Aircraft elastic vibration suppression method and system based on strong tracking Kalman filter

Abstract

The application discloses a kind of aircraft elastic vibration suppression method and system based on strong tracking Kalman filter, wherein the method comprises: obtaining angular acceleration estimation value according to preset gyro data;Obtain the system model of angular velocity and elastic frequency according to angular acceleration estimation value;Rigid body angular velocity estimation result is obtained according to the system model of angular velocity and elastic frequency;Obtain rudder deflection instruction according to rigid body angular velocity estimation result.The application improves the suppression ability of high-frequency signal, and enhances the robustness of stable control system.

Description

Technical Field

[0001] This invention belongs to the field of aircraft stability control technology, and particularly relates to a method and system for suppressing elastic vibrations of aircraft based on strong tracking Kalman filtering. Background Technology

[0002] During actual flight, the fiber optic gyroscope typically deviates from the missile's center of mass, and its sensitive angular velocity measurement consists of two parts: rigid body angular velocity and elastic disturbance angular velocity. If the missile's elastic modes are excited without filtering, the high-frequency signal caused by elastic deformation will be calculated through the control loop and directly reflected in the servo commands. For electric servos, prolonged response to high-frequency signals can lead to a significant increase in current, potentially causing power depletion or irreversible damage to the servo. Therefore, an elastic suppression control strategy is necessary to smoothly input the angular velocity into the control loop.

[0003] In current technology, two or three broadband notch filters are typically connected in series in the control loop to accommodate the elastic frequency differences between ground modal tests and actual flight tests. This sacrifices some amplitude and phase margin, reducing the robustness of the stability control system. Furthermore, when unknown modes appear, the elastic suppression capability of the notch filter is often limited, worsening the rudder system's operating conditions and potentially causing instability in severe cases. Summary of the Invention

[0004] The technical problem solved by this invention is to overcome the shortcomings of the prior art and provide a method and system for suppressing elastic vibration of aircraft based on strong tracking Kalman filtering, thereby improving the ability to suppress high-frequency signals and enhancing the robustness of the stability control system.

[0005] The objective of this invention is achieved through the following technical solution: a method for suppressing elastic vibration of an aircraft based on strong tracking Kalman filtering, comprising: obtaining an angular acceleration estimate based on preset gyroscope data; obtaining a system model of angular velocity and elastic frequency based on the angular acceleration estimate; obtaining a rigid body angular velocity estimate based on the system model of angular velocity and elastic frequency; and obtaining a rudder deflection command based on the rigid body angular velocity estimate.

[0006] In the above-mentioned aircraft elastic vibration suppression method based on strong tracking Kalman filtering, the angular acceleration estimate is obtained by the following formula:

[0007]

[0008] in, This represents the estimated angular acceleration at the current sampling time, where T is the time step, k is the current sampling count, and ω is the current angular acceleration estimate. m (kT-T) represents the angular velocity measurement value at the previous sampling time, ω m (kT-2T) represents the angular velocity measurements at the first two sampling times.

[0009] In the above-mentioned aircraft elastic vibration suppression method based on strong tracking Kalman filtering, the system model of angular velocity and elastic frequency is obtained through the following formula:

[0010]

[0011] Where X(kT) is the state variable at the current sampling time, X(kT-T) is the state variable at the previous sampling time, T is the time step, k is the current sampling count, Φ(kT) is the state transition matrix at the current sampling time, and Ψ(kT) is the control transition matrix at the current sampling time. ω is the estimated angular acceleration at the current sampling time. m (kT) represents the measured angular velocity at the current sampling time, ζ(kT) represents the observation noise at the current sampling time, B is the manipulation matrix, C is the observation matrix, ω1 is the first-order elastic deformation mode frequency, ω2 is the second-order elastic deformation mode frequency, I represents the unit diagonal matrix, the superscript "T" indicates transpose, the superscript "i" indicates power i, "!" indicates factorial, and "Σ" indicates summation.

[0012] In the above-mentioned aircraft elastic vibration suppression method based on strong tracking Kalman filter, the rigid body angular velocity estimation result obtained from the system model of angular velocity and elastic frequency includes: obtaining the predicted value of the state quantity at the current sampling time based on the system model of angular velocity and elastic frequency and the preset state quantity update value at the previous sampling time; obtaining the prediction error covariance matrix based on the preset fading factor and the input noise matrix at the previous sampling time; obtaining the Kalman filter gain based on the prediction error covariance matrix; and obtaining the rigid body angular velocity estimation result based on the Kalman filter gain and the predicted value of the state quantity at the current sampling time.

[0013] In the above-mentioned aircraft elastic vibration suppression method based on strong tracking Kalman filtering, the predicted value of the state variable at the current sampling time is... It can be obtained through the following formula:

[0014]

[0015] Where T is the time step and k is the current sampling number. Let Φ(kT) be the state variable update value at the previous sampling time, Φ(kT) be the state transition matrix at the current sampling time, and Ψ(kT) be the control transition matrix at the current sampling time. This is the estimated angular acceleration value at the current sampling time.

[0016] In the above-mentioned aircraft elastic vibration suppression method based on strong tracking Kalman filtering, the prediction error covariance matrix is ​​obtained by the following formula:

[0017]

[0018] in, Let T be the prediction error covariance matrix at the current sampling time, T be the time step, k be the current sampling number, and n be the... The dimension of the vector is given by Q(kT-T), where Q(kT-T) is the input noise matrix at the previous sampling time, and K(kT-T) is the Kalman filter gain at the previous sampling time. To update the error covariance matrix at the previous sampling time, Let λ(kT-T) be the prediction error covariance matrix at the previous sampling time, and ω be the fading factor at the previous sampling time. m (kT-T) represents the angular velocity measurement at the previous sampling time, and C is the observation matrix. Let Ψ(kT) be the predicted state value at the previous sampling time, Ψ(kT-T) be the control transition matrix at the current sampling time, Ψ(kT-T) be the control transition matrix at the previous sampling time, R be the observation noise matrix, Φ(kT) be the state transition matrix at the current sampling time, Φ(kT-T) be the state transition matrix at the previous sampling time, and I denote the unit diagonal matrix. This is the variance estimate of the information d(kT-T) at the previous sampling time. The superscript "T" indicates transpose, the superscript "2" indicates square, and "Tr(·)" indicates the sum of the diagonal elements of the matrix.

[0019] In the above-mentioned method for suppressing elastic vibrations of aircraft based on strong tracking Kalman filtering, the Kalman filter gain is obtained by the following formula:

[0020]

[0021] Where K(kT) is the Kalman filter gain at the current sampling time, T is the time step, and k is the current sampling count. Let C be the prediction error covariance matrix at the current sampling time, C be the observation matrix, and R be the observation noise matrix.

[0022] In the above-mentioned method for suppressing elastic vibration of aircraft based on strong tracking Kalman filtering, the rigid body angular velocity estimation result is obtained by the following formula:

[0023]

[0024] Where ω(kT) is the estimated rigid body angular velocity. Here, T is the state variable update value, k is the current sampling count, and K(kT) is the Kalman filter gain at the current time. ω is the predicted value of the state variable. m (kT) represents the measured angular velocity at the current sampling time, and C is the observation matrix.

[0025] An aircraft elastic vibration suppression system based on strong tracking Kalman filtering includes: a first module for obtaining an angular acceleration estimate based on preset gyroscope data; a second module for obtaining a system model of angular velocity and elastic frequency based on the angular acceleration estimate; a third module for obtaining a rigid body angular velocity estimate based on the system model of angular velocity and elastic frequency; and a fourth module for obtaining a rudder deflection command based on the rigid body angular velocity estimate.

[0026] An electronic device includes: a memory for storing computer-readable instructions; and a processor for executing the computer-readable instructions to perform an aircraft elastic vibration suppression method based on a strong-tracking Kalman filter.

[0027] Compared with the prior art, the present invention has the following advantages:

[0028] (1) This invention extracts rigid body angular velocity information by setting up a Kalman filter, thus avoiding the problem of insufficient stability margin of the stable control system caused by the notch filter.

[0029] (2) This invention introduces the concept of strong tracking filtering, which makes the dynamic characteristics of the filter's built-in system model closer to the actual situation, thereby improving the ability to suppress high-frequency signals.

[0030] (3) The present invention introduces an adaptive update method for input noise characteristics to suppress filter divergence caused by input angular acceleration delayed by one sampling time and enhance the robustness of the filter itself; it is applicable to situations where the input contains a certain degree of colored noise, and can suppress the elastic vibration of the aircraft when there is a certain difference between the elastic frequencies in ground modal tests and actual flight tests or when unknown modes appear. Attached Figure Description

[0031] Various other advantages and benefits will become apparent to those skilled in the art upon reading the following detailed description of preferred embodiments. The accompanying drawings are for illustrative purposes only and are not intended to limit the invention. Furthermore, the same reference numerals denote the same parts throughout the drawings. In the drawings:

[0032] Figure 1 This is a block diagram illustrating the control principle of the pitch / yaw / roll channel provided in an embodiment of the present invention. Detailed Implementation

[0033] Exemplary embodiments of the present disclosure will now be described in more detail with reference to the accompanying drawings. While exemplary embodiments of the present disclosure are shown in the drawings, it should be understood that the present disclosure may be implemented in various forms and should not be limited to the embodiments set forth herein. Rather, these embodiments are provided to enable a more thorough understanding of the present disclosure and to fully convey the scope of the disclosure to those skilled in the art. It should be noted that, unless otherwise specified, the embodiments and features described herein can be combined with each other. The present invention will now be described in detail with reference to the accompanying drawings and embodiments.

[0034] A Kalman filter is used to extract the rigid body angular velocity information of the aircraft from the angular velocity measurements and is introduced into the control loop to replace the output of the notch filter, thus avoiding the problem of insufficient stability margin in the stable control system caused by the notch filter. Based on the traditional extended Kalman filter, a strong tracking filter concept is introduced, adaptively adjusting the memory length of the Kalman filter according to the current observation data, i.e., the utilization efficiency of old observation data, so that the dynamic characteristics of the system are fully reflected by the current observation information. An adaptive update method for input noise characteristics is introduced to suppress filter divergence caused by angular acceleration delay input and colored noise. This method can also effectively suppress elasticity when there are differences between the actual flight test elastic frequency and ground modal test results, and when unknown modes appear.

[0035] This embodiment provides a method for suppressing elastic vibrations of aircraft based on strong tracking Kalman filtering. The method includes:

[0036] The angular acceleration estimate is obtained based on the preset gyroscope data;

[0037] A system model of angular velocity and elastic frequency is obtained based on the estimated angular acceleration value;

[0038] The rigid body angular velocity estimation results are obtained based on the system model of angular velocity and elastic frequency;

[0039] The rudder deflection command is obtained based on the rigid body angular velocity estimation results.

[0040] This embodiment ensures the suppression of elastic vibration when there are discrepancies between the actual flight test elastic frequency and the ground modal test results, and when unknown modes appear. This technical solution extracts rigid body angular velocity information from the gyroscope output using a Kalman filter, and simultaneously introduces a strong tracking concept and an adaptive update method for input noise characteristics to improve the ability to suppress high-frequency signals and enhance the robustness of the stability control system.

[0041] Specifically, real-time data from the gyroscope is acquired, and the estimated angular acceleration is calculated, as follows:

[0042]

[0043] in, This represents the estimated angular acceleration at the current sampling time, where T is the time step, k is the current sampling count, and ω is the current angular acceleration estimate. m (kT-T) represents the angular velocity measurement value at the previous sampling time, ω m (kT-2T) represents the angular velocity measurements at the first two sampling times.

[0044] Establish a system model for angular velocity and elastic frequency:

[0045] The design of a stable control system needs to pay attention to first-order and second-order elastic deformation. The elastic disturbance angular velocity v(kT) at the current sampling time can be expressed as the sum of two sinusoidal signals of different frequencies.

[0046] v(kT)=v1(kT)+v3(kT)=b1sin(ω1kT+φ1)+b2sin(ω2kT+φ2)

[0047] In the formula, T is the time step, k is the current sampling number, v1(kT) and v3(kT) are the elastic disturbance angular velocities caused by the first and second order elastic deformations at the current sampling time, respectively, b1 and b2 are the amplitudes of the first and second order elastic disturbance angular velocities, respectively, φ1 and φ2 are the phases of the first and second order elastic disturbance angular velocities, respectively, and ω1 and ω2 are the first and second order modal frequencies, respectively.

[0048] Set the state variable X(kT) as:

[0049] X(kT)=[ω(kT)v1(kT)v2(kT)v3(kT)v4(kT)] T

[0050] In the formula, ω(kT) is the rigid body angular velocity at the current sampling time, T is the time step, k is the current sampling number, v1(kT) and v3(kT) are the elastic disturbance angular velocities caused by the first and second order elastic deformations at the current sampling time, respectively, b1 and b2 are the amplitudes of the first and second order elastic disturbance angular velocities, respectively, φ1 and φ2 are the phases of the first and second order elastic disturbance angular velocities, respectively, ω1 and ω2 are the first and second order modal frequencies, respectively, v2(kT) and v4(kT) are the first derivatives of v1(kT) and v3(kT) with respect to time, respectively, and the superscript "T" indicates transpose.

[0051] Establish a system model that considers both input noise and observation noise:

[0052]

[0053] Where X(kT) is the state variable at the current sampling time, X(kT-T) is the state variable at the previous sampling time, T is the time step, k is the current sampling count, Φ(kT) is the state transition matrix at the current sampling time, and Ψ(kT) is the control transition matrix at the current sampling time. This is the estimated angular acceleration value at the current sampling time, which includes the input noise w(kT), ω m (kT) represents the measured angular velocity at the current sampling time, ζ(kT) represents the observation noise at the current sampling time, B is the manipulation matrix, C is the observation matrix, ω1 is the first-order elastic deformation mode frequency, ω2 is the second-order elastic deformation mode frequency, I represents the unit diagonal matrix, the superscript "T" indicates transpose, the superscript "i" indicates power i, "!" indicates factorial, and "Σ" indicates summation.

[0054] Based on the established system model, the strong tracking filtering concept and the method of adaptively updating the input noise characteristics are combined with the traditional extended Kalman filter to obtain the rigid body angular velocity estimation results, and a stable control loop is introduced.

[0055] The rigid body angular velocity estimation results obtained from the system model based on angular velocity and elastic frequency include:

[0056] Based on the system model of angular velocity and elastic frequency and the preset state quantity update value of the previous sampling time, the predicted state quantity value of the current sampling time is obtained;

[0057] The prediction error covariance matrix is ​​obtained based on the preset fading factor and the input noise matrix;

[0058] The Kalman filter gain is obtained from the prediction error covariance matrix;

[0059] The rigid body angular velocity estimation result is obtained based on the Kalman filter gain and the predicted state value at the current sampling time.

[0060] Predicted state values ​​at the current sampling time It can be obtained through the following formula:

[0061]

[0062] Where T is the time step and k is the current sampling number. Let Φ(kT) be the state variable update value at the previous sampling time, Φ(kT) be the state transition matrix at the current sampling time, and Ψ(kT) be the control transition matrix at the current sampling time. This is the estimated angular acceleration value at the current sampling time.

[0063] The strong tracking filter concept is introduced, and the prediction error covariance matrix is ​​calculated based on the fading factor and the input noise matrix, specifically:

[0064]

[0065] in, Let T be the prediction error covariance matrix at the current sampling time, T be the time step, k be the current sampling number, and n be the... The dimension of the vector is given by Q(kT-T), where Q(kT-T) is the input noise matrix at the previous sampling time, and K(kT-T) is the Kalman filter gain at the previous sampling time. To update the error covariance matrix at the previous sampling time, Let λ(kT-T) be the prediction error covariance matrix at the previous sampling time, and ω be the fading factor at the previous sampling time. m (kT-T) represents the angular velocity measurement at the previous sampling time, and C is the observation matrix. Ψ(kT) represents the predicted state value at the previous sampling time, Ψ(kT-T) represents the control transition matrix at the current sampling time, Ψ(kT-T) represents the control transition matrix at the previous sampling time, R represents the observation noise matrix, Φ(kT) represents the state transition matrix at the current sampling time, Φ(kT-T) represents the state transition matrix at the previous sampling time, I represents the unit diagonal matrix, the superscript "T" represents the transpose, the superscript "2" represents the square, and "Tr(·)" represents calculating the sum of the diagonal elements of the matrix.

[0066] In the formula, The variance estimate of the innovation d(kT-T) at the previous sampling time is as follows:

[0067]

[0068] Where T is the time step, k is the current sampling number, and ω m (kT-T) represents the angular velocity measurement at the previous sampling time, and C is the observation matrix. This is the predicted value of the state quantity at the previous sampling time. Σ represents the estimated variance of the innovation at the first two sampling times, W is the length of the sliding estimation window, and "Σ" indicates summation.

[0069] The Kalman filter gain is obtained using the following formula:

[0070]

[0071] Where K(kT) is the Kalman filter gain at the current sampling time, T is the time step, and k is the current sampling count. Let C be the prediction error covariance matrix at the current sampling time, C be the observation matrix, and R be the observation noise matrix.

[0072] The rigid body angular velocity estimation result is obtained by the following formula:

[0073]

[0074] Where ω(kT) is the estimated rigid body angular velocity. Here, T is the state variable update value, k is the current sampling count, and K(kT) is the Kalman filter gain at the current time. ω is the predicted value of the state variable. m (kT) represents the measured angular velocity at the current sampling time, and C is the observation matrix.

[0075] Based on the estimated rigid body angular velocity, the deflection command is calculated, the control surfaces are driven to deflect, and control force and control torque are applied to stabilize the missile attitude.

[0076] This method is illustrated using the pitch channel control of a certain aircraft as an example. The control principle diagram of the pitch channel is shown below. Figure 1 As shown, the specific steps are explained below:

[0077] 1. Acquire real-time gyroscope data, and output the pitch angular velocity measurement value. Considering that the gyroscope output is contaminated by elastic vibration and the angular acceleration needs to be estimated from the gyroscope output, this invention treats the angular acceleration caused by elastic vibration as colored noise, and estimates the angular acceleration by delaying one sampling moment to approximately satisfy the basic assumptions of the Kalman filter regarding noise. The formula is as follows:

[0078]

[0079] In the formula, This represents the estimated pitch acceleration at the current sampling time, where T is the time step, k is the current sampling count, and ω is the current pitch acceleration. m (kT-T) represents the pitch angular velocity measurement value at the previous sampling time, ω m (kT-2T) represents the pitch angular velocity measurements at the first two sampling times.

[0080] 2. Using the estimated pitch acceleration obtained in step 1 as input, establish a system model of the rigid body angular velocity and elastic frequency under the pitch channel. Set the state variable X as:

[0081]

[0082] In the formula, ω(kT) is the estimated pitch rigid body angular velocity at the current sampling time, v1(kT) is the elastic disturbance pitch angular velocity caused by the first-order elastic deformation at the current sampling time, v2(kT) is the first derivative of v1(kT) with respect to time at the current sampling time, v3(kT) is the elastic disturbance pitch angular velocity caused by the second-order elastic deformation at the current sampling time, v4(kT) is the first derivative of v3(kT) with respect to time at the current sampling time, b1 and b2 are the amplitudes of the first-order and second-order elastic disturbance pitch angular velocities, respectively, φ1 and φ2 are the phases of the first-order and second-order elastic disturbance pitch angular velocities, respectively, ω1 and ω2 are the first-order and second-order modal frequencies, respectively, and the superscript "T" indicates transpose.

[0083] The system model is as follows:

[0084]

[0085] Where X(kT) is the state variable at the current sampling time, X(kT-T) is the state variable at the previous sampling time, T is the time step, k is the current sampling count, Φ(kT) is the state transition matrix at the current sampling time, and Ψ(kT) is the control transition matrix at the current sampling time. This is the estimated angular acceleration value at the current sampling time, which includes the input noise w(kT), ω m (kT) represents the measured angular velocity at the current sampling time, ζ(kT) represents the observation noise at the current sampling time, B is the manipulation matrix, C is the observation matrix, ω1 is the first-order elastic deformation mode frequency, ω2 is the second-order elastic deformation mode frequency, I represents the unit diagonal matrix, the superscript "T" indicates transpose, the superscript "i" indicates power i, "!" indicates factorial, and "Σ" indicates summation.

[0086] The settings for ω1 and ω2 can be referenced from the results of ground modal tests.

[0087] 3. Based on the established system model, the strong tracking filter concept and the method of adaptively updating the input noise characteristics are combined with the traditional extended Kalman filter to obtain the rigid body pitch velocity estimation results, and a stable control loop is introduced.

[0088] Set initial parameters:

[0089]

[0090] In the formula, This represents the updated value of the state variable at the initial moment. Let I represent the initial time-time update error covariance matrix, and let I represent the unit diagonal matrix.

[0091] Update the state variable based on the previous sampling time. Predict the state quantity prediction value at the current sampling time. Specifically:

[0092]

[0093] Where T is the time step and k is the current sampling number. Let Φ(kT) be the state variable update value at the previous sampling time, Φ(kT) be the state transition matrix at the current sampling time, and Ψ(kT) be the control transition matrix at the current sampling time. This is the estimated pitch angle acceleration at the current sampling time.

[0094] By introducing the concept of strong tracking filtering and an adaptive update method for the input noise characteristics, the prediction of the error covariance matrix is ​​adjusted, as shown in the following formula:

[0095]

[0096] Where λ(kT-T) is the fading factor, and its initial value can be selected as 1; the initial value of Q and the setting of R can be referenced from historical flight test data, and Q will be updated adaptively thereafter. The variance estimate of the innovation d(kT-T) at the previous sampling time is as follows:

[0097]

[0098] Where T is the time step, k is the current sampling number, and ω m (kT-T) represents the pitch angular velocity measurement at the previous sampling time, and C is the observation matrix. This is the predicted value of the state quantity at the previous sampling time. Σ represents the estimated variance of the innovation at the first two sampling times, where "Σ" indicates summation. W is the length of the sliding estimation window, which needs to be determined by a trade-off between sensitivity and accuracy based on the actual dynamic range. When W is small, the filter is more sensitive to changes in the system's dynamic characteristics, but the accuracy of estimating the input noise characteristics is lower, which can easily lead to filter divergence; when W is large, the filter has lower sensitivity but higher accuracy.

[0099] Calculate the Kalman filter gain and estimate the rigid body pitch velocity, specifically:

[0100]

[0101] Where ω(kT) is the estimated rigid body pitch angular velocity. Here, T is the state variable update value, k is the current sampling count, and K(kT) is the Kalman filter gain at the current time. ω is the predicted value of the state variable. m (kT) represents the measured pitch angular velocity at the current sampling time, and C is the observation matrix.

[0102] 4. Based on the rigid body pitch angular velocity estimate obtained in step 3, calculate the pitch channel rudder deflection command, drive the pitch channel rudder surface to deflect, apply control force and control torque, and stabilize the missile pitch attitude.

[0103] This embodiment also provides an aircraft elastic vibration suppression system based on strong tracking Kalman filtering, including: a first module for obtaining an angular acceleration estimate based on preset gyroscope data; a second module for obtaining a system model of angular velocity and elastic frequency based on the angular acceleration estimate; a third module for obtaining a rigid body angular velocity estimate based on the system model of angular velocity and elastic frequency; and a fourth module for obtaining a rudder deflection command based on the rigid body angular velocity estimate.

[0104] This embodiment also provides an electronic device, including: a memory for storing computer-readable instructions; and a processor for running the computer-readable instructions to execute a method for suppressing elastic vibrations of an aircraft based on a strong-tracking Kalman filter.

[0105] This embodiment extracts rigid body angular velocity information using a Kalman filter, avoiding the insufficient stability margin problem of the stability control system caused by notch filters. This embodiment introduces a strong tracking filter concept, making the dynamic characteristics of the filter's built-in system model closer to the actual situation, thus improving the ability to suppress high-frequency signals. This embodiment introduces a method for adaptively updating input noise characteristics to suppress filter divergence caused by input angular acceleration delayed by one sampling moment, enhancing the filter's robustness. It is suitable for situations where the input contains a certain degree of colored noise, and can suppress aircraft elastic vibrations when there are differences in elastic frequencies between ground modal tests and actual flight tests, or when unknown modes appear.

[0106] Although the present invention has been disclosed above with reference to preferred embodiments, it is not intended to limit the present invention. Any person skilled in the art can make possible changes and modifications to the technical solutions of the present invention by utilizing the methods and techniques disclosed above without departing from the spirit and scope of the present invention. Therefore, any simple modifications, equivalent changes and alterations made to the above embodiments based on the technical essence of the present invention without departing from the content of the technical solutions of the present invention shall fall within the protection scope of the technical solutions of the present invention.

Claims

1. A method for suppressing elastic vibration of an aircraft based on strong tracking Kalman filtering, characterized in that... include: The angular acceleration estimate is obtained based on the preset gyroscope data; A system model of angular velocity and elastic frequency is obtained based on the estimated angular acceleration value; The rigid body angular velocity estimation results are obtained based on the system model of angular velocity and elastic frequency; The rudder deflection command is obtained based on the rigid body angular velocity estimation result; The rigid body angular velocity estimation results obtained from the system model based on angular velocity and elastic frequency include: Based on the system model of angular velocity and elastic frequency and the preset state quantity update value of the previous sampling time, the predicted state quantity value of the current sampling time is obtained; The prediction error covariance matrix is ​​obtained based on the preset fading factor and the input noise matrix at the previous sampling time. The Kalman filter gain is obtained from the prediction error covariance matrix; The rigid body angular velocity estimation result is obtained based on the Kalman filter gain and the predicted state value at the current sampling time.

2. The method for suppressing elastic vibration of aircraft based on strong tracking Kalman filtering according to claim 1, characterized in that: The estimated value of angular acceleration is obtained by the following formula: in, This represents the estimated angular acceleration at the current sampling time, where T is the time step, k is the current sampling count, and ω is the current angular acceleration estimate. m (kT-T) represents the angular velocity measurement value at the previous sampling time, ω m (kT-2T) represents the angular velocity measurements at the first two sampling times.

3. The method for suppressing elastic vibration of aircraft based on strong tracking Kalman filtering according to claim 1, characterized in that: The system model for angular velocity and elastic frequency is obtained through the following formula: Where X(kT) is the state variable at the current sampling time, X(kT-T) is the state variable at the previous sampling time, T is the time step, k is the current sampling count, Φ(kT) is the state transition matrix at the current sampling time, and Ψ(kT) is the control transition matrix at the current sampling time. ω is the estimated angular acceleration at the current sampling time. m (kT) represents the measured angular velocity at the current sampling time, ζ(kT) represents the observation noise at the current sampling time, B is the manipulation matrix, C is the observation matrix, ω1 is the first-order elastic deformation mode frequency, ω2 is the second-order elastic deformation mode frequency, I represents the unit diagonal matrix, the superscript "T" indicates transpose, the superscript "i" indicates power i, "!" indicates factorial, and "Σ" indicates summation.

4. The method for suppressing elastic vibration of aircraft based on strong tracking Kalman filtering according to claim 1, characterized in that: Predicted state values ​​at the current sampling time It can be obtained through the following formula: Where T is the time step and k is the current sampling number. Let Φ(kT) be the state variable update value at the previous sampling time, Φ(kT) be the state transition matrix at the current sampling time, and Ψ(kT) be the control transition matrix at the current sampling time. This is the estimated angular acceleration value at the current sampling time.

5. The method for suppressing elastic vibration of aircraft based on strong tracking Kalman filtering according to claim 1, characterized in that: The prediction error covariance matrix is ​​obtained by the following formula: in, Let T be the prediction error covariance matrix at the current sampling time, T be the time step, k be the current sampling number, and n be the... The dimension of the vector is given by Q(kT-T), where Q(kT-T) is the input noise matrix at the previous sampling time, and K(kT-T) is the Kalman filter gain at the previous sampling time. To update the error covariance matrix at the previous sampling time, Let λ(kT-T) be the prediction error covariance matrix at the previous sampling time, and ω be the fading factor at the previous sampling time. m (kT-T) represents the angular velocity measurement at the previous sampling time, and C is the observation matrix. Let Ψ(kT) be the predicted state value at the previous sampling time, Ψ(kT-T) be the control transition matrix at the current sampling time, Ψ(kT-T) be the control transition matrix at the previous sampling time, R be the observation noise matrix, Φ(kT) be the state transition matrix at the current sampling time, Φ(kT-T) be the state transition matrix at the previous sampling time, and I denote the unit diagonal matrix. This is the variance estimate of the information d(kT-T) at the previous sampling time. The superscript "T" indicates transpose, the superscript "2" indicates square, and "Tr(·)" indicates the sum of the diagonal elements of the matrix.

6. The method for suppressing elastic vibration of aircraft based on strong tracking Kalman filtering according to claim 1, characterized in that: The Kalman filter gain is obtained using the following formula: Where K(kT) is the Kalman filter gain at the current sampling time, T is the time step, and k is the current sampling count. Let C be the prediction error covariance matrix at the current sampling time, C be the observation matrix, and R be the observation noise matrix.

7. The method for suppressing elastic vibration of aircraft based on strong tracking Kalman filtering according to claim 1, characterized in that: The rigid body angular velocity estimation result is obtained by the following formula: Where ω(kT) is the estimated rigid body angular velocity. Here, T is the state variable update value, k is the current sampling count, and K(kT) is the Kalman filter gain at the current time. ω is the predicted value of the state variable. m (kT) represents the measured angular velocity at the current sampling time, and C is the observation matrix.

8. An elastic vibration suppression system for aircraft based on strong tracking Kalman filtering, characterized in that... include: The first module is used to obtain an estimated value of angular acceleration based on preset gyroscope data; The second module is used to obtain a system model of angular velocity and elastic frequency based on the estimated angular acceleration value; The third module is used to obtain the rigid body angular velocity estimation result based on the system model of angular velocity and elastic frequency. The fourth module is used to obtain the rudder deflection command based on the rigid body angular velocity estimation result; The rigid body angular velocity estimation results obtained from the system model based on angular velocity and elastic frequency include: Based on the system model of angular velocity and elastic frequency and the preset state quantity update value of the previous sampling time, the predicted state quantity value of the current sampling time is obtained; The prediction error covariance matrix is ​​obtained based on the preset fading factor and the input noise matrix at the previous sampling time. The Kalman filter gain is obtained from the prediction error covariance matrix; The rigid body angular velocity estimation result is obtained based on the Kalman filter gain and the predicted state value at the current sampling time.

9. An electronic device, characterized in that, include: Memory: Used to store computer-readable instructions; and Processor: for executing the computer-readable instructions to perform the method as described in any one of claims 1 to 7.

Images