A method and apparatus for detecting anomaly observations in precise single-point positioning using a multi-satellite system

Abstract

This invention belongs to the field of radio navigation technology and discloses a method and device for detecting anomaly observations in precise point positioning using multiple satellite systems. The invention first constructs a pre-verification error distribution model for observations from different satellite systems; then, combining the pre-verification error distribution model with actual satellite observation errors, it constructs error verification rules and sets initial values ​​for pre-verification parameters; combining PPP post-verification residual information and PPP absolute positioning accuracy, it constructs a post-verification residual detection standard and sets initial values ​​for post-verification judgment thresholds; finally, for each satellite system, observations with pre-verification errors exceeding the limit are eliminated or downweighted according to the error verification rules, while dynamically adjusting the pre-verification parameters and post-verification judgment thresholds until the post-verification residual detection standard is met. This invention can balance the accuracy and efficiency of anomaly detection in precise point positioning using multi-satellite system fusion.

Description

Technical Field

[0001] This invention belongs to the field of radio navigation technology, and more specifically, relates to a method and device for detecting anomaly observations in precise single-point positioning of a multi-satellite system. Background Technology

[0002] Precise point positioning (PPP) using Global Navigation Satellite System (GNSS) has become a key means of achieving decimeter-to-centimeter level positioning and will be widely used in fields such as intelligent driving and low-altitude economy (e.g., low-altitude manufacturing, low-altitude flight, low-altitude support). However, these applications face challenges such as increased anomalies in satellite observations in complex urban environments, decreased positioning accuracy, and slower processing efficiency. Efficient multi-GNSS observation anomaly detection methods are an important way to improve the performance of PPP positioning in complex environments.

[0003] Traditional anomaly detection methods are divided into two types: pre-detection and post-detection. However, existing methods have the following limitations: (1) Pre-detection directly calculates the error of the observations of each satellite system (BeiDou, GPS, Galileo, GLONASS, etc.) before positioning, compares it with the set threshold, and removes the values ​​with larger errors. It does not consider the differences in clock bias characteristics of multiple systems, and models all system observations in a unified manner. It is difficult to separate systematic gross errors, which can easily lead to the wrong removal of the observations of the entire system. (2) Post-detection judges the size of the residual after positioning and performs iterative processing according to the set threshold and judgment rules. If the proportion of anomalies increases, it is very easy for the positioning results to fail to be output within the sampling interval (usually 1 second), causing calculation congestion. Summary of the Invention

[0004] This invention provides a method and apparatus for detecting anomaly observations in precise single-point positioning using multiple satellite systems, thereby solving the problem of balancing accuracy and efficiency in anomaly detection in existing technologies involving the fusion of multiple satellite systems for precise single-point positioning.

[0005] This invention provides a method for detecting anomaly observations in precise single-point positioning of a multi-satellite system, comprising the following steps:

[0006] Obtain the error space of observations from multiple satellite systems in Precise Point Positioning (PPP), and construct the pre-existing error distribution model for observations from different satellite systems respectively;

[0007] Combining the aforementioned pre-examination error distribution model and the actual satellite observation error, an error verification rule is constructed, and initial values ​​of the pre-examination parameters included in the error verification rule are set. The pre-examination parameters include residual abnormality thresholds and adjustment factors.

[0008] Combining PPP post-verification residual information and PPP absolute positioning accuracy, a post-verification residual detection standard is constructed, and the initial value of the post-verification judgment threshold included in the post-verification residual detection standard is set.

[0009] For each satellite system, observations with errors exceeding the pre-verification limit are eliminated or downweighted according to the error verification rules. At the same time, the pre-verification parameters and the post-verification judgment threshold are dynamically adjusted until the post-verification residual detection standard is established.

[0010] Preferably, the multiple satellite systems include various categories of satellite systems such as BeiDou, GPS, Galileo, and GLONASS.

[0011] By inputting error correction information into the PPP model, the observation error space before positioning calculation is obtained. The observation error space includes the phase observation error space and the pseudorange observation error space.

[0012] Preferably, the PPP model adopts an ionosphere-suppressing combined model;

[0013] The error space of the phase observation value is represented as follows:

[0014]

[0015] The pseudorange observation error space is represented as follows:

[0016]

[0017] In the formula, sys represents a certain type of satellite system, selected from BeiDou, GPS, GALILEO, or GLOANSS; This represents the phase observation error space of the i-th satellite in the satellite system sys. This represents the pseudorange observation error space of the i-th satellite in the satellite system sys. The wavelength of the satellite system sys is indicated. This represents the phase observation value of the i-th satellite in the satellite system sys. This represents the pseudorange observation value of the i-th satellite in the satellite system sys. Let represent the distance from the i-th satellite in the satellite system sys to the ground receiver, and let c represent the speed of light. This indicates the receiver clock bias corresponding to the satellite system sys. This represents the satellite clock bias of the i-th satellite in the satellite system sys. This represents the tropospheric delay of the i-th satellite. This represents the ambiguity portion of the phase observation value of the i-th satellite in the satellite system sys.

[0018] Preferably, based on the observation error space, the residual samples are classified in multiple dimensions according to various feature parameters to form multiple feature subspaces; for the residual sample set in each feature subspace, the prior error distribution model is constructed using statistical modeling methods.

[0019] Preferably, the multiple characteristic parameters include satellite system category, observation type, satellite elevation angle, and signal-to-noise ratio, and the observation type includes pseudorange observations and phase observations.

[0020] Preferably, the error testing rules include testing rules for systematic errors and testing rules for non-systematic errors;

[0021] The test rule for the systematic error is: if the following conditions are met... If the residual term is zero, then the corresponding observation is discarded; where, ;

[0022] The test rule for non-systematic errors is as follows: if either of the following two conditions is met, the observation is determined to be a non-systematic outlier and is subject to weight reduction.

[0023] Condition one: ;

[0024] Condition two: ;

[0025] In the formula, This represents the error space of the i-th satellite in the k-th satellite system after removing systematic errors, where m represents the residual abnormality threshold. This represents the standard deviation of the satellite system's testing; min indicates taking the minimum value. This represents the residual standard deviation of the k-th satellite system. This represents the deviation between the residual term of the k-th satellite system and its median. This represents the arithmetic mean of all observation residuals for the k-th satellite system. This represents the adjustment factor for the absolute deviation of the median. This represents the median absolute deviation corresponding to the k-th satellite system. This represents the adaptive median deviation adjustment factor; The adaptive adjustment factor for the i-th satellite is related to the elevation angle. and signal-to-noise ratio The function.

[0026] Preferably, the systematic error and the non-systematic error are distinguished in the following way:

[0027] If both of the following conditions are met, the error is classified as systematic; otherwise, it is classified as unsystematic.

[0028] Condition one: ;

[0029] Condition two: ;

[0030] In the formula, This represents the average residual of the k-th satellite system. The threshold for judging the average residual. The threshold for judging the standard deviation of the residual.

[0031] Preferably, the post-test residual detection standard is: provided that the PPP absolute positioning accuracy is less than the set positioning accuracy, it must meet the following requirements. ;

[0032] In the formula, This represents the a priori residual of the observations from the i-th satellite in the k-th satellite system. This indicates the post-test threshold. This represents the standard deviation of the post-hoc residuals for the k-th satellite system.

[0033] Preferably, dynamically adjusting the pre-test parameters and the post-test determination threshold includes:

[0034] If the post-test residual detection standard is not met under the current parameters, the post-test judgment threshold is gradually adjusted from the initial value until the post-test residual detection standard is met, and the minimum adjustment value of the post-test judgment threshold is 1.

[0035] If the post-test residual detection standard is still not established when the post-test judgment threshold is adjusted to 1, the pre-test parameters are gradually increased from the initial value until the post-test residual detection standard is established.

[0036] On the other hand, the present invention provides a device for detecting anomaly observations in precise single-point positioning of a multi-satellite system, comprising:

[0037] The pre-approval error distribution model construction unit is used to obtain the error space of observations from multiple satellite systems in Precise Point Positioning (PPP) and to construct pre-approval error distribution models for observations from different satellite systems respectively.

[0038] The error verification rule construction unit is used to construct error verification rules by combining the pre-verification error distribution model and the actual satellite observation error, and to set the initial values ​​of the pre-verification parameters included in the error verification rules. The pre-verification parameters include residual abnormality thresholds and adjustment factors.

[0039] The post-test residual detection standard construction unit is used to combine PPP post-test residual information and PPP absolute positioning accuracy to construct a post-test residual detection standard, and to set the initial value of the post-test judgment threshold included in the post-test residual detection standard.

[0040] The observation anomaly detection and dynamic adjustment unit is used to, for each satellite system, eliminate or downweight observations with excessive pre-verification errors according to the error verification rules, and dynamically adjust the pre-verification parameters and the post-verification judgment threshold until the post-verification residual detection standard is established.

[0041] The multi-satellite system precise single-point positioning anomaly observation detection device is used to perform the steps in the multi-satellite system precise single-point positioning anomaly observation detection method as described above.

[0042] One or more technical solutions provided in this invention have at least the following technical effects or advantages:

[0043] This invention first obtains the error space of observations from multiple satellite systems in Precise Point Positioning (PPP) and constructs pre-verification error distribution models for observations from different satellite systems. Then, combining the pre-verification error distribution models and actual satellite observation errors, error verification rules are constructed, and initial values ​​are set for the pre-verification parameters (including residual abnormality thresholds and adjustment factors) included in the error verification rules. Combining the PPP post-verification residual information and PPP absolute positioning accuracy, a post-verification residual detection standard is constructed, and initial values ​​are set for the post-verification judgment thresholds included in the post-verification residual detection standard. Finally, for each satellite system, observations with pre-verification errors exceeding the limits are eliminated or downweighted according to the error verification rules, while dynamically adjusting the pre-verification parameters and post-verification judgment thresholds until the post-verification residual detection standard is established. This invention addresses the systematic gross errors and misjudgments caused by unified modeling of multiple systems in traditional pre-detection systems, as well as the low computational efficiency resulting from multiple iterations of anomaly removal in post-detection systems. It proposes a pre-detection anomaly detection scheme that handles each system independently. By analyzing the error distribution patterns of observations from different satellite systems during the PPP positioning pre-detection process, it performs one-step anomaly detection on the pre-detection errors of observations from different systems, and performs anomaly removal or dynamic weight reduction processing on each system. Dynamic weight reduction ensures the number of effective satellites while mitigating the impact of gross errors. Based on the characteristics of pre-detection and post-detection errors in a large amount of historical data, this invention establishes a relatively complete pre-detection error model. This model can be directly used as a relatively accurate prior error model. The variance estimate of the prior observations given by this model is relatively accurate, and it can detect anomaly observations. The proportion of cases with excessively large post-detection residuals is lower than when using a pre-detection error model without special processing. This invention, by dynamically adjusting pre-test parameters and post-test thresholds to establish the post-test residual detection standard, eliminates the need for further adjustments to subsequent pre-test and post-test thresholds in anomaly detection. It also eliminates the need for external references and post-test iterative processes. Therefore, this invention improves the efficiency of anomaly detection in multi-satellite system fusion precise point positioning. Overall, this invention balances the accuracy and efficiency of anomaly detection in multi-satellite system fusion precise point positioning. Attached Figure Description

[0044] Figure 1 This is an overall flowchart of a method for detecting anomaly observations in a multi-satellite system for precise single-point positioning, as provided in Embodiment 1 of the present invention. Detailed Implementation

[0045] To better understand the above technical solutions, the following will provide a detailed explanation of the technical solutions in conjunction with the accompanying drawings and specific implementation methods.

[0046] Example 1:

[0047] Example 1 provides a method for detecting anomaly observations in precise single-point positioning using a multi-satellite system, which mainly includes the following steps:

[0048] Obtain the error space of observations from multiple satellite systems in Precise Point Positioning (PPP), and construct the pre-existing error distribution model for observations from different satellite systems respectively;

[0049] Combining the aforementioned pre-examination error distribution model and the actual satellite observation error, an error verification rule is constructed, and initial values ​​of the pre-examination parameters included in the error verification rule are set. The pre-examination parameters include residual abnormality thresholds and adjustment factors.

[0050] Combining PPP post-verification residual information and PPP absolute positioning accuracy, a post-verification residual detection standard is constructed, and the initial value of the post-verification judgment threshold included in the post-verification residual detection standard is set.

[0051] For each satellite system, observations with errors exceeding the pre-verification limit are eliminated or downweighted according to the error verification rules. At the same time, the pre-verification parameters and the post-verification judgment threshold are dynamically adjusted until the post-verification residual detection standard is established.

[0052] The implementation path of Example 1 mainly includes three parts: pre-verification error acquisition (including constructing a pre-verification error distribution model for observations from different satellite systems), anomaly threshold determination (including constructing error verification rules containing pre-verification parameters and constructing a post-verification residual detection standard containing post-verification judgment thresholds), and post-verification feedback weighting (including dynamically adjusting pre-verification parameters and post-verification judgment thresholds). The overall flowchart is as follows: Figure 1 As shown below, each part will be described in detail.

[0053] (1) Obtaining the pre-test error.

[0054] (1.1) Obtain the error space of satellite observations from different systems in PPP.

[0055] The multiple satellite systems mentioned include various types of satellite systems such as BeiDou, GPS, Galileo, and GLONASS. By inputting error correction information into the PPP model, the observation error space before positioning calculation is obtained. This observation error space includes phase observation error space and pseudorange observation error space.

[0056] For example, the satellite systems used in this invention include BeiDou (abbreviated as B), GPS (abbreviated as G), GALILEO (abbreviated as E) and GLONASS (abbreviated as R), and the PPP model adopts an ionospheric desaturation combination model.

[0057] The observation errors of different systems in PPP can be expressed as:

[0058] (1)

[0059] (2)

[0060] In the formula, sys represents a certain type of satellite system, selected from BeiDou, GPS, GALILEO or GLOANSS, that is, sys can be represented as B, G, E or R; This represents the phase observation error space of the i-th satellite in the satellite system sys. This represents the pseudorange observation error space of the i-th satellite in the satellite system sys; the range of i depends on the number of satellites in the same satellite system. The wavelength of the satellite system sys is indicated. This represents the phase observation value (in weeks) of the i-th satellite in the satellite system sys. This represents the pseudorange observation value of the i-th satellite in the satellite system sys. Let represent the distance from the i-th satellite in the satellite system sys to the ground receiver, and let c represent the speed of light. This indicates the receiver clock bias corresponding to the satellite system sys. This represents the satellite clock bias of the i-th satellite in the satellite system sys. This represents the tropospheric delay of the i-th satellite. The ambiguous portion of the phase observation value of the i-th satellite in the satellite system sys.

[0061] (1.2) Construct a priori error distribution model.

[0062] Based on the observation error space, the residual samples are multidimensionally classified according to various feature parameters to form multiple feature subspaces. For the residual sample set in each feature subspace, a statistical modeling method is used to construct the prior error distribution model. The various feature parameters include satellite system category, observation type, satellite elevation angle, and signal-to-noise ratio, etc., and the observation type includes pseudorange observations and phase observations.

[0063] That is, the observation error space obtained in step (1.1) is categorized according to feature dimensions such as satellite system, observation type, satellite elevation angle, and signal-to-noise ratio. Combined with statistical analysis of a large amount of historical data, an error distribution model of the pre-hoc residuals in each feature subspace is constructed. The categorization process can be achieved through rule-based conditional filtering or automatic clustering algorithms (such as...). This can be achieved through hierarchical clustering. Step (1.2) aims to identify and distinguish the main sources of error affecting the residuals, classifying errors into two categories: systematic errors and non-systematic errors, for subsequent anomaly detection of observations.

[0064] For the residual sample set within each feature subspace, a statistical modeling method is used to construct a priori error distribution model. Specifically, the residual sample set can be defined as follows:

[0065]

[0066] In the formula, j represents the j-th feature subspace. Let the residual of the k-th observation be denoted, and the sample mean be calculated. and variance Assume that it approximately follows a normal distribution.

[0067] If the sample distribution exhibits skewness or long-tail characteristics, a Gaussian mixture model or kernel density estimation is further used for fitting to obtain a more accurate residual probability density function.

[0068] Ultimately, a set of mapping relationships can be obtained:

[0069]

[0070] In the formula, This represents the mapping relationship corresponding to the j-th feature subspace, where sys represents the satellite system category and type represents the pseudorange observation or phase observation. This represents the sine value of the satellite's elevation angle. Indicates the signal-to-noise ratio. Let represent the prior error distribution model corresponding to the j-th feature subspace.

[0071] The following method is used to determine systematic errors:

[0072] Calculate the mean of the residuals of each satellite in a certain satellite system within a sliding window. If it shows a smooth trend, it is considered a systematic error.

[0073] First, calculate the mean residual of the satellite system at the current epoch:

[0074] (3)

[0075] In the formula, This represents the average residual of the k-th satellite system. This represents the number of satellites participating in the calculation in the k-th satellite system. Let represent the prior residual of the i-th satellite in the k-th system.

[0076] For example, in the current epoch, if the number of satellites in a certain satellite system is greater than 4 and the residual values ​​of more than 75% of the satellite observations are concentrated, and the average value deviates from 0 overall, it indicates that the error is not caused by individual observations and can be preliminarily identified as a systematic error. After excluding the satellites that deviate, the average residual data of the satellite system can be updated by recalculating equation (3); otherwise, it can be preliminarily identified as a non-systematic error.

[0077] Then, the standard deviation of the residuals is calculated:

[0078] (4)

[0079] In the formula, This represents the residual standard deviation of the k-th satellite system.

[0080] If both of the following conditions are met: and If the error is positive, it is considered a systematic error; otherwise, it is considered a non-systematic error.

[0081] in, The threshold for judging the average residual. The threshold for judging the standard deviation of the residual. The set parameter value is 0.1m to 0.2m. The set parameter values ​​are 0.03m to 0.06m, and the pseudorange observation value is approximately 10 times the phase observation value. The specific values ​​can also be adjusted appropriately based on the proportion of subsequent post-verification residuals that are too large (for example, it can be stipulated that the proportion of post-verification residuals exceeding the limit should not exceed 5%) and the positioning accuracy.

[0082] (2) Determination of abnormal threshold.

[0083] The error testing rules include testing rules for systematic errors and testing rules for non-systematic errors.

[0084] (2.1) Determination of abnormal thresholds and testing rules in systematic errors.

[0085] The test rule for the systematic error is: if the following conditions are met... If so, the observation corresponding to that residual term is discarded.

[0086] in, , , .

[0087] In the formula, This represents the error space of the i-th satellite in the k-th satellite system after removing systematic errors, where m represents the residual abnormality threshold. This represents the standard deviation of the satellite system's testing; min indicates taking the minimum value. This represents the residual standard deviation of the k-th satellite system. This represents the deviation between the residual term of the k-th satellite system and its median, where median represents taking the median.

[0088] Preliminary analogy to statistics In principle, m=3 can be selected first, that is, the initial value of the residual abnormality threshold m is set to 3. Subsequently, the threshold size can be adjusted according to the actual impact of different residual abnormality thresholds on the positioning results. The adjustment principle is to increase the threshold without affecting accuracy, and to avoid erroneous rejection that would reduce the number of available satellites.

[0089] (2.2) Determination of abnormal thresholds and inspection rules in non-systematic errors.

[0090] Non-systematic errors refer to observational anomalies caused by sporadic factors such as transient disturbances, hardware jumps, multipath propagation, and obstruction, and are characterized by high randomness and locality. The key to identifying them lies in determining the anomalies of the residuals of individual satellites after removing systematic errors from the error space of the current epoch.

[0091] The test rule for non-systematic errors is as follows: if either of the following two conditions is met, the observation is determined to be a non-systematic outlier and is subject to weight reduction (i.e., the corresponding adjustment factor is multiplied in the diagonal elements of the covariance matrix during the solution process).

[0092] Condition one: ;

[0093] Condition two: .

[0094] In the formula, This represents the error space of the i-th satellite in the k-th satellite system after removing systematic errors. This represents the arithmetic mean of all observation residuals for the k-th satellite system. This represents the adjustment factor for the absolute deviation of the median. This represents the median absolute deviation corresponding to the k-th satellite system. This represents the adaptive median deviation adjustment factor; The adaptive adjustment factor for the i-th satellite is related to the elevation angle. and signal-to-noise ratio The function.

[0095] Specifically, this invention combines median absolute deviation (MAD), satellite elevation angle weighting, and signal-to-noise ratio (SNR). Thresholds are used to jointly construct robust non-systematic error detection rules, where:

[0096] (5)

[0097] , The initial value selection and adjustment strategy for the three parameters (m, m, and m) is consistent: the initial value of all three parameters can be set to 3, and then the threshold can be increased without affecting the positioning accuracy. Wherein:

[0098] (6)

[0099] In the formula, This represents the elevation angle of the i-th satellite. Let be the signal-to-noise ratio of the observations from the i-th satellite; This represents x raised to the power of y, which is: ; The maximum signal-to-noise ratio is typically set between 35 dB-Hz and 52 dB-Hz, depending on the receiver and antenna.

[0100] (3) Post-verification feedback adjustment.

[0101] The post-test residual detection standard is as follows: Provided the PPP absolute positioning accuracy is less than the set positioning accuracy, it must meet the following requirements. .

[0102] In the formula, This represents the a priori residual of the observations from the i-th satellite in the k-th satellite system. This indicates the post-test threshold. This represents the standard deviation of the post-hoc residuals for the k-th satellite system.

[0103] This invention dynamically adjusts the pre-test parameters and the post-test judgment threshold until the post-test residual detection standard is established.

[0104] The dynamic adjustment of the pre-test parameters and the post-test judgment threshold includes: if the post-test residual detection standard is not met under the current parameters, the post-test judgment threshold is gradually decreased from the initial value until the post-test residual detection standard is met, and the minimum adjustment value of the post-test judgment threshold is 1; if the post-test residual detection standard is still not met when the post-test judgment threshold is adjusted to 1, the pre-test parameters are gradually increased from the initial value until the post-test residual detection standard is met.

[0105] In other words, to preserve observation data to the maximum extent while ensuring positioning accuracy, this invention, after completing pre-approval residual overshoot detection, introduces PPP post-approval residual characteristics and positioning accuracy feedback to dynamically adjust the threshold and weighting factors for gross error judgment, thereby achieving a more robust error suppression strategy. The specific steps are as follows:

[0106] By introducing an external high-precision coordinate reference as a reference, the PPP absolute positioning accuracy WRMS is calculated, and the post-hoc residual standard deviation of each satellite system is obtained. (This can be calculated using formula (4)) and the post-hoc residuals of each satellite observation. (Right now , This represents the post-hoc residual of the i-th satellite in the k-th system. Let WRMS represent the average post-hoc residual of the k-th satellite system. For example, assuming WRMS is less than 3cm (3cm is an empirical value because the system error has been processed before the test, assuming that the impact of the system error on the positioning result has been partially reduced, and the positioning accuracy here is at the centimeter level), the following post-hoc residual test criteria are given:

[0107] (7)

[0108] In this invention To set an adjustable post-hoc decision threshold, since the standard deviation calculated post-hocly will be smaller than that pre-hocly, this can be adjusted based on experience. The initial value is set to 10. If equation (7) holds, then no further adjustment is needed. Conversely, adjust by gradually decreasing the interval by 1. Until equation (7) holds, the minimum adjustment is 1. If If equation (7) still does not hold after being adjusted to 1, then feedback adjustment of m, and After the above operations, no additional adjustments are needed to the pre- and post-verification thresholds. Furthermore, external references are no longer required, and there is no longer a post-verification iterative process, thus enabling autonomous, efficient, and high-precision positioning.

[0109] In summary, this invention first obtains the observation error space before positioning calculation by inputting the current positioning solution status and orbital clock error correction information into the PPP model, and constructs a pre-approval error distribution model for observations of different satellite systems. Then, based on the mean, variance, median, and other information output by the error distribution model, and combined with the actual satellite observation errors, it determines the anomaly judgment threshold and verification rules for different systems. Finally, based on the thresholds, observations with pre-approval errors exceeding the limits are eliminated or downweighted. Simultaneously, the thresholds and verification rules are dynamically adjusted based on the proportion of excessively large post-PPP residuals and positioning accuracy, achieving one-step detection of pre-approval observation errors with independent processing of subsystems and dynamic optimization of anomaly weights. This invention effectively overcomes the limitations of traditional methods in multi-system error separation and anomaly processing efficiency. Furthermore, this invention overcomes the limitations of traditional methods in terms of positioning continuity. If there are gross errors or anomalies in the observations of a certain epoch, the positioning result of that epoch will deviate from the results of other epochs, and the dynamic scene will deviate from the expected trajectory. However, after detecting abnormal observations, this method can eliminate them or assign a variance value that matches their characteristics, minimizing the impact of abnormal observations on the positioning result. After elimination, they no longer affect the positioning, thus ensuring that the positioning result is on the expected trajectory. Based on the above-mentioned abnormal observation detection, the healthy observations retained by each independent system can be merged together for the final high-precision positioning calculation, achieving a synergistic improvement in PPP positioning accuracy and calculation efficiency in complex environments. This invention is applicable to scenarios with stringent requirements for real-time performance and reliability, such as intelligent driving and low-altitude economic applications.

[0110] Example 2:

[0111] Example 2 provides a device for detecting anomaly observations in precise single-point positioning of a multi-satellite system, comprising:

[0112] The pre-approval error distribution model construction unit is used to obtain the error space of observations from multiple satellite systems in Precise Point Positioning (PPP) and to construct pre-approval error distribution models for observations from different satellite systems respectively.

[0113] The error verification rule construction unit is used to construct error verification rules by combining the pre-verification error distribution model and the actual satellite observation error, and to set the initial values ​​of the pre-verification parameters included in the error verification rules. The pre-verification parameters include residual abnormality thresholds and adjustment factors.

[0114] The post-test residual detection standard construction unit is used to combine PPP post-test residual information and PPP absolute positioning accuracy to construct a post-test residual detection standard, and to set the initial value of the post-test judgment threshold included in the post-test residual detection standard.

[0115] The observation anomaly detection and dynamic adjustment unit is used to, for each satellite system, eliminate or downweight observations with pre-verification errors exceeding the limit according to the error verification rules, and dynamically adjust the pre-verification parameters and the post-verification judgment threshold until the post-verification residual detection standard is established.

[0116] The multi-satellite system precise single-point positioning anomaly observation detection device provided in Example 2 is used to perform the steps in the multi-satellite system precise single-point positioning anomaly observation detection method as described in Example 1.

[0117] Since the functions of each unit in the multi-satellite system precise single-point positioning anomaly observation detection device provided in Embodiment 2 correspond to the steps in the multi-satellite system precise single-point positioning anomaly observation detection method provided in Embodiment 1, Embodiment 2 can be understood by referring to the description of Embodiment 1, and will not be repeated here.

[0118] Finally, it should be noted that the above specific embodiments are only used to illustrate the technical solutions of the present invention and not to limit it. Although the present invention has been described in detail with reference to examples, those skilled in the art should understand that modifications or equivalent substitutions can be made to the technical solutions of the present invention without departing from the spirit and scope of the technical solutions of the present invention, and all such modifications or substitutions should be covered within the scope of the claims of the present invention.

Claims

1. A method for detecting anomaly observations using precise single-point positioning in a multi-satellite system, characterized in that, Includes the following steps: Obtain the error space of observations from multiple satellite systems in Precise Point Positioning (PPP), and construct the pre-existing error distribution model for observations from different satellite systems respectively; Combining the aforementioned pre-examination error distribution model and the actual satellite observation error, an error verification rule is constructed, and initial values ​​of the pre-examination parameters included in the error verification rule are set. The pre-examination parameters include residual abnormality thresholds and adjustment factors. Combining PPP post-verification residual information and PPP absolute positioning accuracy, a post-verification residual detection standard is constructed, and the initial value of the post-verification judgment threshold included in the post-verification residual detection standard is set. For each satellite system, observations with errors exceeding the pre-verification limit are eliminated or downweighted according to the error verification rules. At the same time, the pre-verification parameters and the post-verification judgment threshold are dynamically adjusted until the post-verification residual detection standard is established. The dynamic adjustment of the pre-test parameters and the post-test judgment threshold includes: if the post-test residual detection standard is not met under the current parameters, the post-test judgment threshold is gradually decreased from the initial value until the post-test residual detection standard is met, and the minimum adjustment value of the post-test judgment threshold is 1; if the post-test residual detection standard is still not met when the post-test judgment threshold is adjusted to 1, the pre-test parameters are gradually increased from the initial value until the post-test residual detection standard is met.

2. The method for detecting anomaly observations in precise single-point positioning of a multi-satellite system according to claim 1, characterized in that, The multiple satellite systems mentioned include various categories of satellite systems such as BeiDou, GPS, Galileo, and GLONASS. By inputting error correction information into the PPP model, the observation error space before positioning calculation is obtained. The observation error space includes the phase observation error space and the pseudorange observation error space.

3. The method for detecting anomaly observations in precise single-point positioning of a multi-satellite system according to claim 2, characterized in that, The PPP model adopts an ionosphere-free combined model; The error space of the phase observation value is represented as follows: The pseudorange observation error space is represented as follows: In the formula, sys represents a certain type of satellite system, selected from BeiDou, GPS, GALILEO, or GLOANSS; This represents the phase observation error space of the i-th satellite in the satellite system sys. This represents the pseudorange observation error space of the i-th satellite in the satellite system sys. The wavelength of the satellite system sys is indicated. This represents the phase observation value of the i-th satellite in the satellite system sys. This represents the pseudorange observation value of the i-th satellite in the satellite system sys. Let represent the distance from the i-th satellite in the satellite system sys to the ground receiver, and let c represent the speed of light. This indicates the receiver clock bias corresponding to the satellite system sys. This represents the satellite clock bias of the i-th satellite in the satellite system sys. This represents the tropospheric delay of the i-th satellite. This represents the ambiguity portion of the phase observation value of the i-th satellite in the satellite system sys.

4. The method for detecting anomaly observations in precise single-point positioning of a multi-satellite system according to claim 2, characterized in that, Based on the observed error space, the residual samples are classified in multiple dimensions according to various feature parameters to form multiple feature subspaces; for the residual sample set in each feature subspace, the prior error distribution model is constructed using statistical modeling methods.

5. The method for detecting anomaly observations in precise single-point positioning of a multi-satellite system according to claim 4, characterized in that, The various characteristic parameters include satellite system category, observation type, satellite elevation angle, and signal-to-noise ratio. The observation type includes pseudorange observations and phase observations.

6. The method for detecting anomaly observations in precise single-point positioning of a multi-satellite system according to claim 1, characterized in that, The error testing rules include testing rules for systematic errors and testing rules for non-systematic errors; The test rule for the systematic error is: if the following conditions are met... Then, the observations corresponding to the residual terms are discarded; where, ; The test rule for non-systematic errors is as follows: if either of the following two conditions is met, the observation is determined to be a non-systematic outlier and is subject to weight reduction. Condition 1: ; Condition two: ; In the formula, This represents the error space of the i-th satellite in the k-th satellite system after removing systematic errors, where m represents the residual abnormality threshold. This represents the standard deviation of the satellite system's testing; min indicates taking the minimum value. This represents the residual standard deviation of the k-th satellite system. This represents the deviation between the residual term of the k-th satellite system and its median. This represents the arithmetic mean of all observation residuals for the k-th satellite system. This represents the adjustment factor for the absolute deviation of the median. This represents the median absolute deviation corresponding to the k-th satellite system. This represents the adaptive median deviation adjustment factor; The adaptive adjustment factor for the i-th satellite is related to the elevation angle. and signal-to-noise ratio The function.

7. The method for detecting anomaly observations in precise single-point positioning of a multi-satellite system according to claim 6, characterized in that, The systematic errors and the non-systematic errors are distinguished in the following way: If both of the following conditions are met, the error is classified as systematic; otherwise, it is classified as unsystematic. Condition 1: ; Condition two: ; In the formula, This represents the average residual of the k-th satellite system. The threshold for judging the average residual. The threshold for judging the standard deviation of the residual.

8. The method for detecting anomaly observations in precise single-point positioning of a multi-satellite system according to claim 1, characterized in that, The post-test residual detection standard is as follows: Provided the PPP absolute positioning accuracy is less than the set positioning accuracy, it must meet the following requirements. ; In the formula, This represents the a priori residual of the observations from the i-th satellite in the k-th satellite system. This indicates the post-test threshold. This represents the standard deviation of the post-hoc residuals for the k-th satellite system.

9. A device for detecting anomaly observations in precise single-point positioning using a multi-satellite system, characterized in that, include: The pre-approval error distribution model construction unit is used to obtain the error space of observations from multiple satellite systems in Precise Point Positioning (PPP) and to construct pre-approval error distribution models for observations from different satellite systems respectively. The error verification rule construction unit is used to construct error verification rules by combining the pre-verification error distribution model and the actual satellite observation error, and to set the initial values ​​of the pre-verification parameters included in the error verification rules. The pre-verification parameters include residual abnormality thresholds and adjustment factors. The post-test residual detection standard construction unit is used to combine PPP post-test residual information and PPP absolute positioning accuracy to construct a post-test residual detection standard, and to set the initial value of the post-test judgment threshold included in the post-test residual detection standard. The observation anomaly detection and dynamic adjustment unit is used to, for each satellite system, eliminate or downweight observations with excessive pre-verification errors according to the error verification rules, and dynamically adjust the pre-verification parameters and the post-verification judgment threshold until the post-verification residual detection standard is established. The multi-satellite system precise single-point positioning anomaly observation detection device is used to perform the steps in the multi-satellite system precise single-point positioning anomaly observation detection method as described in any one of claims 1 to 8.

Images