A telescope automatic pointing correction method based on star map recognition and attitude solution

Abstract

This invention relates to the field of automatic control technology for astronomical telescopes, and particularly to an automatic telescope pointing correction method based on star chart recognition and attitude calculation. The automatic telescope pointing correction method based on star chart recognition and attitude calculation includes the following steps: S1: Establishing a stable reference optical axis offset, and acquiring the azimuth axis angular velocity, azimuth axis angular acceleration, pitch angle, current optical axis offset, and additional optical axis offset. This invention establishes a continuous evolutionary relationship between load migration state, structural response state, and gravity response capability, enabling dynamic prediction and real-time feedback optimization of the gravity response during azimuth axis movement. This expands automatic pointing from error compensation to active control during the movement process, improving the telescope's automatic pointing accuracy, repeatability consistency, and long-term operational stability, and enhancing the real-time performance and reliability of automatic pointing compensation under dynamic gravity response conditions.

Description

Technical Field

[0001] This invention relates to the field of automatic control technology for astronomical telescopes, and in particular to an automatic pointing correction method for telescopes based on star map recognition and attitude calculation. Background Technology

[0002] With the continuous development of large astronomical telescopes, survey telescopes, and high-precision observation platforms, automatic pointing has become an important component for telescopes to complete target positioning and continuous observation. Existing automatic pointing technologies typically calculate the theoretical attitude based on the target celestial coordinates, drive the azimuth and elevation axes to complete automatic pointing, and then obtain the actual optical axis offset through star chart recognition and attitude calculation, compensating for the deviation between the theoretical and actual attitudes. However, telescopes are large-scale flexible mechanical structures. During the azimuth axis movement, the internal loads of the telescope tube, support structure, bearing housing, and connecting components are continuously redistributed with the movement. The internal load migration has a continuous, phased, and recovery process. Even if the telescope eventually reaches the same azimuth and elevation angles, different azimuth axis movement speeds, acceleration / deceleration methods, and movement paths will still create different structural stress states, resulting in differences in the final gravitational deflection and optical axis offset. This load migration process is difficult to reflect directly through the final attitude error. Most existing technologies focus on establishing static compensation models around the final attitude error, mainly relying on encoder error, pitch attitude, temperature changes, and empirical fitting models to correct the formed optical axis offset. They lack analysis of the dynamic evolution of internal load migration during azimuth axis movement. However, internal load migration has a significant path dependence on subsequent gravity response. The current structural stress state is not only affected by the current motion state but also closely related to the previous load migration history. It is difficult to accurately characterize the actual stress environment of the structure and the subsequent trend of gravity deflection using only the current attitude parameters. Due to the lack of technical means to establish a continuous correlation between motion state, load migration state, structural response state, and gravity response capability, and to optimize the azimuth axis movement process in real time based on prediction results, existing automatic pointing correction methods can usually only compensate after the error is formed. It is difficult to actively suppress the development of internal load migration during movement, which limits the further improvement of the telescope's automatic pointing accuracy, repeatability consistency, and long-term operational stability. Summary of the Invention

[0003] To overcome the drawback of difficulty in predicting the gravitational response caused by load migration, this invention provides an automatic telescope pointing correction method based on star map recognition and attitude calculation.

[0004] The technical implementation scheme of the present invention is as follows: an automatic telescope pointing correction method based on star map recognition and attitude calculation, comprising the following steps: S1: Establish a stable reference optical axis offset, and obtain the azimuth axis angular velocity, azimuth axis angular acceleration, pitch angle, current optical axis offset, and additional optical axis offset; S2: Calculate the motion excitation intensity based on the azimuth axis angular velocity and the azimuth axis angular acceleration, establish a load migration state evolution model based on the motion excitation intensity and the pitch angle, and obtain the load migration state and load migration stage factor; S3: Establish a structural response state evolution model based on the motion excitation intensity, the load migration stage factor, and the pitch angle; establish a gravity response capability evolution model based on the structural response state, the additional optical axis offset, and the historical load migration influence to obtain the gravity response capability. S4: Adjust the azimuth axis movement process based on the gravity response capability feedback, and complete automatic pointing compensation based on the final optical axis offset.

[0005] Preferably, the step of establishing a stable reference optical axis offset and acquiring the azimuth axis angular velocity, azimuth axis angular acceleration, elevation angle, current optical axis offset, and additional optical axis offset includes: during the telescope calibration phase, controlling the telescope to move to multiple preset elevation angle positions according to a preset uniform motion pattern; after the azimuth axis stops moving and a preset stabilization time has elapsed, obtaining the corresponding optical axis offset through star chart recognition and attitude calculation; and repeatedly acquiring multiple sets of optical axis offset data for the same elevation angle. Calculate the stable reference optical axis offset During automatic pointing, the azimuth axis angular velocity is obtained through the azimuth axis encoder. The azimuth axis angular acceleration is calculated based on the angular velocity changes at continuous sampling times. The pitch angle is obtained through the pitch axis encoder. The current optical axis offset is obtained through star map recognition and attitude calculation. The additional optical axis offset is calculated based on the current optical axis offset and the stable reference optical axis offset. .

[0006] Preferably, the step of calculating the motion excitation intensity based on the azimuth axis angular velocity and the azimuth axis angular acceleration includes: calculating the average angular velocity based on historical normal operation samples. and mean angular acceleration The average angular velocity and average angular acceleration are used to standardize the current azimuth axis angular velocity and azimuth axis angular acceleration respectively to obtain the standardized angular velocity. and standardized angular acceleration The degree of motion continuity is calculated based on the change in normalized angular velocity between adjacent sampling periods. The first sampling order The motion excitation intensity is calculated based on the standardized angular velocity, the standardized angular acceleration, and the degree of motion continuity. .

[0007] Preferably, the step of establishing a load migration state evolution model based on the motion excitation intensity and the pitch angle to obtain the load migration state and load migration stage factors includes: calculating the gravity effect factor based on the pitch angle. Initialize load migration state The payload migration state saved at the previous sampling time will be transferred. As input at the current sampling moment; calculate the current load migration state based on the motion excitation intensity, the gravity factor, and the load migration state at the previous sampling moment. The load migration state is limited to the range of 0 to 1; a load migration stage factor is calculated based on the current load migration state. .

[0008] Preferably, the step of establishing a structural response state evolution model based on the motion excitation intensity, the load migration stage factor, and the pitch angle, and establishing a gravity response capability evolution model based on the structural response state, the additional optical axis offset, and the historical load migration influence to obtain the gravity response capability includes: initializing the structural response state. The structural response state saved at the previous sampling time. As input at the current sampling moment; a structural response state evolution model is established based on the motion excitation intensity, the load migration stage factor, and the pitch angle: ,in The gravity effect factor is calculated based on the pitch angle.

[0009] Preferably, the step of establishing a structural response state evolution model based on the motion excitation intensity, the load migration stage factor, and the pitch angle, and establishing a gravity response capability evolution model based on the structural response state, the additional optical axis offset, and the historical load migration influence, to obtain the gravity response capability, includes: using a state cache to save the most recent The load migration status within each sampling period is calculated based on the number of historical load migration states involved in the calculation. Calculate historical influence weights ,in This is the historical impact attenuation coefficient. The historical sampling time is used as the basis for calculating the historical load migration impact based on the historical impact weights and the corresponding load migration states. ,in For the first The load migration state corresponding to each sampling time.

[0010] Preferably, the step of establishing a structural response state evolution model based on the motion excitation intensity, the load migration stage factor, and the pitch angle, and establishing a gravity response capability evolution model based on the structural response state, the additional optical axis offset, and the historical load migration influence, to obtain the gravity response capability, includes: initializing the gravity response capability. The gravity response capability saved at the previous sampling time. As input at the current sampling moment; the initial gravity response capability is calculated based on the structural response state and the additional optical axis offset: ,in For initial gravitational response capability, This represents the structural response state at the current sampling time. Additional optical axis offset at the current sampling time; current gravity response capability corrected based on historical load migration influence: ,in For the corrected gravity response capability, This represents the historical load migration influence coefficient.

[0011] Preferably, the process of adjusting the azimuth axis motion based on the gravity response capability feedback includes: establishing a reference gravity response capability based on historical stable operation samples. ,in, The gravity response capability is calculated under historical stable operating conditions; the gravity response deviation is calculated based on the gravity response capability and the load migration stage factor. ; Calculate the motion adjustment coefficient based on the gravity response deviation: ,in, The motion adjustment coefficient is used to correct the current azimuth axis angular velocity, thus obtaining the azimuth axis angular velocity for the next control cycle. The azimuth axis drive mechanism is controlled to continue moving according to the corrected azimuth axis angular velocity.

[0012] Preferably, the automatic pointing compensation based on the final optical axis offset includes: re-acquiring the currently observed star map after the azimuth axis movement is completed, extracting stellar feature points through star map identification and matching them with a standard star catalog, and obtaining the final optical axis offset using attitude calculation. The attitude compensation amount is calculated based on the gravity response capability and the final optical axis offset. ,in, To stabilize the reference optical axis offset, the azimuth and pitch axes are controlled according to the attitude compensation amount to perform attitude fine-tuning, so that the actual optical axis direction of the telescope coincides with the direction of the target celestial body.

[0013] Preferably, the step of controlling the azimuth and pitch axes to perform attitude fine-tuning based on the attitude compensation amount, so that the actual optical axis direction of the telescope coincides with the direction of the target celestial body, includes: after controlling the azimuth and pitch axes to perform attitude compensation, obtaining the final optical axis offset after attitude compensation again through star map recognition and attitude calculation, as the updated optical axis offset; when the updated optical axis offset meets the preset pointing accuracy threshold, ending the automatic pointing correction; when the updated optical axis offset does not meet the preset pointing accuracy threshold, recalculating the attitude compensation amount based on the updated optical axis offset and continuing to perform attitude fine-tuning; when continuous attitude fine-tuning reaches the preset maximum number of adjustments and still does not meet the preset pointing accuracy threshold, outputting automatic pointing correction failure information and re-executing the automatic pointing process.

[0014] Beneficial Effects: This invention establishes a stable reference optical axis offset, combines star map recognition, attitude calculation, and azimuth axis motion state information to construct motion excitation intensity, achieving a unified characterization of the degree of load disturbance within the structure from the motion state. This allows for a quantitative description of the impact of azimuth axis motion on the redistribution of loads within the structure. By establishing a load migration state evolution model, the load redistribution process within the structure is transformed into a continuous state quantity with cumulative and restorative characteristics. Furthermore, a load migration stage factor is constructed, enabling continuous characterization of structural load changes under different motion stages of the telescope, thus improving the accuracy of structural state analysis. Finally, by establishing a structural response state evolution model, and combining additional optical axis offset and historical load migration influence, a gravity response capability evolution model is constructed, realizing the structural response capability and historical stress state. Co-modeling with the current optical axis change enables the gravity response capability to reflect the impact of the azimuth axis motion path on the subsequent development of gravity deflection, improving the predictive ability of the gravity response development trend. By adjusting the azimuth axis motion speed in real time according to the gravity response capability, motion optimization is introduced into the automatic pointing process. During the gravity deflection formation stage, the internal load migration speed of the structure is actively reduced, weakening the azimuth axis motion's inducing effect on structural deformation, realizing the transformation from traditional ex-post compensation to active control of the motion process. By calculating the attitude compensation amount based on the final optical axis offset and gravity response capability, and combining it with attitude fine-tuning to complete automatic pointing correction, a complete closed loop covering motion state acquisition, structural state evolution, gravity response prediction, motion feedback optimization, and attitude compensation is formed, improving the telescope's automatic pointing accuracy, repeatability consistency, and long-term operational stability. Attached Figure Description

[0015] Figure 1 This is a flowchart of an automatic telescope pointing correction method based on star map recognition and attitude calculation according to the present invention. Figure 2 This is a flowchart illustrating the real-time recursive evolution of the gravity response capability of this invention. Detailed Implementation

[0016] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0017] Example: An automatic telescope pointing correction method based on star map recognition and attitude calculation, such as... Figure 1 and Figure 2 As shown, it includes the following steps: S1: Establish a stable reference optical axis offset, and obtain the azimuth axis angular velocity, azimuth axis angular acceleration, pitch angle, current optical axis offset, and additional optical axis offset; S2: Calculate the motion excitation intensity based on the azimuth axis angular velocity and the azimuth axis angular acceleration, establish a load migration state evolution model based on the motion excitation intensity and the pitch angle, and obtain the load migration state and load migration stage factor; S3: Establish a structural response state evolution model based on the motion excitation intensity, the load migration stage factor, and the pitch angle; establish a gravity response capability evolution model based on the structural response state, the additional optical axis offset, and the historical load migration influence to obtain the gravity response capability. S4: Adjust the azimuth axis movement process based on the gravity response capability feedback, and complete automatic pointing compensation based on the final optical axis offset.

[0018] During the telescope calibration phase, the telescope is controlled to move to multiple preset elevation angle positions according to a preset uniform motion pattern. After the azimuth axis stops moving and a preset stabilization time has elapsed, the corresponding optical axis offset is obtained through star chart recognition and attitude calculation. Multiple sets of optical axis offset data are repeatedly collected for the same elevation angle. Calculate the stable reference optical axis offset During automatic pointing, the azimuth axis angular velocity is obtained through the azimuth axis encoder. The azimuth axis angular acceleration is calculated based on the angular velocity changes at continuous sampling times. The pitch angle is obtained through the pitch axis encoder. The current optical axis offset is obtained through star map recognition and attitude calculation. The additional optical axis offset is calculated based on the current optical axis offset and the stable reference optical axis offset. .

[0019] It should be explained that the stable reference optical axis offset is used to characterize the basic optical axis offset state of the telescope after the load inside the structure has been fully stabilized, and serves as a reference benchmark for subsequent calculation of the additional optical axis offset. After the telescope enters the calibration stage, it is first controlled to move sequentially to multiple preset pitch angle positions according to the preset unified motion mode. The preset unified motion mode uses multiple motion trajectories in the historical normal calibration data to calculate the azimuth axis angular velocity and azimuth axis angular acceleration change law corresponding to each trajectory, and generates a unified calibration motion trajectory based on the average angular velocity curve and the average angular acceleration curve, so that the load migration process corresponding to each pitch angle position is consistent. The historical normal calibration data comes from calibration records in the telescope's historical calibration process that have successfully completed star map recognition, attitude calculation passed consistency verification, and the calibration results meet the accuracy requirements, excluding calibration failures and those with abnormal image quality. A historical normal calibration dataset is established after collecting data on abnormal equipment operation. The preset elevation angle position is divided into multiple calibration elevation angles according to the actual working elevation range of the telescope. For each calibration elevation angle, a corresponding stable reference optical axis offset is established. During automatic pointing, when the current elevation angle is between two adjacent calibration elevation angles, the stable reference optical axis offset corresponding to the two adjacent calibration elevation angles is used as the interpolation endpoint, and linear interpolation is used to calculate the stable reference optical axis offset corresponding to the current elevation angle. After the telescope reaches the corresponding elevation angle, it stops azimuth axis movement and acquires star images after the preset stabilization time. The preset stabilization time is obtained through multiple calibration experiments. In each experiment, the optical axis offset sequence is continuously acquired. When the rate of change of optical axis offset within a continuous sampling period meets the stability criterion, the corresponding waiting time is recorded. The preset stabilization time is obtained by averaging all waiting times.A telescope imaging system acquires images of the night sky, extracts stellar feature points from the star map, and matches them with a standard star catalog. If a match is successful, the attitude calculation is used to calculate the deviation between the actual and theoretical optical axis directions, resulting in the optical axis offset. If the matching fails or the attitude calculation result fails the consistency check, the night sky image is reacquired. The consistency check involves reprojecting the optical axis direction obtained from the attitude calculation onto the current star map and calculating the average reprojection error between the projected stellar positions and the actual identified stellar positions. If the average reprojection error is not greater than a preset reprojection error threshold, the consistency check is considered successful; otherwise, the consistency check is considered unsuccessful. The preset reprojection... The projection error threshold is determined based on historical normal calibration data. Multiple sets of star map matching data with successful star map recognition and correct attitude calculation results are extracted during the historical normal calibration process. The reprojection error after the star map matching is completed for each set of star map matching. The distribution of all reprojection errors is statistically analyzed. All the statistically obtained reprojection errors are arranged in ascending order. Abnormal reprojection errors caused by abnormal image quality, mismatched star points, and abnormal attitude calculation during the calibration process are removed. The maximum value of the remaining reprojection errors is determined as the preset reprojection error threshold, which is used to determine whether the current attitude calculation result meets the accuracy requirements. Multiple sets of optical axis offset data are repeatedly collected for the same pitch angle. According to the formula Calculate the stable reference optical axis offset, where, This indicates the number of stable optical axis offset data sets included in the statistics. Indicates the first Group stabilized optical axis offset, in the formula This indicates the total number of data points included in the statistics. This represents the sum of all stable optical axis offset data; dividing the two is used to reduce the impact of random measurement errors on the reference base. During automatic pointing, the azimuth axis angular position is measured in real time by the azimuth axis encoder, and the azimuth axis angular velocity is obtained by combining the angular position changes at adjacent sampling times. According to the formula Calculate the azimuth axis angular acceleration, where This represents the azimuth axis angular velocity at the previous sampling time. This represents the sampling time interval between adjacent sampling moments; the pitch angle is obtained through the pitch axis encoder. The star map recognition and attitude calculation are performed again to obtain the current optical axis offset. Finally, based on the formula Calculate the additional optical axis offset, where This represents the additional offset caused by the current optical axis offset relative to the stable reference optical axis offset. To reduce the occurrence of calculation anomalies due to an excessively small denominator when the stable reference optical axis offset is close to zero. This represents a pre-set, extremely small positive number, the value of which is determined based on the numerical calculation accuracy of the system, and is used to ensure the numerical stability of the calculation process; the additional optical axis offset is used to characterize the additional optical axis change generated during the azimuth axis movement process relative to the stable force state.

[0020] The average angular velocity was calculated based on historical normal operating samples. and mean angular acceleration The average angular velocity and average angular acceleration are used to standardize the current azimuth axis angular velocity and azimuth axis angular acceleration respectively to obtain the standardized angular velocity. and standardized angular acceleration The degree of motion continuity is calculated based on the change in normalized angular velocity between adjacent sampling periods. The first sampling order The motion excitation intensity is calculated based on the standardized angular velocity, the standardized angular acceleration, and the degree of motion continuity. .

[0021] It should be explained that the average angular velocity... and the mean angular acceleration A reference level is used to characterize the azimuth axis motion state during normal automatic pointing of the telescope, providing a unified benchmark for the standardization of current motion parameters; based on historical normal operation samples, positive and negative azimuth axis motion samples are statistically analyzed, and the average angular velocity in the corresponding directions is calculated. The average angular velocity in the corresponding direction is selected as the standardization benchmark based on the current azimuth axis motion direction. Historical normal operation samples are derived from azimuth axis encoder sampling data recorded during continuous normal observations of the telescope. To establish a historical normal operation sample library, data with drive failures, attitude correction failures, encoder malfunctions, and communication interruptions are first removed. A sample set is then constructed based on data from successfully completing automatic pointing tasks, and the data is processed according to the formula... and Calculate the average angular velocity and the average angular acceleration, wherein, This indicates the number of historically normal operating samples included in the statistics. Indicates the first The azimuth axis angular velocity corresponding to the historical normal operation sample. Indicates the first The azimuth angular acceleration corresponding to a set of historical normal operation samples is averaged to reduce the impact of random motion fluctuations on the reference benchmark. The current azimuth angular velocity and azimuth angular acceleration are standardized using the average angular velocity and average angular acceleration, respectively, with the calculation formulas as follows: and ,in, Represents the standardized angular velocity. Represents the normalized angular acceleration. The smallest positive number is determined based on the accuracy of the system's numerical calculations; the degree of motion continuity is calculated based on the change in standardized angular velocity between adjacent sampling periods. ,in, This represents the normalized angular velocity variation between adjacent sampling periods, reflecting the degree of change in the current motion state. Used to normalize the amplitude of angular velocity change, exponential function This describes the characteristic that motion continuity gradually decreases as angular velocity increases. When the motion state changes little between adjacent sampling periods, the degree of motion continuity remains high. During the first sampling, since the standardized angular velocity from the previous sampling time is not available, let... According to the formula Calculate the intensity of the motion excitation, where, It represents the impact of the current motion amplitude on the structural load input capacity, and uses a squared form to enhance the characterization ability of larger motion amplitudes on load migration; This indicates the degree of motion disturbance caused by angular acceleration. The hyperbolic tangent function is used to describe the change law of the disturbance effect gradually approaching saturation, so that the intensity of motion excitation conforms to the mechanical characteristics of the gradual stabilization of the force change of the telescope's flexible structure. This indicates the moderating effect of motion continuity on the load input process, and through product coupling, it reflects the comprehensive degree of load disturbance input into the telescope structure by azimuth axis motion.

[0022] Calculate the gravity factor based on the pitch angle. Initialize load migration state The payload migration state saved at the previous sampling time will be transferred. As input at the current sampling moment; calculate the current load migration state based on the motion excitation intensity, the gravity factor, and the load migration state at the previous sampling moment. The load migration state is limited to the range of 0 to 1; a load migration stage factor is calculated based on the current load migration state. .

[0023] It should be explained that the gravity effect factor characterizes the degree of influence of the gravity direction on the stress state of the telescope structure under the current pitch attitude, and is determined according to the pitch angle using the formula... Calculations are performed, in which, This indicates the pitch angle at the current sampling time. This represents the projection relationship of the gravity direction corresponding to pitch attitude changes, through... The process maps the gravity factor to a non-negative range, making it a parameter for adjusting the subsequent load migration state evolution model. The load migration state describes the degree of load redistribution within the telescope structure and is initialized when automatic pointing begins. This indicates that the structure is in a stable load state. After each load transition state calculation is completed, the load transition state obtained at the current sampling time is saved to the state cache. The state cache is a pre-allocated circular storage area in the controller memory, used to continuously save the state data corresponding to each sampling time in the order of sampling time. At the beginning of the next sampling period, the load transition state saved at the previous sampling time is read. As input at the current sampling moment, the load migration state maintains continuous evolution characteristics across consecutive sampling periods; the state data includes load migration state, structural response state, and gravity response capability; according to the formula... Calculate the load migration state at the current sampling time, where, This indicates the inheritance effect of the load migration state formed at the previous sampling time on the current state; Indicates the current intensity of exercise stimulation. Factors acting with gravity They jointly induce a new load redistribution, in which The hyperbolic tangent function represents the combined load disturbance caused by the motion input and the current gravitational force. This describes the phenomenon where the rate of load migration growth gradually saturates as the combined load disturbance increases. This reflects that the proportion of remaining transferable load gradually decreases as the degree of load migration increases; This represents the process by which the internal loads of the structure gradually return to stability. The load recovery coefficient represents the amount of load migration state recovered from each adjacent sampling period in the historical normal operation sample. As the dependent variable, the gravity recovery term corresponds to the sampling period. A linear fitting model is established using the load recovery coefficient as the independent variable, and the least squares method is used to solve for the fitting coefficient. , This indicates that the recovery rate changes continuously under increased gravity. The exponential decay function describes the stress release and load rebalancing process of the structural connectors, which conforms to the stress recovery law of the telescope's flexible structure. The load migration state is limited to the range of 0 to 1; according to the formula Calculate the load migration stage factor, where Indicates the current degree of load migration. This indicates that as the load migration is completed, the migration activity gradually decreases. The product coupling is used to reflect the continuous change process of the initial stage, the rapid migration stage and the stable stage of load migration.

[0024] Initialize the structure response state The structural response state saved at the previous sampling time. As input at the current sampling moment; a structural response state evolution model is established based on the motion excitation intensity, the load migration stage factor, and the pitch angle: ,in The gravity effect factor is calculated based on the pitch angle.

[0025] It should be explained that the structural response state characterizes the sensitivity of the telescope structure to additional deformation caused by gravity under the current stress environment; the structural response state is initialized when automatic pointing begins. This indicates that the telescope structure has not yet been induced by the current motion process to form a significant additional structural response; after each calculation of the structural response state, the structural response state obtained at the current sampling time is saved to the state cache, and the structural response state saved at the previous sampling time is read at the beginning of the next sampling period. As input at the current sampling moment, the structural response state continues to evolve across consecutive sampling periods, reflecting the cumulative and restorative nature of the force changes on the telescope's flexible structure; according to the formula Calculate the structural response state at the current sampling time, where This indicates the inheritance effect of the structural response state formed at the previous sampling time on the current structural stress state; This indicates that the current structure still has the remaining ability to generate a structural response, reducing the possibility that the structural response state will increase infinitely with continuous calculations; This indicates the current load transition stage of the structure. The hyperbolic tangent function represents the nonlinear driving effect of the intensity of motion excitation on the growth of the structural response. It describes how the rate of increase in structural response gradually saturates as the motion excitation intensifies, consistent with the mechanical characteristic of a flexible telescope structure undergoing gradual deformation stabilization under sustained load. Together, they represent the increasing effect of the current motion excitation on the structural response state during the active phase of load transfer; This represents the recovery process of the structural response, where, This indicates the degree to which the structure enters the recovery phase after the load transfer is gradually completed. This represents the moderating effect of gravity on the structural recovery rate. An exponential decay function is used to describe the continuous recovery process of the telescope tube, support structure, and connecting components under gravity, characterized by gradual stress release and rebalancing. After calculation, according to... The structural response state is limited to the range of 0 to 1.

[0026] Use state caching to save the most recent The load migration status within each sampling period is calculated based on the number of historical load migration states involved in the calculation. Calculate historical influence weights ,in This is the historical impact attenuation coefficient. The historical sampling time is used as the basis for calculating the historical load migration impact based on the historical impact weights and the corresponding load migration states. ,in For the first The load migration state corresponding to each sampling time.

[0027] It should be explained that the historical load migration influence represents the continuous impact of recent load migration processes on the current structural stress state, enabling the calculation of the gravity response capability to reflect the path dependence and historical memory characteristics of load migration in the telescope's flexible structure; during operation, a state cache is used to save the most recent... The load migration status within each sampling period is updated using a first-in, first-out (FIFO) update method. When the cache reaches a preset capacity, the load migration status corresponding to the latest sampling time overwrites the oldest saved load migration status. The cache length is... The buffer length is determined by the number of consecutive sampling periods during which load migration can continuously affect the structural stress state during a single complete automatic pointing process of the telescope. This is achieved by averaging the sampling periods taken for the load migration state to recover to a stable state in a historical normal operation sample database. When the number of load migration states saved before the current sampling time is insufficient At any time, according to the formula Determine the number of historical load transition states involved in the calculation, among which Indicates the current sampling time sequence number. This indicates the number of historical load migration states actually involved in the calculation; based on the formula... Calculate the historical influence weights corresponding to each historical sampling time, where, Indicates the sequence number of the historical sampling time. The historical influence attenuation coefficient represents the time interval between the historical sampling time and the current sampling time. Based on the historical normal operation sample database, historical load migration state sequences, structural response state sequences, and final gravity response capabilities corresponding to multiple automatic pointing processes were extracted in chronological order; different candidate historical influence attenuation coefficients were set for each. For each candidate historical influence decay coefficient, calculate the corresponding historical influence weight. and historical load migration impact Substituting the values ​​into the gravity response capability evolution model yields the corresponding predicted gravity response capability value. The gravity response capability corresponding to the historical normal operation sample is the calculation result obtained by collecting data at the corresponding sampling time during the historical automatic pointing process, according to the gravity response capability calculation method. The sum of squared prediction errors between the predicted gravity response capability value at each sampling time and the gravity response capability corresponding to each historical normal operation sample is calculated. Taking the minimization of the sum of squared prediction errors as the optimization objective, the corresponding historical influence attenuation coefficient is obtained by solving the least squares method. The historical normal operation samples are from the historical normal operation sample library; the exponential decay function describes the gradual weakening of the influence of historical load migration states on the current structural stress environment over time, which conforms to the law of gradual stress release and load rebalancing within the telescope's flexible structure; according to the formula Calculate the historical load migration impact, where This indicates that the load migration status at each historical sampling time is weighted and accumulated according to the corresponding historical influence weight, comprehensively reflecting the contribution of load migration at different historical stages to the current structural stress state. This represents the sum of all historical influence weights, used to complete the weight normalization process.

[0028] Initialize gravity response capability The gravity response capability saved at the previous sampling time. As input at the current sampling moment; the initial gravity response capability is calculated based on the structural response state and the additional optical axis offset: ,in For initial gravitational response capability, This represents the structural response state at the current sampling time. Additional optical axis offset at the current sampling time; current gravity response capability corrected based on historical load migration influence: ,in For the corrected gravity response capability, This represents the historical load migration influence coefficient.

[0029] It should be explained that the gravity response capability characterizes the trend of the optical axis shift induced by gravity under the current telescope structural state; the gravity response capability is initialized when automatic pointing begins. This indicates that the current structure has not yet formed a gravity response trend caused by this motion process; after each gravity response capability calculation is completed, the gravity response capability obtained at the current sampling time is saved to the state cache, and the gravity response capability saved at the previous sampling time is read at the beginning of the next sampling period. As input at the current sampling moment, the gravity response capability can inherit the state of the previous sampling period and evolve continuously; according to the formula Calculate the initial gravitational response capability, where, This indicates the inheritance effect of existing gravitational response capabilities on the current state; This indicates that the current structure has the remaining capacity to generate a further gravitational response; Indicates the current structural response state. Indicates the current additional optical axis offset. It represents the combined response driving force formed by the structural response capability and the actual additional optical axis offset. The gravitational response capability is enhanced only when the structure is sensitive to gravity and has already generated an additional optical axis offset. The product form is used to reflect the coupling effect between the two. This indicates that as the structure gradually stabilizes, the existing gravitational response gradually diminishes. The exponential decay function describes the continuous change in stress release and gravitational response within the telescope's flexible structure; subsequently, according to the formula... The current gravity response capability is corrected using the historical load migration influence, wherein, The historical load migration influence coefficient represents the amount of historical load migration impact. Based on historical normal operation samples, the historical load migration impact during the historical automatic pointing process is extracted. Structural response state Based on the final actual optical axis offset and the stable reference optical axis offset, the actual gravity response capability corresponding to each historical normal operation sample is calculated; and different candidate historical load migration influence coefficients are set respectively. ,Will Historical load migration impact and structural response state Substituting the values ​​into the gravity response capability correction formula, the predicted gravity response capability values ​​for each sampling time are obtained. The sum of squared prediction errors between the predicted gravity response capability values ​​at each sampling time and the actual gravity response capability of historical samples is calculated. Using the minimum sum of squared prediction errors as the optimization objective, the historical load migration influence coefficient is obtained by solving the least squares method. ; This represents the historical correction amount formed jointly by the historical load migration process and the current structural response state, reflecting the path-dependent nature of the telescope structure's stress state. This is used to limit the growth rate of the modified gravity response capability, ensuring that the growth rate gradually saturates as historical influences increase, consistent with the gradual stabilization of the gravity response of the telescope's flexible structure; after modification, according to... The gravity response capability is limited to the range of 0 to 1.

[0030] Establish a reference gravity response capability based on historical stable operating samples: ,in, The gravity response capability is calculated under historical stable operating conditions; the gravity response deviation is calculated based on the gravity response capability and the load migration stage factor. ; Calculate the motion adjustment coefficient based on the gravity response deviation: ,in, The motion adjustment coefficient is used to correct the current azimuth axis angular velocity, thus obtaining the azimuth axis angular velocity for the next control cycle. The azimuth axis drive mechanism is controlled to continue moving according to the corrected azimuth axis angular velocity.

[0031] It should be explained that the reference gravity response capability is used to characterize the baseline gravity response level of the telescope under historical stable operating conditions, providing a unified reference for judging the additional structural gravity response caused by the current motion process; the historical stable operating samples are derived from data in the historical normal operation samples that successfully pointed automatically and whose final attitude met the pointing accuracy requirements. Before establishing the reference dataset, data with drive anomalies, attitude calculation failures, star map matching failures, and communication anomalies were removed. Then, the gravity response capability corresponding to each historical stable operating sample was extracted, and the result was calculated according to the formula. Calculate the reference gravity response capability, wherein, This indicates the number of historically stable samples included in the statistics. Indicates the first Gravity response capability corresponding to a set of historically stable operating samples. This represents the sum of all historical gravity response capabilities, and its averaging reduces the impact of random fluctuations from different observation missions on the reference benchmark; based on the formula... Calculate the gravity response deviation, where This indicates the degree of deviation of the current gravitational response capability from the stable reference state, reflecting the strengthening or weakening trend of the structure's gravitational response during the current motion process. This represents the current load migration stage factor, used to adjust the impact of different load migration stages on motion feedback; based on the formula... Calculate the motion adjustment coefficient, wherein the motion adjustment coefficient Based on historical normal operation samples, the gravity response deviation and the actual optical axis offset after final attitude compensation corresponding to each historical automatic pointing process were extracted, and different candidate motion adjustment coefficients were set respectively. Substituting each candidate motion adjustment coefficient into the motion adjustment coefficient calculation formula yields the corresponding azimuth axis angular velocity adjustment result. An automatic pointing process simulation is then completed. The pointing error is calculated based on the final optical axis offset obtained from the simulation. Using the minimum final optical axis offset error as the optimization objective, the motion adjustment coefficient is obtained by solving the least squares method. The automatic pointing process simulation is based on the telescope kinematic model and the gravity response capability evolution model; wherein, This describes the nonlinear change in motion adjustment amplitude that gradually levels off as the gravity response deviation increases; based on the formula... Calculate the azimuth axis angular velocity for the next control cycle, where This indicates the azimuth axis angular velocity during the current control cycle. This indicates the velocity correction ratio corresponding to the current motion state; after completing the azimuth axis angular velocity update, The control command is sent to the azimuth axis drive mechanism as the control command for the next control cycle. The updated azimuth axis angular velocity is re-acquired and used in the calculation of the next sampling cycle, forming a closed-loop control process of gravity response capability prediction and motion feedback adjustment.

[0032] After the azimuth axis motion is completed, the current observation star map is acquired again. Star feature points are extracted by star map identification and matched with a standard star catalog. The final optical axis offset is obtained using attitude calculation. The attitude compensation amount is calculated based on the gravity response capability and the final optical axis offset. ,in, To stabilize the reference optical axis offset, the azimuth and pitch axes are controlled according to the attitude compensation amount to perform attitude fine-tuning, so that the actual optical axis direction of the telescope coincides with the direction of the target celestial body.

[0033] It should be explained that after completing the azimuth axis adjustment, the telescope enters the final attitude correction stage. This involves acquiring the current star map, identifying and extracting stellar feature points from the star map, and matching the stellar position distribution with a standard star catalog. Once a match is successful, the attitude calculation is performed to obtain the deviation of the telescope's actual optical axis direction from its theoretical optical axis direction, which is then used as the final optical axis offset. When star map matching fails, the number of stellar feature points participating in the matching is insufficient, or the attitude calculation result fails the consistency check, the current observed star map is re-acquired and star map recognition and attitude calculation are performed again; the final optical axis offset Offset from the stable reference optical axis Using the same angular representation, both have the same dimensions and can be used for difference calculation; using the formula Calculate the attitude compensation amount, where, This indicates the remaining attitude error caused by the current actual optical axis offset relative to the basic optical axis under stable force conditions, reflecting the amount of optical axis offset that still needs to be corrected after this automatic pointing ends; This represents the corrected gravity response capability corresponding to the current sampling period, characterizing the trend of gravity continuing to induce optical axis offset under the current structural state. The remaining attitude error is coupled with the structural gravity response trend using a product form, so that the higher the structural gravity response capability, the more sufficient the compensation amount. As the structure gradually stabilizes, the compensation amplitude decreases accordingly, ensuring that the attitude compensation process is consistent with the actual stress state of the telescope. After the attitude compensation amount is calculated, the controller decomposes the attitude compensation amount into azimuth axis compensation components and pitch axis compensation components based on the telescope kinematic model, and then converts them into control commands for the corresponding drive mechanisms. The azimuth axis drive mechanism and the pitch axis drive mechanism are controlled to perform attitude fine-tuning synchronously, so that the two-axis compensation process is coordinated and the increase in attitude deviation of the other axis caused by single-axis compensation is reduced.

[0034] After attitude compensation is performed on the azimuth and pitch axes, the final optical axis offset after attitude compensation is obtained again through star map recognition and attitude calculation, which is used as the updated optical axis offset. When the updated optical axis offset meets the preset pointing accuracy threshold, the automatic pointing correction ends. When the updated optical axis offset does not meet the preset pointing accuracy threshold, the attitude compensation amount is recalculated based on the updated optical axis offset and attitude fine-tuning continues. When the continuous attitude fine-tuning reaches the preset maximum number of adjustments and still does not meet the preset pointing accuracy threshold, an automatic pointing correction failure message is output and the automatic pointing process is re-executed.

[0035] It should be explained that after completing the attitude compensation, the current observed star map is re-acquired, and the star map recognition and attitude calculation are performed again to obtain the final optical axis offset after attitude compensation. This final optical axis offset is used as the updated optical axis offset, and is used as the verification basis for the attitude compensation effect. If the star map matching fails, the attitude calculation is abnormal, or the updated optical axis offset fails the consistency check, the current observed star map is re-acquired, and the star map recognition and attitude calculation are performed again until a valid updated optical axis offset is obtained. The updated optical axis offset, i.e., the final optical axis offset after attitude compensation, is compared with the preset pointing accuracy threshold. When the updated optical axis offset is not greater than the preset pointing accuracy threshold, it is determined that the actual optical axis direction of the telescope has met the automatic pointing accuracy requirements, and the automatic pointing correction ends. The preset pointing accuracy threshold is determined based on historical stable observation data. The final optical axis offset corresponding to multiple successful automatic pointing and observation images meeting imaging quality requirements is statistically analyzed. The mean and standard deviation of the final optical axis offset are used to construct the allowable offset range, and the upper limit of the offset that meets the preset confidence level is determined as the preset pointing accuracy threshold. The historical stable observation data comes from the historical stable operation sample, which is the automatic pointing data in the historical normal operation sample. Data indicating successful pointing and observation of images meeting imaging quality requirements; the preset confidence level is determined based on historical stable observation data, statistically analyzing multiple historical observations of successful automatic pointing and meeting imaging quality requirements, calculating the corresponding final optical axis offset coverage ratio according to different candidate confidence levels, and using the candidate confidence level corresponding to the coverage ratio stabilizing as the preset confidence level; when the updated optical axis offset is still greater than the preset pointing accuracy threshold, the updated optical axis offset replaces the previous round of compensation results, and a new attitude compensation amount is recalculated according to the attitude compensation amount calculation method, controlling the azimuth and pitch axes to continue performing attitude fine-tuning, reducing attitude error. Gradual convergence; after each attitude fine-tuning, the continuous attitude fine-tuning count is incremented and compared with the preset maximum number of adjustments, which is determined based on historical normal operation samples. The number of attitude fine-tunings required to reach the preset pointing accuracy during successful automatic pointing in the past is counted, and the maximum value of the corresponding statistical result is determined as the preset maximum number of adjustments; when the continuous attitude fine-tuning reaches the preset maximum number of adjustments, if the updated optical axis offset still does not meet the preset pointing accuracy threshold, an automatic pointing correction failure message is output, the state data saved during the current automatic pointing process is cleared, and the automatic pointing process is re-executed.

[0036] The above embodiments are provided for those skilled in the art to implement or use the present invention. Those skilled in the art can make various modifications or changes to the above embodiments without departing from the inventive concept of the present invention. Therefore, the protection scope of the present invention is not limited to the above embodiments, but should be the maximum scope that conforms to the innovative features mentioned in the claims.

Claims

1. A method for automatic telescope pointing correction based on star map recognition and attitude calculation, characterized in that, Includes the following steps: S1: Establish a stable reference optical axis offset, and obtain the azimuth axis angular velocity, azimuth axis angular acceleration, pitch angle, current optical axis offset, and additional optical axis offset; S2: Calculate the motion excitation intensity based on the azimuth axis angular velocity and the azimuth axis angular acceleration, establish a load migration state evolution model based on the motion excitation intensity and the pitch angle, and obtain the load migration state and load migration stage factor; S3: Establish a structural response state evolution model based on the motion excitation intensity, the load migration stage factor, and the pitch angle; establish a gravity response capability evolution model based on the structural response state, the additional optical axis offset, and the historical load migration influence to obtain the gravity response capability. S4: Adjust the azimuth axis movement process based on the gravity response capability feedback, and complete automatic pointing compensation based on the final optical axis offset.

2. The telescope automatic pointing correction method based on star map recognition and attitude calculation according to claim 1, characterized in that, The process of establishing a stable reference optical axis offset and acquiring azimuth axis angular velocity, azimuth axis angular acceleration, elevation angle, current optical axis offset, and additional optical axis offset includes: during the telescope calibration phase, controlling the telescope to move to multiple preset elevation angle positions according to a preset uniform motion pattern; after the azimuth axis stops moving and a preset stabilization time has elapsed, obtaining the corresponding optical axis offset through star chart recognition and attitude calculation; and repeatedly acquiring multiple sets of optical axis offset data for the same elevation angle. Calculate the stable reference optical axis offset During automatic pointing, the azimuth axis angular velocity is obtained through the azimuth axis encoder. The azimuth axis angular acceleration is calculated based on the angular velocity changes at continuous sampling times. The pitch angle is obtained through the pitch axis encoder. The current optical axis offset is obtained through star map recognition and attitude calculation. The additional optical axis offset is calculated based on the current optical axis offset and the stable reference optical axis offset. .

3. The method for automatic telescope pointing correction based on star map recognition and attitude calculation according to claim 1, characterized in that, The calculation of motion excitation intensity based on the azimuth axis angular velocity and the azimuth axis angular acceleration includes: calculating the average angular velocity based on historical normal operation samples. and mean angular acceleration The average angular velocity and average angular acceleration are used to standardize the current azimuth axis angular velocity and azimuth axis angular acceleration respectively to obtain the standardized angular velocity. and standardized angular acceleration The degree of motion continuity is calculated based on the change in normalized angular velocity between adjacent sampling periods. The first sampling order The motion excitation intensity is calculated based on the standardized angular velocity, the standardized angular acceleration, and the degree of motion continuity. .

4. The method for automatic telescope pointing correction based on star map recognition and attitude calculation according to claim 1, characterized in that, The step of establishing a load migration state evolution model based on the motion excitation intensity and the pitch angle to obtain the load migration state and load migration stage factors includes: calculating the gravity effect factor based on the pitch angle. Initialize load migration state The payload migration state saved at the previous sampling time will be transferred. As input at the current sampling moment; calculate the current load migration state based on the motion excitation intensity, the gravity factor, and the load migration state at the previous sampling moment. The load migration state is limited to the range of 0 to 1; a load migration stage factor is calculated based on the current load migration state. .

5. The method for automatic telescope pointing correction based on star map recognition and attitude calculation according to claim 1, characterized in that, The process of establishing a structural response state evolution model based on the motion excitation intensity, the load migration stage factor, and the pitch angle, and establishing a gravity response capability evolution model based on the structural response state, the additional optical axis offset, and the historical load migration influence, to obtain the gravity response capability, includes: initializing the structural response state. The structural response state saved at the previous sampling time. As input at the current sampling moment; a structural response state evolution model is established based on the motion excitation intensity, the load migration stage factor, and the pitch angle: ,in The gravity effect factor is calculated based on the pitch angle.

6. The method for automatic telescope pointing correction based on star map recognition and attitude calculation according to claim 1, characterized in that, The process of establishing a structural response state evolution model based on the motion excitation intensity, the load migration stage factor, and the pitch angle, and establishing a gravity response capability evolution model based on the structural response state, the additional optical axis offset, and the historical load migration influence, to obtain the gravity response capability, includes: using a state cache to store the most recent... The load migration status within each sampling period is calculated based on the number of historical load migration states involved in the calculation. Calculate historical influence weights ,in Here, represents the historical impact attenuation coefficient, and represents the historical sampling time. The historical load migration impact is calculated based on the historical impact weight and the corresponding load migration state. ,in For the first The load migration state corresponding to each sampling time.

7. The method for automatic telescope pointing correction based on star map recognition and attitude calculation according to claim 1, characterized in that, The process involves establishing a structural response state evolution model based on the motion excitation intensity, the load migration stage factor, and the pitch angle; and establishing a gravity response capability evolution model based on the structural response state, the additional optical axis offset, and the historical load migration influence to obtain the gravity response capability, including: initializing the gravity response capability. The gravity response capability saved at the previous sampling time. As input at the current sampling moment; the initial gravity response capability is calculated based on the structural response state and the additional optical axis offset: ,in For initial gravitational response capability, This represents the structural response state at the current sampling time. Additional optical axis offset at the current sampling time; current gravity response capability corrected based on historical load migration influence: ,in For the corrected gravity response capability, This represents the historical load migration influence coefficient.

8. The method for automatic telescope pointing correction based on star map recognition and attitude calculation according to claim 1, characterized in that, The process of adjusting the azimuth axis motion based on the gravity response capability feedback includes: establishing a reference gravity response capability based on historical stable operation samples. ,in, The gravity response capability is calculated under historical stable operating conditions; the gravity response deviation is calculated based on the gravity response capability and the load migration stage factor. ; Calculate the motion adjustment coefficient based on the gravity response deviation: ,in, The motion adjustment coefficient is used to correct the current azimuth axis angular velocity, thus obtaining the azimuth axis angular velocity for the next control cycle. The azimuth axis drive mechanism is controlled to continue moving according to the corrected azimuth axis angular velocity.

9. The method for automatic telescope pointing correction based on star map recognition and attitude calculation according to claim 1, characterized in that, The automatic pointing compensation based on the final optical axis offset includes: re-acquiring the current observation star map after the azimuth axis movement is completed, extracting stellar feature points through star map identification and matching them with a standard star catalog, and obtaining the final optical axis offset using attitude calculation. The attitude compensation amount is calculated based on the gravity response capability and the final optical axis offset. ,in, To stabilize the reference optical axis offset, the azimuth and pitch axes are controlled according to the attitude compensation amount to perform attitude fine-tuning, so that the actual optical axis direction of the telescope coincides with the direction of the target celestial body.

10. The method for automatic telescope pointing correction based on star map recognition and attitude calculation according to claim 9, characterized in that, The step of controlling the azimuth and pitch axes to perform attitude fine-tuning based on the attitude compensation amount, so that the actual optical axis direction of the telescope coincides with the direction of the target celestial body, includes: after controlling the azimuth and pitch axes to perform attitude compensation, obtaining the final optical axis offset after attitude compensation again through star map recognition and attitude calculation, as the updated optical axis offset; when the updated optical axis offset meets the preset pointing accuracy threshold, the automatic pointing correction ends; when the updated optical axis offset does not meet the preset pointing accuracy threshold, the attitude compensation amount is recalculated based on the updated optical axis offset and attitude fine-tuning continues; when continuous attitude fine-tuning reaches the preset maximum number of adjustments and still does not meet the preset pointing accuracy threshold, an automatic pointing correction failure message is output and the automatic pointing process is re-executed.

Images