A method, device, computer and storage medium for correcting a product residual gaussianization envelope

Abstract

The application discloses a correction product residual Gaussian envelope method and device, a computer and a storage medium, and relates to the field of Beidou satellite navigation. The method solves the problems of low availability of envelope results of a traditional zero-mean envelope method and difficulty in adapting to Beidou technical characteristics and application scenarios. The method comprises the following steps: setting a time sliding window, receiving correction product residual data through the time sliding window; constructing a correction product residual sample data set according to the correction product residual data; statistically analyzing the correction product residual sample data set, obtaining correction product residual sample statistical characteristics, performing normalization processing, and obtaining a sample distribution; performing envelope processing on the sample distribution to obtain a unimodal symmetric distribution sample distribution; performing Gaussian processing on the unimodal symmetric distribution sample distribution by using a one-sided envelope to obtain a Gaussian envelope of the sample distribution; and obtaining correction product quality identification information according to the Gaussian envelope of the sample distribution. The method is suitable for the field of satellite navigation high-precision positioning technology.

Description

TECHNICAL FIELD

[0001] The application relates to the field of Beidou satellite navigation, and in particular to a residual error Gaussian envelope method for a PPP-RTK correction product. BACKGROUND

[0002] A global satellite navigation system (GNSS) can provide high-precision positioning information in all-weather and in a global range, and is applied to multiple scenes such as air route navigation, unmanned farms, ship navigation, and automatic driving of agricultural machinery. The positioning accuracy, continuity, integrity, and usability of a satellite navigation system are the most core indicators in navigation service performance, and are increasingly valued by users. The integrity is the ability of the system to timely alarm when a satellite navigation system fails or positioning error exceeds the allowable range. The integrity monitoring process can monitor the entire navigation system, including satellite signals, satellites, and ground receiving equipment, to provide protection for safe operation of the navigation system.

[0003] In recent years, in order to provide high-precision, fast-converging, and wide-coverage location services, the PPP-RTK (Precise Point Positioning-Real Time Kinematic) technology is proposed, which is a new generation of satellite navigation high-precision positioning technology, adopts a cloud-edge collaborative working mode, introduces a satellite navigation enhancement system as a trusted service cloud, and uses correction products and atmospheric enhancement information determined by a global reference station network to realize rapid fixing of ambiguity. In the integrity monitoring process of the PPP-RTK, the residual error of the correction product needs to conform to the assumption of Gaussian distribution, but in the actual situation, due to the correlation of the residual error of the PPP-RTK correction product in the time and space dimensions and the limitation of the number of statistical samples, the real-time PPP-RTK correction product residual error distribution may have multiple non-standard Gaussian distribution characteristics such as non-zero mean, truncation, and multiple peaks.

[0004] In order to clarify the positioning error propagation law of the user terminal and simplify the expression form of the correction product error, the correction product needs to be Gaussian enveloped. In the actual situation, the real-time correction product is affected by multipath, satellite failure, error correlation between monitoring stations, and the limitation of the number of samples, and the statistical distribution of the real-time correction product residual error has multiple non-standard Gaussian characteristics such as multiple peaks, fat tails, non-zero mean, and truncation. For the biased non-Gaussian characteristics, the traditional zero-mean Gaussian envelope method needs a large inflation coefficient to envelope the non-Gaussian characteristics of the correction information residual error, otherwise it is difficult to ensure that the user terminal integrity meets the requirements, but a too large inflation coefficient will cause the availability of the correction information quality identifier to be sharply reduced, and it is difficult to meet the requirements of various application scenarios of the PPP-RTK. SUMMARY

[0005] The application solves the problem that the traditional zero-mean envelope method has low availability of envelope results and is difficult to adapt to the characteristics and application scenarios of Beidou PPP-RTK technology in the case of limited number of product residual samples, influence of correlation, existence of non-standard Gaussian characteristics such as multi-peak, fat tail, non-zero mean and truncation.

[0006] The application provides a product residual Gaussian envelope method, which comprises the following steps:

[0007] S1: setting a time sliding window in a time collection sequence, and receiving real-time product residual data of a trusted service cloud through the time sliding window;

[0008] S2: constructing a product residual sample data set according to the received product residual data;

[0009] S3: statistically analyzing the product residual sample data set, obtaining product residual sample statistical characteristics, dividing the product residual samples into multiple segments for uniform distribution to smooth the empirical distribution, and obtaining sample distribution;

[0010] S4: performing preliminary envelope processing on the sample distribution by using a pair of unimodal symmetric distribution, and obtaining a unimodal symmetric intermediate distribution;

[0011] S5: performing Gaussian processing on the unimodal symmetric intermediate distribution by using a one-sided envelope, obtaining a Gaussian envelope of the sample distribution, and obtaining envelope parameters in the Gaussian envelope of the sample distribution;

[0012] S6: obtaining product quality identification information according to the envelope parameters.

[0013] Further, a preferred embodiment is also provided, wherein the received real-time product residual data of the trusted service cloud comprises satellite clock and orbit correction product residual data, troposphere correction product residual data and ionosphere correction product residual data.

[0014] Further, a preferred embodiment is also provided, wherein the preliminary envelope processing on the sample distribution by using a pair of unimodal symmetric distribution is specifically as follows:

[0015]

[0016] wherein G L is a left envelope, G R is a right envelope, and G a is product residual distribution, and x is product residual.

[0017] Further, a preferred embodiment is also provided, wherein the Gaussian envelope of the sample distribution is obtained by using the following specific method:

[0018] Envelope conditions for planning a one-sided envelope are as follows:

[0019]

[0020] wherein G ob (x) is a Gaussian envelope distribution, G s (x) is a unimodal symmetric intermediate distribution.

[0021] Based on the same inventive concept, the application also provides a correction product residual Gaussian envelope device, which comprises:

[0022] A correction product residual data acquisition unit is configured to set a time sliding window in a time collection sequence, and receive real-time correction product residual data from a trusted service cloud through the time sliding window.

[0023] A correction product residual sample data set acquisition unit is configured to construct a correction product residual sample data set according to the received correction product residual data.

[0024] A sample distribution acquisition unit is configured to statistically analyze the correction product residual sample data set, acquire correction product residual sample statistical characteristics, divide the correction product residual sample into multiple uniform distributions to smooth the empirical distribution, and acquire a sample distribution.

[0025] A unimodal symmetric intermediate distribution acquisition unit is configured to preliminarily envelope process the sample distribution by using a pair of unimodal symmetric distributions, and acquire a unimodal symmetric intermediate distribution.

[0026] A Gaussian envelope of sample distribution acquisition unit is configured to Gaussian process the unimodal symmetric intermediate distribution by using a one-sided envelope, acquire a Gaussian envelope of sample distribution, and acquire envelope parameters in the Gaussian envelope of sample distribution.

[0027] A correction product quality identification information acquisition unit is configured to acquire correction product quality identification information according to the envelope parameters.

[0028] Further, a preferred embodiment of the correction product residual data acquisition unit is provided, which comprises satellite clock and orbit correction product residual data, troposphere correction product residual data, and ionosphere correction product residual data.

[0029] Further, a preferred embodiment of the unimodal symmetric sample distribution acquisition unit is provided, which specifically comprises:

[0030]

[0031] wherein G L is a left envelope, G R is a right envelope, and G a is a correction product residual distribution, and x is a correction product residual.

[0032] Further, a preferred embodiment is also provided, the Gaussian envelope of the sample distribution envelope acquisition unit, specifically:

[0033] The envelope condition of the single-sided envelope is planned:

[0034]

[0035] Wherein, G ob (x) is a Gaussian envelope distribution, G s (x) is a unimodal symmetric intermediate distribution.

[0036] Based on the same inventive concept, the present application also provides a computer device comprising a memory and a processor, the memory having stored therein a computer program, when the processor runs the computer program stored in the memory, the processor executes the method for correcting product residual Gaussian envelope in any one of the above.

[0037] Based on the same inventive concept, the present application also provides a computer readable storage medium, the computer readable storage medium having stored thereon a computer program, the computer program being executed by a processor to perform the method for correcting product residual Gaussian envelope as claimed in any one of the above.

[0038] The present application has the advantages of:

[0039] The present application solves the problem that the traditional zero-mean envelope method has low envelope result availability and is difficult to adapt to the characteristics and application scenarios of Beidou PPP-RTK technology in the case of limited number of corrected product residual samples, affected by correlation, multi-peak, thick tail, non-zero mean, truncation and other non-standard Gaussian characteristics.

[0040] 1. The method for correcting product residual Gaussian envelope, starting from the problem that the corrected product residual samples have multiple non-Gaussian characteristics, proposes a biased Gaussian envelope method based on cumulative distribution function for PPP-RTK technology and application characteristics. The corrected product residual distribution is generated using the corrected product residual data broadcast by the trusted service cloud, and a pair of Gaussian distributions with opposite mean values and same sigma are used to perform Gaussian envelope on the corrected product residual, thereby synchronously ensuring the integrity and availability of user terminal positioning, meeting the needs of various application scenarios of PPP-RTK, and improving the availability of the trusted monitoring system.

[0041] 2. The method for correcting product residual Gaussian envelope, by using Gaussian distribution to envelope the corrected product residual distribution, reduces the user computing burden and improves the efficiency of the trusted monitoring. The problem of using a large envelope parameter to reduce the availability of the system is avoided, which is caused by the traditional Gaussian envelope method using zero-mean Gaussian distribution to envelope the sample distribution.

[0042] The application is suitable for the field of satellite navigation high-precision positioning technology. BRIEF DESCRIPTION OF DRAWINGS

[0043] Figure 1 A flowchart of a Gaussian envelope method for correcting product residual distribution based on CDF for the first embodiment to the fourth embodiment. DETAILED DESCRIPTION

[0044] In order to make the technical solutions and advantages of the application clearer, several embodiments of the application will be further described in detail with reference to the drawings, but the following described embodiments are only several preferred embodiments of the application and are not used to limit the application.

[0045] Embodiment one, see Figure 1 This embodiment describes a method for correcting product residual Gaussian envelope, the method comprises:

[0046] S1: setting a time sliding window in a time collection sequence, receiving real-time correction product residual data of a trusted service cloud through the time sliding window;

[0047] S2: constructing a correction product residual sample data set according to the received correction product residual data;

[0048] S3: statistically analyzing the correction product residual sample data set, obtaining correction product residual sample statistical characteristics, dividing the correction product residual sample into multiple segments of uniform distribution to smooth the empirical distribution, and obtaining sample distribution;

[0049] S4: using a pair of unimodal symmetric distribution to preliminarily envelope process the sample distribution, and obtaining a unimodal symmetric intermediate distribution;

[0050] S5: using a single-sided envelope to Gaussian process the unimodal symmetric intermediate distribution, obtaining a Gaussian envelope of the sample distribution, and obtaining envelope parameters in the Gaussian envelope of the sample distribution;

[0051] S6: obtaining correction product quality identification information according to the envelope parameters.

[0052] Specifically, S3 described obtaining correction product residual sample statistical characteristics, normalizing the correction product residual sample statistical characteristics, and obtaining sample distribution, comprising:

[0053] Resampling the received correction product residual data, first dividing the sample data from the maximum value to the minimum value into N equal-length intervals U(i), i=1, 2, 3…N, the U(i) is a uniformly distributed interval. Record the number of samples in each interval, normalize the number of samples in each interval, preliminarily generate sample statistical characteristics, and obtain sample distribution.

[0054] S6 obtains the correction product quality identification information according to the Gaussian envelope of the sample distribution, specifically: through Gaussian envelope of the correction product, the envelope parameters conforming to Gaussian distribution are generated as quality identification information and broadcast to the user, and the user generates the protection level through convolution of the received envelope parameters to perform integrity monitoring.

[0055] The envelope method of the embodiment is used to solve the problem of multiple non-Gaussian characteristics of the correction product residual under the characteristics of the PPP-RTK technology, and the Gaussian distribution is used to envelope the correction product residual distribution, thereby reducing the user calculation burden and improving the efficiency of the reliability monitoring. The method avoids the problem of using a zero-mean Gaussian distribution to envelope the sample distribution in the traditional Gaussian envelope, which requires a larger envelope parameter and reduces the usability of the system.

[0056] Embodiment two, see Figure 1 This embodiment is a further limitation of the Gaussian envelope method of the correction product residual of the first embodiment. The received real-time correction product residual data of the trusted service cloud includes satellite clock and orbit correction product residual data, troposphere and ionosphere correction product residual data.

[0057] Embodiment three, see Figure 1 This embodiment is a further limitation of the Gaussian envelope method of the correction product residual of the first embodiment. The sample distribution is preliminarily enveloped by using a pair of unimodal symmetric distribution, specifically:

[0058]

[0059] wherein G L is the left envelope, G R is the right envelope, G a is the correction product residual distribution, and x is the correction product residual.

[0060] Specifically, taking the right envelope G R as an example, a unimodal symmetric sample distribution function is generated. The generated N uniform distribution and equal length intervals U(i) are adjusted from right to left. By comparing the sample number of each uniform distribution with the sample number of the previous adjusted uniform distribution, the uniform distribution with a larger sample number is selected as the uniform distribution in the modified interval:

[0061] U t (i)=max(U(i),U t (i+1))

[0062] wherein U tThe adjusted uniform distribution is represented, by adjusting, the unimodality of the middle distribution is ensured, until the adjusted sample number reaches half of the total sample number when the next uniform distribution is reached, the remaining half of the data is obtained by symmetrically adjusting the uniform distribution, ensuring its symmetry, and the remaining data is used as the middle value of the distribution, wherein the generated middle distribution boundary changes the deviation of the unimodal symmetric middle distribution, so that the middle distribution is translated to the right, and G R The above conditions are met.

[0063] The left envelope and the right envelope processing method are the same.

[0064] Embodiment four, see Figure 1 This embodiment is described. This embodiment is a further limitation of the improved product residual Gaussian envelope method described in embodiment one. The Gaussian envelope of the sample distribution is obtained in the following specific method:

[0065] The envelope condition of the unilateral envelope is planned:

[0066]

[0067] Wherein, G ob (x) is a Gaussian envelope distribution function, G s (x) is a middle distribution function.

[0068] The Gaussian envelope of the sample distribution also includes determining the Gaussian envelope boundary:

[0069] The minimum sigma is obtained by Gaussianizing the N uniform distribution intervals of piecewise linearity, and the distribution function of each uniform distribution interval is as follows:

[0070] P(x)=G s (x-m)

[0071] Wherein, m represents the mean of G s , P(x) represents the uniform distribution function of a certain interval. The uniform distribution interval is [x1, x2], and the CDF of the piecewise uniform distribution on the interval is as follows:

[0072]

[0073] P(x1) and P(x2) are the boundary values of the uniform distribution.

[0074] Take the right envelope sigma R For example, a Gaussian distribution is set, and the Gaussian distribution function is a convex function for x>0, so the tangent line of the Gaussian distribution function is taken as the boundary, and then the tangent line equation f(x) is obtained at the tangent point (x1+x2) / 2, and the tangent line equation satisfies:

[0075]

[0076] The variance of the Gaussian distribution is obtained from the tangent equation. The envelope σ is then calculated using the tangent equation at the half-margin points. This search requires an upper and lower bound on the solution. The lower bound is σ. min It is given by the following formula:

[0077] σ min =max(σ1,…,σ i …)0 <i≤N

[0078] σ i Let σ represent the Gaussian distribution calculated from the i-th uniform distribution.

[0079] Determine the upper limit of the Gaussian distribution σ. max x max The lower bound of the last interval of P(x):

[0080]

[0081] Finally, the median value of the upper and lower limits is used as the envelope σ of the final Gaussian distribution function. R .

[0082] The left envelope σ L With right envelope σ R The processing method is the same: compare the left and right Gaussianization parameters, and take the larger value as the final envelope parameter. The final envelope parameter is:

[0083] σ ob =max(σ R ,σ L )

[0084] Implementation Method 5: The Gaussianization envelope device for correcting product residuals described in this implementation method is characterized in that the device comprises:

[0085] The product residual data acquisition unit is used to set a time sliding window in the time acquisition sequence and receive real-time product residual data from the trusted service cloud through the time sliding window.

[0086] The corrected product residual sample dataset acquisition unit is used to construct a corrected product residual sample dataset based on the received corrected product residual data;

[0087] The sample distribution acquisition unit is used to statistically analyze the corrected product residual sample dataset, obtain the statistical characteristics of the corrected product residual samples, divide the corrected product residual samples into multiple uniform distributions to smooth the empirical distribution, and obtain the sample distribution.

[0088] The single-peak symmetric intermediate distribution obtaining unit is configured to obtain a single-peak symmetric intermediate distribution by performing preliminary envelope processing on the sample distribution by using a pair of single-peak symmetric distributions.

[0089] The Gaussian envelope obtaining unit is configured to obtain a Gaussian envelope of the sample distribution by performing Gaussian processing on the single-peak symmetric intermediate distribution by using a one-sided envelope, and obtain envelope parameters in the Gaussian envelope of the sample distribution.

[0090] The correction product quality identification information obtaining unit is configured to obtain correction product quality identification information according to the envelope parameters.

[0091] Embodiment six, the embodiment is a further limitation of the correction product residual Gaussian envelope device of embodiment five, and the correction product residual data obtaining unit comprises: satellite clock orbit, troposphere and ionosphere correction product residual data.

[0092] Embodiment seven, the embodiment is a further limitation of the correction product residual Gaussian envelope device of embodiment five, and the sample distribution of the single-peak symmetric distribution obtaining unit is specifically:

[0093]

[0094] Wherein, G L is a left envelope, G R is a right envelope, G a is a correction product residual distribution, and x is a correction product residual.

[0095] Embodiment eight, the embodiment is a further limitation of the correction product residual Gaussian envelope device of embodiment five, and the Gaussian envelope obtaining unit of the sample distribution is specifically:

[0096] The envelope condition of planning a one-sided envelope is:

[0097]

[0098] Wherein, G ob (x) is a Gaussian envelope distribution function, G s (x) is a single-peak symmetric intermediate distribution function.

[0099] Embodiment nine, the computer device comprises a memory and a processor, and the memory stores a computer program, when the processor runs the computer program stored in the memory, the processor executes the correction product residual Gaussian envelope method in any one of embodiments one to four.

[0100] Embodiment ten, a computer readable storage medium storing a computer program, the computer program, when executed by a processor, performs the method of correcting a product residual Gaussianized envelope according to any one of embodiments one to four.

[0101] Those skilled in the art will appreciate that embodiments of the present application can be supplied as a method, a system, or a computer program product. Accordingly, the present application can take the form of an entirely hardware embodiment, an entirely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, the present application can take the form of a computer program product on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROMs, optical storage devices, etc.) embodying computer readable program code thereon.

[0102] The above detailed description has been described with specific reference to a particular embodiment, but it is clear that no limitation of the scope of the application is intended. Any modification, combination of embodiments, equivalent replacement, and improvement made within the spirit and principle of the application shall be included in the scope of the application.

Claims

1. A method for correcting Gaussianized envelope of product residuals, characterized in that, The method includes: S1: Set a time sliding window in the time acquisition sequence, and receive real-time corrected product residual data from the trusted service cloud through the time sliding window; S2: Construct a sample dataset of corrected product residuals based on the received corrected product residual data; S3: Statistically analyze the corrected product residual sample dataset, obtain the statistical characteristics of the corrected product residual samples, divide the corrected product residual samples into multiple uniform distributions to smooth the empirical distribution, and obtain the sample distribution; S4: Use a pair of unimodal symmetrical distributions to perform preliminary envelope processing on the sample distribution to obtain a unimodal symmetrical intermediate distribution; S5: Gaussianize the unimodal symmetrical intermediate distribution using a one-sided envelope to obtain the Gaussianized envelope of the sample distribution, and obtain the envelope parameters from the Gaussianized envelope of the sample distribution. S6: Obtain the corrected product quality labeling information based on the envelope parameters; The preliminary envelope processing of the sample distribution using a pair of unimodal symmetrical distributions is specifically as follows: ， in, Left envelope, Right envelope, To correct the product residual distribution, To correct product defects; The specific method for obtaining the Gaussianized envelope of the sample distribution is as follows: Envelope conditions for planning a one-sided envelope: ， in, Let Gaussian envelope distribution function be used. It is a unimodal symmetric distribution function; The received trusted service cloud real-time corrected product residual data includes: satellite clock orbit, tropospheric and ionospheric corrected product residual data; The process of obtaining the Gaussianized envelope of the sample distribution also includes determining the Gaussian envelope boundary: For piecewise linear N The minimum value is obtained by Gaussianizing a uniformly distributed interval. σ The distribution function for each uniformly distributed interval is as follows: in, m express G s The mean, P ( x The expression represents the uniform distribution function over a certain interval; the specific interval of the uniform distribution is... Then the piecewise uniformly distributed CDF over this interval is as follows: and The boundary values ​​are uniformly distributed. When the envelope is the right envelope Let there be a Gaussian distribution, and the Gaussian distribution function be given by... x For a function greater than 0, it is a convex function. Therefore, taking the tangent line to the Gaussian distribution function as the boundary, the equation of the tangent line is... f ( x At the tangent point At that point, the equation of the tangent line satisfies: The variance of the Gaussian distribution is obtained from the tangent equation; the envelope σ is calculated using the tangent equation at the half-margin point; such a search requires an upper and lower bound on the solution, the lower bound... It is given by the following formula: σ i Indicates the first i σ of a Gaussian distribution calculated from a uniform distribution; Determine the upper limit of the Gaussian distribution σ. , x max for P ( x The lower limit of the last interval of ) Finally, the median of the upper and lower limits is used as the envelope of the final Gaussian distribution function. ; The left envelope With right envelope The processing method is the same: compare the left and right Gaussianization parameters, and take the larger value as the final envelope parameter. The final envelope parameter is: 。 2. A Gaussianization envelope device for correcting product residuals, characterized in that, The device is implemented based on the method of claim 1, and the device comprises: The product residual data acquisition unit is used to set a time sliding window in the time acquisition sequence and receive real-time product residual data from the trusted service cloud through the time sliding window. The corrected product residual sample dataset acquisition unit is used to construct the corrected product residual sample dataset based on the received corrected product residual data; The sample distribution acquisition unit is used to statistically analyze the corrected product residual sample dataset, obtain the statistical characteristics of the corrected product residual samples, divide the corrected product residual samples into multiple uniform distributions to smooth the empirical distribution, and obtain the sample distribution. The unit for obtaining the unimodal symmetric intermediate distribution is used to perform preliminary envelope processing on the sample distribution using a pair of unimodal symmetric distributions to obtain the unimodal symmetric intermediate distribution. The Gaussianized envelope acquisition unit of the sample distribution is used to perform Gaussianization processing on the single-peaked symmetrical intermediate distribution using a one-sided envelope, to obtain the Gaussianized envelope of the sample distribution, and to obtain the envelope parameters from the Gaussianized envelope of the sample distribution. A corrected product quality labeling information acquisition unit is used to acquire corrected product quality labeling information based on the envelope parameters; The single-peak symmetrical intermediate distribution acquisition unit is specifically as follows: ， in, Left envelope, Right envelope, To correct the product residual distribution, To correct product defects; The Gaussianized envelope acquisition unit for the sample distribution is specifically: Envelope conditions for planning a one-sided envelope: ， in, Let Gaussian envelope distribution function be used. It is a unimodal symmetric distribution function; The corrected product residual data acquisition unit includes: satellite clock orbit, tropospheric and ionospheric corrected product residual data.

3. A computer device, characterized in that: It includes a memory and a processor, wherein the memory stores a computer program, and when the processor runs the computer program stored in the memory, the processor executes the Gaussian envelope correction method for product residuals as described in claim 1.

4. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores a computer program that, when executed by a processor, provides a method for correcting Gaussian envelopes of product residuals as described in claim 1.

Images