A CORS network-based integrity monitoring method for an RTK server end

Abstract

The application discloses a CORS network-based RTK server integrity monitoring method, and relates to the field of satellite navigation, which comprises the following steps: acquiring observation data obtained by a plurality of reference stations in a same region in a CORS network in each epoch in an observation time period; performing fault monitoring on receiver faults, signal quality abnormalities and star clock abnormalities in the observation data of each reference station based on a distributed monitoring scheme and a hierarchical monitoring idea, determining a fault discrimination matrix, comprehensively performing joint discrimination on the fault discrimination matrixes of each reference station, monitoring ephemeris faults, reference station position faults, regional ionospheric abnormalities and regional tropospheric abnormalities by using multi-station observation data, obtaining a first-level monitoring result, selecting a reference station and adjacent reference stations according to a user's approximate position to determine initial differential correction values, monitoring comprehensive faults contained in the observation data of each reference station, and determining differential correction values and integrity parameters to be broadcast to users finally, so that the RTK server integrity can be effectively improved.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of satellite navigation, in particular to a CORS network-based RTK server integrity monitoring method. BACKGROUND

[0002] Providing high-precision and high-integrity positioning services is the development direction of satellite navigation systems for future industry applications. The demand for high-precision and high-integrity positioning technology is increasing in the fields of autonomous driving, unmanned aerial vehicles, maritime affairs, and railways. Real-Time Kinematic (RTK) is the most mature and widely used Global Navigation Satellite System (GNSS) high-precision positioning technology, which can achieve centimeter-level high-precision positioning and has good dynamic characteristics. Compared with Precise Point Positioning (PPP) and PPP-RTK, RTK has irreplaceable advantages and is currently used by most autonomous driving vehicle users. However, the inherent vulnerability of satellite navigation systems limits their application in life safety fields, so the study of RTK integrity is crucial to determine whether it can be applied in autonomous driving and other fields.

[0003] Currently, there are two aspects of RTK integrity research. RTK system operators are actively responding to user demands and promoting the construction of high-precision positioning service performance specifications, but they are still in discussion and have not established a complete theoretical system at the system level to solve the application problems of RTK in autonomous driving. The academic community's research on RTK integrity mainly focuses on the improvement and optimization of user-end algorithms, including fault monitoring and identification methods, protection level (PL) optimization, and integrity monitoring related to the whole-week ambiguity fixing.

[0004] However, current research on RTK integrity remains at the conceptual discussion and user-end algorithm level, with unclear conceptual index definitions and complex algorithms that are not universally applicable. Moreover, there is a lack of systematic analysis framework and solutions, which cannot truly promote the practical application of RTK in life safety fields.

[0005] From the perspective of system construction, detailed and comprehensive integrity monitoring at the server end is the core of ensuring the integrity of corrections, and is also the blind spot in current research and technology. High-reliability corrections and integrity parameters are the key to ensuring positioning integrity. Therefore, there is an urgent need for an RTK server integrity improvement technology. SUMMARY

[0006] The application aims to provide a CORS network-based RTK server integrity monitoring method, which can effectively improve the RTK server integrity.

[0007] To achieve the above-mentioned purpose, the application provides the following solutions.

[0008] In a first aspect, the application provides a CORS network-based RTK server integrity monitoring method, which comprises the following steps.

[0009] Obtaining observation data obtained by a plurality of reference stations in a same region in a CORS network in each epoch in an observation time period; the observation data comprises satellite ephemeris and pseudo-range carrier measurement values.

[0010] Based on a distributed monitoring scheme and a hierarchical monitoring idea, receiver faults, signal quality abnormalities and star clock abnormalities in the observation data of each reference station are monitored, and a fault discrimination matrix of each type of fault is determined.

[0011] Based on a data processing center, joint discrimination is carried out by comprehensively using the fault discrimination matrix corresponding to each reference station, and satellite ephemeris faults, reference station position faults, regional ionospheric abnormalities and regional tropospheric abnormalities are monitored using multi-station observation data, so as to obtain a first-level monitoring result; the first-level monitoring result is a discrimination result for each single fault.

[0012] Based on the first-level monitoring result, a reference station and a nearby reference station are selected according to a user's approximate position, and an initial differential correction number is determined; the initial differential correction number comprises a differential correction number from the reference station to the user and a differential correction number from the reference station to the nearby reference station.

[0013] According to the initial differential correction number, comprehensive faults contained in the observation data of each reference station are monitored, and a final differential correction number and an integrity parameter to be broadcast to the user are determined.

[0014] According to the specific embodiments provided by the application, the following technical effects are disclosed.

[0015] The application adopts a distributed monitoring scheme and a hierarchical monitoring idea to carry out abnormality monitoring. In the distributed monitoring scheme, each reference station independently obtains observation information and carries out partial fault monitoring, i.e. single fault monitoring, and through a data processing center, observation data and monitoring results of each reference station are comprehensively used to comprehensively discriminate faults, so as to effectively discriminate satellite segment faults, propagation segment faults and ground segment faults. In addition, using the hierarchical monitoring idea, after the monitoring of each independent fault is completed, a comprehensive fault source in the observation data is monitored again, so as to exclude the case that multiple small faults are superimposed to cause a large fault, and to ensure that the generated differential correction number and integrity parameter meet the required performance indicators, thereby effectively improving the RTK server integrity. BRIEF DESCRIPTION OF DRAWINGS

[0016] In order to more clearly illustrate the technical solutions in the embodiments of the present application or the prior art, the drawings needed to be used in the embodiments will be briefly introduced. Obviously, the drawings in the following description only constitute some embodiments of the present application, and for those skilled in the art, other drawings can be obtained from these drawings without creative labor.

[0017] Figure 1 The flow chart of the RTK server integrity monitoring method based on the CORS network provided by the present application;

[0018] Figure 2 The flow chart of the reference station hierarchical integrity monitoring provided by the present application;

[0019] Figure 3 The flow chart of the first level monitoring of the RTK server provided by the present application;

[0020] Figure 4 The flow chart of the second level monitoring of the RTK server provided by the present application;

[0021] Figure 5 The flow chart of the signal quality abnormality joint monitoring provided by the present application;

[0022] Figure 6 The flow chart of the ARS star clock abnormality monitoring provided by the present application;

[0023] Figure 7 The flow chart of the Type A ephemeris failure abnormality monitoring provided by the present application;

[0024] Figure 8 The flow chart of the Type B ephemeris failure abnormality monitoring provided by the present application;

[0025] Figure 9 The flow chart of the reference station position abnormality monitoring provided by the present application;

[0026] Figure 10 The flow chart of the regional troposphere residual abnormality monitoring provided by the present application;

[0027] Figure 11 The flow chart of the ionosphere abnormality monitoring provided by the present application. DETAILED DESCRIPTION

[0028] With reference to the drawings of the embodiments of the present application, the technical solutions in the embodiments of the present application will be described clearly and completely. Obviously, the described embodiments are only a part of the embodiments of the present application, but not all the embodiments of the present application. Based on the embodiments of the present application, all the other embodiments obtained by a person of ordinary skill in the art without creative work fall within the scope of the present application.

[0029] The above-mentioned purposes, features and advantages of the present application will be more apparent and understandable, and the present application will be described in further detail with reference to the drawings and specific embodiments.

[0030] Embodiment 1

[0031] The embodiments of the present application provide a CORS network-based RTK server integrity monitoring method. The method is executed by a computer device, specifically, can be executed by a terminal or a server or other computer device alone, or can be executed by a terminal and a server together. In the embodiments of the present application, as shown in the method includes the following steps. Figure 1

[0032] S1: acquiring observation data acquired by a plurality of reference stations in a same region CORS network in each epoch in an observation time period; the observation data includes satellite ephemeris and pseudo-range carrier measurement values.

[0033] S2: based on a distributed monitoring scheme and a hierarchical monitoring idea, monitoring receiver faults, signal quality abnormalities and star clock abnormalities in the observation data of each reference station, and determining fault discrimination matrices of various faults.

[0034] S3: based on a data processing center, comprehensively conducting joint discrimination based on the fault discrimination matrices corresponding to each reference station, and using multi-station observation data to monitor ephemeris faults, reference station position faults, regional ionospheric abnormalities and regional tropospheric abnormalities, to obtain a first-level monitoring result; the first-level monitoring result is a discrimination result for each single fault.

[0035] S4: based on the first-level monitoring result, selecting a reference station and a nearby reference station according to a user's approximate position, and determining initial differential corrections; the initial differential corrections include differential corrections from the reference station to the user and differential corrections from the reference station to the nearby reference station.

[0036] S5: based on the initial differential corrections, monitoring comprehensive faults contained in the observation data of each reference station, and determining differential corrections and integrity parameters to be finally broadcast to the user.

[0037] In actual application, as Figure 2 ​As shown, the present application comprises first-level monitoring and second-level monitoring; wherein the first-level monitoring, i.e. single fault monitoring: for each epoch in the observation time period, each reference station independently carries out fault monitoring for receiver faults, abnormal carrier-to-noise ratio and abnormal satellite clock according to the following steps 101-110, outputs the fault judgment matrix and observation data to the data processing center, and the data processing center gives the final judgment result by comprehensively analyzing the fault judgment matrix of each station. In addition, the data processing center obtains relevant parameters by constructing baseline combination to carry out ionospheric anomaly, tropospheric anomaly and reference station position anomaly monitoring.

[0038] In the single fault monitoring, a distributed monitoring idea is adopted, as shown in Figure 4 , which comprises:

[0039] Step 101: For each epoch in the observation time period, each reference station independently obtains satellite observation data, including ephemeris information, different frequency carrier observation values and pseudo-range observation values. On this basis, a fault judgment matrix is established, wherein each row represents a satellite and each column represents different frequency carrier or pseudo-range observation data. In this embodiment, global positioning system (GPS) and Beidou satellite navigation system (BDS) dual-system dual-frequency observation values are selected, the size of the fault judgment matrix is 96x4, and the initial value of the fault judgment matrix is 0.

[0040] Step 102: Each reference station independently carries out signal quality anomaly judgment according to the carrier-to-noise ratio data. Each reference station first carries out signal quality anomaly monitoring according to the obtained satellite and frequency observation data. For the observation satellite and frequency whose carrier-to-noise ratio is less than the threshold value, an abnormal carrier-to-noise ratio identifier is added to the corresponding position of the fault judgment matrix established in step 101, and in the subsequent processing process, the observation data is not used. The threshold value is established by long-term historical carrier-to-noise ratio data statistics, reflecting the lower limit of the normal carrier-to-noise ratio.

[0041] Step 103: Each reference station independently carries out cycle slip detection. After excluding the observation data with abnormal carrier-to-noise ratio, for each observation satellite in step 102, a no-geometry test statistic T GF is constructed using dual-frequency carrier observation data φ1, φ2, and a MW (Melbourne-Wunnema) test statistic T MW is constructed using dual-frequency carrier and pseudo-range observation data P1, P2 to carry out cycle slip detection. The calculation methods of the above two test statistics are shown in formulas (1) and (2), and the threshold size is obtained according to the experience value and historical data.

[0042] T GF= λ1φ1(t) - λ2φ2(t) - [λ1φ1(t-1) - λ2φ2(t-1)] (1)

[0043]

[0044] After detecting the cycle slip, further distinguish the cycle slip frequency point, and repair the cycle slip.

[0045] Step 104: Each reference station independently performs pseudorange gross error detection. Calculate the pseudorange observation value correction number of each observation channel at the current epoch, use the pseudorange correction number of n epochs (from epoch k-n+1 to current epoch k) to perform quadratic polynomial fitting, obtain the acceleration, slope and step error test statistics for subsequent discrimination, wherein the threshold limit is obtained by long-term historical data statistics.

[0046] If the three test statistics of the observation channel of satellite i frequency point j exceed the threshold limit number is greater than or equal to 2, it is determined that the measured pseudorange has gross error, the pseudorange gross error fault identifier is added in the fault discrimination matrix established in step 101, and in the subsequent processing flow, the observation data is not used.

[0047] Step 105: Each reference station independently performs star clock anomaly monitoring. Calculate the carrier observation value correction number of each observation channel at the current epoch, use the carrier correction number of n epochs (from epoch k-n+1 to current epoch k) to perform quadratic polynomial fitting, obtain the acceleration, slope and step error test statistics, for subsequent discrimination, wherein the threshold limit is obtained by long-term historical data statistics. If for satellite i frequency point j observation channel: if the number of threshold limits is greater than or equal to 2, mark the satellite clock anomaly in the fault discrimination matrix established in step 101; if the number of threshold limits is 1, mark the alarm identifier as 1, and transmit it to the data processing center for multi-station joint discrimination; if the number of threshold limits is 0, mark the satellite as normal.

[0048] Step 106: The data processing center jointly discriminates according to the fault discrimination matrix of each reference station.

[0049] Each reference station sends the processed observation data of each frequency point, ephemeris information, and the fault discrimination matrix constructed according to steps 101-105 to the data processing center. The data processing center first jointly discriminates the signal quality anomaly and the star clock anomaly according to the fault discrimination matrix of each reference station.

[0050] For signal quality anomaly, according to its occurrence in multi-station and multi-star and space-time correlation, it is classified into six categories: single-station single-star accidental fault, single-station single-star blocking fault, single-star signal fault, regional interference, reference station receiver fault and regional ionospheric anomaly. The detailed discrimination process is shown in embodiment 2.

[0051] For the star clock anomaly, for satellite i, if at least one reference station marks its anomaly, or at least two reference stations mark its alarm identification 1, it is determined to be a faulty satellite; otherwise, it is determined to be normal, and the detailed discrimination process is shown in Embodiment 3.

[0052] Step 107: The data processing center carries out Type A ephemeris failure anomaly monitoring. At each epoch in the observation period, the data processing center synthesizes the observation data of each station to construct a multi-baseline ephemeris failure anomaly monitor, obtains the distribution characteristics of the test statistics according to the historical measurement error statistical characteristics and related parameter settings, derives the test threshold according to the allocated false alarm rate index, and evaluates the theoretical missed detection rate to determine whether the theoretical missed detection rate meets the index requirements: if the theoretical missed detection rate cannot meet the index requirements, the test statistics need to be replaced for re-evaluation, and if the index requirements cannot be met after multiple rounds of evaluation, an unusable alarm of the monitor is given; if the theoretical missed detection rate meets the index requirements, the value of the test statistics is calculated, and if it is less than the threshold, the satellite position is marked as normal, otherwise, the satellite position is marked as abnormal. The detailed discrimination process is shown in Embodiment 4. For the satellite with abnormal position, its observation data is not used.

[0053] Step 108: The data processing center carries out Type B ephemeris failure anomaly monitoring. At the ephemeris update time, the data processing center first compares the parameters of the navigation messages of multiple stations to exclude anomalies in the information receiving, processing and decoding processes; secondly, it synchronously carries out ephemeris time correlation monitoring and inter-frequency consistency monitoring to exclude satellite position anomalies caused by ephemeris annotations, and marks alarm identification 1 if it fails the time correlation monitoring and marks alarm identification 2 if it fails the inter-frequency consistency monitoring. Finally, the discrimination results of the ephemeris time correlation monitor and the inter-frequency consistency monitor, i.e., the ephemeris time correlation monitoring result and the inter-frequency consistency monitoring result, are combined, and if neither alarm identification 1 nor 2 exists, the ephemeris is determined to be normal, otherwise, the ephemeris is determined to be abnormal. The detailed discrimination process is shown in Embodiment 4.

[0054] Step 109: The data processing center synchronously carries out reference station position anomaly monitoring, troposphere anomaly monitoring and ionosphere anomaly monitoring. At each epoch in the observation period, the data processing center synthesizes the observation data of each station to select multiple reference station combinations to construct double-difference observation equations. Taking reference stations AB as an example, the station-star double-difference observation equation for satellite i, j and frequency f1 is as follows:

[0055]

[0056] The observation equations of each satellite and frequency are combined to solve the baseline vector x AB , the reference station zenith troposphere delay T r and the double-difference ionosphere delay The baseline vector of multiple groups of reference stations, zenith tropospheric delay and double-difference ionospheric delay values are integrated to perform the following monitoring:

[0057] After obtaining the baseline vector, the difference between the baseline vector solved by multiple reference station combinations and the historical stored data is calculated. The average of all baseline differences related to each reference station is taken as the test statistic. If the test statistic exceeds the threshold limit, the position of the reference station is determined to be abnormal. The detailed process is shown in Example 5.

[0058] After obtaining the zenith tropospheric delay of each reference station, the double-difference tropospheric residual at the minimum elevation angle and the maximum elevation angle of multiple reference station combinations is calculated as the test statistic. If they are all less than the threshold, the troposphere of the service area is marked as normal. Otherwise, the abnormal reference station position is used to determine the abnormal area of the troposphere, and the abnormal area is marked. The detailed process is shown in Example 6.

[0059] After obtaining the double-difference ionospheric delay value, it is determined whether the current time is in the ionospheric active period according to the joint discrimination result of the carrier-to-noise ratio and the geomagnetic parameter. In the active period and the non-active period, linear interpolation and plane fitting methods are used for ionospheric delay value interpolation fitting, respectively, and the integrity parameter is calculated. The detailed process is shown in Example 7.

[0060] Step 110: Data processing center performs comprehensive discrimination and correction number generation. According to the discrimination results of steps 106-109, the abnormal satellite or reference station data is excluded. According to the user's approximate position, reference station A and a nearby reference station B are selected. The difference correction number from the reference station to the user and the difference correction number from the reference station to the selected reference station are calculated using normal observation data. In addition, according to the long-term statistical results, the standard deviation of the pseudorange and carrier measurement error, the standard deviation of the tropospheric correction residual, and the standard deviation of the ionospheric correction residual are given, which are used to calculate the difference correction integrity parameter.

[0061] After completing the single fault monitoring, the probability of occurrence of each type of fault is the missed detection rate of the single fault monitor. Although the probability is low, multiple sub-monitor missed detections may still cause the comprehensive deviation to exceed a certain threshold, so comprehensive fault monitoring is needed.

[0062] The second level of monitoring, i.e., comprehensive fault monitoring: the data processing center selects a reference station and a nearby reference station according to the user's approximate position to generate a difference correction number. Using the observation data of the reference station and the nearby reference station and the difference correction number, the fault detection based on solution separation is performed according to the following steps 201-204 to exclude the comprehensive fault in the observation value and the correction number.

[0063] In the foregoing step 110, two sets of differential corrections are generated after excluding single faults, one set is for the reference station and the user, and the other set is for the reference station A and the reference station B. The present application uses the observation data of the reference station A and the reference station B and the generated corrections to perform a round of fault monitoring on the corrections containing various types of comprehensive faults. If the comprehensive error in the corrections can meet the performance index requirements, it is considered that the generated differential corrections between the reference station A and the user can also meet the integrity index requirements. The known reference station position can provide more redundant information for the monitoring process and improve the monitoring efficiency. The second-level comprehensive fault monitoring is based on the idea of solution separation test, and the process is as shown in Figure 3

[0064] Step 201: According to the differential corrections of the reference station A and the reference station B generated by the data processing center and the processed observation data, the differential observation equation is constructed:

[0065]

[0066] Step 202: Calculate the maximum number of faults to be monitored. The occurrence probability of each type of fault that may be contained in the foregoing observation equation differential corrections is the missed detection rate of the single fault monitor. Although the probability is low, there may still be a situation that multiple sub-monitors miss detection leads to a comprehensive deviation exceeding a certain threshold. Therefore, this paper refers to the ARAIM algorithm idea to allocate P_unmonitored to determine the maximum number of faults to be considered.

[0067] Step 203: Calculate the prior probability of each type of fault combination and perform fault detection. According to the prior probability of each type of fault, the probability of fault combination is calculated. The solution separation idea is used to calculate the positioning solution corresponding to no fault and each type of fault combination. The fault subset with the largest deviation exceeding the threshold is removed, and the test is performed again. Otherwise, repeat the removal process until the number of removed faults reaches the maximum number of faults to be monitored. If it still cannot pass the test at this time, an unusable alarm is given.

[0068] Step 204: Calculate the protection level according to the overall risk allocation and perform discrimination. According to the test result of step 203, the positioning domain protection level is calculated according to the required integrity risk index: if the protection level exceeds the maximum allowable deviation of the given positioning domain, it is marked that the generated corrections are not feasible; otherwise, the generated corrections are trusted and broadcast to the user.

[0069] In an exemplary embodiment, S2 can be replaced with the following steps.

[0070] The carrier-to-noise ratio is used for signal quality monitoring of the observation data, and the observation data with the carrier-to-noise ratio exceeding the threshold is marked. ​

[0071] The pseudo-range rough error detection and the carrier cycle slip detection are performed on the labeled observation data, and the current observation data is ensured.

[0072] The carrier acceleration-slope-step value is calculated based on the carrier observation value of the current observation data and the historical observation data, and the clock failure of the satellite is determined according to the carrier acceleration-slope-step value, and the failure satellite is labeled.

[0073] Based on the failure satellite labeling result, the failure determination matrix corresponding to each reference station and the comprehensive observation data are generated.

[0074] The comprehensive determination matrix is determined according to each failure determination matrix, and the joint determination of the carrier-to-noise ratio of multiple reference stations and the clock failure determination of multiple reference stations are performed based on the comprehensive determination matrix, and the ephemeris failure anomaly monitoring, the reference station position failure anomaly monitoring, the regional troposphere anomaly monitoring and the regional ionosphere anomaly monitoring are performed on the comprehensive observation data.

[0075] The first level monitoring result is determined according to the joint determination result and the anomaly monitoring result.

[0076] In summary, the RTK server integrity monitoring method based on the CORS network provided in the present application includes a signal quality anomaly multi-station joint monitoring method based on the carrier-to-noise ratio (i.e. embodiment 2, signal quality anomaly monitoring), a multi-station joint monitoring method for clock anomalies (i.e. embodiment 3, clock anomaly monitoring), a multi-station joint monitoring method for ephemeris anomalies (i.e. embodiment 4, ephemeris failure anomaly monitoring), a multi-station joint monitoring method for reference station position anomalies (i.e. embodiment 5, reference station position anomaly monitoring), a multi-station joint monitoring method for regional troposphere anomalies (i.e. embodiment 6, regional troposphere anomaly monitoring) and a multi-station joint monitoring method for regional ionosphere anomalies (i.e. embodiment 7, regional ionosphere anomaly monitoring).

[0077] Embodiment 2

[0078] In an exemplary embodiment, the joint determination of the carrier-to-noise ratio of multiple reference stations specifically includes the following steps.

[0079] If the carrier-to-noise ratio of a single station and a single satellite is lower than the first threshold limit and there is no time correlation, it is determined that a single station and a single satellite signal occurs sporadic failure; wherein the single station is a single reference station.

[0080] If the carrier-to-noise ratio of a single station and a single satellite is lower than the first threshold limit and there is time correlation, it is determined that a single station and a single satellite signal occurs shielding failure.

[0081] If the carrier-to-noise ratio of a single satellite is observed by multiple stations simultaneously and is lower than the first threshold limit, it is determined that a single satellite signal failure occurs; wherein the multiple stations are multiple reference stations.

[0082] If the multi-station multi-satellite carrier-to-noise ratio is lower than the first threshold limit, and the multi-station positions are within a set distance range and the number is less than or equal to 3, it is determined that a regional interference failure occurs.

[0083] If a single station observes that the multi-satellite carrier-to-noise ratio is lower than the first threshold limit, it is determined that a reference station receiver failure occurs.

[0084] If the multi-station multi-satellite carrier-to-noise ratio is lower than the first threshold limit, and the multi-station positions are not adjacent or the number of adjacent stations is greater than 3, it is determined that a regional ionospheric anomaly failure occurs; wherein the single-satellite signal failure is a satellite segment failure; the regional ionospheric anomaly failure is a propagation segment failure; the single-station single-satellite signal sporadic failure, the single-station single-satellite blocking failure, the regional interference failure, and the reference station receiver failure are ground segment failures.

[0085] In practical applications, a signal quality anomaly multi-station joint monitoring method based on carrier-to-noise ratio has a process as shown in Figure 5 , which includes:

[0086] Step 301: Each reference station uses historical observation data to establish a carrier-to-noise ratio anomaly model. The present application uses signal carrier-to-noise ratio to carry out signal quality anomaly monitoring. Since the carrier-to-noise ratio level is related to the receiver model, environmental factors, and satellite elevation angle, each reference station needs to use historical carrier-to-noise ratio data for independent modeling. The present application uses an interval modeling method to distinguish all satellite carrier-to-noise ratio data according to a specific elevation angle interval. In each interval, a specific quantile number is selected as a threshold limit according to the index requirements.

[0087] Step 302: At each epoch in the aforementioned observation time period, each reference station independently performs carrier-to-noise ratio anomaly discrimination. According to the satellite elevation angle, the corresponding threshold limit is determined, and a fault discrimination matrix is established for each observed satellite and frequency point according to the method in step 101. For observed satellites and frequency points with carrier-to-noise ratio less than the threshold, add a carrier-to-noise ratio anomaly identifier to the corresponding position of the fault discrimination matrix, and transmit the fault discrimination matrix to the data processing center.

[0088] Step 303: The data processing center comprehensively discriminates the carrier-to-noise ratio discrimination matrix of each observation station. According to the occurrence of signal anomalies in each satellite and each observation station, it is divided into six fault types, and the discrimination process is as follows:

[0089] 1) Determine whether there is only a single-satellite carrier-to-noise ratio anomaly in the current epoch. If yes, go to step 2); if no, go to step 4).

[0090] 2) Determine whether multiple stations observe the anomaly of the satellite. If yes, it is considered that the carrier-to-noise ratio anomaly originates from the satellite end, and the single-satellite signal failure of the satellite is marked, i.e., fault type 3; if no, go to step 3).

[0091] 3) Determine whether the single satellite C / N0 anomaly observed by the single station has time correlation, i.e. the phenomenon occurs at repeated intervals in the historical observation period. If yes, and auxiliary confirmation is made in combination with the actual environment of the observation station, it is considered that the anomaly is caused by the existence of shielding phenomenon around the reference station, and the single station single satellite shielding failure is marked, i.e. failure type 2; if no, it is determined as single station single satellite signal sporadic failure, i.e. failure type 1.

[0092] 4) Determine whether multiple reference stations simultaneously have multiple satellite C / N0 anomalies. If yes, go to step 5; if no, it is considered that the reference station receiver has a failure, resulting in a decrease in the received C / N0 of multiple satellites, and the failure type 5 is marked.

[0093] 5) Determine whether the observation stations observing the multiple satellite C / N0 anomaly are adjacent. If yes, and the number of adjacent observation stations is less than or equal to 3, it is considered that there is regional interference near these observation stations, and the failure type 4 is marked; if no, it is indicated that there is electromagnetic interference in a large range of the entire region, and it is preliminarily determined that the ionospheric anomaly causes the failure, and the failure type 6 is marked.

[0094] Further according to the failure source, the six failure types are classified, among which the failure type 3 belongs to satellite signal quality anomaly, the failure type 6 belongs to ionospheric anomaly in the propagation section, and the rest belong to reference station signal quality anomaly in the ground section.

[0095] Step 304: After distinguishing the various types of failures, different processing is performed. For failure types 1 to 5, the observation data of the abnormal satellite or the abnormal observation station is excluded; for failure type 6, it is necessary to combine the subsequent ionospheric anomaly monitoring to further confirm whether the ionospheric anomaly occurs and the range of the influence, but the discrimination result can be used as an indication index of the subsequent monitor to prompt the possible risk of ionospheric anomaly.

[0096] Embodiment 3

[0097] In an exemplary embodiment, the multi-reference station star clock failure discrimination specifically comprises:

[0098] Each reference station calculates a first test statistic of each satellite multi-epoch; the first test statistic includes carrier acceleration, slope and step value.

[0099] According to the distribution characteristics of the first test statistic and the second false alarm rate index, a second threshold is determined.

[0100] According to the second threshold, the satellite corresponding to any reference station is marked to determine a marking result; the marking result includes satellite anomaly, alarm flag 1 and satellite normal.

[0101] According to the marking results of all reference stations, a satellite is comprehensively discriminated to determine a satellite discrimination result; the satellite discrimination result includes that the satellite is a fault satellite and the satellite is a normal satellite.

[0102] In practical application, a multi-station joint monitoring method for satellite clock abnormality is shown in the flowchart Figure 6 as follows, which comprises the following steps.

[0103] Step 401: Each reference station independently acquires observation data and calculates carrier observation value correction numbers. At each epoch in the aforementioned observation time period, each reference station uses single-frequency carrier observation values to calculate carrier observation value correction numbers of each satellite observation channel: first, the satellite position is calculated according to the verified ephemeris, that is, the satellite clock error correction number, and the station-satellite distance is calculated according to the reference station position, the carrier observation value is subtracted from the station-satellite distance and the satellite clock error correction number to obtain the carrier correction number at the current time; second, the carrier correction number at the reference time is selected, which is subtracted from the carrier correction number at the current epoch to obtain the normalized carrier correction number; finally, the normalized carrier correction numbers of all satellites are averaged, and the normalized carrier correction numbers of each satellite are subtracted, that is, the carrier correction number after eliminating the receiver clock bias is obtained.

[0104] Step 402: Each reference station calculates ARS test statistics of each observed satellite and generates a fault discrimination matrix. At each epoch k in the observation period, the carrier observation value correction number of satellite i is used to perform quadratic polynomial fitting on the carrier correction numbers of n epochs (from epoch k-n+1 to epoch k), and the obtained quadratic term coefficient, linear term coefficient and constant term coefficient are taken as the acceleration A, slope R and step S error test statistics, and the number of statistics exceeding the threshold is counted: if the number of statistics exceeding the threshold is greater than or equal to 2, the satellite clock is marked as abnormal, and the alarm identifier is 2; if the number of statistics exceeding the threshold is 1, the alarm identifier is marked as 1 and transmitted to the data processing center for multi-station joint discrimination; if the number of statistics exceeding the threshold is 0, the satellite is marked as normal.

[0105] Step 403: The data processing center receives the fault matrix of each reference station to carry out comprehensive discrimination. The discrimination criterion is: for satellite i, if at least one reference station marks alarm identifier 2 or at least two reference stations mark alarm identifier 1, it is determined to be a fault satellite; otherwise, it is determined to be normal. For the fault satellite, all carrier and pseudorange observation values are removed.

[0106] Embodiment 4

[0107] In an exemplary embodiment, the satellite ephemeris fault monitoring process specifically comprises the following steps.

[0108] For Type A satellite ephemeris fault anomaly monitoring, a multi-baseline anomaly detector is constructed by comprehensively using multi-station observation data to carry out fault monitoring, which comprises the following steps.

[0109] Based on multi-station observation data, a dual-frequency wide-lane carrier observation equation is constructed for each satellite, and a second test statistic is selected.

[0110] According to the historical measurement error statistical characteristics and the number of wide-lane ambiguity filtering epochs selected, the distribution characteristics of the second test statistic are determined, and a test threshold is determined according to a second false alarm rate index to evaluate a theoretical missed detection rate.

[0111] It is judged whether the theoretical missed detection rate meets the index requirement.

[0112] If yes, the wide-lane ambiguity of the dual-frequency wide-lane carrier observation equation is solved, and the value of the second test statistic is calculated.

[0113] If the value of the second test statistic is less than the test threshold, the satellite position is marked as normal.

[0114] If the value of the second test statistic is not less than the test threshold, the satellite position is marked as abnormal.

[0115] If no, the second test statistic is replaced and the theoretical missed detection rate is re-evaluated until the theoretical missed detection rate meets the index requirement. If after multiple evaluations, the theoretical missed detection rate still does not meet the index requirement, an unusable alarm of the multi-baseline anomaly detector is output.

[0116] When monitoring Type B satellite ephemeris failure anomalies, multi-frequency point ephemeris data and historical ephemeris are used to design a composite monitor based on time correlation and frequency point consistency to carry out failure monitoring, including:

[0117] According to the multi-frequency point ephemeris data, time correlation monitoring and frequency point consistency monitoring are simultaneously carried out.

[0118] In the time correlation monitoring, the current ephemeris and the verified historical ephemeris are used to extrapolate the satellite positions respectively, and the position difference between the two satellite positions is determined as a third test statistic.

[0119] According to the distribution characteristics of the third test statistic and a third false alarm rate index, a third threshold limit is determined.

[0120] According to the third test statistic and the third threshold limit, a first alarm identifier is marked.

[0121] In the frequency point consistency monitoring, different frequency point ephemeris parameters are used to calculate satellite positions, and the position difference between different satellite positions is determined as a fourth test statistic.

[0122] According to the distribution characteristics of the fourth test statistic and a fourth false alarm rate index, a fourth threshold limit is determined.

[0123] According to the fourth test statistic and the fourth threshold, a second alarm identifier is marked.

[0124] According to the first alarm identifier and the second alarm identifier, a comprehensive alarm identifier is determined; the comprehensive alarm identifier includes ephemeris abnormality and ephemeris normality.

[0125] In practical application, a multi-station joint monitoring method for ephemeris abnormality includes monitoring of satellite maneuvering abnormality, i.e., Type A fault, and ephemeris annotation abnormality, i.e., Type B fault.

[0126] The Type A ephemeris fault abnormality monitoring process is as shown in Figure 7 The process includes the following steps.

[0127] Step 501: Construct a dual-frequency wide-lane carrier observation equation and select a test statistic. In each epoch in the aforementioned observation time period, a data processing center synthesizes carrier observation data of each reference station, selects a group of observation stations with a relatively long distance, and constructs a dual-difference wide-lane carrier observation equation. According to the constructed observation equation, the test statistic is selected as: the wide-lane carrier measurement value minus the projection of the baseline on the reference station inter-satellite single-difference direction vector, and then minus the product of the dual-difference wide-lane ambiguity and the wide-lane wavelength.

[0128] Step 502: Obtain the distribution characteristics of the test statistic according to the historical measurement error statistical characteristics and related parameter settings. As can be seen from the expression in step 501, the test statistic contains not only the ephemeris error term, but also the ionospheric residual error, the tropospheric residual error, and the observation noise term. In addition, the ambiguity related term needs to be subtracted in the expression, and errors may exist in the ambiguity fixing process. Therefore, the test statistic does not conform to a simple normal distribution, but presents a "multi-peak" pattern, and different "peaks" reflect ambiguity fixing bias. In order to obtain the threshold of the test statistic, the distribution characteristics of the test statistic need to be obtained, and further the distribution characteristics of the ambiguity need to be obtained.

[0129] The application uses the method of continuous filtering and rounding to fix the wide-lane ambiguity. First, a dual-difference ionosphere-free combined observation value is constructed, and the wide-lane ambiguity is obtained by averaging multi-epoch observation values. It can be known that the estimation error of the wide-lane ambiguity conforms to a normal distribution, and the standard deviation thereof can be calculated according to the original observation noise. If the ambiguity filtering time is relatively long, the probability of fixing errors is relatively low. Therefore, when considering the influence of the test statistic distribution, only the case of bias of ±1 cycle can be considered. At this time, in combination with the formula in step 501, the distribution characteristics of the test statistic can be obtained.

[0130] Step 503: Derive the test threshold according to the allocated false alarm rate index, and evaluate whether the theoretical missed alarm rate meets the index requirements. After obtaining the distribution characteristics of the test statistic, the threshold value can be solved according to the given false alarm rate index. Further, according to the distribution characteristics of the test statistic and the threshold value, the curve of the theoretical false alarm rate changing with the bias can be calculated, by which the performance of the monitor can be evaluated: 1) if the theoretical missed alarm rate meets the performance index requirements, the wide-lane ambiguity is solved and the test statistic value is calculated, and it is judged whether it is less than the threshold value: if it is less than the threshold value, it is marked that the satellite position is normal; otherwise, it is marked that the satellite position is abnormal; 2) if the theoretical missed alarm rate cannot meet the index requirements, the test statistic needs to be replaced for re-evaluation, a double baseline test statistic is constructed by selecting four reference stations, and the performance evaluation is restarted from step 502; if the performance index still cannot be met, the number of baselines is continued to be increased, and so on until the number of baselines exceeds 5; if it still cannot be met, an alarm of the unavailability of the monitor is given.

[0131] The Type B ephemeris failure anomaly monitoring process is shown in Figure 8 , which includes:

[0132] Step 601: At the ephemeris update time, the data processing center first compares the parameters of the multi-station navigation message for consistency, and excludes the anomalies in the information receiving, processing and decoding process. For the navigation ephemeris parameters broadcast by satellite i at frequency point j, if the parameters provided by each reference station are consistent, the test is passed; if not, the principle of majority over minority is adopted to select the parameters provided by the majority of reference stations; if the parameters of all reference stations are inconsistent, an alarm of the unavailability of the ephemeris is directly given, and the alarm identifier is 1.

[0133] Step 602: Time correlation monitoring and frequency consistency monitoring are carried out synchronously for the navigation ephemeris of each satellite to exclude the satellite position anomalies caused by ephemeris annotation.

[0134] Time correlation monitoring process:

[0135] 1) Determine whether there is a verified historical ephemeris, if yes, go to step 2), if not, go to step 3).

[0136] 2) The satellite position is calculated by extrapolation using the current ephemeris and the historical ephemeris respectively, and the three-dimensional position difference calculated by the two is used as the test statistic, if two of the test statistics exceed the threshold value, the alarm identifier 2 is marked; otherwise, the ephemeris data is marked as normal. The test threshold is determined according to the distribution characteristics of the historical test statistic and the false alarm rate index.

[0137] 3) If there is no history-validated ephemeris, no time correlation discrimination is performed, but the current ephemeris is initialized and verified. The specific method is as follows: within 2 minutes after the ephemeris is updated, the new ephemeris parameters are used to select the observation data of a reference station for pseudorange single point positioning, and the average of the three-dimensional positioning errors obtained by the two is used as the test statistic. If two of the test statistics exceed the threshold, the current ephemeris is considered unusable, and the subsequent updated ephemeris also needs to be initialized and verified; if only one or none of the test statistics exceeds the threshold, the current ephemeris is considered usable, and the subsequent updated ephemeris can normally carry out Type B abnormality monitoring.

[0138] Frequency consistency monitoring process:

[0139] 1) Determine whether the ephemeris of different frequencies is complete. If it is complete, go to step 2), if it is not complete, mark alarm identifier 3.

[0140] 2) Calculate the satellite position using ephemeris parameters of different frequencies, and use the three-dimensional position difference calculated by the two as the test statistic. If two of the test statistics exceed the threshold, mark alarm identifier 2; otherwise, mark the ephemeris data as normal. The threshold of the test statistic is determined according to the distribution characteristics of the historical test statistics and the false alarm rate index.

[0141] Step 603: Based on the above discrimination results, if there is any alarm identifier for satellite i, it is determined that the ephemeris of the satellite is abnormal, otherwise it is determined that the ephemeris of the satellite is normal.

[0142] Embodiment 5

[0143] In an exemplary embodiment, the reference station position anomaly monitoring process specifically includes the following steps.

[0144] Select a reference station combination, take the baseline vector, the zenith of each reference station, the process error and the ambiguity as unknowns, and carry out continuous filtering solution to calculate the difference between the baseline vector of the multiple reference station combination and the historical stored data.

[0145] Take the average of all the differences related to each reference station as the fifth test statistic, and take the maximum allowable reference station position deviation allocated in the performance index as the fifth threshold.

[0146] Determine whether the reference station position is abnormal according to the fifth test statistic and the fifth threshold.

[0147] In practical applications, a multi-station joint monitoring method for reference station position anomalies, as shown in the flowchart Figure 9 , includes:

[0148] Step 701: Obtain observation data of each reference station, and construct station-star double difference observation equation. In the aforementioned observation time period, the data processing center synthesizes carrier observation data of each reference station according to one-hour time interval, and constructs multiple sets of station-star double difference observation equations as shown in formula (3). In order to verify the effectiveness, each reference station is constructed with at least three adjacent reference stations to construct observation equations.

[0149] Step 702: Solve each set of baseline vector, and construct verification statistic. According to the constructed observation equation, the baseline vector, zenith troposphere delay value of each station and ambiguity are taken as unknowns to perform continuous filtering solution.

[0150] The embodiment assumes that there are five reference stations A-E, and four sets of baseline vectors are calculated for each reference station. For reference station A, four sets of baseline vectors can be obtained according to the aforementioned solution process: L AB,cal ,L AC,cal ,L AD,cal ,L AE,cal According to the historical baseline vector L AB,res ,L AC,res ,L AD,res ,L AE,res , the baseline AB is taken as an example. According to L AB,res and L AB,cal , the difference δL AB ={δx AB ,δy AB ,δz AB} can be calculated, wherein δx AB , δy AB , δz AB are normally distributed with mean value 0 and standard deviation of the positioning error standard deviation in the direction, that is, The remaining baseline is the same.

[0151] If the position of reference station A is abnormal, the three-dimensional position error will appear in all baselines with A as the reference. If the difference of all baselines with A as the reference is averaged, the following formula can be obtained: wherein ε A ={δx A ,δy A ,δz A} is the offset vector of the stored position of reference station A to the abnormal position. At the same time, since the position error is adjusted during the averaging process, it is assumed that the position error standard deviations of each baseline are the same, and the following formula is obtained:

[0152] And for the baseline vectors centered on the remaining observation stations, only the baseline pointing to reference station A is abnormal. Taking reference station B as an example, the average of the baselines with B as the reference can be obtained:

[0153] In view of the possibility of multiple reference station positions simultaneously having abnormal situations, for the sake of conservation and convenience, the absolute values of each reference station corresponding to the respective baseline horizontal and vertical errors are averaged to serve as a test statistic for judging whether the reference station position is abnormal.

[0154] Step 703: Calculate the test statistic value of each reference station, and judge according to the test threshold: if the test statistic is less than the threshold, mark the reference station position as normal; if the test statistic exceeds the threshold, further check whether the test statistic of the adjacent reference station exceeds the test statistic of the historical normal time, if most of the reference station test statistics exceed the historical normal value, determine that the reference station position is abnormal, need to exclude its observation data, and reposition the position. The threshold size is obtained according to the performance index allocation.

[0155] Embodiment 6

[0156] In an exemplary embodiment, the monitoring process of the troposphere anomaly specifically includes the following steps.

[0157] Select a reference station combination, take the baseline vector, the zenith troposphere error of each reference station and the ambiguity as unknowns, carry out continuous filtering solution, calculate the double difference troposphere residual of each reference station combination at the minimum elevation angle and the maximum elevation angle.

[0158] The double difference troposphere residual is taken as the sixth test statistic, and the maximum allowable troposphere residual allocated in the performance index is taken as the sixth threshold.

[0159] According to the sixth test statistic and the sixth threshold, it is judged whether the troposphere is abnormal.

[0160] In practical application, a multi-station joint monitoring method for regional troposphere anomaly, as shown in the flowchart Figure 10 , includes:

[0161] Step 801: Obtain observation data of each reference station, and construct station-star double difference observation equation as shown in formula (3). At each epoch of the aforementioned observation time period, the data processing center synthesizes the carrier observation data of each reference station to construct multiple sets of station-star double difference observation equations, ensuring that each reference station is at least constructed with an adjacent reference station to construct an observation equation.

[0162] Step 802: Solve each group of baseline vectors, and construct the test statistic. According to the constructed observation equation, continuous filtering solution is carried out to extract zenith tropospheric delay values of each reference station. The extracted zenith tropospheric delay contains dry component and wet component, wherein the dry component accounts for 90% of the total tropospheric delay, and can be accurately corrected by a model; and the wet component has regional characteristics due to different regions and climates, and the wet component of the zenith tropospheric delay is generally taken as an estimated parameter, and is projected to the tilt direction through a projection function. Since the dry component can be accurately corrected, the tropospheric anomaly is more likely to occur in the wet component, and is greatly affected by the elevation angle. Therefore, the tropospheric double-difference residuals are obtained under the conditions of 15° elevation angle and 90° elevation angle, respectively, as the test statistic of the tropospheric anomaly, which is independent of the actual satellite position, and can reflect the maximum value of the tropospheric delay correction residual between the reference stations.

[0163] Step 803: Multi-station joint discrimination of tropospheric anomaly. The maximum value of the double-difference tropospheric correction residual of each group of reference stations is calculated to discriminate the anomaly: if the test statistic of each reference station is less than the threshold limit, the region is marked as normal troposphere; if there is at least one group of test statistics exceeding the threshold limit, the abnormal region is marked according to the distribution of the reference stations, that is, the union of the circular regions with the reference station combination baseline as the diameter is the abnormal region. If the user is located in the abnormal region, additional correction is required for the tropospheric delay value, and the tropospheric residual error standard deviation is amplified.

[0164] Embodiment 7

[0165] In an exemplary embodiment, the ionospheric anomaly monitoring process specifically includes the following steps.

[0166] An observation equation is constructed according to the multi-station observation data.

[0167] The observation equation is solved by filtering and fixing ambiguity to extract double-difference ionospheric delay values.

[0168] According to the joint discrimination result of the carrier-to-noise ratio and the electromagnetic field parameter, it is determined whether the current time is in the ionospheric active period.

[0169] In the ionospheric active period and the ionospheric non-active period, linear interpolation and plane fitting methods are used respectively to interpolate and fit the double-difference ionospheric delay values, and the integrity parameters are calculated.

[0170] A multi-station joint monitoring method for regional ionospheric anomalies, as shown in Figure 11 the flow chart, comprising:

[0171] Step 901: Acquire observation data from each reference station and construct double-difference observation equations. At each epoch of the aforementioned observation period, the data processing center integrates carrier observation data from each reference station, selects multiple sets of reference station combinations to construct double-difference observation equations, ensuring that each reference station constructs an observation equation with at least one neighboring reference station.

[0172] Step 902: Solve for integer ambiguity and extract double-difference ionospheric delay. For the selected observation station combination, firstly, solve for the double-difference wide-lane integer ambiguity, and preliminarily verify the accuracy of the double-difference wide-lane integer ambiguity by checking whether the sum of the double-difference wide-lane integer ambiguities of the closed baseline composed of the three reference stations is equal to 0; secondly, using the relationship satisfied by the ionospheric combination integer ambiguity, the wide-lane double-difference integer ambiguity, and the L1 carrier double-difference integer ambiguity, obtain the floating-point solution of the L1 double-difference ambiguity through filtering, and further solve for the fixed solution; finally, based on the calculated fixed ambiguity solution, estimate the double-difference ionospheric delay using the ionospheric combination formed by the dual-frequency carrier phase measurements.

[0173] Step 903: Determine whether the current time is within an ionospheric active period by comprehensively considering various indicators. After extracting the ionospheric delay, interpolation processing is required to compensate for the ionospheric residual from the user to the base station and to calculate relevant integrity parameters. The appropriate interpolation methods differ between ionospheric active and quiescent periods; therefore, it is necessary to first determine the current ionospheric state. This patent comprehensively uses the aforementioned carrier-to-noise ratio anomaly determination results and geomagnetic index for determination. If the threshold exceeds the set threshold, it is considered to be within an ionospheric active period; otherwise, it is considered to be within an ionospheric quiescent period.

[0174] Step 904: Select different methods to interpolate and fit the ionospheric delay value according to the ionospheric active state, and calculate the integrity parameter.

[0175] If it is determined that the current period is an active ionospheric phase, then a linear fitting method is used for interpolation fitting, and the integrity parameter of the ionospheric correction term is further calculated. The steps are as follows: 1) For satellite j, calculate the difference between the ionospheric vertical delay value estimated based on linear fitting and the vertical delay value calculated based on dual-frequency observations. 2) Within a time interval (m epochs), perform statistical analysis on the vertical error sequence formed by the differences, and calculate the mean and standard deviation of the user's vertical delay value; 3) For the observation data of the m-th epoch, estimate the absolute error deviation of the vertical delay at the user's location based on the absolute value of the IPP residual; 4) For the IPP of each CORS station participating in the positioning, generate a conservative error limit boundary E. i 5) Obtain the integrity parameter GIVE of satellite j, i.e., the estimation error and the maximum error max among the k sequences. k (E k ) and quantization error sum.

[0176] If it is determined that the ionosphere is in a quiet period, a plane fitting method is used for interpolation fitting, and a integrity parameter of the ionosphere correction term is further calculated, and the steps are as follows: 1) the ionosphere vertical delay value is estimated by fitting the IPP of the CORS stations within the threshold around the user station, n Kriging fitting coefficients are found, and linear unbiased minimum mean square error estimation of the ionosphere vertical delay between stations at the ionosphere penetration point is realized; 2) the Kriging variance is calculated according to the estimated Kriging coefficients; 3) the GIVE representing the uncertainty of the ionosphere delay estimation, i.e. the integrity parameter, is obtained according to the required envelope rate index, the integrated Kriging variance, the undersampling variance and the quantization error parameter.

[0177] Step 905: output the ionosphere fitting result at the current time and the integrity parameter.

[0178] The technical features of the above embodiments can be combined in any manner. To make the description concise, all possible combinations of the technical features in the above embodiments are not described, but as long as the combinations of the technical features do not exist contradictions, they should be considered as the scope of the present application.

[0179] The principles and implementation modes of the present application are described by using specific examples in the present application, and the above embodiments are only used to help understand the method of the present application and its core idea; meanwhile, for those skilled in the art, according to the idea of the present application, the specific implementation modes and application ranges will be changed. In conclusion, the content of the present application should not be understood as a limitation.

Claims

1. A method for integrity monitoring of a CORS network-based RTK server, characterized in that, The RTK service end integrity monitoring method based on the CORS network comprises the following steps: Obtaining observation data of each epoch obtained by a plurality of reference stations in a CORS network in a same region within an observation time period, wherein the observation data comprises satellite ephemeris and pseudo-range carrier phase measurement values; Based on a distributed monitoring scheme and a hierarchical monitoring idea, receiver faults, signal quality abnormalities and star clock abnormalities in the observation data of each reference station are monitored to determine fault discrimination matrices of various faults; Based on a data processing center, joint discrimination is carried out by comprehensively using the fault discrimination matrices of each reference station, and satellite ephemeris faults, reference station position faults, regional ionospheric abnormalities and regional tropospheric abnormalities are monitored using multi-station observation data to obtain first-level monitoring results; the first-level monitoring results are discrimination results for each single fault; Based on the first-level monitoring results, reference stations and adjacent reference stations are selected according to the approximate position of a user to determine initial differential corrections; the initial differential corrections comprise differential corrections from the reference stations to the user and differential corrections from the reference stations to the adjacent reference stations; Based on the initial differential corrections, comprehensive faults contained in the observation data of each reference station are monitored to determine differential corrections and integrity parameters to be finally broadcast to the user. 2.The CORS network-based RTK server integrity monitoring method of claim 1, wherein, Based on a distributed monitoring scheme and a hierarchical monitoring idea, receiver faults, signal quality abnormalities and star clock abnormalities in the observation data of each reference station are monitored to determine fault discrimination matrices of various faults, and first-level monitoring results are obtained, which specifically comprise the following steps: Signal quality monitoring is performed on the observation data using carrier-to-noise ratio (CNR) to mark observation data with CNR exceeding a threshold value; Pseudo-range gross error detection and carrier cycle slip detection are performed on the marked observation data to determine current observation data; Carrier acceleration-slope-step values are calculated based on carrier observation values of the current observation data and historical observation values, and satellite faults are discriminated based on the carrier acceleration-slope-step values to mark fault satellites; Based on the fault satellite marking results, fault discrimination matrices corresponding to each reference station and comprehensive observation data are generated; Based on a data processing center, joint discrimination is carried out by comprehensively using the fault discrimination matrices of each reference station, and satellite ephemeris faults, reference station position faults, regional ionospheric abnormalities and regional tropospheric abnormalities are monitored using multi-station observation data to obtain first-level monitoring results, which specifically comprise the following steps: A comprehensive discrimination matrix is determined based on each fault discrimination matrix, and multi-reference station CNR joint discrimination and multi-reference station star clock fault discrimination are performed based on the comprehensive discrimination matrix, and ephemeris fault anomaly monitoring, reference station position fault anomaly monitoring, regional tropospheric anomaly monitoring and regional ionospheric anomaly monitoring are performed on the comprehensive observation data; The first-level monitoring results are determined based on the joint discrimination results and the anomaly monitoring results. 3.The CORS network based RTK server integrity monitoring method of claim 2, wherein, The multi-reference station CNR joint discrimination specifically comprises the following steps: If the CNR of a single satellite of a single reference station is lower than a first threshold value and there is no time correlation, it is determined that a single-station single-satellite signal sporadic fault occurs; wherein the single reference station is a single reference station. If the single-station single-satellite carrier-to-noise ratio is lower than the first threshold and there is time correlation, it is determined that a single-station single-satellite blocking fault occurs; If the single-satellite carrier-to-noise ratio is lower than the first threshold and is observed by multiple stations simultaneously, it is determined that a single-satellite signal fault occurs; wherein the multiple stations are multiple reference stations; If the multi-satellite carrier-to-noise ratio is lower than the first threshold and the positions of the multiple stations are within a set distance range and the number of the multiple stations is less than or equal to 3, it is determined that a regional interference fault occurs; If the multi-satellite carrier-to-noise ratio is lower than the first threshold and is observed by a single station, it is determined that a reference station receiver fault occurs; If the multi-satellite carrier-to-noise ratio is lower than the first threshold and the multiple stations are not adjacent to each other or the number of adjacent stations is greater than 3, it is determined that a regional ionospheric anomaly fault occurs; wherein the single-satellite signal fault is a satellite segment fault; the regional ionospheric anomaly fault is a propagation segment fault; the single-station single-satellite signal occasional fault, the single-station single-satellite blocking fault, the regional interference fault, and the reference station receiver fault are ground segment faults.

4. The CORS network-based RTK server integrity monitoring method according to claim 2, characterized in that, The multi-reference station satellite clock fault determination specifically includes: Each reference station calculates a first test statistic of each satellite multi-epoch; the first test statistic includes carrier acceleration, slope, and step value; According to the distribution characteristics of the first test statistic and a second false alarm rate index, a second threshold is determined; According to the second threshold, a satellite corresponding to any reference station is marked to determine a marking result; the marking result includes satellite anomaly, alarm flag being 1, and satellite normality; According to the marking results of all reference stations, any satellite is comprehensively determined to determine a satellite determination result; the satellite determination result includes a fault satellite and a normal satellite.

5. The CORS network-based RTK server integrity monitoring method according to claim 1, wherein, The satellite ephemeris fault monitoring process specifically includes: For Type A satellite ephemeris fault anomaly monitoring, a multi-baseline anomaly detector is constructed based on multi-station observation data to carry out fault monitoring, including: Based on multi-station observation data, a double-frequency wide-lane carrier observation equation is constructed for each satellite, and a second test statistic is selected; According to historical measurement error statistical characteristics and the number of selected wide-lane ambiguity filtering epochs, the distribution characteristics of the second test statistic are determined, and a test threshold is determined according to a second false alarm rate index to evaluate a theoretical missed detection rate; It is determined whether the theoretical missed detection rate meets the index requirements; according to historical measurement error statistical characteristics and the number of selected wide-lane ambiguity filtering epochs, the distribution characteristics of the second test statistic are determined, and a test threshold is determined according to a second false alarm rate index to evaluate a theoretical missed detection rate; If yes, the wide-lane ambiguity of the double-frequency wide-lane carrier observation equation is solved, and the value of the second test statistic is calculated; If the value of the second test statistic is less than the test threshold, the satellite position is marked as normal; If the value of the second test statistic is not less than the test threshold, the satellite position is marked as abnormal; If no, the second test statistic is replaced and the theoretical missed detection rate is re-evaluated until the theoretical missed detection rate meets the index requirements; if the theoretical missed detection rate still does not meet the index requirements after multiple evaluations, an unusable alarm of the multi-baseline anomaly detector is output. In the monitoring of the satellite ephemeris fault of Type B satellites, multi-frequency ephemeris data and historical ephemeris are used to design a composite monitor based on time correlation and consistency between frequencies, and the fault is monitored, including: According to the multi-frequency ephemeris data, the time correlation monitoring and the consistency monitoring between frequencies are simultaneously carried out; In the time correlation monitoring, the satellite positions are respectively extrapolated by using the current ephemeris and the verified historical ephemeris, and the position difference between the two satellite positions is determined as a third test statistic; According to the distribution characteristics of the third test statistic and a third false alarm rate index, a third threshold is determined; According to the third test statistic and the third threshold, a first alarm mark is marked; In the consistency monitoring between frequencies, the satellite positions are respectively calculated by using the ephemeris parameters of different frequencies, and the position difference between the different satellite positions is determined as a fourth test statistic; According to the distribution characteristics of the fourth test statistic and a fourth false alarm rate index, a fourth threshold is determined; According to the fourth test statistic and the fourth threshold, a second alarm mark is marked; According to the first alarm mark and the second alarm mark, a comprehensive alarm mark is determined; the comprehensive alarm mark includes ephemeris anomaly and ephemeris normal.

6. The CORS network-based RTK server integrity monitoring method according to claim 1, wherein, The monitoring process of the reference station position anomaly specifically includes: A reference station combination is selected, baseline vectors, zenith troposphere error of each reference station and ambiguity are taken as unknowns, continuous filtering solution is carried out, and the difference between the baseline vectors of the multiple reference station combinations and the historical stored data is calculated; The average value of all the differences related to each reference station is taken as a fifth test statistic, and the maximum allowable measurement error allocated in the performance index is taken as a fifth threshold; Whether the reference station position is abnormal is determined according to the fifth test statistic and the fifth threshold.

7. The CORS network-based RTK server integrity monitoring method according to claim 1, wherein, The monitoring process of the troposphere anomaly specifically includes: A reference station combination is selected, baseline vectors, zenith troposphere error of each reference station and ambiguity are taken as unknowns, continuous filtering solution is carried out, and the double-difference troposphere residual at the minimum elevation angle and the maximum elevation angle of each reference station combination is calculated; The double-difference troposphere residual is taken as a sixth test statistic, and the maximum allowable measurement error allocated in the performance index is taken as a sixth threshold; Whether the troposphere is abnormal is determined according to the sixth test statistic and the sixth threshold. 8.The CORS network based RTK server integrity monitoring method of claim 1, wherein, The monitoring process of the ionosphere anomaly specifically includes: An observation equation is constructed according to the observation data of multiple stations; The observation equation is filtered and solved, and the double-difference ionosphere delay value is extracted by fixing the ambiguity; Whether the current time is in the ionosphere active period is determined according to the joint determination result of the carrier-to-noise ratio and the electromagnetic field parameter; In the ionosphere active period and the ionosphere non-active period, the double-difference ionosphere delay value is respectively interpolated and fitted by using the linear interpolation and the plane fitting method, and the integrity parameter is calculated. 9.The CORS network based RTK server integrity monitoring method of claim 1, wherein, According to the initial difference correction, the comprehensive fault contained in the observation data of each reference station is monitored, and the final difference correction and the integrity parameter to be broadcast to the user are determined, specifically including: calculating a maximum number of faults needed to be monitored based on the first level monitoring result and the initial differential correction; combining each fault based on the maximum number of faults needed to be monitored, and calculating a corresponding risk of each fault combination; calculating a protection level based on the corresponding risk of each fault combination; determining a final differential correction and a perfectness parameter to be issued to the user according to the protection level and a maximum allowable error.

Images