Kalman filtering composite axis control method and system based on inertial sensor measurement
The Kalman filter composite axis control method based on inertial sensor measurement solves the problems of weak high-frequency disturbance suppression and model mismatch in composite axis systems, achieving high precision and disturbance resistance stability of photoelectric tracking systems and meeting the application requirements of high-end fields.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2025-11-28
- Publication Date
- 2026-03-10
AI Technical Summary
Existing composite axis control technology suffers from problems in high-end fields, such as weak high-frequency disturbance suppression capability, conflict between inertial sensor installation and fast-reflecting mirror performance, and disturbance compensation being easily affected by model mismatch, making it difficult to achieve high-precision and stable tracking.
A Kalman filter composite axis control method based on inertial sensor measurement is adopted. The inertial measurement data is input into the Kalman filter for fusion and recursive optimal estimation, separating measurement noise and disturbance information. A high-pass feedforward controller is used to suppress mid-to-high frequency disturbances, and dual closed-loop control is achieved by combining coarse tracking and fine tracking modules.
The photoelectric tracking system achieves high tracking accuracy and strong anti-disturbance stability in complex scenarios, meeting the stringent requirements of high-end fields such as aerospace and laser communication.
Smart Images
Figure CN121634825A_ABST
Abstract
Description
TECHNICAL FIELD
[0001] The present application belongs to the technical field of photoelectric tracking, and particularly relates to a Kalman filtering compound axis control method and system based on inertial sensor measurement. BACKGROUND
[0002] With the rapid development of high-end fields such as aerospace, astronomical observation and laser communication, the demand for precision and stability of photoelectric tracking systems is increasingly stringent. Such systems need to overcome two major challenges: first, the wideband, large amplitude mechanical vibration and attitude change brought by the carrier (such as satellite, airplane, ship); second, the need to achieve sustained and stable accurate tracking of high-speed moving targets. To solve the above problems, the traditional single-stage servo tracking system cannot balance tracking accuracy and response bandwidth due to the limitations of mechanical resonance, sensor noise and speed loop bandwidth, therefore, the compound axis control architecture emerges as the times require. The architecture is composed of a coarse tracking module and a fine tracking module in series, the coarse tracking is responsible for large range, low bandwidth tracking and disturbance preliminary suppression, and the fine tracking compensates for residual error and high frequency disturbance, significantly improving the overall tracking accuracy, but the existing technology still has many defects.
[0003] Specifically, the existing compound axis control technology has obvious deficiencies: first, the mid-high frequency disturbance suppression capability is weak, some schemes do not combine inertial sensors, and can only deal with low frequency disturbances; second, in the schemes using inertial sensors, disturbance compensation often relies on accurate models and is realized through disturbance observers, which is easily affected by temperature and load changes, leading to model mismatch and compensation failure; third, if the inertial sensor is directly installed on the fast mirror, the selection is strict due to the space limitation of the light path and may affect the tracking performance; in addition, the traditional disturbance decoupling method based on high-speed filtering will lose low frequency disturbance information and introduce phase lag, while the decoupling method based on internal model principle does not fully consider the disturbance suppression effect of the coarse tracking loop.
[0004] In summary, the existing technology cannot achieve high-precision stable tracking in the presence of model uncertainty, sensor noise and complex disturbance, and a new compound axis control method is needed to solve the above defects to meet the application requirements of high-end fields. SUMMARY
[0005] The purpose of the embodiments of the present application is to provide a Kalman filtering compound axis control method and system based on inertial sensor measurement. To solve the problems of weak mid-high frequency disturbance suppression in the compound axis system, conflict between inertial sensor installation and fast mirror performance, and disturbance compensation being easily affected by model mismatch.
[0006] To achieve the above purpose, the following technical solutions are adopted in the present application: In a first aspect, a Kalman filtering compound axis control method based on inertial sensor measurement is provided, comprising the following steps: inertial measurement data reflecting the coarse tracking visual axis motion, the inertial measurement data comprising at least angular velocity data and angular acceleration data; Step S2: inputting the inertial measurement data into a Kalman filter taking angle, angular velocity and angular acceleration as state vectors; fusing and recursively optimally estimating the angular velocity and angular acceleration data containing noise through the Kalman filter to separate measurement noise and extract optimal angular velocity estimation values containing disturbance information; Step S3: generating a velocity loop control signal of the coarse tracking system based on the optimal angular velocity estimation values; and generating a position loop control signal of the coarse tracking system based on coarse tracking miss distance information; Step S4: generating a control signal of the fine tracking system based on fine tracking miss distance information; Step S5: performing feedforward processing on the optimal angular velocity estimation values to generate a feedforward compensation signal, and synthesizing the feedforward compensation signal with the control signal of the fine tracking system to suppress medium and high frequency disturbances.
[0007] In a possible implementation, the Kalman filter is implemented based on the following equations: a system discretization state equation: X(k)=F×X(k-1)+W(k), where X(k) is a state vector at time k, F is a state transition matrix, X(k-1) is a state vector at time k-1, and W(k) is process noise; the measurement equation is: Z(k)=H×X(k)+V(k), where Z(k) is an observation at time k, H is a measurement matrix, and V(k) is an inertial sensor measurement noise sequence.
[0008] In a possible implementation, the expression of the state transition matrix F is:
[0009] where T is a sampling time interval of the Kalman filter.
[0010] In a possible implementation, the expression of the measurement matrix H is:
[0011] The observation Z(k) includes angular velocity measurement values and angular acceleration measurement values collected by the inertial sensor.
[0012] In a possible implementation, the recursive optimal estimation process of the Kalman filter includes: The current frame attitude is predicted based on the last frame state estimation value, a prediction error covariance matrix is calculated, a Kalman filter gain is updated, and the system state estimation and the estimation error covariance matrix are updated in combination with the inertial sensor measurement value.
[0013] In a possible implementation, in step S1, the inertial measurement data is collected by an inertial sensor mounted on a coarse tracking camera structure; the inertial sensor is a three-axis gyroscope, a three-axis accelerometer, or an inertial measurement unit integrating both.
[0014] In a possible implementation, in step S5, the feedforward processing is implemented by a feedforward controller, and a transfer function of the feedforward controller is in the form of a high-pass filter:
[0015] where Ks is a gain coefficient, ω f is a cutoff angular frequency, used for filtering low-frequency disturbances and compensating only for medium-high frequency disturbances.
[0016] In a second aspect, a Kalman filter compound axis control system based on inertial sensor measurement is provided, comprising: an inertial measurement unit mounted on a coarse tracking camera structure, configured to collect angular velocity and angular acceleration data of a coarse tracking visual axis; a Kalman filter module configured to construct a discretized state equation and a measurement equation with angles, angular velocities, and angular accelerations as state vectors, to fuse the data collected by the inertial measurement unit, to separate measurement noise and disturbance information, and to output optimal angular velocity estimation values containing disturbance information; a coarse tracking control module including a velocity loop controller and a position loop controller, the velocity loop controller being configured to construct a velocity feedback loop based on the optimal angular velocity estimation values, and the position loop controller being configured to implement position control in combination with coarse tracking miss distance information, to form a position-velocity double closed-loop control; a fine tracking control module including a fine tracking controller, configured to construct a fine tracking closed-loop control and to compensate for residual errors; a feedforward control module adopting a high-pass filter structure, configured to feed forward the optimal angular velocity estimation values containing disturbance information output by the Kalman filter module to an output end of the fine tracking controller, to suppress medium-high frequency disturbances.
[0017] In a possible implementation, an actuator of the coarse tracking control module is a U-shaped frame two-dimensional turntable, and an actuator of the fine tracking control module is a fast mirror.
[0018] In a possible implementation, in the Kalman filter module, the process noise W(k) and the measurement noise V(k) are mutually independent Gaussian white noises, the covariance matrix of the process noise is Q(k), and the covariance matrix of the measurement noise is R(k).
[0019] Compared with the prior art, the application has the following beneficial effects: The Kalman filter composite axis control method based on inertial sensor measurement provided in the application deploys the inertial sensor on a coarse tracking camera structure, thereby avoiding the installation space limitation of the fast reaction mirror and ensuring accurate collection of the line-of-sight inertial information. Meanwhile, the inertial measurement data is input into the Kalman filter for fusion and recursive optimal estimation, effectively separating the measurement noise and disturbance information, and the output optimal angular velocity estimation value supports the stable control of the coarse tracking position-velocity double closed loop and is targeted to suppress the medium and high frequency disturbances through the high-pass feedforward controller. The method combines the wide-range capture capability of coarse tracking and the high-precision compensation characteristics of fine tracking, effectively solves the problems of narrow disturbance suppression bandwidth and model mismatch affecting compensation accuracy in the traditional composite axis system, and finally realizes high tracking precision and strong anti-disturbance stability of the photoelectric tracking system in the complex scene of carrier vibration and target high-speed motion, meeting the stringent requirements of high-end fields such as aerospace observation and laser communication.
[0020] In a possible implementation, the processing of the inertial measurement data by the Kalman filter realizes effective separation of the measurement noise and disturbance information, and the output optimal angular velocity estimation value supports the coarse tracking velocity closed loop and the fine tracking feedforward compensation at the same time. The feedforward control module adopts a high-pass filter structure, which can accurately filter low-frequency interference and focus on medium and high frequency disturbance suppression, finally enabling the composite axis system to have wide-range capture capability of coarse tracking and realize high-precision compensation effect of fine tracking in the complex scene of carrier vibration and model parameter change, significantly improving the comprehensive tracking precision and anti-disturbance stability of the photoelectric tracking system, and meeting the stringent application requirements of high-end fields such as aerospace observation and laser communication. BRIEF DESCRIPTION OF DRAWINGS
[0021] Figure 1 A control block diagram of the Kalman filter composite axis control method based on inertial sensor measurement provided in the application is shown in the figure; Figure 2 A tracking error comparison between the traditional composite axis control method and the method of the application is shown in the figure. DETAILED DESCRIPTION
[0022] In the following, only certain exemplary embodiments are simply described. As those skilled in the art can recognize, the described embodiments can be modified in various different ways without departing from the spirit or scope. Therefore, the drawings and descriptions are to be regarded as illustrative in nature rather than restrictive.
[0023] In the description of the present application, it needs to be understood that the terms "center", "longitudinal", "lateral", "length", "width", "thickness", "upper", "lower", "front", "back", "left", "right", "vertical", "horizontal", "top", "bottom", "inner", "outer", "clockwise", "counterclockwise", "axial", "radial", "circumferential" and the like indicate the orientation or positional relationship based on the orientation or positional relationship shown in the drawings, which are only for the convenience of describing the present application and simplifying the description, and do not indicate or imply that the device or element referred to must have a particular orientation, be constructed and operated in a particular orientation, and therefore cannot be understood as a limitation on the present application.
[0024] In addition, the terms "first", "second" are only for descriptive purposes and cannot be understood as indicating or implying relative importance or implicitly indicating the number of the technical features indicated. Therefore, the features defined with "first", "second" can explicitly or implicitly include one or more of the features. In the description of the present application, the meaning of "a plurality of" is two or more, unless otherwise explicitly specified and limited. The specific embodiments of the present application are further described in detail below with reference to the accompanying drawings.
[0025] In the present application, unless otherwise explicitly specified and limited, the terms "mounting", "connecting", "connecting", "fixing" and the like should be understood broadly, for example, it can be fixed connection, or detachable connection, or integral; it can be mechanical connection, or electrical connection, or communication; it can be directly connected, or indirectly connected through intermediate medium, or the internal communication of two elements or the interaction relationship between two elements. For those skilled in the art, the specific meaning of the above terms in the present application can be understood according to the specific circumstances.
[0026] In the present application, unless otherwise explicitly specified and limited, the first feature "on" or "under" the second feature can include the first and second features directly contacting, or the first and second features not directly contacting but contacting through another feature between them. Moreover, the first feature "on", "above" and "on" the second feature includes the first feature directly above and obliquely above the second feature, or only indicates that the horizontal height of the first feature is higher than that of the second feature. The first feature "under", "below" and "under" the second feature includes the first feature directly above and obliquely above the second feature, or only indicates that the horizontal height of the first feature is less than that of the second feature.
[0027] The technical solutions in the embodiments of the present application will be described clearly and completely below in combination with the drawings in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, rather than all the embodiments. Based on the embodiments in the present application, all the other embodiments obtained by those skilled in the art without creative work are within the scope of protection of the present application.
[0028] As shown in Figure 1 and Figure 2 , the present application is a Kalman filter composite axis control method based on inertial sensor measurement, comprising the following steps: Step S1: obtaining inertial measurement data reflecting the motion of the coarse tracking visual axis, specifically, the inertial sensor can be installed on the coarse tracking camera structure to collect the angular velocity and angular acceleration data of the coarse tracking visual axis.
[0029] Step 2: inputting the inertial measurement data into a Kalman filter with angle, angular velocity and angular acceleration as state vectors; the Kalman filter is used to fuse and recursively estimate the angular velocity and angular acceleration data containing noise, so as to separate the measurement noise and extract the optimal angular velocity estimate containing disturbance information.
[0030] Specifically, since the inertial sensor has drift error and measurement noise, the Kalman filter is used for data fusion and optimal estimation.
[0031] The angle, angular velocity and angular acceleration are selected as the state vectors, denoted as The discrete state equation of the system is as follows:
[0032] In the formula:
[0033] denotes the state transition matrix, denotes the sampling time interval, denotes the process noise.
[0034] The observation is denoted as The measurement equation of the system is: ; In the formula:
[0035] denotes the system measurement matrix, denotes the inertial sensor measurement noise sequence.
[0036] In the discrete state equation and the measurement equation, and are mutually independent Gaussian white noises, which satisfy the following statistical properties: wherein, denotes the desired operator, denotes the covariance matrix of process noise, denotes the covariance matrix of measurement noise .
[0037] According to the Kalman filtering algorithm, the optimal state estimation is performed.
[0038] According to the state estimation of the previous frame , the attitude prediction of the current frame is performed to obtain the state prediction value:
[0039] According to the estimation error covariance matrix of the previous frame , the prediction error covariance matrix is calculated:
[0040] According to the measurement value Z(k) of the inertial sensor, the system state estimation is updated:
[0041] The estimation error covariance matrix is updated:
[0042] Step 3: Based on the optimal angular velocity estimation value, the velocity loop control signal of the coarse tracking system is generated; and based on the coarse tracking miss distance information, the position loop control signal of the coarse tracking system is generated.
[0043] The system angular velocity estimated by the Kalman filter is taken as the velocity feedback, and according to the characteristics of the two-dimensional turntable , the velocity loop controller of the coarse tracking system is designed . In combination with the position loop controller designed based on the coarse tracking miss distance information , the coarse tracking position-velocity double closed-loop control is completed.
[0044] Step 4: Based on the fine tracking miss distance information, the control signal of the fine tracking system is generated.
[0045] According to the characteristics of the fast reflecting mirror , the fine tracking controller is designed, and the fine tracking closed-loop control is constructed; Step 5: The optimal angular velocity estimation value is fed forward to generate a feedforward compensation signal, and the feedforward compensation signal is combined with the control signal of the fine tracking system to suppress the medium and high frequency disturbances.
[0046] The feedforward controller is designed The optimal angular velocity estimation value containing disturbance information is fed forward to the output end of the fine tracking controller, the feedforward controller adopts a high-pass filter form for suppressing medium-high frequency disturbances, in which:
[0047] In a possible embodiment, by deploying the inertial sensor on the coarse tracking camera structure, both the installation space limitation of the fast return mirror is avoided and the accurate acquisition of the line-of-sight inertia information is ensured; at the same time, by taking angle, angular velocity and angular acceleration as the state vector in the Kalman filtering model, the effective separation of the measurement noise and the disturbance information is realized, and the output optimal angular velocity estimation value not only supports the stable control of the coarse tracking position-velocity double closed loop, but also specifically suppresses the medium-high frequency disturbance through the high-pass feedforward controller.
[0048] The method combines the wide-range capture capability of coarse tracking with the high-precision compensation characteristics of fine tracking, effectively solves the problems of narrow disturbance suppression bandwidth and model mismatch affecting compensation precision in the traditional compound axis system, and finally realizes the high tracking precision and strong anti-disturbance stability of the photoelectric tracking system under the complex scene of carrier vibration and high-speed target motion, meeting the stringent demands of high-end fields such as aerospace observation and laser communication.
[0049] In a possible embodiment, a Kalman filtering compound axis control system based on inertial sensor measurement is provided, including an inertial measurement unit (IMU), a Kalman filtering module, a coarse tracking control module, a fine tracking control module and a feedforward control module.
[0050] The inertial measurement unit is installed on the coarse tracking camera structure and is used to collect angular velocity and angular acceleration data of the coarse tracking line-of-sight.
[0051] The Kalman filtering module is used to construct a discretized state equation and a measurement equation with angle, angular velocity and angular acceleration as the state vector, fuse the data collected by the inertial measurement unit, separate the measurement noise and the disturbance information, and output an optimal angular velocity estimation value containing disturbance information.
[0052] The coarse tracking control module includes a velocity loop controller and a position loop controller, the velocity loop controller constructs a velocity feedback loop based on the optimal angular velocity estimation value, and the position loop controller realizes position control in combination with coarse tracking miss distance information, forming a position-velocity double closed loop control.
[0053] The fine tracking control module includes a fine tracking controller, which is used to construct a fine tracking closed loop control and compensate residual errors.
[0054] The feedforward control module adopts a high-pass filter structure, which is used for feeding forward the optimal angular velocity estimation value containing disturbance information output by the Kalman filter module to an output end of the fine tracking controller, so as to suppress the medium-high frequency disturbance.
[0055] The principle of the control system is as follows: The Kalman filter composite axis control system based on the measurement of the inertial sensor is combined with the signal flow direction and the module interaction, and the core control principle is as follows: 1. Principle of the coarse tracking control subsystem Target deviation sensing and position loop starting: after the external target signal is input, the CCD image sensor collects the angle deviation of the target and the coarse tracking visual axis, inputs the deviation signal into the position loop controller, and generates a coarse tracking speed instruction.
[0056] Speed closed loop and inertial information feedback: the speed instruction enters the speed loop controller , and drives the coarse tracking actuator to act; meanwhile, the inertial measurement unit installed in the coarse tracking camera structure collects the angular velocity / acceleration of the coarse tracking visual axis in real time, is filtered through the Filter module, is denoised through the Kalman filter module, is fed back to the speed loop controller as speed feedback, and constitutes a coarse tracking speed closed loop; another path is integrated through an integral element 1 / s, and the position data containing the medium-high frequency disturbance information are introduced into the fine tracking position closed loop through the feedforward controller, so that the external disturbance is compensated . The influence of the external disturbance on the system performance is limited.
[0057] The double closed loop control of the coarse tracking can realize large-range target capture and low-frequency disturbance suppression. However, the suppression capability for the medium-high frequency disturbance is limited, so the system performance needs to be further improved through the fine tracking loop.
[0058] 2. Principle of the fine tracking control subsystem Residual error compensation starting: the residual deviation of the output angle of the coarse tracking visual axis and the target angle is collected by another CCD image sensor and input into the fine tracking controller .
[0059] High-precision error correction: the fine tracking controller generates a micro-adjustment instruction, drives the fine tracking actuator to act, and compensates for the residual error of the coarse tracking in a high frequency and high precision.
[0060] Disturbance cooperative suppression: the disturbance information output by the inertial measurement unit in the coarse tracking loop is synchronously transmitted to the input end of the fine tracking controller through the feedforward path, so that the fine tracking subsystem responds to the medium-high frequency disturbance in advance, and the coarse tracking subsystem forms low-frequency and high-frequency disturbance suppression in the full frequency band.
[0061] 3. Core role of Kalman filter module The inertial data collected by the inertial measurement unit is first input into the Kalman filter module. Discrete equations are constructed with angle, angular velocity, and angular acceleration as state vectors. The optimal angular velocity estimation value output by the filter algorithm separates measurement noise and disturbance information, supports the feedback of the coarse tracking speed loop, and provides accurate disturbance input for the feedforward controller, realizing the cooperative control of "inertial information-double closed loop-feedforward".
[0062] Specifically, the coarse tracking control module is a U-shaped two-dimensional turntable, and its model is assumed to be The fine tracking control module is a fast response mirror, and its model is assumed to be .
[0063] The specific steps are as follows: A three-axis inertial measurement unit containing three-axis fiber-optic gyroscopes and three-axis quartz accelerometers is installed on the coarse tracking U-shaped frame to measure the angular velocity and angular acceleration of the coarse tracking visual axis.
[0064] Data fusion is performed using the Kalman filter algorithm.
[0065] Angle, angular velocity, and angular acceleration are selected as state vectors, denoted as ; State transition matrix:
[0066] Sampling period , Measurement matrix:
[0067] Covariance matrix of process noise , covariance matrix of measurement noise .
[0068] According to the Kalman filter algorithm, the optimal state estimation is performed.
[0069] According to the state estimation of the previous frame , the attitude of the current frame is predicted to obtain the state prediction value:
[0070] According to the estimation error covariance matrix of the previous frame , the prediction error covariance matrix is calculated:
[0071] Update the Kalman filter gain:
[0072] According to the measurement value of the inertial sensor, the system state estimation is updated:
[0073] The estimation error covariance matrix is updated:
[0074] The system angular velocity estimated by the Kalman filter is taken as the velocity feedback According to the two-dimensional turntable characteristics, a velocity loop controller of the coarse tracking system is designed:
[0075] The position loop controller is designed in combination with the coarse tracking off-target amount information:
[0076] Thus, the coarse tracking position-velocity double closed loop control is completed.
[0077] According to the fast reflecting mirror characteristics The fine tracking controller is designed:
[0078] The feedforward controller is designed The angular velocity (containing disturbance information) estimated by the Kalman filter is fed forward to the output of the fine tracking controller, so that the anti-disturbance ability of the system is improved, and the medium and high frequency disturbances are suppressed.
[0079] Under the same disturbance condition, the traditional compound axis control method and the Kalman filter compound axis control method based on the measurement of the inertial sensor are compared. For example Figure 2 is a tracking error comparison diagram, and it can be obviously seen that the method of the application can effectively improve the disturbance suppression ability of the system, and greatly reduce the tracking residual error.
[0080] In the embodiments of the application, the inertial measurement unit is arranged in the coarse tracking camera structure, so that the installation space limitation of the fast reflecting mirror is avoided, and the accurate collection of inertial information is ensured; the Kalman filter module is combined to realize effective separation of the measurement noise and the disturbance information, and the optimal angular velocity estimation value output supports the coarse tracking velocity closed loop and the fine tracking feedforward compensation; the feedforward control module adopts a high-pass filter structure, can accurately filter low-frequency interference and focus on medium and high frequency disturbance suppression, and finally makes the compound axis system have the large-range capture ability of coarse tracking and realize the high-precision compensation effect of fine tracking in the complex scene of carrier vibration and model parameter change, significantly improves the comprehensive tracking precision and anti-disturbance stability of the photoelectric tracking system, and can meet the strict application requirements of high-end fields such as aerospace observation and laser communication.
[0081] It should be noted that the hardware platform for implementing the control method is not limited in the present application, and the method can be implemented on various hardware platforms such as embedded hardware, special chips, PLC, computer, etc., and the specific implementation manner thereof on any hardware platform belongs to the protection scope of the present patent.
[0082] Finally, it should be noted that: the above embodiments are only used to illustrate the technical solutions of the present application, and not to limit them; although the present application has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that the modification of the technical solutions recorded in the foregoing embodiments, or the equivalent replacement of part or all of the technical features, does not make the essence of the corresponding technical solution deviate from the scope of the technical solutions of the embodiments of the present application.
Claims
1. A Kalman filter compound axis control method based on inertial sensor measurements, characterized by, The method comprises the following steps: Step S1: acquiring inertial measurement data reflecting coarse tracking visual axis motion, the inertial measurement data comprising at least angular velocity data and angular acceleration data; Step S2: inputting the inertial measurement data into a Kalman filter with angle, angular velocity and angular acceleration as state vectors; fusing and recursively optimally estimating the angular velocity and angular acceleration data containing noise through the Kalman filter to separate measurement noise and extract optimal angular velocity estimation values containing disturbance information; Step S3: generating a velocity loop control signal of the coarse tracking system based on the optimal angular velocity estimation values; and generating a position loop control signal of the coarse tracking system based on coarse tracking miss distance information; Step S4: generating a control signal of the fine tracking system based on fine tracking miss distance information; Step S5: performing feedforward processing on the optimal angular velocity estimation values to generate a feedforward compensation signal, and synthesizing the feedforward compensation signal with the control signal of the fine tracking system to suppress medium and high frequency disturbances.
2. The method of claim 1, wherein, The Kalman filter is implemented based on the following equations: A system discretization state equation: X(k) = F × X(k-1) + W(k), wherein X(k) is a state vector at k time, F is a state transition matrix, X(k-1) is a state vector at k-1 time, and W(k) is process noise; The measurement equation is: Z(k) = H × X(k) + V(k), wherein Z(k) is an observation at k time, H is a measurement matrix, and V(k) is an inertial sensor measurement noise sequence.
3. The method of claim 2, wherein, The expression of the state transition matrix F is: wherein T is a sampling time interval of the Kalman filter.
4. The method of claim 2, wherein, The expression of the measurement matrix H is: The observation Z(k) comprises angular velocity measurement values and angular acceleration measurement values.
5. The method of claim 1, wherein, The recursive optimal estimation process of the Kalman filter comprises: Based on the last frame state estimation value, a current frame attitude is predicted, a prediction error covariance matrix is calculated, a Kalman filter gain is updated, and the system state estimation and estimation error covariance matrix are updated in combination with the inertial sensor measurement value.
6. The method of claim 1, wherein, In step S1, the inertial measurement data is collected by an inertial sensor installed on a coarse tracking camera structure; the inertial sensor is a three-axis gyroscope, a three-axis accelerometer or an inertial measurement unit integrating both.
7. The method of claim 1, wherein, In step S5, the feedforward processing is realized by a feedforward controller, and the transfer function of the feedforward controller is in the form of a high-pass filter: where Ks is a gain coefficient, ω f is the cut-off angular frequency, used to filter low-frequency disturbances and only compensate for mid-high frequency disturbances.
8. A Kalman filter compound axis control system based on inertial sensor measurements, characterized by, It comprises: an inertial measurement unit installed on a coarse tracking camera structure for collecting angular velocity and angular acceleration data of a coarse tracking visual axis; a Kalman filter module for constructing a discretization state equation and a measurement equation with angle, angular velocity and angular acceleration as state vectors, fusing the data collected by the inertial measurement unit, separating measurement noise and disturbance information, and outputting optimal angular velocity estimation values containing disturbance information; a coarse tracking control module comprising a velocity loop controller and a position loop controller, the velocity loop controller constructing a velocity feedback loop based on the optimal angular velocity estimation values, and the position loop controller realizing position control in combination with coarse tracking miss distance information to form a position-velocity double closed loop control. The fine tracking control module comprises a fine tracking controller, which is used to construct a fine tracking closed-loop control to compensate for residual error. The feedforward control module adopts a high-pass filter structure and is used to feed forward the optimal angular velocity estimation value containing disturbance information output by the Kalman filter module to an output end of the fine tracking controller to suppress medium and high frequency disturbance.
9. The system of claim 8, wherein, The actuator of the coarse tracking control module is a U-shaped two-dimensional turntable, and the actuator of the fine tracking control module is a fast mirror.
10. The system of claim 8, wherein, In the Kalman filter module, the process noise W(k) and the measurement noise V(k) are mutually independent Gaussian white noises, the covariance matrix of the process noise is Q(k), and the covariance matrix of the measurement noise is R(k).