An intelligent method for suppressing micro-vibration of flywheel assemblies on-orbit satellites

Abstract

A method for intelligently suppressing micro-vibrations of flywheel assemblies for on-orbit satellites. The method relates to the technical field of spacecraft systems, and in particular to the technical field of intelligently suppressing micro-vibrations of flywheel assemblies for on-orbit satellites. The method effectively solves the problem of the lack of quantitative indicators of vibration isolation effects in the temperature control process of a small speed condition coverage of the flywheel assembly, thereby improving computational efficiency and accuracy. The method comprises the following steps: extracting frequency domain disturbance data of the flywheel at a specified speed; obtaining the displacement response of the sensitive component in the frequency domain under unit disturbance; obtaining the displacement response of the sensitive component in the time domain under the disturbance of the flywheel assembly; calculating the actual image shift value; calculating the quantitative characterization of the performance of the vibration isolator under the specified speed condition through the actual image shift value; constructing and training a "speed condition-isolator performance" proxy model; and intelligently controlling the intelligent suppression of micro-vibrations of the flywheel assembly of the on-orbit satellite.

Description

Technical Field

[0001] The present invention relates to the technical field of spacecraft systems, and in particular to the technical field of intelligent suppression of micro-vibration of a flywheel assembly for an on-orbit satellite. Background Art

[0002] With the rapid development of space remote sensing technology, the resolution, pointing accuracy, and attitude stability of remote sensing satellites are constantly improving, and the requirements for satellite micro-vibration suppression are also becoming increasingly stringent. During the on-orbit operation phase, the flywheel (or group), as the main attitude control actuator on the satellite, produces small-amplitude, wide-bandwidth micro-vibrations, which can cause the optical axis of sensitive components such as optical payloads to deviate from the target pointing, thereby seriously affecting the payload imaging quality. Therefore, it is necessary to effectively suppress the micro-vibration disturbances generated by the flywheel operation to ensure the control accuracy and imaging quality of the optical remote sensing camera.

[0003] At present, vibration isolation devices for satellites during the rocket launch phase and the on-orbit operation phase are mainly divided into two categories. The first category is passive vibration isolation devices with materials as the core, such as silicone rubber flexible legs, wire rope flexible legs, etc. Its advantages are simple structure and high reliability, and its disadvantages are low vibration isolation efficiency, and usually only one fixed vibration isolation frequency (for second-order systems). The second category is semi-active and active vibration isolation devices based on various driving modes, such as electromagnetic mechanisms and Stewart mechanisms. Its advantages are high vibration isolation efficiency, but its disadvantages are obvious: complexity, high cost, low reliability, limited use in the rocket launch phase, and as the driving process is continuously executed, there is a risk of increased cumulative errors in the system and reduced calculation accuracy.

[0004] The existing technology has the following problems:

[0005] The Chinese invention patent application, "An Intelligent System for Suppressing Microvibrations in Orbit for Satellites" (CN116119037A), proposes a microvibration suppression system suitable for the real-world operating environment of on-orbit satellites. This system consists of a central control unit, a power distribution and thermal control unit, and a variable-parameter satellite isolator. This system can adjust the isolator's temperature control strategy based on the satellite's flywheel assembly speed, thereby varying the temperature of the isolator's filler polymer to alter its isolation performance. However, this method requires a "flywheel assembly speed-frequency-disturbance amplitude" data table generated through ground testing to calculate the target temperature. This method has two drawbacks: first, the massive workload of ground testing inevitably prevents it from covering all speed conditions of the satellite's flywheel assembly; second, using the disturbance amplitude as thermal control guidance ignores the correlation between flywheel disturbances and satellite imaging quality. Summary of the Invention

[0006] In response to the above problems, the present invention discloses an intelligent suppression method for flywheel group micro-vibrations of on-orbit satellites, which relates to the field of spacecraft system technology. It effectively solves the problems of small modeling data scale, small coverage of flywheel group speed operating conditions, and lack of quantitative evaluation indicators for vibration isolation effect in the temperature control process of variable parameter satellite vibration isolators, thereby improving calculation efficiency and accuracy.

[0007] The method comprises the following steps:

[0008] S1, extract the frequency domain disturbance data of the flywheel n at a specified speed;

[0009] S2. Obtain the displacement response of the sensitive component in the frequency domain under unit disturbance;

[0010] S3. Obtain the time-domain displacement response of the sensitive component under the flywheel assembly disturbance based on the frequency-domain disturbance data of the flywheel n at a specified speed and the frequency-domain displacement response of the sensitive component under unit disturbance;

[0011] S4. Calculate the actual image displacement value by the time domain displacement response of the sensitive component under the flywheel group disturbance;

[0012] S5. Calculate the quantitative characterization of the vibration isolator performance under specified speed conditions by using actual image shift values;

[0013] S6. Build and train the “speed condition-isolator performance” agent model;

[0014] S7. Intelligent control of micro-vibration of on-orbit satellite flywheel assembly.

[0015] Furthermore, the extracting of frequency domain disturbance data of flywheel n at a specified speed is specifically as follows: extracting the frequency domain measurement results of flywheel n according to the actual speed operating condition information of the satellite flywheel group, and obtaining the frequency domain disturbance data of flywheel n at the specified speed, wherein n=[1,2,…,N], and N represents the total number of flywheels in the flywheel group.

[0016] Furthermore, the obtaining of the displacement response of the sensitive component in the frequency domain under unit disturbance is specifically: obtaining the displacement response of the sensitive component in the frequency domain under unit disturbance through frequency response analysis under unit disturbance, wherein the sensitive component includes an optical payload of the remote sensing satellite.

[0017] Furthermore, the method of obtaining the time domain displacement response of the sensitive component under the flywheel group disturbance is specifically as follows: in the time domain, the displacement response information of the sensitive component caused by the flywheel n disturbance is superimposed with the frequency domain displacement response of the sensitive component under the unit disturbance to obtain the time domain displacement response of the sensitive component under the flywheel group disturbance; the method of obtaining the displacement response information of the sensitive component caused by the flywheel n disturbance is as follows: the frequency domain disturbance data of the flywheel n at the specified speed and the frequency domain displacement response of the sensitive component under the unit disturbance are multiplied, summed, and Fourier transformed in sequence to obtain the displacement response information of the sensitive component caused by the flywheel n disturbance.

[0018] Furthermore, the calculation of the actual image shift value is specifically as follows: the comprehensive image shift caused by the flywheel group disturbance in the time domain is calculated through the time domain displacement response of the sensitive component under the flywheel group disturbance; the comprehensive image shift caused by the flywheel group disturbance in the time domain is obtained by Fourier transforming the comprehensive image shift caused by the flywheel group disturbance in the frequency domain; the comprehensive image shift caused by the flywheel group disturbance in the frequency domain is the actual image shift.

[0019] Furthermore, the specific calculation formula for quantitatively characterizing the performance of the vibration isolator under the specified speed condition is: The v represents the quantitative characterization of the vibration isolator performance under the specified speed condition, and the Δp i Represents the actual image displacement value corresponding to the i-th frequency sampling point, represents the allowable image shift value corresponding to the i-th frequency sampling point, where i=[1, 2, ..., I], and I represents the total number of frequency sampling points.

[0020] Furthermore, the construction and training of the "speed condition-vibration isolator performance" proxy model is specifically as follows: using an artificial intelligence algorithm to establish a regression learner as the "speed condition-vibration isolator performance" model to be trained, using the speed vector Ω of the flywheel n as the input sample set X to be labeled, adding the label value f(X) of the input sample set to the input sample set X, obtaining the input sample set after the label value is added, dividing the input sample set after the label value is added into a training set, a test set and a validation set in a ratio of 3:1:1, and when the determination coefficient is greater than 0.9, ending the training to obtain the "speed condition-vibration isolator performance" proxy model; the determination coefficient represents the determination coefficient between the predicted value and the true value of the test set, and the label value f(X) of the sample set is a quantitative representation of the vibration isolator performance under the specified speed condition.

[0021] Furthermore, the intelligent suppression of micro-vibrations of the on-orbit satellite flywheel assembly is specifically achieved by calculating target parameters through the "speed condition-isolator performance" proxy model, converting the target parameters into temperature control instructions, and then adjusting the vibration isolation performance according to the temperature control instructions, thereby achieving intelligent suppression of micro-vibrations of the on-orbit satellite flywheel assembly. The beneficial effects of the present invention are:

[0022] (1) The present invention provides a complete and universal analysis method for the intelligent suppression of flywheel micro-vibration during the satellite's on-orbit phase. Based on the traditional method, it proposes a superposition method for multi-source disturbances of the flywheel group, a quantitative characterization and rapid prediction method for the performance of the vibration isolator, and an intelligent adjustment strategy for the variable parameter vibration isolator.

[0023] (2) The idea of ​​quantitatively characterizing the vibration isolation performance of the satellite flywheel assembly isolator proposed in the present invention enables the flywheel assembly speed conditions to be used for training machine learning models in the form of feature vectors, effectively introducing artificial intelligence algorithms with high data-driven capabilities into the field of on-orbit satellite vibration suppression.

[0024] (3) The present invention can realize the rapid and accurate prediction and evaluation of the performance of the vibration isolator under multi-flywheel disturbance. The calculation time for a single speed condition is shortened from 1s level of the traditional method to 1ms level, and the relative error can be controlled within 5% and the determination coefficient can be controlled above 0.95.

[0025] The method of the present invention can be applied in the fields of remote sensing satellite assembly and manufacturing, remote sensing satellite vibration suppression, and remote sensing satellite imaging system design. BRIEF DESCRIPTION OF THE DRAWINGS

[0026] Figure 1 This is a flow chart of the performance analysis of the vibration isolator under multi-flywheel disturbance in an embodiment of the present invention;

[0027] Figure 2 This is a flow chart of intelligent adjustment of a satellite isolator with variable parameters under variable working conditions in an embodiment of the present invention. DETAILED DESCRIPTION

[0028] The technical solution of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the embodiments described are only some embodiments of the present invention, not all embodiments. All other embodiments obtained by ordinary technicians in this field based on the embodiments of the present invention without making any creative efforts shall fall within the scope of protection of the present invention.

[0029] This embodiment provides an intelligent suppression method for micro-vibration of a flywheel assembly on an on-orbit satellite. The intelligent suppression method includes two stages. The first stage is the performance analysis of the vibration isolator under multi-flywheel disturbance. The process is as follows: Figure 1 As shown, it includes steps S1-S5. The second stage is to construct an intelligent control model to suppress the micro-vibration of the flywheel assembly of the on-orbit satellite. The process is as follows Figure 2 As shown, steps S6-S7 are included.

[0030] S1. Extract frequency domain disturbance data of flywheel n at a specified speed.

[0031] The relevant operations of step S1 are introduced with a specific example:

[0032] The implementation object selected in this embodiment is a certain type of optical remote sensing satellite. First, the perturbation dynamic characteristics test is carried out on each flywheel of the satellite's orthogonal flywheel group in turn.

[0033] The time domain test results of the disturbance force and disturbance torque of flywheel n at a specified speed are obtained through a force measuring platform, and are converted into corresponding frequency domain measured results through the Fourier transform method. The frequency domain measured results of flywheel n are then extracted based on the actual speed operating condition information of the satellite flywheel group to obtain the frequency domain disturbance data of flywheel n at the specified speed, where n = [1, 2, …, N], and N represents the total number of flywheels in the flywheel group.

[0034] S2. Obtain the displacement response of the sensitive component in the frequency domain under unit disturbance.

[0035] The relevant operations of step S2 are introduced with a specific example:

[0036] Through frequency response analysis under unit disturbance, the frequency domain displacement response of the sensitive component under unit disturbance is obtained. The sensitive component refers to the optical payload of the remote sensing satellite. Micro-vibration can cause the target pointing of its optical axis to shift, thus seriously affecting the imaging quality of the payload. The unit disturbance represents the action of the unit force and torque at the flywheel position n, obtained through the satellite's unit frequency response analysis.

[0037] S3. Obtain the time-domain displacement response of the sensitive component under the flywheel assembly disturbance according to the frequency-domain disturbance data of the flywheel n at the specified speed and the frequency-domain displacement response of the sensitive component under the unit disturbance.

[0038] The relevant operations of step S3 are introduced with a specific example:

[0039] In the time domain, the frequency domain disturbance data of flywheel n at the specified speed and the frequency domain displacement response of the sensitive component under unit disturbance are multiplied to obtain the displacement response under the action of disturbance forces and moments in different directions. The displacement responses under the action of disturbance forces / torques in different directions are then summed and superimposed by Fourier transform to obtain the displacement response information of the sensitive component caused by the disturbance of flywheel n. The displacement response information of the sensitive component caused by the disturbance of flywheel n is superimposed with the frequency domain displacement responses of the sensitive component under multiple unit disturbances to obtain the time domain displacement response of the sensitive component under the disturbance of the flywheel group, thereby completing the superposition of multi-source disturbances of the flywheel group.

[0040] S4. Calculate the actual image displacement value through the time domain displacement response of the sensitive component under the flywheel group disturbance.

[0041] The relevant operations of step S4 are introduced with a specific example:

[0042] The translational displacement and rotational displacement of each disturbance-sensitive component under multi-flywheel disturbance are calculated by the time-domain displacement response of the sensitive component under the flywheel group disturbance. On this basis, the comprehensive image shift caused by the flywheel group disturbance in the time domain is calculated by the transfer characteristics of the optical system. The comprehensive image shift represents the offset of the central image point. The comprehensive image shift caused by the flywheel group disturbance in the time domain is obtained by Fourier transform to obtain the comprehensive image shift caused by the flywheel group disturbance in the frequency domain. The comprehensive image shift caused by the flywheel group disturbance in the frequency domain is the actual image shift.

[0043] S5. Quantitative characterization of the isolator performance under specified speed conditions is calculated by actual image shift value.

[0044] The relevant operations of step S5 are introduced with a specific example:

[0045] According to the design index of satellite micro-vibration pixel offset, the maximum ratio of the actual image motion of each frequency sampling point to the sinusoidal image motion tolerance is used as the evaluation index of the satellite vibration isolator performance under the specified speed condition, that is, The v represents the quantitative characterization of the vibration isolator performance under the specified speed condition, and the Δp i Represents the actual image displacement value corresponding to the i-th frequency sampling point, represents the allowable image shift value corresponding to the i-th frequency sampling point, where i=[1, 2, ..., I], and I represents the total number of frequency sampling points.

[0046] S6. Build and train the “speed condition-isolator performance” agent model.

[0047] The relevant operations of step S6 are introduced with a specific example:

[0048] An artificial intelligence algorithm is used to establish a regression learner as a "speed condition-isolator performance" model to be trained. The data sample acquisition method of the artificial intelligence algorithm in this embodiment is as follows: a certain number of speed conditions are randomly collected as sample input values, the speed vector Ω of the flywheel n is used as the input sample set X to be labeled, the label value f(X) of the input sample set is added to the input sample set X to obtain an input sample set with the added label value, and the input sample set with the added label value is divided into a training set, a test set, and a validation set in a ratio of 3:1:1. Based on the above-mentioned analysis of the vibration isolator performance under multi-flywheel disturbance, the corresponding vibration isolation performance characteristic value is calculated as the output sample set; a machine learning model for regression analysis is selected as the regression learner of the proxy model; and indicators such as the determination coefficient and the root mean square error are used to quantitatively evaluate the prediction performance. When the determination coefficient is greater than 0.9, the training is terminated to obtain the "speed condition-isolator performance" proxy model; the determination coefficient represents the determination coefficient between the predicted value and the true value of the test set, and the label value f(X) of the sample set is a quantitative representation of the vibration isolator performance under the specified speed condition.

[0049] Results from random numerical experiments demonstrate that the coefficient of determination between the predicted and true values ​​of the constructed "speed condition-vibration isolation performance" proxy model is no less than 0.95, with a relative error of no more than 5%. This approach provides a highly accurate approximation of vibration isolation performance evaluation indicators across all speed conditions for in-orbit satellites. Furthermore, for a single speed condition, the proxy model can reduce the calculation time from 1 second to 1 millisecond, significantly improving the computational efficiency of the flywheel isolator's vibration isolation performance.

[0050] S7. Intelligent control of micro-vibration of on-orbit satellite flywheel assembly.

[0051] The relevant operations of step S7 are introduced with a specific example:

[0052] The established proxy model library is used to calculate the optimal vibration isolator parameters under the speed conditions of the on-orbit satellite flywheel assembly according to the real-time speed conditions of the satellite in the on-orbit phase. The optimal vibration isolator parameters are used as target parameters, and the target parameters are converted into temperature control instructions based on the dynamic thermomechanical analysis test results of the vibration isolator filling material and the current temperature state of the vibration isolator. The temperature of the vibration isolator filling material is adjusted according to the temperature control instructions to change the vibration isolation performance, thereby realizing intelligent suppression of the variable parameter satellite vibration isolator, and thus making the variable parameter satellite vibration isolator achieve the best performance in suppressing microvibrations under the current speed conditions.

Claims

1. An intelligent method for suppressing micro-vibration of a flywheel assembly of an on-orbit satellite, characterized in that: The method comprises the following steps: S1, extract the flywheel at the specified speed Frequency domain disturbance data of S2. Obtain the displacement response of the sensitive component in the frequency domain under unit disturbance; S3, flywheel at specified speed The frequency domain disturbance data and the frequency domain displacement response of the sensitive component under unit disturbance are used to obtain the time domain displacement response of the sensitive component under the flywheel group disturbance; S4. Calculate the actual image displacement value by the time domain displacement response of the sensitive component under the flywheel group disturbance; S5. Calculate the quantitative characterization of the vibration isolator performance under specified speed conditions by using actual image shift values; S6. Build and train the "speed condition-isolator performance" agent model; S7. Intelligent control of micro-vibration of on-orbit satellite flywheel assembly.

2. The intelligent suppression method for micro-vibration of a flywheel assembly for an on-orbit satellite according to claim 1, characterized in that: The extraction is specified under the flywheel speed The frequency domain disturbance data is specifically: extract the flywheel according to the actual speed condition information of the satellite flywheel group The frequency domain measurement results of the flywheel at the specified speed are obtained. The frequency domain disturbance data, , Indicates the total number of flywheels in the flywheel set.

3. The intelligent suppression method for micro-vibration of a flywheel assembly for an on-orbit satellite according to claim 1, characterized in that: The method of obtaining the displacement response of the sensitive component in the frequency domain under unit disturbance is specifically: obtaining the displacement response of the sensitive component in the frequency domain under unit disturbance through frequency response analysis under unit disturbance, wherein the sensitive component includes the optical payload of the remote sensing satellite.

4. The intelligent suppression method for micro-vibration of a flywheel assembly for an on-orbit satellite according to claim 3, characterized in that: The method of obtaining the time domain displacement response of the sensitive component under the disturbance of the flywheel group is as follows: in the time domain, the flywheel The displacement response information of the sensitive component caused by the disturbance is superimposed on the displacement response of the sensitive component in the frequency domain under the unit disturbance to obtain the displacement response of the sensitive component in the time domain under the disturbance of the flywheel group; the flywheel The process of obtaining the displacement response information of the sensitive component caused by the disturbance is as follows: The frequency domain disturbance data and the displacement response of the sensitive component in the frequency domain under unit disturbance are multiplied, summed and Fourier transformed in sequence to obtain the flywheel Displacement response information of sensitive components caused by disturbance.

5. The intelligent method for suppressing micro-vibration of a flywheel assembly for an on-orbit satellite according to claim 4, characterized in that: The calculation of the actual image shift value is specifically as follows: the comprehensive image shift caused by the flywheel group disturbance in the time domain is calculated through the time domain displacement response of the sensitive component under the flywheel group disturbance; the comprehensive image shift caused by the flywheel group disturbance in the time domain is obtained by Fourier transforming the comprehensive image shift caused by the flywheel group disturbance in the frequency domain; the comprehensive image shift caused by the flywheel group disturbance in the frequency domain is the actual image shift.

6. The intelligent method for suppressing micro-vibration of a flywheel assembly for an on-orbit satellite according to claim 5, characterized in that: The specific calculation formula for quantitative characterization of the vibration isolator performance under the specified speed condition is: ; It represents the quantitative characterization of the performance of the vibration isolator under the specified speed condition. Indicates the The actual image displacement value corresponding to the frequency sampling point is Indicates the The allowable image shift value corresponding to the frequency sampling point is , Indicates the total number of frequency sampling points.

7. The intelligent method for suppressing micro-vibration of a flywheel assembly for an on-orbit satellite according to claim 6, characterized in that: The construction and training of the "speed condition-isolator performance" agent model is specifically as follows: using artificial intelligence algorithm to establish a regression learner as the "speed condition-isolator performance" to be trained model, using the flywheel Speed ​​vector As the input sample set to be labeled , to the input sample set Add the label value of the input sample set , obtain the input sample set after adding the label value, divide the input sample set after adding the label value into training set, test set and validation set in a ratio of 3:1:1, and when the determination coefficient is greater than 0.9, end the training to obtain the "speed condition-vibration isolator performance" proxy model; the determination coefficient represents the determination coefficient between the predicted value and the true value of the test set, and the label value of the sample set It is a quantitative representation of the performance of the vibration isolator under specified speed conditions.

8. The intelligent method for suppressing micro-vibration of a flywheel assembly for an on-orbit satellite according to claim 7, characterized in that: The intelligent suppression of micro-vibrations of the on-orbit satellite flywheel assembly is specifically achieved by calculating target parameters through the "speed condition-isolator performance" proxy model, converting the target parameters into temperature control instructions, and then adjusting the vibration isolation performance according to the temperature control instructions to achieve intelligent suppression of micro-vibrations of the on-orbit satellite flywheel assembly.

Images