A generalized data detection and reduction method, system, and apparatus for improving the accuracy of space object positioning

Abstract

The present application relates to spaceflight measurement and control technical field, especially relate to a kind of generalized data detection and simplification method, system and equipment for improving the precision of space target positioning, comprising: obtaining the observation data of space target, construct extended Gaussian-Markov model containing potential modeling error, generate multiple hypothesis set by different combination of modeling error parameter;Utilize likelihood ratio test criterion to construct the test statistic between different hypotheses, according to the observation environment requirement self-adaptive selection executes generalized data detection process or generalized data simplification process;Determine the optimal model hypothesis of current observation by multiple hypothesis testing, based on the optimal model hypothesis parameter adaptive estimation is carried out, and the position parameter of space target is solved.This application guarantees high-precision requirement by forward selection strategy, enhances robustness under complex environment by backward elimination strategy, realizes the effective identification and elimination of multiple modeling error, significantly improves the precision and reliability of space target positioning.

Description

Technical Field

[0001] This invention relates to the field of aerospace telemetry and control technology, and in particular to a generalized data detection and simplification method, system and equipment for improving the positioning accuracy of space targets. Background Technology

[0002] In space target surveillance and measurement missions (such as precise orbit determination of low-Earth orbit satellites and space debris cataloging), massive amounts of observation data are typically acquired using ground-based radar, optical telescopes, or spaceborne GNSS receivers. Least square adjustment is then performed using a Gauss-Markov model to calculate the target's position, velocity, and other state parameters. However, least squares estimation is only the optimal unbiased estimate when the observation error follows a zero-mean Gaussian distribution.

[0003] In actual space target observation, modeling errors inevitably become mixed into the observation data due to electromagnetic interference, sensor malfunctions, or the influence of complex space environments. These errors are mainly divided into two categories: one is sudden gross errors, usually manifested as outliers; the other is systematic errors that are not modeled, such as atmospheric propagation delay residues, gravitational perturbation biases, or multipath effects. If these errors are ignored directly, the results of least squares estimation will be severely biased, and may even lead to orbit divergence.

[0004] To address these issues, Barda proposed the classic Data Probing (DS) method, using the w-test statistic to identify individual outliers. For cases with multiple outliers, iterative data probing (IDS) was subsequently developed, purifying the data by systematically removing outliers. However, the IDS method is susceptible to masking and flooding effects among multiple outliers, leading to missed detections or false positives. Furthermore, its inherent assumption that all data is initially normal is ineffective when outliers are dense. Therefore, researchers have proposed the Data Reduction (DR) method, which uses reverse logic to treat initial data as outliers and gradually recover normal data, thereby enhancing robustness.

[0005] Although existing DS and DR methods have achieved some success in handling gross errors, they still have significant limitations when facing complex space target observation tasks: First, existing methods are mainly designed for gross errors and are difficult to effectively detect and distinguish systematic biases common in space measurements; second, both gross errors and systematic errors can be regarded as model biases mathematically, but there is a lack of a generalized statistical testing framework to handle different types and quantities of errors simultaneously; finally, space target tasks have diverse requirements, sometimes requiring extremely high orbit determination accuracy, and sometimes emphasizing robustness in harsh environments, and existing single strategies are difficult to meet these two very different needs. Summary of the Invention

[0006] The purpose of this invention is to overcome the shortcomings of existing technologies and provide a generalized data detection and simplification method for improving the positioning accuracy of space targets, comprising the following steps: S1: Obtain observation data of space targets, construct an extended Gauss-Markov model containing potential modeling errors based on the observation data, and generate a set of multiple hypotheses by combining the modeling error parameters in different ways. S2: Using the likelihood ratio test criterion, construct the test statistics between different hypotheses with nested relationships in the multiple hypothesis set, and then adaptively select to execute the generalized data probing process or the generalized data simplification process according to the needs of the observation environment. S3: The set of multiple hypotheses is filtered through multiple hypothesis testing to determine the optimal model hypothesis under the current observation scenario. Based on the optimal model hypothesis, the relevant parameters of the space target are adaptively estimated to obtain the position parameters of the space target.

[0007] Preferably, in step S1, constructing an extended Gauss-Markov model containing potential modeling errors based on the observed data includes: The extended Gauss-Markov model is established based on the observation data, which includes the observation vector of the space target, random noise, and the position parameters to be estimated. , Where y is the space target observation vector, A is the design matrix, x is the position parameter to be estimated, e is random noise, f is the modeling error parameter vector, and F is the modeling error design matrix. The modeling error includes one or at least two combinations of gross errors in space target observation and atmospheric propagation delay interference. The modeling error parameter vector f is decomposed as follows: , in, This is a diagonal pattern matrix, where the diagonal elements indicate the presence of various modeling errors. This is an error magnitude vector used to describe the specific numerical value of the existing modeling error.

[0008] Preferably, in step S1, generating a multiple hypothesis set includes: Define the complete set of all potential modeling error indices. ; Construct the complete set All subsets Each subset corresponds to a specific error combination hypothesis; For each subset From the modeled error design matrix Extract the corresponding columns to construct a simplified error matrix. Let the corresponding error magnitude vector be denoted as Thus, each hypothesis This is expressed as including a specific modeling error term in the observation expectation, as shown below: .

[0009] Preferably, in step S2, constructing the test statistics among different hypotheses with nested relationships in the multiple hypothesis set includes: Calculate the hypotheses in the multiple hypothesis set. The residual vector below And its weighted norm square, as shown below: , in, The cofactor matrix of the observations; For nested assumptions When the observed noise variance Given the given information, construct the test statistic that follows a chi-square distribution, as shown below: ; When the observed noise variance When the unknown is known, use assumptions. The variance estimate is constructed according to the following parameters. The test statistic for the distribution, which quantifies the difference in goodness of fit between two nested hypotheses, and the variance estimate, are shown below: , Where n is the total number of observations, and t is the number of location parameters to be estimated. Assumption The number of error parameters in the model.

[0010] Preferably, in step S2, the generalized data probing process is selected to be executed, including: When the observation conditions of the space target are good and there is a high requirement for the estimation accuracy of the positioning results, the generalized data detection process should be selected. The generalized data probing process adopts a forward selection strategy, gradually increasing the hypothesis dimension from 0. In the t-th iteration, the low-dimensional hypothesis that was not rejected in the previous round is compared with all the high-dimensional hypotheses in the current round by a likelihood ratio test. If the test statistic exceeds the critical value, the low-dimensional hypothesis is rejected, and the hypothesis with the largest test statistic is retained from the candidate hypothesis set of the current round as a new optimal hypothesis candidate. Repeat the iteration until the current hypothesis cannot be rejected or the maximum number of iterations is reached. Output the final unrejected hypothesis as the optimal model hypothesis.

[0011] Preferably, in step S2, selecting to execute the generalized data simplification process includes: When the space target observation environment is complex, there are multiple modeling error interferences, and there are high requirements for the robustness of the positioning results, the generalized data simplification process should be selected. The generalized data simplification process adopts a backward elimination strategy, which gradually reduces the assumed dimension starting from the maximum dimension. In the t-th iteration, the likelihood ratio test is performed between the high-dimensional hypothesis accepted in the previous round and all low-dimensional hypotheses in the current round. If the test statistics of all low-dimensional hypotheses are not significant relative to the high-dimensional hypotheses, the set of low-dimensional hypotheses is accepted, and the hypothesis with the smallest test statistic is selected from the set of acceptable low-dimensional hypotheses as the new optimal hypothesis candidate. Repeat the iteration until no lower-dimensional hypothesis can be accepted or the maximum number of iterations is reached, and output the finally accepted hypothesis as the optimal model hypothesis.

[0012] Preferably, in step S3, the optimal model hypothesis for the current observation is determined through multiple hypothesis testing, and adaptive parameter estimation is performed, including: Based on the selected generalized data probing process or the generalized data simplification process, the optimal model hypothesis is determined through iterative probing and identification steps. Extract its corresponding error design matrix ; Based on the optimal model assumptions The extended observation equations are used to jointly solve for the estimated position parameters using the least squares principle. And error parameter estimation The extended observation equation is as follows: ; By subtracting the estimated modeling error from the observations, the corrected observation data is obtained, and the final space target position parameters are output. To eliminate the influence of modeling errors, the corrected observation data are as follows: .

[0013] Preferably, the construction conforms to The test statistics for the distribution include: .

[0014] Based on the same concept, the present invention also provides a generalized data detection and simplification system for improving the positioning accuracy of space targets, comprising: The model building and hypothesis generation module is used to acquire observation data of space targets, construct an extended Gauss-Markov model containing potential modeling errors based on the observation data, and generate multiple hypothesis sets by combining the modeling error parameters in different ways. The statistical testing and strategy selection module is used to construct test statistics between different hypotheses with nested relationships in the multiple hypothesis set using the likelihood ratio test criterion, and then adaptively select to execute the generalized data probing process or the generalized data simplification process according to the needs of the observation environment. The hypothesis screening and parameter estimation module is used to screen the set of multiple hypotheses through multiple hypothesis testing, determine the optimal model hypothesis under the current observation scenario, and adaptively estimate the relevant parameters of the space target based on the optimal model hypothesis to obtain the position parameters of the space target.

[0015] Based on the same concept, the present invention also provides a computer device, including a memory and a processor, wherein the memory stores computer-readable instructions, which, when executed by the processor, cause the processor to perform the steps of a generalized data detection and simplification method for improving the positioning accuracy of space targets as described in any one of the embodiments.

[0016] Compared with the prior art, the beneficial effects of the present invention are: (1) This invention constructs an extended Gauss-Markov model containing potential modeling errors and generates multiple hypothesis sets by combining the modeling error parameters in different ways. This enables the inclusion of various error types such as gross errors and atmospheric propagation delay interference in space target observation into a unified mathematical framework for processing. This overcomes the limitation of traditional methods that can only handle a single type of error and lays a complete theoretical foundation for subsequent error identification and elimination.

[0017] (2) This invention constructs test statistics between different hypotheses using the likelihood ratio test criterion, and adaptively selects to execute the generalized data probing process or the generalized data simplification process according to the needs of the observation environment, thus realizing flexible configuration of the strategy: when the observation conditions are good, the generalized data probing process is selected, and the forward selection strategy introduces as few additional parameters as possible to ensure that the variance of the location parameter estimation is minimized, thereby meeting the high-precision positioning requirements; when the observation environment is complex and there are multiple error interferences, the generalized data simplification process is selected, and the backward elimination strategy identifies and eliminates as many modeling errors as possible to ensure that the deviation of the estimation results is minimized, thereby enhancing the robustness and effectiveness of the system.

[0018] (3) This invention determines the optimal model hypothesis for the current observation through multiple hypothesis testing, and performs adaptive parameter estimation based on the optimal model hypothesis. After deducting the estimated modeling error from the observation value, the spatial target position parameters are calculated, which realizes the accurate identification and effective elimination of multiple modeling errors. It significantly alleviates the masking effect and flooding effect in the traditional iterative data detection method, effectively improves the accuracy and reliability of spatial target positioning, and has high reliability, flexibility and wide engineering applicability. Attached Figure Description

[0019] Various other advantages and benefits will become apparent to those skilled in the art upon reading the following detailed description of preferred embodiments. The accompanying drawings are for illustrative purposes only and are not intended to limit the invention.

[0020] Figure 1 This is a flowchart of a generalized data detection and simplification method for improving the positioning accuracy of space targets according to the present invention; Figure 2 This is a comparison of different methods when two gross errors in ranging occur. Detailed Implementation

[0021] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative of the invention and are not intended to limit the invention. Obviously, the described embodiments are only some, not all, of the embodiments described in this application. All other embodiments obtained by those skilled in the art based on the embodiments in this application without creative effort are within the scope of protection of this application.

[0022] Those skilled in the art will understand that, unless otherwise stated, the singular forms “a” and “an” used herein, and “the”, may also include the plural forms. It should be further understood that the term “comprising” as used in this specification means the presence of the stated features, integers, steps, operations, elements, and / or components, but does not exclude the presence or addition of one or more other features, integers, steps, operations, elements, components, and / or groups thereof.

[0023] First Embodiment Please see Figure 1 As shown, this embodiment provides a generalized data detection and simplification method for improving the positioning accuracy of space targets, including the following steps: S1: Acquire observation data of space targets, construct an extended Gauss-Markov model containing potential modeling errors based on the observation data, and generate multiple hypothesis sets by combining the modeling error parameters in different ways. Specifically, in this embodiment, the raw observation values ​​(such as ranging, angle measurement, carrier phase, etc.) of space targets (such as satellites and debris) are acquired through ground-based radar, optical telescopes or spaceborne GNSS receivers.

[0024] Preferably, in step S1, constructing an extended Gauss-Markov model containing potential modeling errors based on the observed data includes: An extended Gauss-Markov model is established based on observation data including the space target observation vector, random noise, and the position parameters to be estimated. , Where y is the spatial target observation vector, A is the design matrix, x is the position parameter to be estimated, and e is random noise. Considering the various non-random errors that may exist in the observation, a modeling error term is introduced. f is the modeling error parameter vector, and F is the modeling error design matrix. Specifically, in this embodiment, the modeling error types include gross errors and systematic errors. Gross errors refer to outliers or sudden anomalies in the observed data, while systematic errors refer to residual biases that have not been modeled, including atmospheric propagation delay residuals, gravitational perturbation biases, or multipath effects. The modeling error parameter vector f is decomposed as follows: , in, Let be a diagonal pattern matrix, indicating whether the i-th type of error exists (1 indicates existence, 0 indicates non-existence). The diagonal elements of the diagonal pattern matrix are used to indicate the existence status of various modeling errors. This is an error magnitude vector used to describe the specific numerical value of the existing modeling error.

[0025] Preferably, in step S1, generating a multiple hypothesis set includes: Define the complete set of all potential modeling error indices. ; Build the complete set All subsets Each subset corresponds to a specific error combination hypothesis; For each subset From the modeled error design matrix Extract the corresponding columns to construct a simplified error matrix. (dimension is) ), and denote the corresponding error magnitude vector as Thus, each hypothesis This is expressed as including a specific modeling error term in the observation expectation, as shown below: , in For corresponding The error magnitude parameter vector.

[0026] S2: Using the likelihood ratio test criterion, construct test statistics between different hypotheses with nested relationships in the multiple hypothesis set, and then adaptively select to execute the generalized data probing process or the generalized data simplification process according to the needs of the observation environment.

[0027] Preferably, in step S2, constructing test statistics among different hypotheses with nested relationships in the multiple hypothesis set includes: Calculate the hypotheses in the multiple hypothesis set The residual vector below And its weighted norm square, as shown below: , in, The cofactor matrix of the observations; specifically, in this embodiment, for the multiple hypothesis set. Any of the assumptions in The corresponding extended observation model is: By applying the least squares principle, the parameter estimates under this assumption can be obtained. and This leads to the residual vector, and the weighted norm square of the residual vector (i.e., the sum of squared residuals) is: The weighted norm square of the residual vector (i.e., the sum of squared residuals) is: ; For nested assumptions When the observed noise variance Given the given information, construct the test statistic that follows a chi-square distribution, as shown below: , Specifically, in this embodiment, under the original assumption Under the condition that it is true, Obeying the degree of freedom Chi-square distribution ( ),in = Assumption The number of standard error parameters; When the observed noise variance When the unknown is known, use assumptions. The variance estimate is constructed according to the following parameters. The test statistic of the distribution is used to quantify the difference in goodness of fit between two nested hypotheses, and the variance estimate is shown below: , Where n is the total number of observations, and t is the number of location parameters to be estimated. Assumption The number of error parameters in the model.

[0028] Preferably, under the null hypothesis Under the condition that it is true, Obeying the degree of freedom The F-distribution is constructed to follow the... The test statistics for the distribution include: .

[0029] Preferably, in step S2, the generalized data detection process or the generalized data simplification process is adaptively selected according to the requirements of the observation environment, including: Calculate the residual sequence based on the initial observation data, and calculate the kurtosis coefficient or skewness coefficient of the residual sequence; When the kurtosis coefficient is lower than the set threshold or the absolute value of the skewness coefficient is lower than the set threshold, the observation conditions are deemed to be good, and the generalized data detection process is selected to be executed. When the kurtosis coefficient is higher than the set threshold or the absolute value of the skewness coefficient is higher than the set threshold, it is determined that there are multiple modeling error interferences in the observation environment, and the generalized data simplification process is selected to be executed.

[0030] Preferably, in step S2, the generalized data probing process is selected to be executed, including: When the space target observation conditions are good and there is a high requirement for the estimation accuracy of the positioning results, the generalized data detection process (GDS) is selected to be executed. Specifically, in this embodiment, a good environment refers to stable space target observation conditions, low electromagnetic interference, normal sensor operation, and sparse and small-scale modeling errors in the observation data. The generalized data probing process employs a forward selection strategy, gradually increasing the hypothesis dimension from 0. Specifically, in this embodiment, the current hypothesis dimension q=0, which is the current optimal hypothesis. (No error assumption); In the t-th iteration, the low-dimensional hypotheses that were not rejected in the previous round are compared with all the high-dimensional hypotheses in the current round using a likelihood ratio test. If the test statistic exceeds the critical value, the low-dimensional hypothesis is rejected, and the hypothesis with the largest test statistic is retained from the candidate hypothesis set of the current round as a new optimal hypothesis candidate. Specifically, in this embodiment, in the current iteration t, a candidate hypothesis set is generated based on the current hypothesis dimension q and the selected strategy. , and That is, all nested hypotheses that have one more error parameter than the current hypothesis and contain the current hypothesis, for the candidate hypothesis set. Each hypothesis in Calculate its relationship with the current optimal hypothesis. Test statistic between and with critical value Comparison, test statistic Reflecting from Expand to The degree of improvement in fit, if there is at least one Make If the result is positive, it indicates that introducing additional error parameters significantly improves the model fit, and the process proceeds to the identification step; otherwise, the iteration terminates and the model is accepted. As a final assumption, from all satisfying Candidates In the process, the hypothesis with the largest test statistic (i.e. the smallest improvement in fit) is selected as the new optimal hypothesis; Repeat the iteration until the current hypothesis cannot be rejected or the maximum number of iterations is reached. Output the final unrejected hypothesis as the optimal model hypothesis.

[0031] Preferably, in step S2, the generalized data simplification process is selected to be executed, including: When the space target observation environment is complex, there are multiple modeling error interferences, and there are high requirements for the robustness of the positioning results, the generalized data simplification process is selected. Specifically, in this embodiment, a complex environment refers to the presence of strong electromagnetic interference, multipath effects, intermittent sensor failures, or space weather disturbances, and the observation data may contain multiple, mixed modeling errors (such as multiple gross errors and systematic errors coexisting). The Generalized Data Reduction (GDR) process employs a backward elimination strategy, gradually reducing the hypothesis dimension starting from the maximum dimension. Specifically, in this embodiment, the current hypothesis dimension q = p (the maximum dimension), and the current optimal hypothesis... (Including assumptions for all potential errors); In the t-th iteration, a likelihood ratio test is performed between the high-dimensional hypothesis accepted in the previous round and all low-dimensional hypotheses in the current round. If the test statistics of all low-dimensional hypotheses relative to the high-dimensional hypotheses are not significant, then the set of low-dimensional hypotheses is accepted, and the hypothesis with the smallest test statistic is selected from the set of acceptable low-dimensional hypotheses as a new optimal hypothesis candidate. Specifically, in this embodiment, in the current iteration t, a candidate hypothesis set is generated based on the current hypothesis dimension q and the selected strategy. , and For the candidate hypothesis set Each hypothesis in Calculate its relationship with the current optimal hypothesis. Test statistic between and with critical value Comparison, test statistic Reflecting from Expand to The degree of fitting loss, if for all All This indicates that removing the error parameter does not significantly worsen the fit, therefore all candidate low-dimensional hypotheses are acceptable, and the process proceeds to the identification step. If a certain... Make This indicates that removing this error parameter would severely damage the fit, therefore this simplification direction is unacceptable, but other candidates need to be examined further; if all candidates lead to significant losses, the iteration is terminated, and the remaining candidate is retained. As a final assumption, from all satisfying Candidates In this process, the hypothesis with the smallest test statistic (i.e. the smallest fitting loss) is selected as the new optimal hypothesis. Repeat the iteration until no lower-dimensional hypothesis can be accepted or the maximum number of iterations is reached, and output the finally accepted hypothesis as the optimal model hypothesis.

[0032] S3: The set of multiple hypotheses is filtered through multiple hypothesis testing to determine the optimal model hypothesis under the current observation scenario. Based on the optimal model hypothesis, the relevant parameters of the space target are adaptively estimated, and the position parameters of the space target are obtained.

[0033] Preferably, in step S3, the optimal model hypothesis for the current observation is determined through multiple hypothesis testing, and adaptive parameter estimation is performed, including: Based on the selected generalized data probing or generalized data simplification process, the optimal model hypothesis is determined through iterative probing and identification steps. Extract its corresponding error design matrix ; Based on the optimal model assumptions The extended observation equations are used to jointly solve for the estimated position parameters using the least squares principle. And error parameter estimation The extended observation equation is as follows: , Specifically, in this embodiment, the normal equation is established based on the extended observation equation: ; By subtracting the estimated modeling error from the observations, the corrected observation data is obtained, and the final space target position parameters are output. To eliminate the influence of modeling errors and correct the observed data, as shown below: , Specifically, in this embodiment, at this time, It should approximately satisfy the ideal Gauss-Markov model. Based on this, solve again This allows us to obtain the final spatial target position parameters.

[0034] like Figure 2 The figure shows a performance comparison of the least squares method (LS), the generalized data detection method (GDS), and the generalized data reduction method (GDR) under different noise conditions. The first row corresponds to the case where the standard deviation of the observation noise is known, and the second row corresponds to the case where the standard deviation of the observation noise is unknown. The four columns represent the changes in the correct detection probability (Pcp), the correct identification probability (Pcr), the floating root mean square error (RMSE), and the fixed root mean square error (RMSE) as a function of the interference intensity parameter (s).

[0035] Depend on Figure 2 It can be seen that, when the noise standard deviation is known, both GDS and GDR can rapidly improve the correct detection probability and correct identification probability with the increase of interference intensity, and control the floating-point solution and fixed solution errors at a low level, with overall performance significantly better than the LS method. In contrast, the positioning error of the LS method continuously increases with the increase of S.

[0036] Furthermore, when the noise standard deviation is unknown, the correct detection probability and correct identification probability of the GDS method decrease significantly, approaching zero. Its floating-point and fixed-point solution errors also deteriorate significantly with increasing S, and its performance is close to or even worse than that of the LS method. In contrast, the GDR method maintains a high correct detection probability and correct identification probability under the same conditions, and consistently maintains a low positioning error, demonstrating stronger robustness and better error suppression capabilities.

[0037] Therefore, Figure 2 This demonstrates that the GDR method proposed in this embodiment has a more significant performance advantage than the GDS and LS methods under conditions of incomplete noise statistics and multiple modeling error interferences, and is more suitable for high-precision positioning tasks in complex space target observation scenarios.

[0038] Second Embodiment Based on the same concept, the present invention also provides a generalized data detection and simplification system for improving the positioning accuracy of space targets, comprising: The model building and hypothesis generation module is used to acquire observation data of space targets, construct an extended Gauss-Markov model containing potential modeling errors based on the observation data, and generate multiple hypothesis sets by combining the modeling error parameters in different ways. The statistical testing and strategy selection module is used to construct test statistics between different hypotheses with nested relationships in the multiple hypothesis set using the likelihood ratio test criterion, and then adaptively select to execute the generalized data probing process or the generalized data simplification process according to the needs of the observation environment. The hypothesis screening and parameter estimation module is used to screen the set of multiple hypotheses through multiple hypothesis testing, determine the optimal model hypothesis under the current observation scenario, and adaptively estimate the relevant parameters of the space target based on the optimal model hypothesis to obtain the position parameters of the space target.

[0039] Third Embodiment Based on the same concept, this embodiment also provides a computer device, including a memory and a processor. The memory stores computer-readable instructions, which, when executed by the processor, cause the processor to perform the steps of a generalized data detection and simplification method for improving the positioning accuracy of space targets as described in the embodiment.

[0040] Based on the same concept, the present invention also provides a storage medium storing computer-readable instructions, which, when executed by one or more processors, cause the one or more processors to perform the steps of a generalized data detection and simplification method for improving the positioning accuracy of space targets as described in any one embodiment.

[0041] It is understood that, for the aforementioned generalized data detection and simplification method for improving the positioning accuracy of space targets, if all of them are implemented as software functional modules and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, in essence, or the part that contributes to the prior art, or all or part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer server or a network device, etc.) to execute all or part of the steps of the methods of the various embodiments of the present invention. The aforementioned storage medium includes: USB flash drive, mobile hard drive, read-only memory (ROM), random access memory (RAM), magnetic disk or optical disk, and other media capable of storing program code.

[0042] Computer-readable storage media may include data signals propagated in baseband or as part of a carrier wave, carrying readable program code. Such propagated data signals may take various forms, including but not limited to electromagnetic signals, optical signals, or any suitable combination thereof. A readable storage medium may also be any readable medium other than a readable storage medium that can transmit, propagate, or transfer a program for use by or in connection with an instruction execution system, apparatus, or device. The program code contained on the readable storage medium may be transmitted using any suitable medium, including but not limited to wireless, wired, optical fiber, RF, etc., or any suitable combination thereof.

[0043] The above description is merely a preferred embodiment of the present invention. The scope of protection of the present invention is not limited to the above embodiments. All technical solutions falling within the scope of the present invention's concept are within the scope of protection of the present invention. It should be noted that for those skilled in the art, any improvements and modifications made without departing from the principles of the present invention should also be considered within the scope of protection of the present invention.

Claims

1. A generalized data detection and simplification method for improving the positioning accuracy of space targets, characterized in that, Includes the following steps: S1: Obtain observation data of space targets, construct an extended Gauss-Markov model containing potential modeling errors based on the observation data, and generate a set of multiple hypotheses by combining the modeling error parameters in different ways. S2: Using the likelihood ratio test criterion, construct the test statistics between different hypotheses with nested relationships in the multiple hypothesis set, and then adaptively select to execute the generalized data probing process or the generalized data simplification process according to the needs of the observation environment. S3: The set of multiple hypotheses is filtered through multiple hypothesis testing to determine the optimal model hypothesis under the current observation scenario. Based on the optimal model hypothesis, the relevant parameters of the space target are adaptively estimated to obtain the position parameters of the space target.

2. The generalized data detection and simplification method for improving the positioning accuracy of space targets according to claim 1, characterized in that, In step S1, an extended Gauss-Markov model incorporating potential modeling errors is constructed based on the observed data, including: The extended Gauss-Markov model is established based on the observation data, which includes the observation vector of the space target, random noise, and the position parameters to be estimated. ， Where y is the space target observation vector, A is the design matrix, x is the position parameter to be estimated, e is random noise, f is the modeling error parameter vector, and F is the modeling error design matrix. The modeling error includes one or at least two combinations of gross errors in space target observation and atmospheric propagation delay interference. The modeling error parameter vector f is decomposed as follows: ， in, This is a diagonal pattern matrix, where the diagonal elements indicate the presence of various modeling errors. This is an error magnitude vector used to describe the specific numerical value of the existing modeling error.

3. The generalized data detection and simplification method for improving the positioning accuracy of space targets according to claim 1, characterized in that, In step S1, a multiple hypothesis set is generated, including: Define the complete set of all potential modeling error indices. ; Construct the complete set All subsets Each subset corresponds to a specific error combination hypothesis; For each subset From the modeled error design matrix Extract the corresponding columns to construct a simplified error matrix. Let the corresponding error magnitude vector be denoted as Thus, each hypothesis This is expressed as including a specific modeling error term in the observation expectation, as shown below: 。 4. The generalized data detection and simplification method for improving the positioning accuracy of space targets according to claim 1, characterized in that, In step S2, the test statistics among different hypotheses with nested relationships in the multiple hypothesis set are constructed, including: Calculate the hypotheses in the multiple hypothesis set. The residual vector below And its weighted norm square, as shown below: ， in, The cofactor matrix of the observations; For nested assumptions When the observed noise variance Given the given information, construct the test statistic that follows a chi-square distribution, as shown below: ； When the observed noise variance When the unknown is known, use assumptions. The variance estimate is constructed according to the following parameters. The test statistic for the distribution, which quantifies the difference in goodness of fit between two nested hypotheses, and the variance estimate, are shown below: ， Where n is the total number of observations, and t is the number of location parameters to be estimated. Assumption The number of error parameters in the model.

5. A generalized data detection and simplification method for improving the positioning accuracy of space targets according to claim 4, characterized in that, In step S2, the generalized data probing process is selected, including: When the observation conditions of the space target are good and there is a high requirement for the estimation accuracy of the positioning results, the generalized data detection process should be selected. The generalized data probing process adopts a forward selection strategy, gradually increasing the hypothesis dimension from 0. In the t-th iteration, the low-dimensional hypothesis that was not rejected in the previous round is compared with all the high-dimensional hypotheses in the current round by a likelihood ratio test. If the test statistic exceeds the critical value, the low-dimensional hypothesis is rejected, and the hypothesis with the largest test statistic is retained from the candidate hypothesis set of the current round as a new optimal hypothesis candidate. Repeat the iteration until the current hypothesis cannot be rejected or the maximum number of iterations is reached. Output the final unrejected hypothesis as the optimal model hypothesis.

6. A generalized data detection and simplification method for improving the positioning accuracy of space targets according to claim 4, characterized in that, In step S2, the generalized data simplification process is selected to be executed, including: When the space target observation environment is complex, there are multiple modeling error interferences, and there are high requirements for the robustness of the positioning results, the generalized data simplification process should be selected. The generalized data simplification process adopts a backward elimination strategy, which gradually reduces the assumed dimension starting from the maximum dimension. In the t-th iteration, the likelihood ratio test is performed between the high-dimensional hypothesis accepted in the previous round and all low-dimensional hypotheses in the current round. If the test statistics of all low-dimensional hypotheses are not significant relative to the high-dimensional hypotheses, the set of low-dimensional hypotheses is accepted, and the hypothesis with the smallest test statistic is selected from the set of acceptable low-dimensional hypotheses as the new optimal hypothesis candidate. Repeat the iteration until no lower-dimensional hypothesis can be accepted or the maximum number of iterations is reached, and output the finally accepted hypothesis as the optimal model hypothesis.

7. The generalized data detection and simplification method for improving the positioning accuracy of space targets according to claim 1, characterized in that, In step S3, the optimal model hypothesis for the current observations is determined through multiple hypothesis testing, and adaptive parameter estimation is performed, including: Based on the selected generalized data probing process or the generalized data simplification process, the optimal model hypothesis is determined through iterative probing and identification steps. Extract its corresponding error design matrix ; Based on the optimal model assumptions The extended observation equations are used to jointly solve for the estimated position parameters using the least squares principle. And error parameter estimation The extended observation equation is as follows: ； By subtracting the estimated modeling error from the observations, the corrected observation data is obtained, and the final space target position parameters are output. To eliminate the influence of modeling errors, the corrected observation data are as follows: 。 8. A generalized data detection and simplification method for improving the positioning accuracy of space targets according to claim 4, characterized in that, The construction conforms to The test statistics for the distribution include: 。 9. A generalized data detection and simplification system for improving the positioning accuracy of space targets, characterized in that, include: The model building and hypothesis generation module is used to acquire observation data of space targets, construct an extended Gauss-Markov model containing potential modeling errors based on the observation data, and generate multiple hypothesis sets by combining the modeling error parameters in different ways. The statistical testing and strategy selection module is used to construct test statistics between different hypotheses with nested relationships in the multiple hypothesis set using the likelihood ratio test criterion, and then adaptively select to execute the generalized data probing process or the generalized data simplification process according to the needs of the observation environment. The hypothesis screening and parameter estimation module is used to screen the set of multiple hypotheses through multiple hypothesis testing, determine the optimal model hypothesis under the current observation scenario, and adaptively estimate the relevant parameters of the space target based on the optimal model hypothesis to obtain the position parameters of the space target.

10. A computer device, characterized in that, The system includes a memory and a processor, wherein the memory stores computer-readable instructions that, when executed by the processor, cause the processor to perform the steps of a generalized data detection and simplification method for improving the positioning accuracy of space targets as described in any one of claims 1 to 8.

Images