A laser communication terminal pointing control method based on a meta-model disturbance observer

Abstract

This invention relates to a laser communication terminal pointing control method based on a meta-model interference observer. Addressing the high-precision pointing control requirements of laser communication terminals under the influence of satellite platform vibration and friction, and tackling the problem of difficulty in accurately characterizing and estimating multi-source interference, the method includes: First, analyzing the sources of interference and constructing a terminal dynamic model incorporating the interference. A data-mechanism mapping is established using historical interference data and state-space Kriging modeling to obtain a meta-model of the interference. Second, designing a meta-model interference observer to achieve real-time estimation of the interference. Finally, using the real-time interference estimate, a robust backstepping controller is designed to suppress interference estimation errors and achieve desired pointing angle tracking, thus completing a laser communication terminal pointing control method based on a meta-model interference observer. The control method designed in this invention features simple implementation, high interference estimation accuracy, and strong robustness.

Description

Technical Field

[0001] This invention belongs to the field of aerospace control, specifically relating to a laser communication terminal pointing control method based on a meta-model interference observer. Background Technology

[0002] The pointing control performance of laser communication terminals is a key factor affecting the link establishment speed, communication quality, and efficiency of inter-satellite laser communication. In the microgravity environment of space, multi-source heterogeneous interference, such as vibration and friction of the satellite platform, is transmitted through the structure, causing fluctuations in the torque of the laser communication terminal's rotating shaft, thus leading to pointing angle deviation. Specifically, from the perspective of interference sources and characteristics, satellite platform vibration includes various resonant interferences such as attitude motion-induced vibration of flexible attachments and attitude actuator vibration. Furthermore, nonlinear friction directly acts on the corresponding rotating shaft of the laser communication terminal and affects other axes through coupling torque transmission. Therefore, anti-interference pointing control of laser communication terminals has become one of the key technologies for improving laser communication performance. Traditional control methods based on interference suppression are highly conservative, requiring a trade-off between control performance and anti-interference capability. Interference observer-based methods can further utilize interference model information, thereby reducing the conservatism of interference estimation. Accurately characterizing the meta-model (mechanistic model) of the interference is crucial for improving the accuracy of the observer estimation. Relying on ground testing environments, flexible attachments such as solar panels and antennas can experimentally obtain some resonant frequencies and damping information, while attitude control actuators such as flywheels contain harmonic interference related to the rotor frequency. Furthermore, by utilizing the known rotational speed and angle information of the laser communication terminal, test data on multi-source interference experienced by each axis can be obtained. However, how to use the interference test data to design a meta-model interference observer and achieve high-precision pointing control of the laser communication terminal remains a key technology that urgently needs to be overcome in the engineering application of inter-satellite laser communication networks.

[0003] Currently, research on pointing control of laser communication terminals rarely considers interference characteristics, focusing instead on interference suppression or conservative estimation. Chinese patent application CN201510515353.X proposes a coarse tracking control system for a space laser communication terminal based on a linear piezoelectric motor, considering electromagnetic interference to meet the dynamic performance requirements of the laser communication terminal tracking control system. This system uses the error calculation between the current angular position and the target angular position of the two axes of the laser communication terminal to generate control signals. However, this method does not fully consider the impact of torque interference on control performance. Chinese patent application CN202111614886.5 proposes a finite-time anti-interference control method for pointing control of a wideband inertial reference unit in laser communication. This method uses a finite-time extended state observer to estimate interference and combines it with a finite-time controller to reduce the impact of interference on the output. However, this method is relatively conservative in its handling of interference and does not fully utilize some known information about the interference. The literature "Application of Improved Internal Model Control in Airborne Laser Communication Systems" addresses the pointing control problem of the azimuth axis of a laser communication terminal. Based on traditional internal model control, it proposes a two-degree-of-freedom internal model control algorithm based on an observer, improving the system's anti-interference capability. However, this method does not fully analyze the relationship between interference characteristics and observer parameter selection, and its handling of interference is somewhat conservative.

[0004] In summary, under the influence of multi-source interference such as vibration and friction of satellite platforms, there is a lack of anti-interference pointing control methods for laser communication terminals that fully utilize prior known data and information about interference. It is urgent to overcome the problem of high-precision pointing control of periscope-type laser communication terminals based on data-driven meta-model interference observers. Summary of the Invention

[0005] To address the coarse pointing control requirements of periscope-type laser communication terminals under the influence of multi-source interference such as vibration and friction of satellite platforms, and to overcome the problem of insufficient utilization of prior known information on multi-source interference in existing methods, this invention provides a pointing control method for laser communication terminals based on a meta-model interference observer. The method uses data to achieve a meta-model representation of the interference, designs a meta-model interference observer, and uses the interference estimate to design a robust backstepping controller, thereby improving the accuracy of interference estimation and achieving high-precision, highly robust pointing control of the laser communication terminal.

[0006] To achieve the above objectives, the present invention adopts the following technical solution:

[0007] A laser communication terminal pointing control method based on a meta-model interference observer includes the following steps:

[0008] The first step is to establish a dynamic model of the laser communication terminal that includes interference. By means of ground frequency sweeping and vibration testing, the actual operating environment is simulated to obtain historical data of interference on the azimuth and pitch axes of the laser communication terminal. Then, the continuous element model of the interference is obtained by using the state space kriging modeling method.

[0009] The second step involves using the laser communication terminal dynamics model established in the first step and the obtained continuous element model of interference to design an element model interference observer, thereby achieving real-time and accurate estimation of multi-source interference.

[0010] The third step involves using the real-time accurate estimates of the multi-source interference obtained in the second step to design a robust backstepping controller, which completes interference feedforward compensation and interference estimation error feedback suppression, thereby achieving high-precision target angle tracking.

[0011] Furthermore, the first step includes:

[0012] (1) Establish a dynamic model of the laser communication terminal containing interference:

[0013] ,

[0014] in, q = [ q 1 , q 2 ] T This represents the pointing angle vector of the laser communication terminal. and These represent the azimuth and pitch angles of the laser communication terminal, respectively, with the upper right corner labeled "". " indicates solving for the transpose of a matrix or vector; and These represent the pointing angle vectors of the laser communication terminal, respectively. First and second derivatives with respect to time; This represents the control input torque vector; This represents the interference torque vector experienced by the laser communication terminal; Represents the inertia matrix of the laser communication terminal; This represents the Coriolis force and centrifugal force coefficient matrix of the laser communication terminal. Specifically, assuming the center of mass of each axis of the laser communication terminal is at the center of the connecting rod, the inertia matrix... With coefficient matrix They can be represented as:

[0015] M ( q ) = [ I Z 1 + I X 2 sin 2 ( q 2 ) + I Y 2 cos 2 ( q 2 ) + L 2 2 m 2 4 0 0 I Z 2 ] C ( q , q ˙ ) = [ ( I X 2 − I Y 2 ) sin ( q 2 ) cos ( q 2 ) q ˙ 2 ( I X 2 − I Y 2 ) sin ( q 2 ) cos ( q 2 ) q ˙ 1 ( I Y 2 − I X 2 ) sin ( q 2 ) cos ( q 2 ) q ˙ 1 0 ] ,

[0016] in, This represents the moment of inertia of the azimuth axis along its Z-axis. , and These represent the moments of inertia of the pitch axis along its X, Y, and Z axes, respectively. and These represent the angular velocities of the azimuth and pitch axes of the laser communication terminal, respectively. Indicates the length of the pitch axis linkage of the laser communication terminal; Represent the pitch axis mass of the laser communication terminal; define state variables. State variables The state-space model is obtained as follows:

[0017] ,

[0018] in, express The inverse matrix, This represents the inertia matrix of the laser communication terminal in the state-space model. This represents the matrix of Coriolis and centrifugal force coefficients of the laser communication terminal in the state-space model. , State variables , The first derivative with respect to time; d = M ( x 1 ) − 1 f d = [ d 1 , d 2 ] T , This represents the multi-source interference moment vectors along the azimuth and pitch axes of the laser communication terminal, derived from the multi-source interference along the azimuth axis. and pitch axis multi-source interference Composition. Through ground-based frequency sweeping and vibration testing, the actual operating environment is simulated. Using known control input torque, laser communication terminal pointing angle vector, and angular velocity vector data, the multi-source interference torque vectors of the laser communication terminal's azimuth and pitch axes are calculated. The solution results are recorded to obtain historical data of interference.

[0019] (2) Using the obtained historical disturbance data and the state-space kriging modeling method, the discrete element model of the disturbance is obtained:

[0020] ,

[0021] in, and Let represent the state vectors of the disturbance at time k and time k+1, respectively; and These represent the interference discrete state transition matrix and the output matrix obtained using historical interference data, respectively. This represents the disturbance calculated using the discrete element model of the disturbance at time k. The state transition matrix is ​​explained in detail below. The calculation method. Assume the scalar dynamics of the state transition matrix to be solved is... , definition includes The vector form of the step delay is z k = [ y k , y k − 1 , ⋯ , y k − n p + 1 ] T ,in This represents the data to be solved measured at time k. "This indicates that the same principle applies. Define the time delay vector." The successor vector form is z k + = [ y k + 1 , y k , ⋯ , y k − n p + 2 ] T Define a matrix containing historical data of an unknown dynamic system:

[0022] D = [ z ¯ 1 z ¯ 2 ⋯ z ¯ N ] ,

[0023] D + = [ z ¯ 1 + z ¯ 2 + ⋯ z ¯ N + ] ,

[0024] in, The number of data sets to be saved; , and They represent the 1st, 2nd and... Group-stored time delay vector Data vectors with the same structure; Input stored Historical data; It is a matrix The successor vector, i.e. , and They are respectively , and The successor vector. Input dynamics. The state transition matrix can be solved using the following formula:

[0025] { H = 2 ( H 1 + H 2 ) C = [ D T , 1 T ] T , C + = [ ( D + ) T , 1 T ] T A = 2 H − 1 H 2 + H − 1 C T ( CH − 1 C T ) − 1 ( C + − 2 CH − 1 H 2 ) ,

[0026] Among them, the upper right corner is marked with " " indicates that the inverse matrix is ​​being solved; and It is a positive definite weight matrix. This is the total weight matrix; To satisfy the requirement of a row vector consisting entirely of 1s in the matrix dimension; and For auxiliary parameter matrix; Input dynamics calculated from historical data The discrete state transition matrix. The input is dynamically taken as... and The discrete state transition matrices for azimuth and pitch axis disturbances can be obtained separately. and and historical data matrix and .structure and ,in This indicates the construction of a block diagonal matrix. and Represent matrices respectively and The first row of elements completes the construction of the discrete interference meta-model. Further, the discrete meta-model of the interference is transformed into a continuous meta-model:

[0027] ,

[0028] in, A continuous state vector representing the disturbance. express The first derivative with respect to time; This represents the discrete state transition matrix obtained using historical interference data. The continuous state transition matrix; This represents the disturbance calculated using the continuous element model.

[0029] The second step is to design the meta-model disturbance observer:

[0030] ,

[0031] in, This represents the state vector of the perturbed observer in the meta-model. Represents the state vector of the perturbed observer in the meta-model. The first derivative with respect to time; This represents the gain matrix of the disturbance observer in the meta-model, and its selection must ensure that the matrix... Negative definite; This represents the multi-source interference moment vectors representing the azimuth and elevation axes of the laser communication terminal. The estimated value.

[0032] The third step involves designing a robust backstepping controller to perform interference feedforward compensation and feedback to suppress interference estimation errors.

[0033] ,

[0034] in, For the laser communication terminal to track the pointing angle vector, and These represent the desired tracking pointing angle vectors of the laser communication terminal. First and second derivatives with respect to time; and This is the state tracking error vector; Represents the virtual control vector. Virtual control vector The first derivative with respect to time; Represents the auxiliary control input vector; and The gain matrix of the robust backstepping controller can be selected as follows: and ,in and It is a constant value greater than zero. This is the identity matrix for the corresponding dimension.

[0035] The beneficial effects of this invention are as follows:

[0036] This invention employs a meta-model construction method based on historical interference data, combined with a meta-model interference observer and a robust backstepping controller, to achieve high-precision pointing control of laser communication terminals under the influence of multi-source heterogeneous interference. It features simple implementation, high interference estimation accuracy, and strong robustness, and is suitable for pointing control systems of laser communication terminals with multi-source interference.

[0037] The method of this invention for laser communication terminal pointing control enables the learning and modeling of complex interference characteristics by incorporating ground interference test data. The dynamic and steady-state performance of the laser communication terminal's pointing can be adjusted by modifying the gain matrices of the observer and controller. Compared to controllers that do not utilize interference models or historical data, the designed controller enhances the overall system's anti-interference capability, reduces the conservatism of interference estimation, and improves the accuracy and speed of laser communication terminal pointing control. Attached Figure Description

[0038] Figure 1 This is a flowchart of a laser communication terminal pointing control method based on a meta-model interference observer according to the present invention. Detailed Implementation

[0039] The present invention will now be described in detail with reference to the accompanying drawings and embodiments.

[0040] like Figure 1 As shown, the laser communication terminal pointing control method based on a meta-model interference observer of the present invention includes the following steps:

[0041] The first step is to establish a dynamic model of the laser communication terminal that includes interference:

[0042] ,

[0043] in, q = [ q 1 , q 2 ] T This represents the pointing angle vector of the laser communication terminal. and These represent the azimuth and pitch angles of the laser communication terminal, respectively, with the upper right corner labeled "". " indicates solving for the transpose of a matrix or vector; and These represent the pointing angle vectors of the laser communication terminal, respectively. First and second derivatives with respect to time; This represents the control input torque vector; This represents the interference torque vector experienced by the laser communication terminal; Represents the inertia matrix of the laser communication terminal; This represents the Coriolis force and centrifugal force coefficient matrix of the laser communication terminal. Specifically, assuming the center of mass of each axis of the laser communication terminal is at the center of the connecting rod, the inertia matrix... With coefficient matrix They can be represented as:

[0044] M ( q ) = [ I Z 1 + I X 2 sin 2 ( q 2 ) + I Y 2 cos 2 ( q 2 ) + L 2 2 m 2 4 0 0 I Z 2 ] C ( q , q ˙ ) = [ ( I X 2 − I Y 2 ) sin ( q 2 ) cos ( q 2 ) q ˙ 2 ( I X 2 − I Y 2 ) sin ( q 2 ) cos ( q 2 ) q ˙ 1 ( I Y 2 − I X 2 ) sin ( q 2 ) cos ( q 2 ) q ˙ 1 0 ] ,

[0045] in, This represents the moment of inertia of the azimuth axis along its Z-axis. , and These represent the moments of inertia of the pitch axis along its X, Y, and Z axes, respectively. and These represent the angular velocities of the azimuth and pitch axes of the laser communication terminal, respectively. Indicates the length of the pitch axis linkage of the laser communication terminal; Represent the pitch axis mass of the laser communication terminal; define state variables. State variables The state-space model is obtained as follows:

[0046] ,

[0047] in, express The inverse matrix, This represents the inertia matrix of the laser communication terminal in the state-space model. This represents the matrix of Coriolis and centrifugal force coefficients of the laser communication terminal in the state-space model. , State variables , The first derivative with respect to time; d = M ( x 1 ) − 1 f d = [ d 1 , d 2 ] T , This represents the multi-source interference moment vectors along the azimuth and pitch axes of the laser communication terminal, derived from the multi-source interference along the azimuth axis. and pitch axis multi-source interference Composition. Through ground-based frequency sweeping and vibration testing, the actual operating environment is simulated. Using known control input torque, laser communication terminal pointing angle vector, and angular velocity vector data, the multi-source interference torque vectors of the laser communication terminal's azimuth and pitch axes are calculated. The solution results are recorded to obtain historical data of interference.

[0048] Using the obtained historical disturbance data and the state-space kriging modeling method, a discrete element model of the disturbance is obtained:

[0049] ,

[0050] in, and Let represent the state vectors of the disturbance at time k and time k+1, respectively; and These represent the interference discrete state transition matrix and the output matrix obtained using historical interference data, respectively. This represents the disturbance calculated using the discrete element model of the disturbance at time k. The state transition matrix is ​​explained in detail below. The calculation method. Assume the scalar dynamics of the state transition matrix to be solved is... , definition includes The vector form of the step delay is z k = [ y k , y k − 1 , ⋯ , y k − n p + 1 ] T ,in This represents the data to be solved measured at time k. "This indicates that the same principle applies. Define the time delay vector." The successor vector form is z k + = [ y k + 1 , y k , ⋯ , y k − n p + 2 ] T Define a matrix containing historical data of an unknown dynamic system:

[0051] D = [ z ¯ 1 z ¯ 2 ⋯ z ¯ N ] ,

[0052] D + = [ z ¯ 1 + z ¯ 2 + ⋯ z ¯ N + ] ,

[0053] in, The number of data sets to be saved; , and They represent the 1st, 2nd and... Group-stored time delay vector Data vectors with the same structure; Input stored Historical data; It is a matrix The successor vector, i.e. , and They are respectively , and The successor vector. Input dynamics. The state transition matrix can be solved using the following formula:

[0054] { H = 2 ( H 1 + H 2 ) C = [ D T , 1 T ] T , C + = [ ( D + ) T , 1 T ] T A = 2 H − 1 H 2 + H − 1 C T ( CH − 1 C T ) − 1 ( C + − 2 CH − 1 H 2 ) ,

[0055] Among them, the upper right corner is marked with " " indicates that the inverse matrix is ​​being solved; and It is a positive definite weight matrix. This is the total weight matrix; To satisfy the requirement of a row vector consisting entirely of 1s in the matrix dimension; and For auxiliary parameter matrix; Input dynamics calculated from historical data The discrete state transition matrix. The input is dynamically taken as... and The discrete state transition matrices for azimuth and pitch axis disturbances can be obtained separately. and and historical data matrix and .structure and ,in This indicates the construction of a block diagonal matrix. and Represent matrices respectively and The first row of elements completes the construction of the discrete interference meta-model. Further, the discrete meta-model of the interference is transformed into a continuous meta-model:

[0056] ,

[0057] in, A continuous state vector representing the disturbance. express The first derivative with respect to time; This represents the discrete state transition matrix obtained using historical interference data. The continuous state transition matrix; This represents the disturbance calculated using the continuous element model.

[0058] The second step is to design the meta-model disturbance observer:

[0059] ,

[0060] in, This represents the state vector of the perturbed observer in the meta-model. Represents the state vector of the perturbed observer in the meta-model. The first derivative with respect to time; This represents the gain matrix of the disturbance observer in the meta-model, and its selection must ensure that the matrix... Negative definite; This represents the multi-source interference moment vectors representing the azimuth and elevation axes of the laser communication terminal. The estimated value.

[0061] The third step is to design a robust backstepping controller to perform interference feedforward compensation and feedback to suppress interference estimation errors.

[0062] ,

[0063] in, For the laser communication terminal to track the pointing angle vector, and These represent the desired tracking pointing angle vectors of the laser communication terminal. First and second derivatives with respect to time; and This is the state tracking error vector; Represents the virtual control vector. Virtual control vector The first derivative with respect to time; Represents the auxiliary control input vector; and The gain matrix of the robust backstepping controller can be selected as follows: and ,in and It is a constant value greater than zero. This is the identity matrix for the corresponding dimension.

[0064] The specific embodiments described above further illustrate the purpose, technical solution, and beneficial effects of the present invention. It should be understood that the above descriptions are merely specific embodiments of the present invention and are not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A laser communication terminal pointing control method based on a meta-model interference observer, characterized in that, Includes the following steps: The first step is to establish a dynamic model of the laser communication terminal that includes interference. By using ground-based frequency sweeping and vibration testing, the actual operating environment is simulated to obtain historical interference data for the azimuth and pitch axes of the laser communication terminal. Then, using the state-space kriging modeling method, a continuous element model of the interference is obtained. The second step involves using the laser communication terminal dynamics model established in the first step and the obtained continuous element model of interference to design an element model interference observer, thereby achieving real-time and accurate estimation of multi-source interference. The third step is to use the real-time accurate estimates of the multi-source interference obtained in the second step to design a robust backstepping controller, complete the interference feedforward compensation and interference estimation error feedback suppression, and achieve high-precision desired pointing angle tracking. The first step includes: (1) Establish a dynamic model of the laser communication terminal containing interference: ， in, This represents the pointing angle vector of the laser communication terminal; and These represent the pointing angle vectors of the laser communication terminal, respectively. First and second derivatives with respect to time; This represents the control input torque vector; This represents the interference torque vector experienced by the laser communication terminal; Represents the inertia matrix of the laser communication terminal; Represent the Coriolis force and centrifugal force coefficient matrix of the laser communication terminal; define the state variables. State variables The state-space model is obtained as follows: ， in, express The inverse matrix, This represents the inertia matrix of the laser communication terminal in the state-space model. This represents the matrix of Coriolis and centrifugal force coefficients of the laser communication terminal in the state-space model. , State variables , The first derivative with respect to time; , This represents the multi-source interference torque vectors along the azimuth and pitch axes of the laser communication terminal; it simulates the actual operating environment through ground-based frequency sweeping and vibration testing, utilizing known control input torque vectors. Using the pointing angle vector and angular velocity vector data of the laser communication terminal, the multi-source interference torque vectors of the azimuth and pitch axes of the laser communication terminal are calculated. Record the solution results to obtain historical interference data; (2) Using the obtained historical disturbance data and the state-space kriging modeling method, the discrete element model of the disturbance is obtained: ， in, and Let represent the state vectors of the disturbance at time k and time k+1, respectively; and These represent the interference discrete state transition matrix and the output matrix obtained using historical interference data, respectively. This represents the disturbance calculated using the discrete element model of the disturbance at time k; the discrete element model of the disturbance is then converted into a continuous element model: ， in, A continuous state vector representing the disturbance. express The first derivative with respect to time; This represents the discrete state transition matrix obtained using historical interference data. The continuous state transition matrix; This represents the interference calculated using the continuous element model; The second step includes: Design a meta-model interference observer: ， in, This represents the state vector of the perturbed observer in the meta-model. Represents the state vector of the perturbed observer in the meta-model. The first derivative with respect to time; This represents the gain matrix of the perturbation observer in the meta-model. This represents the multi-source interference moment vectors representing the azimuth and elevation axes of the laser communication terminal. The estimated value; The third step includes: Design a robust backstepping controller to perform disturbance feedforward compensation and feedback to suppress disturbance estimation errors: ， in, For the laser communication terminal to track the pointing angle vector, and These represent the desired tracking pointing angle vectors of the laser communication terminal. First and second derivatives with respect to time; and This is the state tracking error vector; Represents the virtual control vector. Virtual control vector The first derivative with respect to time; Represents the auxiliary control input vector; and This is the gain matrix of the robust backstepping controller.

Images