Random model construction method and device considering dependence precision factors between satellite and station

Abstract

The invention discloses a random model construction method and device considering a satellite and station dependence precision factor, and the method comprises the steps: selecting all available observation residual errors in a sliding window in each observation time period, calculating the precision factor of observation station dependence according to a pseudo-range residual error and a phase residual error, and calculating the precision factor of satellite dependence according to the phase residual error; and constructing a time-varying random model considering the precision factor dependent between observation stations and the precision factor dependent on satellites. According to the time-varying random model constructed by the method, the precision factor dependent on observation stations and the precision factor dependent on satellites are considered, so that the precision of satellite clock error estimation is improved.

Description

Technical Field

[0001] The present invention relates to the field of satellite navigation and positioning technology, and in particular to a method and device for constructing a random model taking into account satellite and station dependent precision factors. Background Art

[0002] With the development of real-time global navigation satellite system products, real-time precise point positioning has become an important tool and has been widely used in many scientific and engineering fields. Real-time satellite clock error is one of the key products for user-side real-time precise point positioning and has attracted much attention in recent years. Currently, the International GNSS Service (IGS) publicly provides real-time satellite products, including satellite clock error. Since the satellite clock error predicted for a few minutes using the commonly used quadratic polynomial model is not negligible, the error may exceed 0.2 nanoseconds when predicting 20 minutes, and the error may be even greater when predicting 1 hour using a complex wavelet neural network model. Satellite clock error is usually estimated in real-time mode using a globally / regionally distributed reference station observation stream to control the error of satellite clock error prediction. In order to meet the demand for high-precision clock error estimation, people have designed a variety of strategies to increase the data processing speed by utilizing the capabilities of high-performance computing devices and designing accelerated computing methods. However, there are relatively few studies on the stochastic model of real-time satellite clock error estimation, which includes the expected value and discreteness of observation noise.

[0003] Similar to precise point positioning and relative positioning, observations are usually weighted according to the satellite elevation angle or carrier-to-noise ratio, but these methods cannot fully reflect the magnitude of the observation noise. Although the observation stations used for satellite clock error estimation are usually equipped with geodetic receivers, the observation quality of each station varies due to differences in instrument design and observation environment. If the differences between observation stations are ignored, the accuracy of the estimated satellite clock error will be affected.

[0004] Although the GNSS observation function model has been significantly improved in recent years, there are still some unresolved systematic errors, which are particularly evident in GNSS data processing due to the different development stages of different satellite systems. In addition, the accuracy and stability of the satellite orbits are updated every few hours and then used to predict subsequent satellite clock error estimates. In real-time satellite clock error estimation, the estimated satellite clock error can partially absorb the satellite orbit error, but the actual size varies with the size of the reference network and the type of satellite. The above errors should all be optimally considered in the stochastic model, otherwise they will affect the data processing accuracy. By comparing the solutions of different weighting strategies in real-time precise single-point positioning, it is found that appropriately weighting the system based on the quality of satellite products can shorten the convergence time and improve coordinate reproducibility. However, due to the delay problem of reference satellite products, this method is not applicable in actual real-time analysis.

[0005] In summary, random models play an important role in precise GNSS data processing, but the existing traditional weighting strategy based on satellite elevation angle and carrier-to-noise ratio cannot fully represent the contribution of observations between each observation station and satellite in network data processing, resulting in low accuracy of real-time satellite clock error estimation. Summary of the invention

[0006] The purpose of the present invention is to overcome the above-mentioned defects and problems existing in the prior art, and to provide a method and device for constructing a random model taking into account the precision factor of satellite and station dependence. The method takes into account the precision factor of dependence between observation stations and the precision factor of satellite dependence, thereby improving the accuracy of real-time satellite clock error estimation.

[0007] To achieve the above objectives, the technical solution of the present invention is: a random model construction method considering satellite and station dependent precision factors, comprising:

[0008] In each observation period, all available observation residuals are selected in the sliding window, and the inter-station dependent precision factor is calculated based on the pseudorange residuals and phase residuals, and the satellite dependent precision factor is calculated based on the phase residuals;

[0009] A time-varying random model is constructed that takes into account the inter-station dependent dilution of precision and the satellite dependent dilution of precision.

[0010] The time-varying random model is:

[0011]

[0012] σ ori =sσ0 / sin(ele);

[0013] In the formula, σ new is the standard deviation; w r is the precision factor of the inter-station dependence; w s is the satellite-dependent precision factor; σ ori is the standard deviation calculated by the original random model; σ0 is the standard deviation of the unprocessed phase observation in the zenith direction; ele is the satellite elevation angle; s is the ratio of the pseudorange to the phase observation error.

[0014] The calculation method of the precision factor of the inter-observation station dependence is:

[0015] Divide all pseudorange and phase observations by s / sin(ele) to obtain normalized observation residuals;

[0016] The root mean square error for each observation station is calculated using the normalized observation residuals associated with each observation station in the sliding window;

[0017] Calculate the overall root mean square error of all normalized observation residuals in the sliding window;

[0018] Calculate the precision factor Ratio of the observation site r :

[0019] Ratio r =RMS r / RMS all (STA);

[0020] Where, RMS r is the root mean square error; RMS all (SAT) is the overall root mean square error of all normalized observation residuals;

[0021] The precision factor of inter-observation station dependence is calculated based on the precision factor of observation station dependence.

[0022] The precision factor of the inter-station dependency is:

[0023]

[0024] When calculating the dilution of precision of inter-station dependence, the sliding window length is set to 1 day.

[0025] The satellite-dependent dilution of precision is calculated as follows:

[0026] Calculate the RMS error of the unnormalized phase residuals for a station-satellite pair

[0027]

[0028] In the formula, v i is the i-th unnormalized phase residual in the sliding window; n is the number of phase residuals in the sliding window;

[0029] Calculate the effect of satellite correlation errors on observations between station-satellite pairs

[0030]

[0031] In the formula, is the total RMS error of all normalized phase residuals; ave(·) is the average value of the phase residuals in the sliding window;

[0032] Calculate the satellite-dependent dilution of precision w s :

[0033]

[0034] Where m is the number of observation stations that observed the satellite within the sliding window; obsnum r is the phase residual number of the observation station.

[0035] When calculating satellite-dependent DOP, the sliding window length is set to 30 minutes.

[0036] A random model construction device considering satellite and station-dependent precision factors, the device is applied to the above method, the device comprising:

[0037] The precision factor calculation module is used for each observation period to select all available observation residuals in the sliding window, and calculate the precision factor of inter-observation station dependence based on the pseudorange residuals and phase residuals, and calculate the precision factor of satellite dependence based on the phase residuals;

[0038] The random model building module is used to build a time-varying random model that takes into account the precision factor of inter-observation station dependence and the precision factor of satellite dependence.

[0039] A random model building device considering satellite and station dependent precision factors, including a memory and a processor;

[0040] The memory is used to store computer program code and transmit the computer program code to the processor;

[0041] The processor is used to execute the method according to the instructions in the computer program code.

[0042] A computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the method described above is implemented.

[0043] Compared with the prior art, the present invention has the following beneficial effects:

[0044] In a random model construction method and device considering satellite and station-dependent precision factors of the present invention, the method selects all available observation residuals in a sliding window for each observation period, and calculates the precision factor of observation station dependence based on pseudorange residuals and phase residuals, and calculates the precision factor of satellite dependence based on phase residuals, and constructs a time-varying random model considering the precision factor of observation station dependence and the precision factor of satellite dependence. The present invention considers the precision factor of observation station dependence and the precision factor of satellite dependence, thereby improving the accuracy of satellite clock error estimation. BRIEF DESCRIPTION OF THE DRAWINGS

[0045] Figure 1 The present invention is a flowchart of a random model construction method that takes into account satellite and station-dependent precision factors.

[0046] Figure 2 It is a structural block diagram of a random model building device of the present invention that takes into account satellite and station-dependent precision factors.

[0047] Figure 3 It is a structural block diagram of a random model building device that takes into account satellite and station-dependent precision factors of the present invention. DETAILED DESCRIPTION

[0048] The present invention is further described in detail below in conjunction with the accompanying drawings and specific implementation methods.

[0049] See also Figure 1 The present invention provides a random model construction method considering satellite and station dependent precision factors, comprising:

[0050] S1. In each observation period, all available observation residuals are selected in the sliding window, and the inter-station dependent precision factor is calculated based on the pseudorange residuals and phase residuals, and the satellite dependent precision factor is calculated based on the phase residuals;

[0051] S2. Construct a time-varying random model that considers the precision factor dependent between observation stations and the precision factor dependent on satellites.

[0052] Satellite clock error estimation is performed continuously and all observation residuals are calculated after parameter estimation.

[0053] Furthermore, the time-varying random model is:

[0054]

[0055] σ ori =sσ0 / sin(ele);

[0056] In the formula, σ new is the standard deviation; w r is the precision factor of the inter-station dependence; w s is the satellite-dependent precision factor; σ ori is the standard deviation calculated by the original random model, which adopts a highly dependent weighting strategy; σ0 is the standard deviation of the unprocessed GPS / BDS phase observation value in the zenith direction; ele is the satellite elevation angle; S is the ratio of the pseudorange to the phase observation error. Since the measurement accuracy of the phase is 1 / 100 of the pseudorange, for the pseudorange, s = 100, and for the phase, s = 1.

[0057] Furthermore, the calculation method of the precision factor of the dependence between the observation stations is:

[0058] Divide all pseudorange and phase observations by s / sin(ele) to obtain normalized observation residuals;

[0059] The root mean square error for each observation station is calculated using the normalized observation residuals associated with each observation station in the sliding window;

[0060] Calculate the overall root mean square error of all normalized observation residuals in the sliding window;

[0061] Calculate the precision factor Ratio of the observation site r :

[0062] Ratio r =RMS r / RMS all (STA);

[0063] Where, RMS r is the root mean square error; RMS all (SAT) is the overall root mean square error of all normalized observation residuals;

[0064] The precision factor of inter-observation station dependence is calculated based on the precision factor of observation station dependence.

[0065] Furthermore, the precision factor of the inter-observation station dependency is:

[0066]

[0067] Furthermore, since the precision factor of the dependence between observation stations is mainly determined by the quality of the instrument and the observation environment, the sliding window length is set to 1 day when calculating the precision factor of the dependence between observation stations.

[0068] Furthermore, the calculation method of the satellite-dependent precision factor is:

[0069] Calculate the RMS error of the unnormalized phase residuals for a station-satellite pair

[0070]

[0071] In the formula, v i is the i-th unnormalized phase residual in the sliding window; n is the number of phase residuals in the sliding window;

[0072] Calculate the effect of satellite correlation errors on observations between station-satellite pairs

[0073]

[0074] In the formula, is the total RMS error of all normalized phase residuals; ave(·) is the average value of the phase residuals in the sliding window;

[0075] Calculate the satellite-dependent dilution of precision w s :

[0076]

[0077] Where m is the number of observation stations that observed the satellite within the sliding window; obsnum r is the phase residual number of the observation station.

[0078] Furthermore, the satellite-dependent precision factor in the real-time satellite clock error estimation is mainly determined by the satellite orbit error, which changes every time the satellite orbit is updated. Therefore, when calculating the satellite-dependent precision factor, the sliding window length is set to 30 minutes.

[0079] To account for the differences in accuracy of all satellites, the satellite-dependent DOP is calculated for each satellite separately. To reliably separate the station-dependent DOP from the satellite-dependent DOP, a real-time GPS satellite clock error estimation process is performed, during which the satellite orbits are fixed to the post-processed GNSS orbits and the station-dependent DOP is calculated at each moment to assess its evolution over time. In practice, due to its stability, the station-dependent DOP may be updated every few days in post-processing mode, after which the GPS / BDS satellite clock errors are estimated in real-time mode while the satellite-dependent DOP is calculated. It is beneficial to calculate the satellite-dependent DOP simultaneously with the satellite products in server-side data processing and provide them to the user together with the orbit and clock products.

[0080] The present invention proposes an improved method for a time-varying random model that takes into account the precision factors between observation stations and satellite dependence. Compared with the traditional weighted strategy, the improved random model takes into account the differences between the observation precisions of different observation stations and satellites, thereby improving the accuracy of real-time satellite clock error estimation. The data analysis results based on 7 days of GPS / BDS observations show that there are differences in the precision factors between different observation stations. In addition, the precision factors between satellites vary greatly and are closely related to the imported satellite orbits. When using GNSS orbits updated every hour, the accuracy of GPS, BDS IGSO satellites and BDS MEO satellite clock errors is improved by 39.8%, 43.3% and 28.5%, respectively.

[0081] See also Figure 2 A random model construction device considering satellite and station-dependent precision factors, the device is applied to the random model construction method considering satellite and station-dependent precision factors as described above, the device comprising:

[0082] The precision factor calculation module is used for each observation period to select all available observation residuals in the sliding window, and calculate the precision factor of inter-observation station dependence based on the pseudorange residuals and phase residuals, and calculate the precision factor of satellite dependence based on the phase residuals;

[0083] The random model building module is used to build a time-varying random model that takes into account the precision factor of inter-observation station dependence and the precision factor of satellite dependence.

[0084] See also Figure 3 ,The present invention also provides a random model construction device considering satellite and station dependent precision factors, including a memory and a processor;

[0085] The memory is used to store computer program code and transmit the computer program code to the processor;

[0086] The processor is used to execute the above-mentioned random model construction method considering the satellite and station-dependent precision factors according to the instructions in the computer program code.

[0087] The present invention also provides a computer-readable storage medium having a computer program stored thereon. When the computer program is executed by a processor, the method for constructing a random model that takes into account the satellite and station-dependent precision factors is implemented.

[0088] Generally speaking, the computer instructions for implementing the method of the present invention may be carried in any combination of one or more computer-readable storage media. Non-transitory computer-readable storage media may include any computer-readable media, except for the signal itself that is temporarily propagating.

[0089] The computer-readable storage medium may be, for example, but not limited to, an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, device or device, or any combination thereof. More specific examples of computer-readable storage media (a non-exhaustive list) include: an electrical connection with one or more wires, a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EKROM or flash memory), an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination thereof. In the present invention, a computer-readable storage medium may be any tangible medium containing or storing a program that may be used by or in conjunction with an instruction execution system, device, or device.

[0090] Computer program code for performing the operation of the present invention can be written in one or more programming languages ​​or a combination thereof, including object-oriented programming languages ​​such as Java, Smalltalk, C++, and conventional procedural programming languages ​​such as "C" or similar programming languages, in particular, Python suitable for neural network computing and platform frameworks based on TensorFlow, PyTorch, etc. can be used. The program code can be executed entirely on the user's computer, partially on the user's computer, as an independent software package, partially on the user's computer and partially on a remote computer, or entirely on a remote computer or server. In the case of a remote computer, the remote computer can be connected to the user's computer or to an external computer (for example, using an Internet service provider to connect via the Internet) through any type of network, including a local area network (LAN) or a wide area network (WAN).

[0091] The above-mentioned device and non-temporary computer-readable storage medium can refer to the specific description of a random model construction method considering satellite and station-dependent precision factors and its beneficial effects, which will not be repeated here.

[0092] Although the embodiments of the present invention have been shown and described above, it should be understood that the above embodiments are exemplary and should not be construed as limitations on the present invention. A person skilled in the art may change, modify, replace and vary the above embodiments within the scope of the present invention.

Claims

1. A random model construction method considering satellite and station dependent precision factors, characterized in that: include: In each observation period, all available observation residuals are selected in the sliding window, and the inter-station dependent precision factor is calculated based on the pseudorange residuals and phase residuals, and the satellite dependent precision factor is calculated based on the phase residuals; A time-varying random model is constructed that takes into account the inter-station dependent dilution of precision and the satellite dependent dilution of precision.

2. A random model construction method considering satellite and station-to-station dependent precision factors according to claim 1, characterized in that: The time-varying random model is: s ori =sσ0 / sin(ele); In the formula, σ new is the standard deviation; w r is the precision factor of the inter-station dependence; w s is the satellite-dependent precision factor; σ ori is the standard deviation calculated by the original random model; σ0 is the standard deviation of the unprocessed phase observation in the zenith direction; ele is the satellite elevation angle; S is the ratio of the pseudorange to the phase observation error.

3. The method for constructing a random model considering satellite and station-to-station dependent precision factors according to claim 2, characterized in that: The calculation method of the precision factor of the inter-observation station dependence is: Divide all pseudorange and phase observations by s / sin(ele) to obtain normalized observation residuals; The root mean square error for each observation station is calculated using the normalized observation residuals associated with each observation station in the sliding window; Calculate the overall root mean square error of all normalized observation residuals in the sliding window; Calculate the precision factor Ratio of the observation site r : Ratio r =RMS r / RMS all (STA) Where, RMS r is the root mean square error; RMS all (SAT) is the overall root mean square error of all normalized observation residuals; The precision factor of inter-observation station dependence is calculated based on the precision factor of observation station dependence.

4. The method for constructing a random model considering satellite and station-dependent precision factors according to claim 3, characterized in that: The precision factor of the inter-station dependency is:

5. The method for constructing a random model considering satellite and station-to-station dependent precision factors according to claim 3, characterized in that: When calculating the dilution of precision of inter-station dependence, the sliding window length is set to 1 day.

6. The method for constructing a random model considering satellite and station-dependent precision factors according to claim 2, characterized in that: The satellite-dependent dilution of precision is calculated as follows: Calculate the RMS error of the unnormalized phase residuals for a station-satellite pair In the formula, v i is the i-th unnormalized phase residual in the sliding window; n is the number of phase residuals in the sliding window; Calculate the effect of satellite correlation errors on observations between station-satellite pairs In the formula, is the total RMS error of all normalized phase residuals; ave(·) is the average value of the phase residuals in the sliding window; Calculate the satellite-dependent dilution of precision w s : Where m is the number of observation stations that observed the satellite within the sliding window; obsnum r is the phase residual number of the observation station.

7. The method for constructing a random model considering satellite and station-dependent precision factors according to claim 6, characterized in that: When calculating satellite-dependent DOP, the sliding window length is set to 30 minutes.

8. A random model construction device considering satellite and station dependent precision factors, characterized in that: The device is applied to the method described in any one of claims 1 to 7, and the device comprises: The precision factor calculation module is used for each observation period to select all available observation residuals in the sliding window, and calculate the precision factor of inter-observation station dependence based on the pseudorange residuals and phase residuals, and calculate the precision factor of satellite dependence based on the phase residuals; The random model building module is used to build a time-varying random model that takes into account the precision factor of inter-observation station dependence and the precision factor of satellite dependence.

9. A random model construction device considering satellite and station dependent precision factors, characterized in that: including memory and processor; The memory is used to store computer program code and transmit the computer program code to the processor; The processor is configured to execute the method according to any one of claims 1 to 7 according to instructions in the computer program code.

10. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the method according to any one of claims 1 to 7 is implemented.

Images