Pseudorange coarse error detection method, electronic device and storage medium

Abstract

A pseudo-range gross error detection method, an electronic device and a storage medium. The method is applied to a Beidou satellite navigation and positioning system, comprising: acquiring satellite observation data of a rover station and a reference station; determining a geometry-free ionosphere-free combined observation value of double-difference pseudo-range for a plurality of frequency point combinations of a satellite signal respectively according to the satellite observation data; performing standard residual error processing on the geometry-free ionosphere-free combined observation value of double-difference pseudo-range; using a DBSCAN clustering algorithm with an adaptive neighborhood radius to perform gross error detection on the standardized residual error values corresponding to the plurality of frequency point combinations respectively, to determine a gross error point satellite set corresponding to each frequency point combination; and filtering gross error points from the satellite observation data according to the gross error point satellite set corresponding to the plurality of frequency point combinations. The scheme provided in the embodiments of the present application can effectively identify gross error points and filter observation data to retain observation data meeting pseudo-range quality requirements, thereby laying a data foundation for subsequent precise positioning.

Description

Technical Field

[0001] This article relates to, but is not limited to, the field of satellite positioning technology, and in particular to a pseudorange gross error detection method, electronic equipment, and storage medium. Background Technology

[0002] Global Navigation Satellite System (GNSS) provides global users with three-dimensional, all-weather, high-quality positioning, navigation, and timing (PNT) services, making it a crucial positioning tool for industries such as transportation, communication, and surveying. Positioning accuracy is one of the core performance indicators of a GNSS system. Relative positioning technology, through inter-station and inter-satellite double-difference processing, effectively eliminates or significantly reduces common errors such as clock bias, hardware delay, orbital error, and atmospheric delay, achieving high-precision positioning calculations and becoming the mainstream technology in the current satellite positioning field.

[0003] Common relative positioning modes include differential positioning based on pseudorange observations, and real-time kinematic (RTK) positioning, which further integrates high-precision carrier phase observations, and its extended form—Network RTK (NRTK). Although pseudorange observations are significantly less accurate than carrier phase observations, they still play an irreplaceable role in relative positioning: on the one hand, they provide initial estimates for resolving carrier phase integer ambiguities; on the other hand, they enhance the observation geometry and improve model strength when the number of visible satellites is limited; simultaneously, pseudorange can be used to construct combined observations to assist in ambiguity fixation and the detection and repair of phase cycle slips. Therefore, the quality of pseudorange observations affects the success rate of ambiguity fixation and the accuracy and reliability of the final positioning results. Improving the quality of observation data through gross error detection is a key step in improving positioning accuracy. Summary of the Invention

[0004] This application provides a pseudorange gross error detection method, electronic device, and storage medium. Based on geometrically and ionospherically inverse combined observations of double-difference pseudoranges from multiple frequency points, the method utilizes the DBSCAN clustering algorithm for gross error detection and then filters the satellite observation data. This effectively identifies gross error points and filters the observation data to retain only those meeting pseudorange quality requirements, laying a data foundation for subsequent precise positioning.

[0005] This application provides a pseudorange gross error detection method, applied to the BeiDou satellite navigation and positioning system, including: Acquire satellite observation data from the rover and reference station, and based on the satellite observation data, determine the geometrically and ionospherically inverse combination observation values ​​of double-difference pseudorange for multiple frequency point combinations of satellite signals; The geometrically and ionospherically unaffected combined observations of the double-difference pseudorange are processed using standardized residuals. The DBSCAN clustering algorithm, which uses density-based spatial clustering with adaptive neighborhood radius, is used to detect gross errors in the standardized residual values ​​corresponding to the multiple frequency point combinations, thereby determining the set of gross error satellites corresponding to each frequency point combination. Based on the set of coarse-impaired satellites corresponding to the multiple frequency point combinations, the satellite observation data is filtered for coarse impairments. Each of the frequency point combinations includes: at least one frequency point in a high-frequency point set and at least one frequency point in a low-frequency point set; the high-frequency point set includes frequency points B1C and B1I, and the low-frequency point set includes frequency points B3I, B2b, B2(a+b), and B2a.

[0006] This application also provides an electronic device, including: one or more processors; and a storage device for storing one or more programs, which, when executed by the one or more processors, cause the one or more processors to implement the pseudorange gross error detection method as described in any embodiment of this disclosure.

[0007] This application also provides a computer-readable storage medium storing a computer program thereon, which, when executed by a processor, implements the pseudorange gross error detection method as described in any embodiment of this disclosure.

[0008] Other features and advantages of this application will be set forth in the following description, and will be apparent in part from the description, or may be learned by practicing the application. Other advantages of this application can be realized and obtained by means of the embodiments described in the description and the accompanying drawings. Attached Figure Description

[0009] The accompanying drawings are used to provide an understanding of the technical solutions of this application and constitute a part of the specification. They are used together with the embodiments of this application to explain the technical solutions of this application and do not constitute a limitation on the technical solutions of this application.

[0010] Figure 1 The flowchart of a pseudorange gross error detection method provided in this disclosure embodiment; Figure 2 A process for gross error detection using the DBSCAN clustering algorithm with adaptive neighborhood radius is provided in this embodiment of the disclosure; Figure 3 Another method for detecting gross errors using the DBSCAN clustering algorithm is provided in this embodiment of the disclosure; Figure 4 This is a schematic diagram of a post-hoc pseudorange differential positioning deviation diagram of the original measured data S001-S002 provided in an embodiment of this disclosure; Figure 5 A schematic diagram of simulated gross error time series plots added to C19 and C20 of S001 provided in an embodiment of this disclosure; Figure 6 This is a schematic diagram of a post-hoc pseudorange differential positioning deviation diagram provided by an embodiment of the present disclosure, consisting of S001 simulated data and S002 original measured data without the use of gross error detection. Figure 7 This is a schematic diagram of a post-hoc pseudorange differential positioning deviation map using gross error detection for S001 simulated data and S002 original measured data, provided in an embodiment of this disclosure. Detailed Implementation

[0011] This application describes several embodiments, but these descriptions are exemplary and not limiting, and it will be apparent to those skilled in the art that many more embodiments and implementations are possible within the scope of the embodiments described herein. Although many possible combinations of features are shown in the drawings and discussed in the detailed description, many other combinations of the disclosed features are also possible. Unless specifically limited, any feature or element of any embodiment may be used in combination with, or may replace, any feature or element of any other embodiment.

[0012] This application includes and contemplates combinations of features and elements known to those skilled in the art. The embodiments, features, and elements disclosed in this application can also be combined with any conventional features or elements to form unique inventive solutions. Any feature or element of any embodiment can also be combined with features or elements from other inventive solutions to form another unique inventive solution. Therefore, it should be understood that any feature shown and / or discussed in this application can be implemented individually or in any suitable combination. Therefore, the embodiments are not limited except by the limitations imposed by the appended claims and their equivalents. Furthermore, various modifications and changes can be made within the scope of the appended claims.

[0013] Furthermore, in describing representative embodiments, the specification may have presented methods and / or processes as a specific sequence of steps. However, the method or process should not be limited to the specific order of steps described herein, to the extent that it does not depend on such a specific order. As will be understood by those skilled in the art, other sequences of steps are also possible. Therefore, the specific order of steps set forth in the specification should not be construed as a limitation of the claims. Moreover, the claims relating to the method and / or process should not be limited to the steps performed in the order written, and those skilled in the art will readily understand that these orders can be varied and still remain within the spirit and scope of the embodiments of this application.

[0014] During satellite signal propagation and reception, pseudorange observations are highly susceptible to various error sources such as multipath effects, ionospheric scintillation, and signal interference, leading to gross errors (outliers). These outliers, mixed into the solution process, severely interfere with ambiguity, prolong convergence time, and can even cause positioning deviations. Therefore, establishing an efficient and robust pseudorange gross error detection mechanism has become a crucial prerequisite for ensuring the availability, continuity, and reliability of high-precision GNSS relative positioning, possessing significant research and engineering value.

[0015] In 2020, the BeiDou-3 Global Navigation Satellite System was officially completed and put into operation, marking the beginning of a new era of global service for the BeiDou system, providing stable and reliable navigation, positioning, and timing services to users around the world. Compared with BeiDou-2, BeiDou-3 not only significantly increased the number of satellites, but also achieved a major upgrade in its signal system: while retaining the original B1I and B3I signals, the B2I signal was eliminated, and four new signals, B1C, B2a, B2b, and B2(a+b), were added.

[0016] Relative positioning technology, through inter-station and inter-satellite double-difference processing, has become the mainstream technique in the current satellite positioning field. In some feasible gross error detection schemes, most methods are based on the residual values ​​obtained during the positioning process. However, since pseudorange observations containing gross errors have already been used in parameter estimation, and there is correlation between satellite observations after double-difference, the gross error observations often affect the residuals of other satellites after adjustment and filtering calculations, thus limiting the accuracy and reliability of the detection. The addition of these new satellites and signals in the BeiDou-3 system has brought new ideas, methods, and system-level solutions to the field of high-precision navigation and positioning.

[0017] The BeiDou-3 satellite system can broadcast signals at six different frequency points: B1C, B1I, B3I, B2b, B2(a+b), and B2a. Based on the relative magnitude of each frequency point, they can be roughly divided into two categories: the first category is a set of higher-frequency signals, including the B1C and B1I frequency points, collectively referred to as the high-frequency signal set in this disclosure, and the corresponding set of frequency points is called the high-frequency point set; the second category is a set of lower-frequency signals, covering the B3I, B2b, B2(a+b), and B2a frequency points, collectively referred to as the low-frequency signal set in this disclosure, and the corresponding set of frequency points is called the low-frequency point set. The classification of BeiDou-3 system frequency points and their frequency information are shown in Table 1 below: Table 1. Frequency Classification and Frequency Information of BeiDou-3 System This application provides a pseudorange gross error detection method applied to the BeiDou satellite navigation and positioning system. It proposes a novel scheme for gross error detection using a combination of multiple frequency points, leveraging the newly added frequency signals of the BeiDou-3 system. The method is as follows: Figure 1 As shown, it includes: Step 110: Obtain satellite observation data from the rover and the reference station. Based on the satellite observation data, determine the geometrically and ionospherically inverse combination observation values ​​of the double-difference pseudorange for multiple frequency point combinations of the satellite signal. Step 120: Perform standardized residual processing on the geometrically and ionospherically insensitive combined observations of the double-difference pseudorange; Step 130: Using the DBSCAN (Density-Based Spatial Clustering of Applications with Noise) clustering algorithm with adaptive neighborhood radius, gross error detection is performed on the standardized residual values ​​corresponding to the multiple frequency point combinations to determine the set of gross error satellites corresponding to each frequency point combination. Step 140: Perform coarse error filtering on the satellite observation data based on the coarse error satellite set corresponding to the multiple frequency point combinations; Each of the frequency point combinations includes: at least one frequency point in a high-frequency point set and at least one frequency point in a low-frequency point set; the high-frequency point set includes frequency points B1C and B1I, and the low-frequency point set includes frequency points B3I, B2b, B2(a+b), and B2a.

[0018] Each frequency combination includes high-frequency and low-frequency points. By using observations from multiple frequency combinations without geometry or ionosphere, and employing clustering algorithms for gross error detection, the detection accuracy and efficiency can be improved.

[0019] In some exemplary embodiments, for the BeiDou-3 system, in order to limit the noise amplification of the pseudorange without geometry and without ionosphere within an acceptable range, each frequency combination includes three frequency points: the first frequency point Second frequency point and the third frequency For example, in combination method one: the first frequency point is a frequency point in the high-frequency point set, and the second and third frequency points are frequency points in the low-frequency point set, that is, a frequency point is selected from the high-frequency point set in Table 1 as... Two frequency points are selected from the low-frequency set as ,total =12 combinations. Alternatively, combination method two: the first frequency point is a frequency point from the low-frequency point set, and the second and third frequency points are frequency points from the high-frequency point set; that is, select one frequency point from the low-frequency set in Table 1 as... Two frequencies are selected from the high-frequency set as ,total =4 combinations.

[0020] In some exemplary embodiments, when each frequency point combination includes 3 frequency points, the plurality of frequency point combinations may include all 16 combinations mentioned above. Alternatively, to reduce computational complexity, some combinations may be selected. In some exemplary embodiments, the frequency points included in the partial combinations may cover all 6 frequency points.

[0021] In some exemplary embodiments, the method further includes step 100, determining the plurality of frequency point combinations based on the frequency points corresponding to the satellite observation data; wherein the second and third frequency points included in the plurality of frequency point combinations cover all frequency points corresponding to the satellite observation data.

[0022] It is understood that the satellite observation data refers to the actual satellite data observed. In the BeiDou Navigation Satellite System, a receiver can typically observe 10-16 satellites. The actual number of satellites observed may vary depending on factors such as the receiver's current location or obstruction. Observation data is generally understood as the core raw data used for satellite positioning calculations, including: pseudorange, carrier phase (referred to as carrier), and Doppler shift (referred to as Doppler), also corresponding to pseudorange observations, carrier phase observations, and Doppler shift observations; or, it may also include: navigation messages, including: ephemeris, almanacs, etc. Optionally, other data may be included in the observation data as needed, not limited to the aspects listed above. More detailed aspects are not listed here. It is understood that due to factors such as the number of satellites and signal quality, the receiver may only receive signals from some of the six frequency points. For example, the total number of frequency points corresponding to the satellite observation data may be 5 or 4. In this case, multiple frequency point combinations are constructed within the range of these 5 or 4 frequency points.

[0023] In some exemplary embodiments, the method further includes step 100, determining the multiple frequency point combinations based on the frequency points corresponding to the satellite observation data and required for positioning calculation; wherein the second and third frequency points included in the multiple frequency point combinations cover all frequency points corresponding to the satellite observation data and required for positioning calculation. That is, the multiple frequency point combinations determined in step 100 are selected from all frequency points corresponding to the satellite observation data, based on the needs of positioning calculation, and multiple frequency point combinations are constructed based on the selected frequency points. For example, based on the current receiver computing resource load, memory status, or the complexity of the positioning calculation algorithm, some frequency points are selected from all frequency points corresponding to the observation data. For example, the satellite observation data includes data from all 6 frequency points, but during the positioning calculation process, considering CPU load and memory consumption, B1C, B1I, B3I, and B2a are selected for positioning calculation, and multiple frequency point combinations are constructed based on these 4 frequency points, each combination including 3 frequency points.

[0024] In some exemplary embodiments, the number of frequency point combinations determined in step 100 is the minimum number of combinations that satisfy the condition that the second and third frequency points included in the frequency point combination cover all the selected frequency points. For example, if satellite observation data corresponds to 6 frequency points, and the positioning calculation uses all 6 frequency points, then 3 frequency point combinations are determined: (B1I, B3I, B2b), (B1I, B2(a+b), B2a), and (B2a, B1C, B1I); or, 3 frequency point combinations: (B1C, B3I, B2b), (B1C, B2(a+b), B2a), and (B2a, B1C, B1I), where the second and third frequency points included in the 3 frequency point combinations cover all 6 frequency points. As another example, if satellite observation data corresponds to 6 frequency points, and the positioning calculation uses 5 of them, then the second and third frequency points included in the 3 frequency point combinations are determined to cover all 5 frequency points. More examples are not listed here.

[0025] In some exemplary embodiments, step 110, based on the satellite observation data, determines geometrically and ionospherically inverse combination observations of the double-difference pseudorange for multiple frequency point combinations of the satellite signal, including: Based on the satellite observation data, for each satellite, the geometrically and ionospherically indeterminate pseudorange combined observation value for each frequency combination is determined using the following method. GFIF : ; in, This represents the double difference operator between stations and between satellites; u，r These represent the rover and the reference station, respectively. i, jThese represent non-reference stars and reference stars, respectively. Represents pseudorange observations. These represent the three frequency points included in the frequency point combination. These represent the double-difference pseudorange observations at three frequency points. The combination coefficients; ; These are the frequencies of the three frequency points.

[0026] It can be seen that the observation equations after double difference between stations and between satellites are as follows: (1); in: This represents the double difference operator between stations and between satellites; u，r These represent the rover and the reference station, respectively. i, j These represent non-reference stars and reference stars, respectively. Indicates the frequency point number; This represents pseudorange observations, in meters. Represents frequency-independent geometric terms, including receiver-to-satellite distance, tropospheric delay, ephemeris error, etc., in meters; Indicates the first The ionospheric coefficient at each frequency point, i.e.: ( (representing frequency); This represents the ionospheric delay term at the first frequency, in meters. The noise term representing pseudodistance, in meters.

[0027] when At that time, the double-difference pseudorange observations at three frequency points It is possible to construct geometrically and ionospherically free combined observations with double-difference pseudoranges. The formula is as follows: (2); in, This represents the noise term of the double-difference pseudorange combination without geometry and ionospheric delay. It can be seen that after eliminating both the geometric correlation term and the ionospheric delay term, only the combined noise remains in a combination of three frequency points. That is, the double-difference pseudorange observations without geometry and ionospheric delay at the three frequency points. GFIF Also known as the combined residual term, it is equal to the noise term of the double-difference pseudo-moment combination without geometry or ionosphere. In other words, the observations of the dual-difference pseudorange without geometry or ionosphere are... GFIF It only includes combined noise terms.

[0028] As can be seen, based on the calculation method provided in this application, it is possible to calculate the geometrically and ionospherically unaffected combined observations of double-difference pseudoranges without performing parameter estimation or adjustment calculations. This significantly reduces computational complexity and avoids the impact of gross errors on other observations during parameter estimation and adjustment.

[0029] Three combination coefficients as follows: (3); Assuming the noise standard deviation is equal at the three frequencies, that is: According to the error propagation law, the noise standard deviation of the double-difference pseudorange geometrically and ionosphere-free combination can be obtained. : (4); set up This is called the noise amplification factor of a geometrically and ionosphere-free combination.

[0030] Table 2 shows the 12 3-frequency combinations, combination coefficients, and noise amplification coefficients obtained from Method 1: Table 2. Frequency combinations, combination coefficients, and noise amplification coefficients for Method 1. The four 3-frequency combinations, combination coefficients, and noise amplification coefficients obtained from Method 2 are shown in Table 3: Table 3. Frequency combinations, combination coefficients, and noise amplification coefficients for Method 2. By observing Tables 2 and 3, we can see that various combination types... Since the coefficients are relatively small, when the gross error of the pseudorange at the first frequency point is small, it is difficult to detect using a single combination. In other words, the coefficient of the first frequency point in the GFIF combination is small, and it is not sensitive to small gross errors, so multiple combinations are needed to compensate for each other. For example, (B1I, B3I, B2a) is sensitive to gross errors at frequencies B3I and B2a, but not to gross errors at B1I; (B1a, B1C, B1I) is sensitive to gross errors at frequencies B1C and B1I, but not to gross errors at B1a. Combining these two combinations to judge gross errors makes them sensitive to B1C, B1I, B3I, and B2a, making it easy to detect gross errors. Therefore, based on the frequency points included in the actual observation data, multiple geometrically and ionospherically free combinations are appropriately selected from the entire set of possible combinations. Their complementary characteristics are used to achieve synchronous joint detection of gross errors at all frequency points of the BeiDou-3 satellite, effectively improving the system's coverage of multi-frequency anomalous signals and increasing the efficiency of gross error detection.

[0031] In some exemplary embodiments, step 120 involves standardizing the geometrically and ionospherically unaffected combined observations of the double-difference pseudorange, including: For each frequency combination of each satellite, the residuals are standardized using the following method: (5); in, These are the standardized residuals of double-difference pseudorange observations with no geometry and no ionosphere. GFIF These are observations of a geometrically and ionospherically incompatible combination of double-difference pseudoranges. The noise standard deviation of the geometrically and ionosphere-free combination of double-difference pseudo-ranges; (4); This represents the noise variance of the double-difference pseudorange for each satellite.

[0032] In some exemplary embodiments, the noise variance of the double-difference pseudorange of each satellite is calculated based on an elevation angle stochastic model.

[0033] In some exemplary embodiments, step 130 employs the DBSCAN clustering algorithm with adaptive neighborhood radius to perform gross error detection on the standardized residual values ​​corresponding to the multiple frequency point combinations, including: For each frequency combination, perform the following steps for gross error detection: Using the normalized residual value of the double-difference pseudorange of each satellite under this frequency combination as a data point, a distance matrix of multiple data points is constructed. Based on the distance matrix, the neighborhood radius of the DBSCAN clustering algorithm is determined using the K-distance graph method. For all data points corresponding to all satellites under this frequency combination, the DBSCAN clustering algorithm is executed to detect gross errors and identify gross error points.

[0034] In some exemplary embodiments, step 140 includes: using the satellites included in the union of the coarse difference satellite sets corresponding to the multiple frequency point combinations as the filtering target satellites; Data from the target satellite being filtered out is removed from the satellite observation data.

[0035] It is known that each identified gross error point corresponds to a satellite, and multiple gross error points correspond to satellites that constitute a gross error satellite set, including at least one satellite. Step 140 involves taking the union of the sets of gross error points (i.e., outliers) marked for each frequency combination, and integrating the gross error point (i.e., outlier) satellites marked for different frequency combinations. After removing data from these satellites from the satellite observation data, satellite observation data with pseudorange quality that meets the requirements is obtained. Pseudorange data from satellite observation data with pseudorange quality that meets the requirements can be used for pseudorange differential positioning, calculating ambiguity based on pseudorange observation values, and real-time dynamic positioning including pseudorange equations.

[0036] In some exemplary embodiments, step 110 includes: Step 110-1: Obtain pseudorange observations of BeiDou-3 satellites from the rover and reference stations. Using formula (6), construct inter-station and inter-satellite double-difference pseudorange observations for each satellite. (6); Step 110-2: Based on the number of frequency points of the double-difference observations and the needs of positioning calculation, construct the geometrically and ionospherically free combination observations of multiple (>=2) frequency point combinations for each satellite according to formula (2). GFIF For example, three frequency combinations: (B1I, B3I, B2b), (B1I, B2(a+b), B2a), and (B2a, B1C, B1I).

[0037] In some exemplary embodiments, step 120 includes: determining the noise variance based on the elevation angle stochastic model, and performing standardized residual processing on the geometrically and ionospherically unspatial combined observations of the double-difference pseudorange.

[0038] Among them, the noise variance of the double-difference pseudorange of each satellite is calculated using the formula (7) of the elevation angle stochastic model; (7); in, Represents the constant error term; This represents the error term related to the elevation angle; The elevation angle is in radians. For example, =0.3m, =0.3m.

[0039] The noise standard deviation of geometrically non-ionospheric combinations with different double-difference pseudo-magnetic moments is calculated according to formula (4). .

[0040] Based on formula (5), the geometrically and ionospherically insensitive combination observations of the double-difference pseudorange for different frequency combinations of each satellite are obtained. GFIFStandardized residual processing is performed to obtain the standardized residual values ​​of the geometrically and ionospherically unaffected combined observations of the double-difference pseudo-range. .

[0041] Based on this, for each observed satellite and its corresponding combination of three frequency points, the standardized residual values ​​of the geometrically and ionospherically indeterminate pseudorange observations were calculated. .

[0042] Accordingly, step 130 includes: for each frequency combination, using the DBSCAN clustering algorithm with adaptive neighborhood radius to detect gross errors in the standardized residual values ​​of the double-difference pseudorange without geometry or ionosphere. In this step, a frequency combination includes the standardized residual values ​​of the double-difference pseudorange without geometry or ionosphere of multiple satellites, and gross errors need to be detected sequentially for the standardized residual values ​​of the double-difference pseudorange without geometry or ionosphere of multiple satellites in each frequency combination.

[0043] For each frequency combination, such as Figure 2 As shown, perform the following steps: Step 130-1: Using the normalized residual value of the double-difference pseudorange of each satellite under this frequency point combination without geometry and without ionosphere as a data point, construct a distance matrix of multiple data points; The distance matrix includes the distances from each data point to all data points. The expression is as follows: (8); in, This represents the data points corresponding to the 1st satellite, the 2nd satellite, ..., the nth satellite, i.e., the combined standardized residual values. Let be the distance between the i-th satellite and the j-th satellite, where i, j = 1, ..., n.

[0044] Step 130-2: Set the minimum number of data points within the neighborhood. For example, the minimum number of data points in the neighborhood can be set to twice the data dimension. Since the data points are one-dimensional, the minimum number of data points in the neighborhood can be set to 2, i.e. .

[0045] Step 130-3, based on the distance matrix The K-distance graph method is used to determine the neighborhood radius of the DBSCAN clustering algorithm; that is, the neighborhood radius of DBSCAN is obtained by finding inflection points based on drawing a k-distance graph, including: Step 130-3-1: Using the distance matrix, sort the distances in each row from smallest to largest. Each data point is stored in a vector. middle; Step 130-3-2, convert the vector Sort the elements in the vector in ascending order to obtain the vector. ; Step 130-3-3, calculate the vector according to formula (9). The second-order difference vector ; (9); in, Representing vectors The m-th element, Contains n elements, It contains n-2 elements.

[0046] Step 130-3-4, Compare Find the position corresponding to the largest element among all elements. ; Step 130-3-5, determine the vector The Each element is an initial value of the neighborhood radius. ; Step 130-3-6: Determine a more reliable and robust final value for the neighborhood radius according to formula (10). .

[0047] (10); in, This represents the lower bound of the neighborhood radius. This represents the upper limit of the neighborhood radius.

[0048] Step 130-4: For all data points corresponding to all satellites under this frequency point combination, execute the DBSCAN clustering algorithm to detect gross errors and identify gross error points.

[0049] Among them, the DBSCAN clustering algorithm is used for coarse error detection, such as Figure 3 As shown, it includes: Step 130-4-1: Mark all data points as unvisited. Step 130-4-2: Select one of the unvisited data points. Mark as visited, and calculate the neighborhood radius of the data point. Inside points ; Step 130-4-3, if the neighborhood radius of the data point Inside points Mark data points Create a new cluster with the core focus. Continue to step 130-4-4 to sub-step 130-4-5; if Mark data points This is a gross error (i.e., an anomaly), and the process returns to step 130-4-2; Step 130-4-4, regarding the core points The neighborhood points are recursively expanded until the cluster is formed. It cannot be expanded further; Step 130-4-5: Repeat steps 130-4-2 to 130-4-5 until all points have been visited.

[0050] As can be seen, by applying the DBSCAN clustering algorithm based on machine learning, the standardized residual values ​​of the combined observations are used as a set of data points for outlier detection. This can identify gross errors in multiple satellites at once, overcoming the limitations of traditional methods that process each satellite individually or can only detect one error at a time. It has the ability to detect multiple (points) gross errors simultaneously, thus enhancing the anti-interference capability against multiple gross errors.

[0051] In some exemplary embodiments, two stations S001 and S002, approximately 60 km apart, are used. S001 is used as the rover station and S002 as the reference station for pseudorange differential localization based on post-simulation gross data. Since the original actual observation data of S001 and S002 do not include data for B2(a+b), three different pseudorange frequency combinations without geometry or ionosphere are selected here: (B2a, B1C, B1I), (B1C, B3I, B2a), and (B1C, B2b, B2a).

[0052] Figure 4 This is a post-hoc pseudorange differential positioning deviation diagram using the original measured data from stations S001 and S002.

[0053] Figure 5 To add simulated gross error time series plots to the B1I, B2a, and B3I frequency points of the C19 and C20 satellite measured data at rover S001 every 5 minutes.

[0054] Figure 6 To utilize the simulated gross error data from S001 and the original measured data from S002, a post-hoc pseudorange differential positioning deviation map is used without using gross error detection.

[0055] Figure 7 To utilize the simulated gross error data in S001 and the original measured data in S002, the post-hoc pseudorange differential positioning deviation map of the gross error detection described in this patent is used.

[0056] contrast Figure 6 and Figure 4Without using gross error detection, a "flying point" phenomenon occurs when gross errors are added in the simulation, indicating that the simulated pseudorange gross errors cause deviations in the positioning results, resulting in positioning accuracy that does not meet requirements. (Comparison) Figure 7 and Figure 6 After using the gross error detection method provided in this application embodiment, the "flying point" at the moment when gross errors are added in the simulation no longer exists; then compare... Figure 7 and Figure 4 The results showed that the accuracy of the two methods was basically the same, indicating that the pseudorange quality gross error detection method for BeiDou-3 relative positioning provided in this application is feasible and effective.

[0057] This application also provides an electronic device, including: one or more processors; and a storage device for storing one or more programs, which, when executed by the one or more processors, cause the one or more processors to implement the pseudorange gross error detection method as described in any embodiment of this application.

[0058] In some exemplary embodiments, the electronic device includes: a BeiDou satellite navigation signal receiver, or a BeiDou satellite navigation smart terminal, or a BeiDou satellite navigation vehicle terminal, etc.

[0059] This application also provides a computer-readable storage medium storing a computer program thereon, which, when executed by a processor, implements the pseudorange gross error detection method as described in any embodiment of this application.

[0060] This application proposes a gross error detection method based on multiple double-difference pseudorange frequency point combinations without geometry or ionosphere, coupled with an adaptive DBSCAN clustering algorithm, for observation data available from the BeiDou-3 system at multiple frequencies. Utilizing the complementary characteristics among the combined observations of multiple double-difference pseudorange frequency points without geometry or ionosphere, the method first constructs geometrically and ionosphere-free observations of different combinations of inter-station and inter-satellite double-difference pseudorange. Then, it performs standardized residual processing on the combined observations and employs an adaptive neighborhood radius DBSCAN clustering algorithm to accurately detect satellites with gross errors. This method can effectively identify gross errors on multiple satellites and multiple frequencies simultaneously, improving detection efficiency and accuracy. When using the observation data after removing gross errors for relative positioning, it can accelerate the positioning convergence process, improve the accuracy of the fixed solution and the overall positioning reliability, and further enhance the system's stability and positioning performance.

[0061] It will be understood by those skilled in the art that all or some of the steps, systems, or apparatuses disclosed above, and their functional modules / units, can be implemented as software, firmware, hardware, or suitable combinations thereof. In hardware implementations, the division between functional modules / units mentioned above does not necessarily correspond to the division of physical components; for example, a physical component may have multiple functions, or a function or step may be performed collaboratively by several physical components. Some or all components may be implemented as software executed by a processor, such as a digital signal processor or microprocessor, or as hardware, or as an integrated circuit, such as an application-specific integrated circuit (ASIC). Such software may be distributed on a computer-readable medium, which may include computer storage media (or non-transitory media) and communication media (or transient media). As is known to those skilled in the art, the term "computer storage medium" includes volatile and non-volatile, removable and non-removable media implemented in any method or technology for storing information (such as computer-readable instructions, data structures, program modules, or other data). Computer storage media include, but are not limited to, RAM, ROM, EEPROM, flash memory or other memory technologies, CD-ROM, digital versatile disc (DVD) or other optical disc storage, magnetic cartridges, magnetic tape, disk storage or other magnetic storage devices, or any other medium that can be used to store desired information and can be accessed by a computer. Furthermore, it is well known to those skilled in the art that communication media typically contain computer-readable instructions, data structures, program modules, or other data in modulated data signals such as carrier waves or other transmission mechanisms, and may include any information delivery medium.

[0062] Although embodiments of this application have been shown and described above, it is understood that the above embodiments are exemplary and should not be construed as limiting this application. Those skilled in the art can make changes, modifications, substitutions and variations to the above embodiments within the scope of this application.

Claims

1. A pseudorange gross error detection method, applied to the BeiDou satellite navigation and positioning system, characterized in that, include: Acquire satellite observation data from the rover and reference station, and based on the satellite observation data, determine the geometrically and ionospherically inverse combination observation values ​​of double-difference pseudorange for multiple frequency point combinations of satellite signals; The geometrically and ionospherically unaffected combined observations of the double-difference pseudorange are processed using standardized residuals. The DBSCAN algorithm, a density-based spatial clustering algorithm with adaptive neighborhood radius, is used to detect gross errors in the standardized residual values ​​corresponding to the multiple frequency point combinations, thereby determining the set of gross error satellites corresponding to each frequency point combination. Based on the set of coarse-impaired satellites corresponding to the multiple frequency point combinations, the satellite observation data is filtered for coarse impairments. Each of the frequency point combinations includes: a first frequency point, a second frequency point, and a third frequency point; The first frequency point is a frequency point in the high-frequency point set, and the second and third frequency points are frequency points in the low-frequency point set. The high-frequency point set includes at least one of the following frequency points: B1C and B1I, and the low-frequency point set includes at least two of the following frequency points: B3I, B2b, B2ab, and B2a; or, the first frequency point is a frequency point in the low-frequency point set, and the second and third frequency points are frequency points in the high-frequency point set. The high-frequency point set includes the frequency points B1C and B1I, and the low-frequency point set includes at least one of the following frequency points: B3I, B2b, B2ab, and B2a. The density-based spatial clustering algorithm with noisy features, employing adaptive neighborhood radius, performs gross error detection on the standardized residual values ​​corresponding to the multiple frequency point combinations, including: For each frequency combination, perform the following steps for gross error detection: The standardized residual value of the double-difference pseudorange of each satellite under this frequency point combination is taken as a data point. For the data points corresponding to all satellites under this frequency point combination, the DBSCAN algorithm with the adaptive neighborhood radius is used to detect anomalies, realize gross error detection, and identify the anomaly points as gross error points.

2. The method according to claim 1, characterized in that, The second and third frequency points included in the combination of multiple frequency points cover all frequency points corresponding to the satellite observation data; or, The second and third frequency points included in the combination of multiple frequency points cover the satellite observation data and are all the frequency points required for multi-frequency point positioning calculation.

3. The method according to claim 1 or 2, characterized in that, The step of determining geometrically and ionospherically inverse pseudorange combinations of observations based on the satellite observation data for multiple frequency point combinations of satellite signals includes: Based on the satellite observation data, for each satellite, the geometrically and ionospherically indeterminate pseudorange combined observation value for each frequency combination is determined using the following method. GFIF : ； in, This represents the double difference operator between stations and between satellites; u，r These represent the rover and the reference station, respectively. i, j These represent non-reference stars and reference stars, respectively. Represents pseudorange observations. These represent the three frequency points included in the frequency point combination. These represent the double-difference pseudorange observations at three frequency points. The combination coefficients; ； These are the frequencies of the three frequency points.

4. The method according to claim 3, characterized in that, The standardization of residuals for the geometrically and ionospherically insensitive combined observations of the double-difference pseudorange includes: For each frequency combination of each satellite, the residuals are standardized using the following method: ； in, These are the standardized residuals of double-difference pseudorange observations without geometry or ionosphere. GFIF These are observations of a geometrically and ionospherically incompatible combination of double-difference pseudoranges. The noise standard deviation of the geometrically undecoupled and ionosphere-free combination of double-difference pseudo-ranges; ； This represents the noise variance of the double-difference pseudorange for each satellite.

5. The method according to claim 4, characterized in that, The noise variance of the double-difference pseudorange for each satellite is calculated based on the elevation angle stochastic model.

6. The method according to claim 1 or 2, characterized in that, The density-based spatial clustering algorithm with noisy features, employing adaptive neighborhood radius, performs gross error detection on the standardized residual values ​​corresponding to the multiple frequency point combinations, including: For each frequency combination, perform the following steps for gross error detection: Using the standardized residual value of the double-difference pseudorange of each satellite under this frequency combination as a data point, a distance matrix of multiple data points is constructed. Based on the distance matrix, the neighborhood radius of the DBSCAN algorithm is determined using the K-distance graph method. For all data points corresponding to all satellites under this frequency combination, the DBSCAN algorithm is executed to detect gross errors and identify gross error points.

7. The method according to any one of claims 1-2, characterized in that, The step of filtering the satellite observation data for coarse errors based on the set of coarse error satellites corresponding to the multiple frequency point combinations includes: The satellites included in the union of the coarse difference satellite sets corresponding to the multiple frequency point combinations are used as the filtering target satellites; Data from the target satellite being filtered out is removed from the satellite observation data.

8. An electronic device, characterized in that, include: One or more processors; Storage device for storing one or more programs. When the one or more programs are executed by the one or more processors, the one or more processors implement the pseudorange gross error detection method as described in any one of claims 1-7.

9. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the program is executed by the processor, it implements the pseudorange gross error detection method as described in any one of claims 1-7.

Images