Cascade satellite navigation anomaly detection and elimination system and method based on inter-frequency pseudo-range deviation distribution

Abstract

The invention discloses a cascade satellite navigation anomaly detection and elimination system and method based on inter-frequency pseudo-range deviation distribution, and the system comprises a positioning residual detection module and an inter-frequency pseudo-range deviation distribution detection module, and the positioning residual detection module comprises an anomaly detection module and an anomaly recognition module; the inter-frequency pseudo-range deviation distribution detection module is used for carrying out anomaly judgment on each pseudo-range and outputting an anomaly detection result of each pseudo-range; the positioning residual error detection module is used for carrying out total anomaly detection on the integrity of the satellite navigation system, and the anomaly detection module carries out residual error chi-square detection according to the positioning residual error and judges whether an anomaly exists at the moment or not; and the anomaly identification module constructs an identification subset by means of a result of the inter-frequency pseudo-range deviation distribution detection module, so that fault identification is realized, and the accuracy of fault identification is ensured.

Description

Technical Field

[0001] This invention belongs to the field of satellite navigation system integrity detection technology, and specifically relates to a cascaded satellite navigation anomaly detection and elimination system and method based on inter-frequency pseudorange deviation distribution. Background Technology

[0002] Global Navigation Satellite System (GNSS) offers the advantage of providing absolute positioning services to users worldwide, 24 / 7. However, throughout the entire process from signal generation and propagation to user reception, satellite navigation signals can be distorted due to satellite anomalies, environmental factors, and receiver interference. A satellite navigation system can only enter the practical application stage of navigation and positioning if it meets certain performance requirements, which mainly include accuracy, integrity, continuity, and availability. Among these, integrity is the most critical performance requirement. For satellite navigation and positioning systems, the level of integrity performance determines the security and reliability of navigation and positioning services. To address this issue, monitoring methods are needed to detect and eliminate faults, protecting users from catastrophic information misleading caused by satellite signal anomalies or distortions. Integrity monitoring primarily involves the detection and identification of satellite faults, and in the fault detection process of satellite navigation systems, monitoring at the user end is paramount.

[0003] User-side monitoring, specifically Receiver Autonomous Integrity Monitoring (RAIM) technology, refers to the receiver utilizing received satellite signals to acquire various useful information through real-time signal processing and performing consistency checks using redundant observations. When the positioning error exceeds a preset detection threshold, this technology can promptly issue an alarm to the user, thereby ensuring the safety of navigation services. Ideally, RAIM should be able to detect, isolate, and eliminate faulty measurement sources during navigation calculations to ensure the validity and reliability of positioning results. For situations where a single observation is disturbed, this method typically has a high accuracy rate. However, in practical applications, anomalies may simultaneously appear in multiple observations at a single frequency, significantly increasing the difficulty of accurate anomaly identification. It is worth noting that current receivers can typically receive signals from multiple frequencies, and observations of the same satellite at different frequencies should normally remain consistent. Based on this characteristic, new anomaly identification methods can be explored from the perspective of pseudorange observation consistency across frequencies, thereby improving the accuracy of anomaly identification.

[0004] Therefore, this invention addresses the aforementioned problems by proposing a cascaded satellite navigation anomaly detection and removal system and method based on inter-frequency pseudorange deviation distribution. This method primarily utilizes residual detection, supplemented by pseudorange inter-frequency deviation consistency detection, to fuse and judge receiver observations, thereby effectively improving the accuracy of satellite navigation anomaly detection and enhancing the system's reliability and integrity, thus possessing significant engineering application value. Summary of the Invention

[0005] This invention provides a cascaded satellite navigation anomaly detection and elimination system and method based on inter-frequency pseudorange deviation distribution to solve the problems in the prior art.

[0006] To achieve the above objectives, the technical solution adopted by the present invention is as follows:

[0007] A cascaded satellite navigation anomaly detection and elimination system based on inter-frequency pseudorange deviation distribution includes an inter-frequency pseudorange deviation distribution detection module and a positioning residual detection module. The positioning residual detection module includes an anomaly detection module and an anomaly identification module.

[0008] The inter-frequency pseudorange deviation distribution detection module is used to judge the anomalies of each pseudorange of each satellite in the satellite navigation system and output the anomaly detection results of each pseudorange.

[0009] The positioning residual detection module is used for overall anomaly detection of the integrity of the satellite navigation system.

[0010] The anomaly detection module performs chi-square detection on the residual based on the positioning residual to determine whether there is an anomaly at that moment;

[0011] The anomaly identification module uses the results of the inter-frequency pseudorange deviation distribution detection module to construct an identification subset, thereby achieving fault identification and ensuring the accuracy of fault identification.

[0012] As a preferred embodiment, the input of the inter-frequency pseudorange deviation distribution detection module is the set of pseudorange observations at the main frequency point and the set of observations at the secondary frequency point for each satellite at the current epoch, and the output of the inter-frequency pseudorange deviation distribution detection module is a set of pseudorange observation quantum sets that are distinguished as normal and abnormal.

[0013] The inter-frequency pseudorange deviation distribution detection module includes an inter-frequency pseudorange difference and distribution detection unit, wherein...

[0014] The inter-frequency pseudorange differential unit performs inter-frequency pseudorange differential processing: in the inter-frequency pseudorange differential stage, the dual-frequency pseudorange observations of each satellite are subtracted to obtain the inter-frequency pseudorange deviation for that epoch. This deviation mainly reflects the variation characteristics of the inter-frequency link error and observation noise. Theoretically, this detection quantity follows a preset statistical distribution model (such as a normal distribution); this deviation quantity is then input as a detection quantity to the distribution detection unit.

[0015] The distribution detection unit performs distribution detection on the detection quantity based on the statistical characteristics of the observation noise of each satellite (such as the standard normal distribution detection algorithm), determines whether there are any anomalies in the corresponding observations, and divides the input pseudorange observations into normal subsets and abnormal subsets, providing a reliable data foundation for subsequent anomaly removal and navigation calculation.

[0016] The anomaly identification module takes as input the set of pseudorange observations classified according to the results of the inter-frequency pseudorange deviation distribution detection module for that epoch, and outputs a set of pseudorange observations that have passed the detection, which can be used for the next stage of localization or fusion. The anomaly identification module mainly completes three parts: constructing the subset to be detected, anomaly detection, and observation removal and output. This module first performs residual detection on the entire set of observations (e.g., using the least squares residual method). If no anomalies are found, the next step of localization calculation is directly performed. If anomalies exist, the module enters the inter-frequency pseudorange deviation distribution detection module for preliminary classification, using the normal subset as the base subset. Anomaly observations are added cyclically for anomaly information detection. If an anomaly fails the detection, the probabilistic anomaly signal added in this detection is removed; otherwise, it is considered normal information. This process is repeated until all probabilistic anomaly signals have been detected.

[0017] A method for cascaded satellite navigation anomaly detection and removal based on inter-frequency pseudorange deviation distribution includes the following steps:

[0018] S1. Read the set of observation information for the current epoch's main frequency point from the satellite observation files. , where: superscript Indicates the first Each satellite, with the subscript 1 indicating the dominant frequency point. For the first Pseudorange observations at the main frequency of a satellite ;

[0019] S2, Set of observation information for the current epoch dominant frequency point Perform residual detection. If the detection passes, output all observations directly to the positioning solution. If the detection fails, proceed to step S3.

[0020] S3. Read the set of sub-frequency observation information from the satellite observation file at the current epoch. Where: subscript 2 indicates the sub-frequency point, For the first Pseudorange observations at the secondary frequency of each satellite, and the set of observation information at the current epoch main frequency. and the set of observation information of secondary frequency points in the current epoch. The observations in the process are obtained by performing inter-frequency pseudorange difference. , For the first Pseudorange observations at the main and secondary frequencies of a single satellite and difference;

[0021] S4, to Each test metric in the constellation is tested using a standard normal distribution algorithm. Observations with normal test results are assigned to the corresponding set of normal observations for that constellation. Conversely, they are included in the set of anomalous observations corresponding to that constellation. ;

[0022] S5. Determine whether the number of constellation observations in the intact subset at this point meets the requirement: , , ,in: and These are the number of GPS and BDS observations in the normal observation set, respectively; if the requirements are met, proceed to step S6; otherwise, return to the anomaly detection method based on least squares residuals.

[0023] S6, Merge and get ,merge and get ;Will The set as a basic subset, for Perform residual detection to ensure the integrity of the subset again. If the detection passes, proceed to step S7; otherwise, return to the anomaly detection method based on least squares residuals.

[0024] S7, from One pseudorange observation is selected sequentially for residual detection. If the detection passes, the observation is classified as... If an observation fails the test, it is removed from the collection.

[0025] Cycle step S7 to The set average is subjected to residual detection, and the final result is output. Once the location calculation is complete, the identification process ends.

[0026] As a preferred embodiment, in step S3, the inter-frequency pseudorange difference is:

[0027] For the same satellite Pseudorange observations at the main frequency and secondary frequency and The difference is calculated as follows:

[0028] (1)

[0029] Among them: superscript Indicates the first Each satellite has a subscript 1 indicating the primary frequency and a subscript 2 indicating the secondary frequency.

[0030] As a preferred embodiment, in step S4, the standard normal distribution detection algorithm is used:

[0031] For a certain satellite The detection volume at this epoch Assuming no anomalies, it should conform to the expected value. The variance is normal distribution ,Right now:

[0032] (2)

[0033] Constructing standardized statistics As shown below:

[0034] (3)

[0035] in, , Let be the number of stationary observation samples, and i be the i-th stationary observation sample; with a false alarm rate of When, the corresponding bilateral detection threshold is :

[0036] (4)

[0037] in, is the probability density function of the standard normal distribution;

[0038] The algorithm's decision criteria are expressed as follows:

[0039] (5)

[0040] As a preferred embodiment, step S5, the anomaly detection method based on least squares residuals, includes the following steps:

[0041] Pseudorange residuals are typically used for anomaly detection based on least squares residuals. The pseudorange observation equation is as follows:

[0042] (6)

[0043] Where: x is the three-dimensional position change and receiver clock error, y is the pseudorange observation of the current n visible stars, i.e., the pseudorange observation value, H is an n×4 coefficient matrix, and ε is the pseudorange observation noise;

[0044] Based on the least squares method, the least squares solution of the state variables is obtained. :

[0045] (7)

[0046] Where: T represents the matrix transpose operation;

[0047] The deviation vector of the observed quantity can be obtained. :

[0048] (8)

[0049] Least square residual vector w:

[0050] ε(9)

[0051] Construct the residual sum of squares (SSE):

[0052] (10)

[0053] Given a false alarm rate, the following probability equation holds:

[0054] (11)

[0055] in, The probability of a certain random event occurring. The variance of pseudorange observation noise. For degrees of freedom The chi-square distribution density function, Indicates the number of stars currently visible; To set the false alarm rate; it is calculated according to the formula. For the classic RAIM algorithm, a uniform detection value is constructed. At this time, the threshold for:

[0056] (12)

[0057] When the detected quantity V is within the threshold, the observation is considered to be consistent; otherwise, the observation is considered to be abnormal.

[0058] When an anomaly is detected, the observations are cyclically removed until a set of observations below the observation threshold is found, which is considered a reliable set of observations.

[0059] Compared with the prior art, the present invention has the following beneficial effects:

[0060] (1) This invention proposes a cascaded satellite navigation anomaly detection and elimination system and method based on inter-frequency pseudorange deviation distribution. This method introduces multi-frequency observation information. This strategy enhances the reliability of detection and effectively improves the accuracy of satellite navigation anomaly identification.

[0061] (2) This method utilizes the inherent correlation between dual-frequency observations to construct the detection quantity, eliminates some common error sources and enhances the response of abnormal features through inter-frequency difference, improves the utilization rate and detection sensitivity of observation information without increasing external auxiliary information, and realizes the preliminary identification of abnormal observations;

[0062] (3) This method introduces the inter-frequency pseudorange deviation distribution detection step into the detection process to preliminarily identify and screen each satellite observation. It can remove obviously abnormal observation data at the front-end stage, thereby reducing the computational burden of subsequent statistical detection and parameter estimation.

[0063] (4) By introducing inter-frequency observation difference features, this method can not only identify single observation anomalies, but also effectively deal with the situation where there are correlation errors or systematic biases among multiple frequency observations, thereby improving the ability to identify complex anomaly types.

[0064] (5) This invention is based on software algorithms and has the advantages of great flexibility and low cost. Attached Figure Description

[0065] Figure 1 This is a single-epoch structure diagram of a cascaded satellite navigation anomaly detection and elimination system and method based on inter-frequency pseudorange deviation distribution according to the present invention.

[0066] Figure 2 This is a flowchart of a cascaded satellite navigation anomaly detection and elimination system and method based on inter-frequency pseudorange deviation distribution, according to the present invention. Detailed Implementation

[0067] The present invention will be further described below with reference to embodiments.

[0068] Example 1

[0069] like Figure 1 As shown, a cascaded satellite navigation anomaly detection and elimination system based on inter-frequency pseudorange deviation distribution includes an inter-frequency pseudorange deviation distribution detection module and a positioning residual detection module. The positioning residual detection module includes an anomaly detection module and an anomaly identification module.

[0070] The inter-frequency pseudorange deviation distribution detection module is used to judge the anomalies of each pseudorange of each satellite in the satellite navigation system and output the anomaly detection results of each pseudorange.

[0071] The positioning residual detection module is used for overall anomaly detection of the integrity of the satellite navigation system.

[0072] The anomaly detection module performs chi-square detection on the residual based on the positioning residual to determine whether there is an anomaly at that moment;

[0073] The anomaly identification module uses the results of the inter-frequency pseudorange deviation distribution detection module to construct an identification subset, thereby achieving fault identification and ensuring the accuracy of fault identification.

[0074] As a preferred embodiment, the input of the inter-frequency pseudorange deviation distribution detection module is the set of pseudorange observations at the main frequency point and the set of observations at the secondary frequency point for each satellite at the current epoch, and the output of the inter-frequency pseudorange deviation distribution detection module is a set of pseudorange observation quantum sets that are distinguished as normal and abnormal.

[0075] The inter-frequency pseudorange deviation distribution detection module includes an inter-frequency pseudorange difference and distribution detection unit, wherein...

[0076] The inter-frequency pseudorange differential unit performs inter-frequency pseudorange differential processing: in the inter-frequency pseudorange differential stage, the dual-frequency pseudorange observations of each satellite are subtracted to obtain the inter-frequency pseudorange deviation for that epoch. This deviation mainly reflects the variation characteristics of the inter-frequency link error and observation noise. Theoretically, this detection quantity follows a preset statistical distribution model (such as a normal distribution); this deviation quantity is then input as a detection quantity to the distribution detection unit.

[0077] The distribution detection unit performs distribution detection on the detection quantity based on the statistical characteristics of the observation noise of each satellite (such as the standard normal distribution detection algorithm), determines whether there are any anomalies in the corresponding observations, and divides the input pseudorange observations into normal subsets and abnormal subsets, providing a reliable data foundation for subsequent anomaly removal and navigation calculation.

[0078] The anomaly identification module takes as input the set of pseudorange observations classified according to the results of the inter-frequency pseudorange deviation distribution detection module for that epoch, and outputs a set of pseudorange observations that have passed the detection, which can be used for the next stage of localization or fusion. The anomaly identification module mainly completes three parts: constructing the subset to be detected, anomaly detection, and observation removal and output. This module first performs residual detection on the entire set of observations (e.g., using the least squares residual method). If no anomalies are found, the next step of localization calculation is directly performed. If anomalies exist, the module enters the inter-frequency pseudorange deviation distribution detection module for preliminary classification, using the normal subset as the base subset. Anomaly observations are added cyclically for anomaly information detection. If an anomaly fails the detection, the probabilistic anomaly signal added in this detection is removed; otherwise, it is considered normal information. This process is repeated until all probabilistic anomaly signals have been detected.

[0079] Example 2

[0080] like Figure 2 As shown, a method for cascaded satellite navigation anomaly detection and removal based on inter-frequency pseudorange deviation distribution includes the following steps:

[0081] S1. Read the set of observation information for the current epoch's main frequency point from the satellite observation files. , where: superscript Indicates the first Each satellite, with the subscript 1 indicating the dominant frequency point. For the first Pseudorange observations at the main frequency of a satellite ;

[0082] S2, Set of observation information for the current epoch dominant frequency point Perform residual detection. If the detection passes, output all observations directly to the positioning solution. If the detection fails, proceed to step S3.

[0083] S3. Read the set of sub-frequency observation information from the satellite observation file at the current epoch. Where: subscript 2 indicates the sub-frequency point, For the first Pseudorange observations at the secondary frequency of each satellite, and the set of observation information at the current epoch main frequency. and the set of observation information of secondary frequency points in the current epoch. The observations in the process are obtained by performing inter-frequency pseudorange difference. , For the first Pseudorange observations at the main and secondary frequencies of a single satellite and difference;

[0084] As a preferred embodiment, in step S3, the inter-frequency pseudorange difference is:

[0085] For the same satellite Pseudorange observations at the main frequency and secondary frequency and The difference is calculated as follows:

[0086] (1)

[0087] Among them: superscript Indicates the first Each satellite has a subscript 1 indicating the primary frequency and a subscript 2 indicating the secondary frequency.

[0088] As a preferred embodiment, in step S4, the standard normal distribution detection algorithm is used:

[0089] For a certain satellite The detection volume at this epoch Assuming no anomalies, it should conform to the expected value. The variance is normal distribution ,Right now:

[0090] (2)

[0091] Constructing standardized statistics As shown below:

[0092] (3)

[0093] in, , Let be the number of stationary observation samples, and i be the i-th stationary observation sample; with a false alarm rate of When, the corresponding bilateral detection threshold is :

[0094] (4)

[0095] in, is the probability density function of the standard normal distribution;

[0096] The algorithm's decision criteria are expressed as follows:

[0097] (5).

[0098] S4, to Each test metric in the constellation is tested using a standard normal distribution algorithm. Observations with normal test results are assigned to the corresponding set of normal observations for that constellation. Conversely, they are included in the set of anomalous observations corresponding to that constellation. ;

[0099] S5. Determine whether the number of constellation observations in the intact subset at this point meets the requirement: , , ,in: and These are the number of GPS and BDS observations in the normal observation set, respectively; if the requirements are met, proceed to step S6; otherwise, return to the anomaly detection method based on least squares residuals.

[0100] As a preferred embodiment, step S5, the anomaly detection method based on least squares residuals, includes the following steps:

[0101] Pseudorange residuals are typically used for anomaly detection based on least squares residuals. The pseudorange observation equation is as follows:

[0102] (6)

[0103] Where: x is the three-dimensional position change and receiver clock error, y is the pseudorange observation of the current n visible stars, i.e., the pseudorange observation value, H is an n×4 coefficient matrix, and ε is the pseudorange observation noise;

[0104] Based on the least squares method, the least squares solution of the state variables is obtained. :

[0105] (7)

[0106] Where: T represents the matrix transpose operation;

[0107] The deviation vector of the observed quantity can be obtained. :

[0108] (8)

[0109] Least square residual vector w:

[0110] ε(9)

[0111] Construct the residual sum of squares (SSE):

[0112] (10)

[0113] Given a false alarm rate, the following probability equation holds:

[0114] (11)

[0115] in, The probability of a certain random event occurring. The variance of pseudorange observation noise. For degrees of freedom The chi-square distribution density function, Indicates the number of stars currently visible; To set the false alarm rate; it is calculated according to the formula. For the classic RAIM algorithm, a uniform detection value is constructed. At this time, the threshold for:

[0116] (12)

[0117] When the detected quantity V is within the threshold, the observation is considered to be consistent; otherwise, the observation is considered to be abnormal.

[0118] When an anomaly is detected, the observations are cyclically removed until a set of observations below the observation threshold is found, which is considered a reliable set of observations.

[0119] S6, Merge and get ,merge and get ;Will The set as a basic subset, for Perform residual detection to ensure the integrity of the subset again. If the detection passes, proceed to step S7; otherwise, return to the anomaly detection method based on least squares residuals.

[0120] S7, from One pseudorange observation is selected sequentially for residual detection. If the detection passes, the observation is classified as... If an observation fails the test, it is removed from the collection.

[0121] Cycle step S7 to The set average is subjected to residual detection, and the final result is output. Once the location calculation is complete, the identification process ends.

[0122] The above description is only a preferred embodiment of the present invention. It should be noted that for those skilled in the art, several improvements and modifications can be made without departing from the principle of the present invention, and these improvements and modifications should also be considered within the scope of protection of the present invention.

Claims

1. A cascaded satellite navigation anomaly detection and elimination system based on inter-frequency pseudorange deviation distribution, characterized in that, It includes an inter-frequency pseudorange deviation distribution detection module and a positioning residual detection module, wherein the positioning residual detection module includes an anomaly detection module and an anomaly identification module; The inter-frequency pseudorange deviation distribution detection module is used to judge the anomalies of each pseudorange of each satellite in the satellite navigation system and output the anomaly detection results of each pseudorange. The positioning residual detection module is used for overall anomaly detection of the integrity of the satellite navigation system. The anomaly detection module performs chi-square detection on the residual based on the positioning residual to determine whether there is an anomaly at that moment; The anomaly identification module uses the results of the inter-frequency pseudorange deviation distribution detection module to construct an identification subset, thereby achieving fault identification and ensuring the accuracy of fault identification.

2. The cascaded satellite navigation anomaly detection and elimination system based on inter-frequency pseudorange deviation distribution according to claim 1, characterized in that, The input of the inter-frequency pseudorange deviation distribution detection module is the set of pseudorange observations at the main frequency point and the set of pseudorange observations at the secondary frequency point for each satellite in the current epoch, and the output of the inter-frequency pseudorange deviation distribution detection module is a set of pseudorange observation quantum sets that are distinguished as normal and abnormal. The inter-frequency pseudorange deviation distribution detection module includes an inter-frequency pseudorange difference and distribution detection unit, wherein... The inter-frequency pseudorange differential unit performs inter-frequency pseudorange differential processing: in the inter-frequency pseudorange differential stage, the dual-frequency pseudorange observations of each satellite are subtracted to obtain the inter-frequency pseudorange deviation for that epoch. The deviation reflects the variation characteristics of the inter-frequency link error and observation noise, and the detection quantity follows a preset statistical distribution model; the deviation is input as a detection quantity to the distributed detection unit. The distribution detection unit performs a standard normal distribution detection algorithm on the detection quantity based on the statistical characteristics of the observation noise of each satellite to determine whether there are any anomalies in the corresponding observations. This divides the input pseudorange observations into normal subsets and abnormal subsets, providing a reliable data foundation for subsequent anomaly removal and navigation calculation. The anomaly identification module takes as input a set of pseudorange observations classified according to the results of the inter-frequency pseudorange deviation distribution detection module at each epoch, and outputs a set of pseudorange observations that have passed the detection, which can be used for the next stage of localization or fusion. The anomaly identification module first performs residual detection on the entire set of observations. If no anomalies are found, it directly proceeds to the next step of localization calculation. If anomalies are found, it enters the inter-frequency pseudorange deviation distribution detection module for preliminary classification, using the normal subset as the base subset, and cyclically adding anomaly observations for anomaly information detection. Once an anomaly fails the detection, the probability anomaly signal added in this detection is removed; otherwise, the probability anomaly information is considered normal information. This cyclic process continues until all probability anomaly signals have been detected.

3. A method for anomaly detection and removal in cascaded satellite navigation based on inter-frequency pseudorange deviation distribution, characterized in that, Includes the following steps: S1. Read the set of observation information for the current epoch's main frequency point from the satellite observation files. , where: superscript Indicates the first Each satellite, with the subscript 1 indicating the dominant frequency point. For the first Pseudorange observations at the main frequency of a satellite ; S2, Set of observation information for the current epoch dominant frequency point Perform residual detection. If the detection passes, output all observations directly to the positioning solution. If the detection fails, proceed to step S3. S3. Read the set of sub-frequency observation information from the satellite observation file at the current epoch. Where: subscript 2 indicates the sub-frequency point, For the first Pseudorange observations at the secondary frequency of each satellite, and the set of observation information at the current epoch main frequency. and the set of observation information of secondary frequency points in the current epoch. The observations in the process are obtained by performing inter-frequency pseudorange difference. , For the first Pseudorange observations at the main and secondary frequencies of a single satellite and The difference; S4, to Each test metric in the dataset is tested using a standard normal distribution algorithm. Observations with normal test results are assigned to the constellation's corresponding normal observation set. Conversely, they are classified into the set of anomalous observations corresponding to a constellation. ; S5. Determine whether the number of constellation observations in the intact subset at this point meets the requirement: , , ,in: and These are the number of GPS and BDS observations in the normal observation set, respectively; if the requirements are met, proceed to step S6; otherwise, return to the anomaly detection method based on least squares residuals. S6, Merge and get ,merge and get ;Will The set as a basic subset, for Perform residual detection to ensure the integrity of the subset again. If the detection passes, proceed to step S7; otherwise, return to the anomaly detection method based on least squares residuals. S7, from One pseudorange observation is selected sequentially for residual detection. If the detection passes, the observation is classified as... If an observation fails the test, it is removed from the collection. Cycle step S7 to The set average is subjected to residual detection, and the final result is output. Once the location calculation is complete, the identification process ends.

4. The method for cascaded satellite navigation anomaly detection and removal based on inter-frequency pseudorange deviation distribution according to claim 3, characterized in that, In step S3, the inter-frequency pseudorange difference is: For the same satellite Pseudorange observations at the main frequency and secondary frequency and The difference is calculated as follows: （1） Among them: superscript Indicates the first Each satellite has a subscript 1 indicating the primary frequency and a subscript 2 indicating the secondary frequency.

5. The method for cascaded satellite navigation anomaly detection and removal based on inter-frequency pseudorange deviation distribution according to claim 4, characterized in that, In step S4, the standard normal distribution detection algorithm is as follows: For the One satellite, the number of detections at this epoch. Assuming no anomalies, it should conform to the expected value. The variance is normal distribution ,Right now: （2） Constructing standardized statistics As shown below: （3） in, , Let be the number of stationary observation samples, and i be the i-th stationary observation sample; with a false alarm rate of When, the corresponding bilateral detection threshold is : （4） in, is the probability density function of the standard normal distribution; The algorithm's decision criteria are expressed as follows: （5）。 6. The method for cascaded satellite navigation anomaly detection and elimination based on inter-frequency pseudorange deviation distribution according to claim 5, characterized in that, In step S5, the anomaly detection method based on least squares residuals includes the following steps: Pseudorange residuals are typically used for anomaly detection based on least squares residuals. The pseudorange observation equation is as follows: （6） Where: x is the three-dimensional position change and receiver clock error, y is the pseudorange observation of the current n visible stars, i.e., the pseudorange observation value, H is an n×4 coefficient matrix, and ε is the pseudorange observation noise; Based on the least squares method, the least squares solution of the state variables is obtained. : （7） Where: T represents the matrix transpose operation; The deviation vector of the observed quantity can be obtained. : （8） Least square residual vector w: e (9) Construct the residual sum of squares (SSE): （10） Given a false alarm rate, the following probability equation holds: （11） in, The probability of a certain random event occurring. The variance of pseudorange observation noise. For degrees of freedom The chi-square distribution density function, Indicates the number of stars currently visible; To set the false alarm rate; it is calculated according to the formula. For the classic RAIM algorithm, a uniform detection value is constructed. At this time, the threshold for: （12） When the detected quantity V is within the threshold, the observation is considered to be consistent; otherwise, the observation is considered to be abnormal. When an anomaly is detected, the observations are cyclically removed until a set of observations below the observation threshold is found, which is considered a reliable set of observations.

Images