Automatic identification method of telemetry data abnormal value, electronic equipment and medium

Abstract

The invention discloses an automatic identification method for a telemetry data abnormal value, which comprises the following steps of: firstly, calculating a residual error between GNSS telemetry data acquired in real time and an extrapolation orbit, and then judging the category of the residual error to identify the abnormal value. According to the method, the abnormal value in the GNSS telemetry data can be automatically identified in real time without establishing a kinetic model and an observation value model, and the overall algorithm is simple and high in efficiency.

Description

Technical Field

[0001] The present invention relates to the field of aerospace technology, and in particular to an automatic recognition method, electronic equipment and medium for abnormal values ​​of telemetry data. Background Art

[0002] Detecting outliers in GNSS (Global Navigation Satellite System) data is a common problem in data processing. In GNSS observation data, outliers may be caused by factors such as multipath effects, satellite signal interruption, or equipment failure.

[0003] According to the classification of processing procedures, the methods for identifying GNSS data outliers are divided into preprocessing and postprocessing. Preprocessing is to detect data outliers through a simple verification method, which can only identify a part of the outliers, while most of the outliers need to be identified through least squares adjustment and its extension method. The outlier identification method based on post-processing establishes an accurate model, solves the algorithm equation, and then identifies the outliers in the GNSS observations based on the statistical analysis of the pseudo-post residuals. The calculation process of the post-processing method is complex and the real-time performance is poor. At the same time, due to the harsh space environment, satellite-borne GNSS receivers are more likely to produce outliers, which increases the difficulty of post-processing to a certain extent. Especially for giant low-orbit satellite constellations, the existing methods for identifying outliers based on least squares or sequential adjustment can hardly meet their computing power and real-time requirements. Summary of the invention

[0004] In view of some or all of the problems in the prior art, the present invention provides a method for automatically identifying abnormal values ​​of telemetry data in a first aspect, comprising:

[0005] Calculate the residuals between the real-time acquired GNSS telemetry data and the extrapolated orbit; and

[0006] The residuals are classified to identify outliers.

[0007] Further, determining the category of the residual includes:

[0008] Obtaining data features of the residual; and

[0009] The data feature is compared with data features of preset K residual subclasses to determine the category of the residual.

[0010] Furthermore, the preset K residual subclasses are obtained by the following steps:

[0011] Obtain reference orbit based on historical GNSS telemetry data;

[0012] calculating residuals of the reference orbit and the historical GNSS telemetry data, wherein the residuals include outliers; and

[0013] Through multiple iterations of cluster analysis, the residual is divided into K subclasses, at least one of which is an outlier.

[0014] Furthermore, obtaining the reference orbit includes:

[0015] Based on the historical GNSS telemetry data, dynamic fitting is performed to obtain the orbit parameters of the initial epoch; and

[0016] Based on the orbital parameters of the initial epoch and the dynamical model adopted for dynamical fitting, the reference orbit is obtained by integration.

[0017] Furthermore, K-Means algorithm is used to implement cluster analysis.

[0018] Furthermore, the number of residual subclasses is determined by the elbow method.

[0019] Furthermore, the automatic identification method further comprises:

[0020] If the number of outliers identified continuously is greater than the preset value, the data is clustered again to update the data features.

[0021] Based on the automatic identification method as described above, the second aspect of the present invention also provides an electronic device for automatically identifying abnormal values ​​of telemetry data, including a memory and a processor, wherein the memory is configured to store a computer program, and the computer program executes the automatic identification method as described above when the processor is running.

[0022] A third aspect of the present invention also provides a computer-readable storage medium for automatically identifying abnormal values ​​of telemetry data, which stores a computer program, and when the computer program is run on a processor, it executes the automatic identification method as described above.

[0023] Compared with the prior art, the present invention provides an automatic identification method for telemetry data outliers. It does not need to establish a dynamic model and an observation value model, avoids calculating a large normal equation, and can automatically identify outliers in GNSS telemetry data in real time. The automatic identification method has a simple algorithm, high efficiency, and is easy to explain. It can realize real-time automatic identification of outliers and is applicable to outlier identification of massive GNSS data of low-orbit giant constellations. BRIEF DESCRIPTION OF THE DRAWINGS

[0024] To further illustrate the above and other advantages and features of various embodiments of the present invention, a more specific description of various embodiments of the present invention will be presented with reference to the accompanying drawings. It will be understood that these drawings only depict typical embodiments of the present invention and are therefore not to be considered as limiting the scope thereof. In the accompanying drawings, for clarity, identical or corresponding parts will be represented by identical or similar reference numerals.

[0025] Figure 1 A schematic flow chart showing a method for automatically identifying abnormal values ​​in telemetry data according to an embodiment of the present invention;

[0026] Figures 2A-2C A schematic diagram showing the results of cluster analysis of each iteration of a method for automatically identifying abnormal values ​​of telemetry data according to an embodiment of the present invention; and

[0027] Figure 3 A schematic diagram showing the results of outlier identification according to an automatic identification method for telemetry data outliers based on an embodiment of the present invention is shown. DETAILED DESCRIPTION

[0028] In the following description, the present invention is described with reference to various embodiments. However, those skilled in the art will recognize that the various embodiments can be implemented without one or more specific details or with other replacement and / or additional methods or components. In other cases, well-known structures or operations are not shown or described in detail to avoid blurring the inventive point of the present invention. Similarly, for the purpose of explanation, specific numbers and configurations are set forth to provide a comprehensive understanding of the embodiments of the present invention. However, the present invention is not limited to these specific details. In addition, it should be understood that the various embodiments shown in the drawings are illustrative representations and are not necessarily drawn in correct proportions.

[0029] In this specification, reference to "one embodiment" or "the embodiment" means that a particular feature, structure, or characteristic described in conjunction with the embodiment is included in at least one embodiment of the present invention. The phrase "in one embodiment" appearing in various places in this specification does not necessarily all refer to the same embodiment.

[0030] It should be noted that the embodiments of the present invention describe the method steps in a specific order, but this is only for the purpose of illustrating the specific embodiment, rather than limiting the order of the steps. On the contrary, in different embodiments of the present invention, the order of the steps can be adjusted according to actual needs.

[0031] In the present invention, each module of the system according to the present invention can be implemented using software, hardware, firmware or a combination thereof. When the module is implemented using software, the function of the module can be implemented by a computer program flow, for example, the module can be implemented by a code segment (such as a code segment of a language such as C, C++) stored in a storage device (such as a hard disk, a memory, etc.), wherein the corresponding function of the module can be implemented when the code segment is executed by a processor. When the module is implemented using hardware, the function of the module can be implemented by setting a corresponding hardware structure, for example, the function of the module can be implemented by hardware programming a programmable device such as a field programmable gate array (FPGA), or the function of the module can be implemented by designing an application-specific integrated circuit (ASIC) including electronic devices such as a plurality of transistors, resistors and capacitors. When the module is implemented using firmware, the function of the module can be written into a read-only memory such as an EPROM or EEPROM of the device in the form of a program code, and the corresponding function of the module can be implemented when the program code is executed by a processor. In addition, some functions of the module may need to be implemented by separate hardware or by collaboration with the hardware, for example, the detection function is implemented by corresponding sensors (such as proximity sensors, acceleration sensors, gyroscopes, etc.), the signal transmission function is implemented by corresponding communication devices (such as Bluetooth devices, infrared communication devices, baseband communication devices, Wi-Fi communication devices, etc.), the output function is implemented by corresponding output devices (such as displays, speakers, etc.), and so on.

[0032] The purpose of the present invention is to design an automatic and real-time GNSS telemetry data outlier identification system and method, which does not require the establishment of complex mathematical and physical models and solution algorithm equations, but instead extracts GNSS telemetry data features based on machine learning to identify outliers.

[0033] Satellites in orbit are affected by orbital perturbations, and their orbits show certain regularities. Therefore, the observation data of the GNSS receivers on the satellites also have certain orbit-related characteristics. By subtracting the corresponding reference orbit from the GNSS data acquired by the onboard receiver, the orbital characteristics in the GNSS data are deducted. After deducting the orbital characteristics in the data, ideally, the residual data only has statistical characteristics, and theoretically, the residual should show a normal distribution. Therefore, automatic identification of outliers can be achieved through simple statistical analysis, that is, by performing cluster analysis on a large amount of residual data, extracting the characteristics of GNSS telemetry data, dividing the data into K subclasses, and then automatically identifying outliers by judging which subclass the real-time data belongs to.

[0034] The technical solution of the present invention is further described below in conjunction with the accompanying drawings of the embodiments.

[0035] Figure 1A flow chart showing a method for automatically identifying abnormal values ​​in telemetry data according to an embodiment of the present invention is shown. Figure 1 As shown, a method for automatically identifying abnormal values ​​of telemetry data includes:

[0036] First, in step 101, residual subclasses are obtained. As mentioned above, it is necessary to perform machine learning based on historical data, perform cluster analysis on the residual data, extract the features of the GNSS telemetry data, and divide the data into K residual subclasses. In one embodiment of the present invention, a reference orbit is first obtained based on historical GNSS telemetry data, and then the residual of the reference orbit and the historical GNSS telemetry data is calculated. Finally, through multiple iterations of cluster analysis, the residual is divided into K subclasses, wherein the residual includes an outlier, so the K residual subclasses include at least one subclass that is an outlier.

[0037] In one embodiment of the present invention, the reference orbit is obtained through an orbit fitting and dynamic integration process. Specifically, first, dynamic fitting is performed based on historical GNSS telemetry data to obtain the orbit parameters of the initial epoch, and then the reference orbit is obtained by integration based on the orbit parameters of the initial epoch and the dynamic model used in the dynamic fitting. It should be understood that in some other embodiments of the present invention, other common methods in the field can also be used to calculate the reference orbit. For example, the orbit data can be output by the orbit module, which will not be repeated here.

[0038] In one embodiment of the present invention, considering the complexity of on-orbit GNSS telemetry data, the automatic identification method adopts the K-Means method to divide the residual data into K subclasses, so that the data within the subclass has good statistical characteristics. At the same time, the data between subclasses have obvious characteristic differences, so that outliers can be identified. The number of residual subclasses, that is, the selection of K value, is the key to the K-Means method. If the K value is small, it may lead to the inability to separate outliers, while a large K value will increase the calculation burden and affect real-time performance. Therefore, in one embodiment of the present invention, a threshold is set according to the performance of the satellite receiver, and the K value is automatically calculated in combination with the El bow method. In one embodiment of the present invention, multiple iterations can be performed until the outliers in the residual are separated. It should be noted that before each iteration, the outliers identified in the previous iteration should be eliminated;

[0039] Next, in step 102, a real-time residual is calculated. The residual between the real-time acquired GNSS telemetry data and the extrapolated orbit is calculated, wherein the method for acquiring the extrapolated orbit is similar to that of the reference orbit and will not be described in detail here; and

[0040] Finally, in step 103, outliers are identified. The category of the residual is determined to identify outliers. As mentioned above, the subclasses obtained by the K-Means method have good statistical characteristics. Therefore, in one embodiment of the present invention, the data features of the residual are first calculated, and then the data features are compared with the data features of the preset K residual subclasses to determine the category of the residual, and to determine and remove outliers in the residual.

[0041] In one embodiment of the present invention, the GNSS telemetry data characteristics are monitored in real time, and if multiple outliers are continuously identified, cluster analysis is performed on the data again to update the data characteristics, that is, steps 101 to 103 are repeated.

[0042] In order to verify the effect of the automatic identification method. The 48-hour GNSS telemetry data of an experimental satellite was used for testing, and the orbit sampling rate of the telemetry data was 1 second. The first 24 hours of data were used for model training, and the last 24 hours of data were used for testing. First, the orbital characteristics were deducted according to the method described above. Then outliers were identified. Considering that the residual data characteristics are relatively simple after deducting the orbital characteristics, the maximum number of subclasses is set to 10, and the normalized variance is 0.9. The K value is calculated using the Elbow algorithm. Since the experimental satellite is equipped with a single-frequency receiver, it is significantly affected by the ionosphere. In addition, affected by the atmospheric drag in the low-orbit satellite dynamics model, the error of the dynamic integral orbit is also large. Therefore, according to the mission planning requirements, the threshold of the residual is set to 500m. The K-Means method is used for cluster analysis. The results of the first cluster analysis are as follows. Figure 2A As shown in the figure, there are 4 subclasses in total. The black “*” indicates the center point of the subclass. It can be seen that the first cluster analysis identified 3 groups of outliers, all with magnitudes of 10 6 After removing the outliers, the remaining residual data is subjected to a second cluster analysis. The results are as follows Figure 2B As shown in the figure, there are 3 subclasses in total, and the black "*" indicates the center point of the subclass. It can be seen that the second cluster analysis identified 2 outlier subclasses with an order of magnitude of 10 3 m. Repeat the above process and iterate until all outliers are identified. Figure 2C As shown in the figure, the last iteration of cluster analysis eliminated all outliers, and there were 6 subclasses in total. The black "*" represents the center point of the subclass. Since the observations of single-frequency receivers are greatly affected by the ionosphere, there is an obvious system deviation in the radial direction of the orbit. Therefore, the observation error of the onboard GNSS receiver is large. According to the design indicators of this type of receiver, the observation noise is large. This also proves that the calculation results of the automatic identification method are consistent with the actual situation. Finally, the orbit residual data of the last 24 hours is substituted to determine which subclass the data belongs to, so as to realize the automatic identification of GNSS data outliers. The results are shown in the figure. Figure 3As shown in the figure, the line-bound part represents the data after removing outliers, and the black “*” represents Figure 2C The six sub-class centers in . For clarity, the outliers are not marked in the figure. Theoretically, "normal" data should be distributed in Figure 2C It can be seen from the center points of the six subclasses that the automatic recognition method has a high recognition rate for outliers.

[0043] Based on the automatic identification method as described above, the present invention also provides an automatic identification system for abnormal values ​​of telemetry data, which comprises:

[0044] A cluster analysis module, which is used to determine the subclasses of the residuals after the track characteristics are proposed based on historical data according to the method as described above, that is, cluster analysis is performed by the K-Meas algorithm;

[0045] The orbit calculation module is used to determine the extrapolated orbit or reference orbit based on the GNSS telemetry data. That is, firstly, the dynamic fitting is performed according to the GNSS telemetry data to obtain the orbit parameters of the initial epoch, and then the reference orbit is obtained by integration based on the orbit parameters of the initial epoch and the dynamic model adopted by the dynamic fitting;

[0046] A residual calculation module, which is used to calculate the residual between the GNSS telemetry data and the extrapolated orbit; and

[0047] The identification module is used to determine the subclass of the residual to identify outliers.

[0048] Book

[0049] Based on the automatic identification method as described above, the present invention also provides an electronic device for automatically identifying abnormal values ​​of telemetry data, which includes a memory and a processor, wherein the memory is configured to store a computer program, and the computer program executes the automatic identification method as described above when the processor is running.

[0050] The present invention further provides a computer-readable storage medium for automatically identifying abnormal values ​​of telemetry data, which stores a computer program, and the computer program executes the automatic identification method as described above when running on a processor. The computer-readable storage medium can be a tangible device that can hold and store instructions used by an instruction execution device, such as an electrical storage device, a magnetic storage device, an optical storage device, an electromagnetic storage device, a semiconductor storage device, or any suitable combination of the above. Specifically, the computer-readable storage medium includes, but is not limited to: a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), a static random access memory (SRAM), a portable compact disk read-only memory (CD-ROM), a digital versatile disk (DVD), a memory stick, a floppy disk, a mechanical encoding device, a punch card or a protruding structure in a groove on which instructions are stored, and any suitable combination of the above.

[0051] In an embodiment of the present invention, the computer program can be downloaded from a computer-readable storage medium to each computing / processing device, or downloaded to an external computer or external storage device through a network, such as the Internet, a local area network, a wide area network, and / or a wireless network. The network may include copper transmission cables, optical fiber transmission, wireless transmission, routers, firewalls, switches, gateway computers, and / or edge servers. The network adapter card or network interface in each computing / processing device receives the computer program from the network and forwards it for storage in a computer-readable storage medium in each computing / processing device.

[0052] In an embodiment of the present invention, the computer program may be an assembly instruction, an instruction set architecture (ISA) instruction, a machine instruction, a machine-related instruction, a microcode, a firmware instruction, a state setting data, or a source code or an object code written in any combination of one or more programming languages, the programming language including an object-oriented programming language, such as Smalltalk, C++, etc., and a conventional procedural programming language, such as "C" language or a similar programming language. The computer program may be executed entirely on a user's computer, partially on a user's computer, as an independent software package, partially on a user's computer, partially on a remote computer, or entirely on a remote computer or server. In the case of a remote computer, the remote computer may be connected to the user's computer via any type of network, including a local area network (LAN) or a wide area network (WAN), or may be connected to an external computer (e.g., connected via the Internet using an Internet service provider). In some embodiments, an electronic circuit, such as a programmable logic circuit, a field programmable gate array (FPGA), or a programmable logic array (PLA), may be customized by using the state information of the computer program, and the electronic circuit may execute the computer program.

[0053] Although various embodiments of the present invention are described above, it should be understood that they are presented as examples only and not as limitations. It is obvious to those skilled in the relevant art that various combinations, modifications and changes can be made thereto without departing from the spirit and scope of the present invention. Therefore, the breadth and scope of the present invention disclosed herein should not be limited by the exemplary embodiments disclosed above, but should only be defined according to the attached claims and their equivalents.

Claims

1. A method for automatically identifying abnormal values ​​in telemetry data, characterized in that: Includes steps: Calculate the residuals between the real-time acquired GNSS telemetry data and the extrapolated orbit; and The residuals are classified to identify outliers.

2. The automatic identification method according to claim 1, characterized in that: Determining the category of the residual includes the steps of: Obtaining data features of the residual; and The data feature is compared with data features of preset K residual subclasses to determine the category of the residual.

3. The automatic identification method according to claim 2, characterized in that: The preset K residual subclasses are obtained by the following steps: Obtain reference orbit based on historical GNSS telemetry data; calculating residuals of the reference orbit and the historical GNSS telemetry data, wherein the residuals include outliers; and Through multiple iterations of cluster analysis, the residual is divided into K subclasses, at least one of which is an outlier.

4. The automatic identification method according to claim 3, characterized in that: Obtaining the reference track includes the following steps: Based on the historical GNSS telemetry data, dynamic fitting is performed to obtain the orbit parameters of the initial epoch; as well as Based on the orbital parameters of the initial epoch and the dynamical model adopted for dynamical fitting, the reference orbit is obtained by integration.

5. The automatic identification method according to claim 3, characterized in that: The K-means algorithm was used for cluster analysis.

6. The automatic identification method according to claim 3, characterized in that: The number of residual subclasses is determined by the elbow method.

7. The automatic identification method according to claim 1, characterized in that: The automatic identification method further comprises the steps of: If the number of outliers identified continuously is greater than the preset value, the data is clustered again to update the data features.

8. An electronic device for automatically identifying abnormal values ​​of telemetry data, characterized in that: The method comprises a memory and a processor, wherein the memory is configured to store a computer program, and the computer program executes the automatic identification method according to any one of claims 1 to 7 when the processor is running.

9. A computer-readable storage medium for automatically identifying abnormal values ​​in telemetry data, characterized in that: A computer program is stored, and when the computer program is run on a processor, the automatic identification method according to any one of claims 1 to 7 is executed.

Images