interference fringes of the atomic gravimeter, aiming to improve the global search ability, local development accuracy and stability of the vibration compensation algorithm to meet the high-precision application requirements in the laboratory; A method combining Logistic chaotic mapping and Latin Hypercube Sampling (LHS) is used to initialize the whale population; A non-linear Sigmoid control factor and an adaptive factor are introduced to achieve the balance between global search and local optimization and enhance the convergence performance of the algorithm; The ability of the algorithm to jump out of the local optimal solution is enhanced through the Gaussian distribution mutation operator to expand the search space; The mirror reflection method is adopted to solve the boundary problem to ensure the continuity and stability of the search solution. Description of the Drawings

[0045] Figure 1 It is a schematic flowchart of the present invention;

[0046] Figure 2 It is a schematic diagram of the iteration curve;

[0047] Figure 3 It is a comparative curve graph before and after compensation;

[0048] Figure 4 It is the output U of the seismometer s and the actual vibration velocity v of the mirror ref There is a schematic diagram of the transfer function between them;

[0049] Figure 5 It is a schematic diagram of a spiral attacking prey. Detailed Embodiments

[0050] Next, the technical solution of the present invention will be described in detail through specific embodiments. Many specific details are set forth in the following description in order to fully understand the present invention. However, the present invention can be implemented in many other ways different from those described herein, and those skilled in the art can make similar improvements without departing from the connotation of the present invention. Therefore, the present invention is not limited by the specific embodiments disclosed below.

[0051] As Figures 1 to 5 shown, the vibration compensation method for atomic interferometric gravimeters proposed by the present invention includes the following steps:

[0052] Step 1: Obtain the atomic interference signal P and the vibration voltage signal U output by the seismometer s ;

[0053] The atomic absolute gravimeter is based on the principle of cold atom matter wave interference. By applying laser pulses to cold atoms, the splitting, coherent manipulation, and interference process of atomic wave packets are realized. Usually, the π / 2-π-π / 2 three-pulse laser scheme is adopted, and its function is similar to that of an optical Mach-Zehnder interferometer. Assuming that the atom is initially in the ground state |g>, under the action of the first π / 2 laser pulse, the atomic wave packet is equally likely to be divided into two coherent states |g> and |e>. After a free evolution time T, the atoms in the two states propagate along different paths; under the action of the π pulse, the energy levels and momentum directions are exchanged, and they continue to propagate along their respective paths; after another free evolution time T, the second π / 2 pulse causes the separated atomic wave packets to be superposed again, forming an interference signal. Satisfying:

[0054] P = A + B cosΔΦ; (1)

[0055] In the formula, A and B are the contrast ratios of the interference fringes, which can be obtained by cosine fitting of the interference fringes; ΔΦ is the atomic interference phase, as follows:

[0056]

[0057] In the formula, is the interference phase generated by vibration noise, that is, the term that needs to be vibration-compensated; α is the Raman light chirp rate; k eff is the equivalent wave vector of the Raman light; T is the time interval between two adjacent Raman light pulses.

[0058] From formula (1) and formula (3), the gravitational acceleration g can be determined:

[0059]

[0060] During the gravity measurement process, vibration noise severely limits the measurement accuracy, and the vibration of the Raman mirror is one of the important factors. The atomic interference signal P and the vibration velocity signal U output by the seismometer s are recorded simultaneously. The vibration voltage signal U output by the seismometer s is used to calculate the phase shift caused by the vibration of the mirror (formula 6); the atomic interference signal P is used to generate the objective function (the root mean square error of the compensated interference fringes).

[0061] The Raman mirror is placed directly above the seismometer. U s can approximately reflect the vibration v of the Raman mirror ref . However, due to the limitations of the mechanical structure and the performance of the seismometer, there is a transfer function between the seismometer output U s and the true vibration v of the mirror ref , and their relationship is as Figure 4 shown.

[0062] In Figure 4Among them, T1 represents the transfer function from the ground vibration velocity to the vibration velocity of the Raman mirror, T2 represents the transfer function from the ground vibration velocity to the vibration velocity measured by the seismometer, and T3 represents the transfer function of the seismometer itself. Therefore, the actual vibration velocity v of the Raman mirror ref and the vibration voltage signal U output by the seismometer s have the following transfer function between them:

[0063]

[0064] Therefore, it is necessary to restore the true vibration of the mirror as accurately as possible through the vibration voltage signal Us output by the seismometer to achieve the ultimate goal of vibration compensation for the interference fringes. The key point of restoring the true vibration lies in correctly solving the transfer function H, and the accuracy of H directly affects the effect of vibration compensation. However, in the actual measurement process, due to complex factors such as the earthquake propagation path, the internal mechanical structure and dynamic characteristics of the seismometer, it is difficult to directly measure T1 and T2, and T3 is not an ideal linear transfer function. Therefore, it is necessary to reasonably simplify the existing research. Referring to the optical gravimeter, the gain coefficient K and the delay coefficient τ are introduced to simplify the transfer function, and the simplified transfer function model is as follows:

[0065]

[0066] Among them, K s represents the nominal sensitivity of the seismometer, which can be found in the instrument operation manual. v s is the velocity signal measured by the seismometer. Given the gain coefficient K and the delay coefficient τ, the phase shift caused by the vibration of the mirror is calculated through the following formula for vibration compensation:

[0067]

[0068] Among them, t represents time, and g s (t) is the sensitivity function of the atomic gravimeter with π / 2 - π - π / 2 Raman pulses, which is expressed as follows:

[0069]

[0070] Theoretically, when the gain coefficient K and the delay coefficient τ are closer to the true values, the more it can reflect the true vibration of the mirror, and the more beneficial it is to the later vibration compensation of the interference fringes.

[0071] Step 2: Initialize the population parameters of the improved whale optimization algorithm, including the population size, the maximum number of iterations, and the search spaces of the gain coefficient K and the delay coefficient τ;

[0072] Each sample point in the whale population represents a pair (K, τ), which are two optimization objectives used to correct the vibration transfer function (Formula (5)). Different value combinations of the sample points (K, τ) within the preset search space, where K ∈ [K L , K U , and τ ∈ [τ L , τ U . By improving the whale optimization algorithm to iteratively optimize (K, τ), the objective function (RMSE) is minimized, thereby achieving precise compensation for vibration noise.

[0073] Step 3: Generate the initial whale population by combining the Logistic chaotic mapping and Latin hypercube sampling LHS;

[0074] Since the traditional whale optimization algorithm is sensitive to the distribution of the initial population. If the initial population distribution is unreasonable, it may lead to incomplete coverage of the search space, thus affecting the algorithm performance. Therefore, this embodiment proposes a method that combines the Logistic chaotic mapping and Latin hypercube sampling LHS to initialize the whale population. Combining the strong global search ability of the Logistic chaotic mapping and the uniform coverage of the Latin hypercube sampling LHS not only increases the population diversity but also improves the global search ability.

[0075] Divide the population N particles into two parts. The population size N logistic generated by the Logistic chaotic mapping is β · N particles , and the population size N lhs generated by the Latin hypercube sampling LHS is (1 - β) · N particles , where β is the proportionality coefficient, preferably β equals 0.3. Initialize N logistic and N lhs respectively, and then merge the two generated sequences to obtain the initial whale population.

[0076] Sample points generated by the Logistic chaotic mapping:

[0077]

[0078] Among them, a and b are the minimum and maximum values of the parameters respectively, r is the chaotic parameter of the Logistic mapping, usually taking the value of 4, x i ∈ (0, 1) is the Logistic chaotic sequence, is the parameter value after x i is mapped;

[0079] Divide the parameter range [a, b] evenly into N intervals. Sample points generated by the Latin hypercube sampling LHS:

[0080]

[0081] Among them, is the interval width, m is the index of the current interval width, rand represents a random number in the range [0, 1), is the parameter value after mapping the j-th interval width.

[0082] Finally, the mapped parameter value y i is expressed as follows:

[0083]

[0084] Step 4: Calculate the phase shift caused by vibration noise and the atomic interference phase ΔΦ. Taking the root mean square error of the compensated interference fringes as the objective function, calculate the fitness of the whale individuals;

[0085] Using the population positions initialized in Step 3, calculate the atomic interference phase ΔΦ through formula (2), and calculate the interference phase generated by vibration noise through formula (6) Compensate the calculated interference phase into the atomic interference phase ΔΦ, and calculate the root mean square error value during the cosine fitting of the compensated interference fringes;

[0086] Step 5: Iteratively update the population positions through the improved whale optimization algorithm, output the optimal gain coefficient and the optimal delay coefficient, and obtain the atomic interference signal after vibration compensation, that is, realize the vibration compensation of the atomic absolute gravimeter.

[0087] According to Steps 1 to 5, this embodiment is used to estimate the simplified model of the Raman reflector transfer function and is used for the correction of the interference fringes of the atomic gravimeter, aiming to improve the global search ability, local development accuracy and stability of the vibration compensation algorithm to meet the high-precision application requirements of the laboratory; combining the Logistic chaos mapping and the Latin hypercube sampling LHS method to initialize the whale population; introducing a non-linear Sigmoid control factor and an adaptive factor to achieve the balance between global search and local optimization and enhance the convergence performance of the algorithm; enhancing the ability of the algorithm to jump out of the local optimal solution and expanding the search space through the Gaussian distribution mutation operator; using the mirror reflection method to solve the boundary problem and ensure the continuity and stability of the search solution.

[0088] In one embodiment, in Step 5, iteratively update the population positions through the improved whale optimization algorithm, specifically including S1 to S5:

[0089] S1. Use a non-linear Sigmoid control factor and an adaptive factor to dynamically adjust the search parameters for the vector, balancing global search and local optimization;

[0090] In this embodiment, the whale searches for prey in three ways: surrounding the prey, spiral attack, or random search strategy, specifically as follows (a1) to (a3):

[0091] (a1) Surround the prey;

[0092] Taking the current optimal candidate solution as the target prey, the humpback whale can identify the position of the prey and surround it, and this behavior is represented by the following equation:

[0093]

[0094] where A and C are coefficient vectors, is the position vector of the currently obtained optimal solution, is the current position vector. It should be noted here that should be updated in each iteration.

[0095] The calculation formulas for vectors A and C are as follows:

[0096]

[0097] where a is a control factor that linearly decreases from 2 to 0 during the iteration process (in the exploration and exploitation phases), and r is a random vector in [0, 1].

[0098] This embodiment proposes the expressions of the non - linear Sigmoid control factor and the adaptive factor as follows:

[0099]

[0100] where a Sigmoid is the non - linear Sigmoid control factor, a Adaptive is the adaptive factor, k is the Sigmoid steepness, iter is the current iteration number, and iter Max is the maximum iteration number.

[0101] Then the calculation formula for vector A becomes:

[0102]

[0103] The non - linear control factor a Sigmoid provides non - linear dynamic regulation, and through the curve characteristics of the Sigmoid function, endows the algorithm with different search capabilities in different stages. The adaptive factor a Adaptive provides linear dynamic regulation and gradually narrows the search range. The combination of the two enables the whale individuals to conduct a large - range search within the entire search space in the initial stage, covering more possible solution regions. As the number of iterations increases, a Adaptive decreases accordingly, and a SigmoidAs it increases, the global search ability and local development ability gradually balance. In the later stage, the closer it is to the maximum number of iterations iter Max , a Sigmoid The closer it is to 1, the search range focuses on fine-tuning near the optimal solution.

[0104] (a2) Spiral to attack the prey;

[0105] Simulate the spiral process of the humpback whale, that is, the bubble net attack behavior, which is to move along a spiral path closer to the prey. As Figure 5 shown, this method first calculates the distance between the whale located at (X,Y) and the prey located at (X * ,Y * ). Then a spiral equation is established between the positions of the whale and the prey to mimic the spiral movement of the humpback whale, as follows:

[0106]

[0107] where, represents the distance from the i-th whale to the prey (optimal solution), c is a constant defining the spiral shape, and l is a random number within the range of [-1,1].

[0108] In this embodiment, the humpback whale swims around the prey within a shrinking circle while swimming along a spiral path. IWOA randomly selects two behaviors of surrounding the prey and spiral attacking by introducing a probability parameter p to simulate this simultaneous behavior. The mathematical model is as follows:

[0109]

[0110] where the probability parameter p is a random number between [0,1].

[0111] (a3) Randomly search for the prey;

[0112] The adaptive change of the search vector A allows the IWOA algorithm to smoothly transition between exploration and exploitation. When |A|>1, the algorithm focuses on exploration, otherwise it is used for exploitation. The mathematical model is as follows:

[0113]

[0114] where, is a random position vector (random whale) selected from the current population.

[0115] S2. Update the population position by selecting the strategies of surrounding the prey, spiral attacking or randomly searching according to the probability parameter p and the search vector A;

[0116] where the specific details of the strategies of surrounding the prey, spiral attacking or randomly searching are shown in step S1.

[0117] S3. Randomly apply Gaussian distribution mutation to the updated whale positions to increase the population diversity in the next iteration;

[0118] Generate perturbation values consistent with the spectral characteristics of vibration noise using Gaussian distribution, and randomly mutate the updated whale positions to avoid the optimization process falling into local optima and increase the population diversity in the next iteration. When the random number rand ∈ [0, 1) < 0.2, the position update with perturbation is as follows:

[0119]

[0120] where, is the whale position after introducing perturbation, and N(.) is the normal distribution or Gaussian distribution.

[0121] S4. Use the mirror reflection boundary treatment method to perform reflection mapping on the variables that exceed the boundary and return to the search space; the mathematical model is as follows:

[0122] When x is less than the lower bound x min , x = x min +(x min -x);

[0123] When x is greater than or equal to the upper bound x max , x = x max -(x max -x);

[0124] where, x is the whale position, and x min , x max are the lower and upper limit values of the search space, respectively.

[0125] Regarding the problem of particle out-of-bounds, on the basis of retaining a certain degree of randomness, the mirror reflection method avoids invalid calculations caused by out-of-bounds, effectively prevents falling into boundary solutions, ensures the legality of the solutions and the continuity in the optimization process, and at the same time enhances the search ability near the boundary.

[0126] S5. Iteratively enter the step of calculating the fitness of whale individuals until the maximum number of iterations is reached.

[0127] Through steps S1 to S5, the improved whale optimization algorithm simulates the social behavior of humpback whales and realizes the application in vibration compensation of atomic interferometric gravimeters. Apply this embodiment to the vibration compensation of atomic interferometric gravimeters in the laboratory environment. By minimizing the root mean square error RMSE, find the whale position corresponding to the minimum RMSE and output the optimal gain coefficient K opt and the optimal delay coefficient τ opt information contained in this whale position. Calculate K opt and τ optAt this time, the vibration speed of the Raman mirror during each atomic interference is obtained, and then the phase change caused by the vibration is obtained, and the atomic interference fringes are corrected, that is, the phase shift caused by the vibration noise is efficiently optimized and compensated, and the influence of the vibration interference on the measurement accuracy is reduced. Compared with the traditional algorithm, this embodiment has been improved in terms of global search ability, local optimization accuracy, boundary handling robustness, and convergence speed, and a more efficient and accurate vibration compensation solution is achieved.

[0128] In one embodiment, the vibration compensation method of the atomic interferometer gravimeter is specifically as follows:

[0129] Step1. Set the basic parameters and determine the population size and the maximum number of iterations, and set the search space of the gain coefficient K and the delay coefficient τ. Import the vibration voltage signal U s and the atomic interference signal P;

[0130] Step2. Initialize the whale population using the Logistic-LHS chaotic mapping, where the Logistic-LHS chaotic mapping is a method that combines the Logistic chaotic mapping and the Latin hypercube sampling LHS;

[0131] Step3. Using the initialized population positions, perform vibration compensation through formulas (6) and (2), and calculate the fitness of the whale individuals with the minimum RMSE value of the fitted fringes after compensation as the objective function.

[0132] Step4. Update the positions of the whale population according to formulas (8)-(16), and select the attack method of the whale according to the parameters |A| and p.

[0133] Among them, the method of selecting the attack method of the whale according to the parameters |A| and p is specifically as follows:

[0134] When p < 0.5 and |A| < 1, update the population position using the strategy of surrounding the prey;

[0135] When p < 0.5 and |A| ≥ 1, update the population position using the random search strategy;

[0136] When p ≥ 0.5, update the population position using the spiral attack.

[0137] Step5. If rand(.) < 0.2, randomly mutate the updated whale positions according to formula (17), and check whether the population exceeds the boundary. If it exceeds, process it by the mirror reflection method.

[0138] Step6. Save the memories of these individuals and save the optimal individual in the current population.

[0139] Step 7. Iterate back to Step 3. If the algorithm meets the termination condition, that is, the maximum number of iterations is reached as set, stop the iteration and output the optimal position;

[0140] Step 8. Finally, output the optimal gain k opt and the delay time τ opt , obtain the RMSE value of the corrected fringe and its cosine fitting, so as to obtain the atom interference signal after vibration compensation, and realize the vibration compensation of the atom absolute gravimeter.

[0141] Finally, through Steps 1 to 8, the iteration curve as Figure 2 shown can be obtained. It can be seen that the fitness value after the eighth iteration is 0.066982 and the obtained optimal proportional gain Kopt is equal to 1.07, and the optimal delay time τopt is equal to 0.028 s. Finally, combined with Figure 3 the comparison chart before and after compensation shown, IWOA has the characteristic of fast convergence in the process of optimizing the transfer function parameters, which proves that IWOA is applicable to quickly search for the optimal parameter combination of the simplified transfer function.

[0142] As mentioned above, it is only the preferred specific implementation manner of the present invention, but the protection scope of the present invention is not limited thereto. Any person skilled in the art within the technical scope disclosed by the present invention, according to the technical solution and inventive concept of the present invention, makes equivalent replacement or change, and should be covered by the protection scope of the present invention.

Claims

1. An improved vibration compensation method applicable to atomic interferometric gravimeters, characterized in that, Including: Obtaining the atomic interference signal and the vibration voltage signal output by the seismometer; Initializing the population parameters of the improved whale optimization algorithm, including the population size, the maximum number of iterations, the search space of the gain coefficient K and the delay coefficient τ; Generating an initial whale population by combining the Logistic chaotic map and the Latin hypercube sampling LHS; Calculating the phase shift caused by vibration noise and the atomic interference phase ΔΦ, using the root mean square error of the compensated post-interference fringes as the objective function, and calculating the fitness of individual whales; Iteratively updating the population position through the improved whale optimization algorithm, outputting the optimal gain coefficient and the optimal delay coefficient, obtaining the atomic interference signal after vibration compensation, and realizing the vibration compensation of the atomic absolute gravimeter.

2. The vibration compensation method according to claim 1, characterized in that, In the method of generating an initial whale population by combining the Logistic chaotic map and the Latin hypercube sampling LHS, specifically: Divide the population N particles into two parts, the population quantity N generated by the Logistic chaotic map logistic = β·N particles , and the population quantity N generated by the Latin hypercube sampling LHS lhs = (1 - β)·N particles , where β is the proportionality coefficient; Initialize N logistic and N lhs respectively, and merge the two generated sequences to obtain the initial whale population.

3. The vibration compensation method according to claim 1, wherein Sample points generated by the Logistic chaotic map: Among them, a and b are the minimum and maximum values of the parameters respectively, r is the chaotic parameter of the Logistic map, usually taking the value of 4, and x i ∈(0,1) is the Logistic chaotic sequence, is the parameter value after the mapping of x i ; Dividing the parameter range [a, b] evenly into N intervals, and sample points generated by the Latin hypercube sampling LHS: wherein, is the interval width, m is the index of the current interval width, and rand represents a random number in the range [0, 1), is the parameter value after mapping of the j-th interval width.

4. The vibration compensation method according to claim 1, characterized in that In the process of iteratively updating the population position through the improved whale optimization algorithm, specifically: Using a non-linear Sigmoid control factor and an adaptive factor to dynamically adjust the search parameters of the vector, and balancing global search and local optimization; Selecting the strategies of surrounding the prey, spiral attack or random search to update the population position according to the probability parameter p and the search vector A; Randomly applying Gaussian distribution mutation to the updated whale positions to increase the population diversity in the next iteration; Using the mirror reflection boundary treatment method to perform reflection mapping on the variables exceeding the boundary and returning to the search space; Iteratively entering the step of calculating the fitness of individual whales until the maximum number of iterations is reached.

5. The vibration compensation method according to claim 4, wherein In the process of using a non-linear Sigmoid control factor and an adaptive factor to dynamically adjust the search parameters of the vector, the expressions of the non-linear Sigmoid control factor and the adaptive factor are as follows: Among them, a Sigmoid is a non-linear Sigmoid control factor, a Adaptive is an adaptive factor, k is the Sigmoid steepness, iter is the current iteration number, iter Max is the maximum iteration number.

6. The vibration compensation method according to claim 4, wherein In the process of selecting the strategies of surrounding the prey, spiral attack or random search to update the population position according to the probability parameter p, specifically: When p < 0.5 and |A| < 1, use the strategy of surrounding the prey to update the population position; When p < 0.5 and |A| ≥ 1, use the random search strategy to update the population position; When p ≥ 0.5, use the spiral attack to update the population position.

7. The vibration compensation method according to claim 4, characterized in that In the process of randomly applying Gaussian distribution mutation to the updated whale positions, specifically: when rand[0, 1] < 0.2, apply Gaussian distribution mutation to the updated whale positions, and the position update introducing perturbation is as follows: Among them, is the position of the whale after introducing the perturbation, is the updated position of the whale, iter is the current iteration number, and a Sigmoid is the non-linear Sigmoid control factor.

8. The vibration compensation method according to claim 4, wherein In the process of using the mirror reflection boundary treatment method to perform reflection mapping on the variables exceeding the boundary, the mirror reflection boundary treatment method is specifically as follows: When x is less than the lower bound x min then, x = x min +(x min - x); When x is greater than or equal to the upper bound x max , x = x max -(x max - x); where x is the current position of the whale, x min , x max are the lower and upper bounds of the search space, respectively.

9. A computer device, comprising a memory, a processor, and a computer program stored on the memory and executable on the processor, characterized in that, When the processor executes the computer program, it realizes the atomic interference gravimeter vibration compensation method described in any one of claims 1-8.

10. A computer-readable storage medium, characterized in that, Several classification programs are stored on the computer-readable storage medium, and the several classification programs are used to be called by the processor and execute the atomic interference gravimeter vibration compensation method described in any one of claims 1-8.

Images