Deep learning-based high-precision positioning and orientation method and system for single Beidou satellite signal receiver

Abstract

The invention discloses a high-precision positioning and orientation method and system for a single Beidou satellite signal receiver based on deep learning. The method comprises the following steps: S1, acquiring data of an original observation value of a base station and an original observation value of a moving station; step S2, preprocessing the acquired data; s3, constructing a frequency mixing single-difference model, and performing ambiguity fixation by using the model to complete baseline solution; and S4, high-precision positioning and high-precision orientation are realized. Through the satellite original observation data high-quality processing and frequency mixing difference ambiguity fixing technology based on deep learning, the influence of various interference factors in weak signal environments such as urban complex environments and dense trees is eliminated, high-precision positioning and orientation are achieved, the second-level convergence speed is achieved, and the anti-interference capacity is high.

Description

Technical Field

[0001] The present invention relates to the technical field of satellite signal data processing, and particularly to a high-precision positioning and orientation method and system for a single Beidou satellite signal receiver based on deep learning, which is specifically applied to fields such as single Beidou satellite positioning and navigation, and is particularly suitable for use in weak signal environments such as complex urban environments and dense forests. Background Art

[0002] Currently, single Beidou receivers generally adopt a dual-antenna design and preferentially use the Beidou B1 and B2 dual-frequency positioning and orientation mode to achieve positioning and orientation. This mode has the advantages of fast speed, high ambiguity fixation rate, and high precision. However, this technology is applicable to the ideal situation where satellite navigation signals are received well and both Beidou-3 B1 and B2 dual-frequency signals are available, and is not applicable to complex environments such as cities.

[0003] In weak signal environments such as complex urban environments and dense forests, the following main interference factors that affect positioning accuracy and positioning time exist:

[0004] 1. The frequency point conditions of the satellite observation data obtained by the receiver are very complex. There are differences in the tracking and acquisition of different frequency points of the same satellite, some frequency points are lost in tracking, and single, dual, and triple-frequency data may exist for all satellites in the same epoch.

[0005] 2. In weak signal environments such as complex urban environments and dense forests, the total number of visible satellites decreases, the multipath effect, and the quality of satellite observation data deteriorate, etc., all of which will have a negative impact on the on-chip dynamic real-time RTK positioning accuracy. In particular, the obvious influence of the Beidou satellite hybrid constellation (GEO / IGSO / MEO) on the geometric distribution relationship of visible satellites makes it relatively more complex to implement a stable RTK positioning technology in a single Beidou system compared to GNSS multi-systems. Summary of the Invention

[0006] In view of this, one of the objectives of the present invention is to provide a high-precision positioning and orientation method for a single Beidou satellite signal receiver based on deep learning. Through high-quality processing of satellite raw observation data and mixed-frequency differential ambiguity fixation technology based on deep learning, the influence of various interference factors in weak signal environments such as complex urban environments and dense forests is eliminated, achieving high-precision positioning and orientation, a second-level convergence speed, and strong anti-interference ability.

[0007] One of the objectives of the present invention is achieved through the following technical solutions:.

[0008] The high-precision positioning and orientation method for a single Beidou satellite signal receiver based on deep learning includes the following steps:

[0009] Step S1: Obtain the raw observation values of the reference station and the raw observation value data of the rover station;

[0010] Step S2: Preprocess the acquired data;

[0011] Step S3: Construct a mixed-frequency single-difference model and use this model to fix the ambiguity to complete baseline solution;

[0012] Step S4: Achieve high-precision positioning and high-precision orientation.

[0013] Furthermore, in the said Step S2, it includes preprocessing the data by using a deep learning-based quality control technique, and the deep learning-based quality control technique includes cycle slip detection based on the recurrent neural network (RNN) and chi-square detection based on the convolutional neural network (CNN);

[0014] The RNN-based cycle slip detection includes:

[0015] Extract the residual, signal strength, and Doppler frequency change from the original data as the input of the recurrent neural network (RNN); use the labeled data to train the model, adjust the model parameters to minimize the prediction error, and use cross-validation techniques to improve the generalization ability of the model; use the trained model to predict new observation data and identify cycle slip events; for the identified cycle slip events, take corrective, elimination, or downweighting measures to ensure the accuracy of the positioning result;

[0016] The CNN-based chi-square detection includes:

[0017] Use the convolutional neural network (CNN) and use the labeled data to train the model. The training data includes data that conforms to the expected distribution and data that does not conform to the expected distribution to effectively distinguish between the two types of data; eliminate and downweight the data that does not conform to the chi-square distribution.

[0018] Furthermore, after preprocessing the data by using a deep learning-based quality control technique, it also includes: constructing single-difference observation equations and double-difference observation equations, then performing extended Kalman filtering, and after the filtering is completed, fixing the wide-lane ambiguity. If the fixation is successful, then fix the narrow-lane ambiguity. If the fixation fails, enter the next epoch.

[0019] Furthermore, the construction and solution method of the mixed-frequency single-difference model is as follows:

[0020] Step S31: Use the data of single-frequency satellites, common-frequency satellites, and multi-frequency satellites to construct a mixed-frequency single-difference model;

[0021] Step S32: Obtain the sequential filtering floating-point solution and the floating-point ambiguity;

[0022] Step S33: Perform posterior quality control. If the verification fails, eliminate the observation value;

[0023] Step S34: If the verification passes, perform ambiguity fixing for the full set and subsets.

[0024] Further, in the step S31, the data includes carrier phase observations, pseudorange observations, and broadcast ephemeris data.

[0025] Further, in the step S32, the sequential filtering uses sequential filtering methods such as Kalman filtering to filter the single-difference model to obtain the floating-point solution and floating-point ambiguities.

[0026] Further, in the step S34,

[0027] Full-set ambiguity fixing is to fix the single-difference ambiguities of all satellites and use an algorithm to fix the floating-point ambiguities to integers;

[0028] Subset ambiguity fixing is to first fix the ambiguities of some satellites and gradually expand to full-set ambiguity fixing.

[0029] Further, when further processing the result data output by the mixed-frequency single-difference model, a mixed-frequency differential ambiguity fixing method that combines multiple combinations of partial ambiguities and wide and narrow lanes is adopted to comprehensively improve the fixing effect. The specific steps of the fixing method are as follows:

[0030] Determine whether it is multi-frequency fixing or single-frequency fixing. If it is multi-frequency fixing, the following method one is adopted. If it is single-frequency fixing, the following method two is adopted;

[0031] Method one: First, perform wide-lane ambiguity fixing. If the fixing is successful, output the corresponding result data for partial wide-lane ambiguity fixing. If it fails, end. If the partial wide-lane ambiguity fixing is successfully completed, output the corresponding result data for narrow-lane ambiguity fixing. If it fails, end. If the narrow-lane ambiguity fixing is successfully completed, output the corresponding result data for partial narrow-lane ambiguity fixing. If it fails, end. If the partial narrow-lane ambiguity fixing is successfully completed, the fixed ambiguities and positioning and orientation gains are achieved. If not, end.

[0032] Method two: First, perform non-combination ambiguity fixing. If the fixing is successful, output the corresponding result data for partial non-combination ambiguity fixing. If it fails, end. If the partial non-combination ambiguity fixing is successfully completed, the fixed ambiguities and positioning and orientation gains are achieved. If not, end.

[0033] Another object of the present invention is to provide a high-precision positioning and orientation system for a single Beidou satellite signal receiver based on deep learning. The system includes an ARM subsystem, a radio frequency circuit unit, a power supply unit, an interface unit, a clock unit, a WIFI unit, and a Beidou satellite signal receiving antenna;

[0034] The Beidou satellite signal received by the antenna first passes through a low-noise amplifier and a power divider, and then enters the RF circuit unit through a band-pass filter. After digitizing the signal, the RF circuit unit sends it to the ARM small system, and realizes the high-precision positioning and orientation method of the single Beidou satellite signal receiver based on deep learning as described above by executing the computer program deployed on the operation module in the ARM small system, and completes the real-time high-precision positioning and orientation calculation.

[0035] The beneficial effects of the present invention are as follows:

[0036] (1) Efficient signal processing algorithm: The present invention adopts a mixing single-difference model solution algorithm based on deep learning in a single Beidou receiver. Compared with traditional multi-satellite system receivers, it can more effectively achieve centimeter-level positioning accuracy, sub-degree-level orientation accuracy, second-level convergence speed, and strong anti-interference ability;

[0037] (2) Strong anti-interference ability: For the reception and processing of Beidou signals, the present invention adopts a mixing single-difference model solution algorithm based on deep learning, which can maintain good stability and accuracy in weak signal environments such as complex urban environments and dense trees, and achieve high-precision and fast positioning and orientation;

[0038] (3) High integration, low power consumption, and obvious cost advantages: This processing board adopts a highly integrated design, and realizes the receiver design based on a single-chip ARM CPU small system + necessary peripheral circuits, reduces the dependence on external components, reduces the system complexity, and achieves the capabilities of low power consumption, small volume, and low cost for a single board;

[0039] (4) Easy to develop and integrate: The present invention can be well compatible and docked with existing Beidou application systems, and supports multiple development environments and application platforms. It supports USB, serial port, and network port connections, enabling developers to upgrade products and expand functions more conveniently;

[0040] (5) Wide application range: The present invention is applicable to a variety of single Beidou application scenarios, including but not limited to single Beidou receivers, driverless assistance devices, personal navigation devices, and Internet of Things devices, etc.

[0041] Other advantages, objectives, and features of the present invention will be described to some extent in the subsequent specification, and to some extent, will be obvious to those skilled in the art based on the study of the following text, or can be taught from the practice of the present invention. The objectives and other advantages of the present invention can be realized and obtained through the following specification and the foregoing claims. Description of the Drawings

[0042] In order to make the objectives, technical solutions, and advantages of the present invention clearer, the present invention will be further described in detail below with reference to the drawings, where:

[0043] Figure 1 Schematic diagram of the method flow of the present invention;

[0044] Figure 2 Block diagram of the method principle of the present invention;

[0045] Figure 3 Block diagram of the principle of quality control technology based on deep learning;

[0046] Figure 4 Block diagram of the principle of the construction and solution method of the mixing single-difference model;

[0047] Figure 5 Quantitative comparison table of the effects of using the mixing single-difference model;

[0048] Figure 6 Flow chart of the mixing differential ambiguity fixing;

[0049] Figure 7 Schematic diagram of the overall architecture of the system;

[0050] Figure 8 Schematic diagram of the ARM small system architecture;

[0051] Figure 9 Implementation block diagram of the radio frequency circuit module;

[0052] Figure 10 Schematic diagram of the architecture of the power supply system;

[0053] Figure 11 Schematic diagram of the architecture of the WIFI unit;

[0054] Figure 12 Schematic diagram of the architecture of the clock unit;

[0055] Figure 13 Schematic diagram of the front panel interface unit circuit of the interface unit;

[0056] Figure 14 Schematic diagram of the rear panel interface unit circuit of the interface unit. Specific implementation manners

[0057] Hereinafter, preferred embodiments of the present invention will be described in detail with reference to the accompanying drawings. It should be understood that the preferred embodiments are only for illustrating the present invention, rather than limiting the protection scope of the present invention.

[0058] As Figure 1 and Figure 2 shown, the high-precision positioning and orientation method of the single Beidou satellite signal receiver based on deep learning of the present invention includes the following steps:

[0059] Step S1: Obtain the original observation values of the reference station and the original observation value data of the rover station;

[0060] Step S2: Preprocess the acquired data;

[0061] Step S3: Construct a mixed-frequency single-difference model and use this model for ambiguity fixing to complete baseline solution;

[0062] Step S4: Achieve high-precision positioning and high-precision orientation.

[0063] The above steps will be further described below. First, in Step S2, it includes preprocessing the data by using a deep learning-based quality control technique. The deep learning-based quality control technique includes cycle slip detection based on the recurrent neural network (RNN) and chi-square detection based on the convolutional neural network (CNN);

[0064] The cycle slip detection based on RNN includes:

[0065] Considering the consistency between the carrier phase and the pseudorange observation value, when the two are inconsistent, a cycle slip can be determined to have occurred. Therefore, cycle slip detection first extracts the residual, signal strength, and Doppler frequency change from the original data as the input of the recurrent neural network (RNN); uses the labeled data to train the model, adjusts the model parameters to minimize the prediction error, and uses techniques such as cross-validation to improve the generalization ability of the model; uses the trained model to predict new observation data and identify cycle slip events; for the identified cycle slip events, take measures such as correction, rejection, or downweighting to ensure the accuracy of the positioning result;

[0066] The chi-square detection based on CNN includes:

[0067] In the traditional chi-square detection, by comparing the difference between the observed value and the expected value, the accuracy is not high, and abnormal observed values cannot be effectively identified. The present invention utilizes the powerful non-linear fitting ability of the convolutional neural network (CNN) to construct a more complex statistical test method, uses the labeled data to train the model, and the training data includes data that conforms to the expected distribution and data that does not conform to the expected distribution, so as to effectively distinguish the two types of data. Reject and downweight the data that does not conform to the chi-square distribution.

[0068] In weak signal environments such as complex urban environments and dense trees, the total number of visible satellites decreases, multipath effects occur, and the quality of satellite observation data deteriorates, all of which will have a negative impact on the on-chip dynamic real-time RTK positioning accuracy. In particular, the obvious influence of the Beidou satellite hybrid constellation (GEO / IGSO / MEO) on the geometric distribution relationship of visible satellites makes it more complex to achieve stable RTK positioning technology in a single Beidou system compared to GNSS multi-systems. In view of the problem that conventional data preprocessing techniques cannot effectively and timely detect cycle slips and perform chi-square tests, the present invention creatively realizes high-quality processing of satellite raw observation data based on deep learning, greatly improving the effectiveness of cycle slip detection and chi-square tests, and achieving on-chip single-Beidou RTK real-time centimeter-level positioning in weak signal environments such as complex urban environments and dense trees.

[0069] As Figure 3 shown, after preprocessing the data using the quality control technology based on deep learning, it further includes: constructing single-difference observation equations and double-difference observation equations, then performing extended Kalman filtering. After the filtering is completed, the wide-lane ambiguity is fixed. If the fixation is successful, the narrow-lane ambiguity is fixed; if not, it proceeds to the next epoch.

[0070] As Figure 4 shown, in step S3, the construction and solution method of the mixed-frequency single-difference model are as follows:

[0071] Step S31: Construct a mixed-frequency single-difference model using the data of single-frequency satellites, co-frequency satellites, and multi-frequency satellites;

[0072] Step S32: Obtain the sequential filtering floating-point solution and floating-point ambiguity;

[0073] Step S33: Perform posterior quality control. If the verification fails, the observation value is rejected;

[0074] Step S34: If the verification passes, perform full-set and subset ambiguity fixing.

[0075] In the above step S31, a mixed-frequency single-difference model is constructed using the data of single-frequency satellites, co-frequency satellites, and multi-frequency satellites. The data sources include using the data of single-frequency satellites, co-frequency satellites, and multi-frequency satellites to construct a mixed-frequency single-difference model. These data include carrier phase observations, pseudorange observations, and broadcast ephemeris, etc. For multi-frequency satellites, using carrier phase observations of different frequencies can better eliminate errors such as ionospheric delay. Error processing is that when the user is close to the base station, the single-difference ionospheric delay and tropospheric delay can be approximated to zero. For longer baselines, the ionospheric delay effect can be eliminated using the dual-frequency ionosphere-free combination.

[0076] In step S32, sequential filtering uses sequential filtering methods such as Kalman filtering to filter the single-difference model, obtaining a floating-point solution and floating-point ambiguities. Further, smoothing processing can be used to perform multi-epoch smoothing on the floating-point ambiguities, reducing the influence of pseudorange noise and multipath errors and improving the accuracy of ambiguity estimation.

[0077] In step S33, posterior quality control is performed. If the verification fails, the observations are rejected; the verification method is to perform posterior quality control on the floating-point ambiguities and their residuals to check for outliers or errors. The anomaly handling includes: if the verification fails, the abnormal observations are rejected and filtering and ambiguity estimation are performed again; for satellite signals with frequent anomalies, the observation data of this satellite can be considered to be temporarily excluded to ensure the stability of the solution.

[0078] In step S34, the full-set ambiguity fixing is to fix the single-difference ambiguities of all satellites. Using algorithms such as integer least squares method (LAMBDA method), the floating-point ambiguities are fixed to integers. The fixed ambiguities are used to further improve the positioning accuracy and support high-precision positioning applications.

[0079] The subset ambiguity fixing is that in some cases, the ambiguities of some satellites can be fixed first and gradually extended to the full-set ambiguity fixing. The subset ambiguity fixing can improve the fixing success rate, especially when the observation conditions are poor or the number of satellites is small.

[0080] This technology realizes centimeter-level positioning accuracy, sub-degree-level orientation accuracy, second-level convergence speed, and strong anti-interference ability through the efficient error suppression of the mixed-frequency single-difference model and the geometric strengthening of the baseline constraint, providing a reliable solution for high-precision navigation in weak signal environments such as complex urban environments and dense trees. The ability improvements brought by the implementation of the technology are as Figure 5 shown in the list.

[0081] When further processing the result data output by the mixed-frequency single-difference model, in view of the problems of high difficulty and low fixing rate existing in the currently mainstream full-ambiguity fixing method, by adopting a mixed-frequency differential ambiguity fixing method that combines partial ambiguities and various combinations of wide and narrow lanes, the fixing effect is further comprehensively improved. The flow chart of the mixed-frequency differential ambiguity fixing is as Figure 6 shown. In this embodiment, the specific steps of the fixing method are as follows:

[0082] Determine whether it is multi-frequency fixing or single-frequency fixing. If it is multi-frequency fixing, method one below is adopted; if it is single-frequency fixing, method two below is adopted;

[0083] Method 1: First, perform wide-lane ambiguity fixing. If the fixing is successful, output the corresponding result data for partial wide-lane ambiguity fixing. If it fails, end. If partial wide-lane ambiguity fixing is successfully completed, output the corresponding result data for narrow-lane ambiguity fixing. If it fails, end. If narrow-lane ambiguity fixing is successfully completed, output the corresponding result data for partial narrow-lane ambiguity fixing. If it fails, end. If partial narrow-lane ambiguity fixing is successfully completed, the ambiguity is fixed and the positioning and orientation gain is achieved. If not, end.

[0084] Method 2: First, perform non-combination ambiguity fixing. If the fixing is successful, output the corresponding result data for partial non-combination ambiguity fixing. If it fails, end. If partial non-combination ambiguity fixing is successfully completed, the ambiguity is fixed and the positioning and orientation gain is achieved. If not, end.

[0085] The present invention combines partial ambiguity selection with wide and narrow lane hierarchical fixing, increasing the success rate of ambiguity fixing from 50%-70% of the traditional FAR (false recognition rate) to 85%-95%, shortening the convergence time by 70%-90%, achieving a dynamic scene positioning accuracy of 2-5 cm (RMS), and significantly enhancing the robustness in occluded and interfered environments. This technology provides an efficient and reliable ambiguity resolution scheme for high-precision positioning in urban complex environments and weak signal environments such as dense trees.

[0086] In the algorithm design of the present invention, in combination with the practical problems in actual applications, innovative designs are specifically carried out in key modules such as data preprocessing, mixed-frequency single-difference model construction, and ambiguity fixing in the algorithm: adopting high-quality processing technology for satellite raw observation data based on deep learning, and under the mixed-frequency single-difference algorithm model, adopting a mixed-frequency differential ambiguity fixing method combining partial ambiguity and various combinations of wide and narrow lanes, achieving centimeter-level positioning accuracy, sub-degree-level orientation accuracy, second-level convergence speed, and strong anti-interference ability in urban complex environments and weak signal environments such as dense trees.

[0087] Based on the design concept of the aforementioned positioning and orientation method, as Figure 7 shown, the present invention also provides a high-precision positioning and orientation system for a single Beidou satellite signal receiver based on deep learning. The system includes an ARM subsystem, a radio frequency circuit unit, a power supply unit, an interface unit, a clock unit, a WIFI unit, and a Beidou satellite signal receiving antenna;

[0088] The Beidou satellite signal received by the antenna first passes through a low-noise amplifier and a power divider, and then enters the RF circuit unit through a band-pass filter. After digitizing the signal, the RF circuit unit sends it to the ARM subsystem. Through the high-precision positioning and orientation method of the single Beidou satellite signal receiver based on deep learning deployed on the operation module in the ARM subsystem, real-time high-precision positioning and orientation calculation is completed. The following will further explain each component of the system:

[0089] (I) ARM Subsystem

[0090] In this embodiment, as Figure 8 shown, the minimum ARM peripheral system mainly includes 6 parts: ARM CPU, DDR3 memory, NAND Flash, USB interface, debug serial port, and debug network port.

[0091] a) ARM: With the ARM chip MCIMX6G2CVM05AB as the core, combined with peripheral clock, reset, DDR3 memory, NAND Flash, NOR Flash, USB interface, debug serial port, and debug network port, etc., it provides a control platform for the entire module. The ARM chip has a built-in operation unit and deploys the mixing single-difference model calculation algorithm based on deep learning on the local machine.

[0092] b) DDR3 memory: It is the large-capacity external memory of the ARM, enabling the ARM to meet the requirements of data operation storage and data exchange;

[0093] c) NAND Flash data storage: The NAND Flash with an MMC interface is mainly used to store the system startup program, system program, and user program, and the designed capacity is 32GB;

[0094] d) USB interface: Two USB interfaces, one is connected to the rear panel for PC debugging download / storing data download, and the other is connected to the baseband board for communication;

[0095] e) Debug serial port / communication serial port: It is composed of a TTL serial port provided by the ARM chip and then through an RS232 level conversion chip, and is mainly used for communication with the PC serial port and downloading programs during debugging;

[0096] f) Debug / communication network port: There are two network ports in total. One is connected to the rear panel and is mainly used for communication with the PC network port and downloading programs during debugging; when not used for debugging, it is used as a communication Ethernet port. The other is connected to the baseband board for communication.

[0097] (II) RF Circuit Unit

[0098] The RF circuit accesses two Beidou satellite signals. Each Beidou satellite signal first reaches the antenna module. The weak satellite signal received by the antenna is amplified by a low-noise amplifier and then sent to a power divider, and then the satellite signal is split through the corresponding filters. The multiple split satellite signals enter the RF circuit. After down-conversion, amplification, and analog-to-digital conversion inside the RF circuit, digital signals are output. The implementation block diagram is as shown in Figure 9 shown.

[0099] (3) Power supply system

[0100] The power supply system mainly includes an overvoltage protection circuit, a charging management circuit, a combiner monitor, and a power supply monitoring circuit, a total of 4 parts. The functions of each part are as follows:

[0101] a) Overvoltage protection circuit: When the input power supply exceeds the set threshold, cut off the power supply path;

[0102] b) Charging management circuit: Manage the charging of the device's lithium battery;

[0103] c) Combiner monitor: Select the input path of the power supply (external power supply or battery), and respond to external power-on and power-off signals;

[0104] d) Power supply monitoring circuit: Control battery charging and receive the shutdown command from the software.

[0105] The power supply system architecture of this embodiment is as shown in Figure 10 shown.

[0106] (4) WIFI unit

[0107] The WIFI unit designed in the present invention supports 802.11b / g / n and supports AP. When designing, the SDHC2 controller and UART3 controller of the CPU are used to communicate with the WIFI chip WL1837MOD to implement the WIFI function. This chip supports dual-band RF signals of 2.4G and 5.0G, integrates a radio frequency, power amplifier, clock, RF switch, filter, passive device, and power management unit, and the operating temperature is -40°C to 85°C; it supports WLAN processors and RF transceivers of IEEE standards 802.11a, 802.11b, 802.11g, and 802.11n, supports a 4-bit SDIO host interface, and supports simultaneous configuration of WLink 8 through STA and AP; it supports Bluetooth 4.1 and CSA2, and performs Bluetooth data transmission with the host through the UART interface; it supports coexistence of WIFI and Bluetooth; to meet the requirements of industrial environment use, a temperature-compensated crystal oscillator is integrated inside the chip. In the low-power mode, the CPU can turn off the power of the WIFI. The module block diagram is as shown in Figure 11 shown, meeting the protocol requirements.

[0108] (5) Clock unit

[0109] The 12MHz clock serves as the operating clock for the STM32F100C8 and is implemented using a crystal oscillator; the 32.768KHz serves as the low-speed clock for the ARM and WIFI chips and is implemented using crystal oscillators; the 24MHz serves as the system clock for the ARM chip and is implemented using a crystal oscillator; the 50MHz serves as the reference clock for the PHY chip and is provided by the CPU; the 26MHz is the reference clock for the WIFI chip and is integrated inside the chip. The schematic diagram of the clock scheme is as Figure 12 shown.

[0110] (VI) Interface Unit

[0111] The interface unit of this receiver includes two parts: the front panel interface and the rear panel interface.

[0112] As Figure 13 shown, the front panel interface unit circuit mainly includes three parts: 6-way LED drive signal output, 6-way key signal input, and 1-way I2C interface.

[0113] As Figure 14 shown, the rear panel interface circuit mainly includes four parts: 1-way USB interface, 2-way RS232 interfaces, 1-way RS485 interface, and 1-way 10M / 100M Ethernet interface.

[0114] It should be recognized that the embodiments of the present invention can be implemented or carried out by computer hardware, a combination of hardware and software, or by computer instructions stored in a non-transitory computer-readable memory. The methods can be implemented in a computer program using standard programming techniques - including a non-transitory computer-readable storage medium configured with a computer program, where the storage medium so configured causes the computer to operate in a specific and predefined manner - according to the methods and drawings described in the specific embodiments. Each program can be implemented in a high-level procedural or object-oriented programming language to communicate with the computer system. However, if desired, the program can be implemented in assembly or machine language. In any case, the language can be a compiled or interpreted language. Additionally, for this purpose, the program can run on a dedicated integrated circuit programmed for this purpose.

[0115] Furthermore, the operations of the processes described herein can be performed in any suitable order, unless otherwise indicated herein or otherwise clearly contradicted by the context. The processes described herein (or variations and / or combinations thereof) can be executed under the control of one or more computer systems configured with executable instructions, and can be implemented as code (e.g., executable instructions, one or more computer programs, or one or more applications) executed jointly on one or more processors, by hardware, or a combination thereof. The computer program includes multiple instructions executable by one or more processors.

[0116] Further, the method may be implemented in any type of computing platform operably connected to a suitable one, including but not limited to personal computers, minicomputers, mainframes, workstations, network or distributed computing environments, separate or integrated computer platforms, or communicating with charged particle tools or other imaging devices, etc. Aspects of the present invention may be implemented in machine-readable code stored on a non-transitory storage medium or device, whether removable or integrated into the computing platform, such as a hard disk, optical read and / or write storage medium, RAM, ROM, etc., such that it is readable by a programmable computer and, when read by the computer, can be used to configure and operate the computer to perform the processes described herein. In addition, the machine-readable code, or portions thereof, may be transmitted via a wired or wireless network. When such media includes instructions or programs that implement the steps described above in conjunction with a microprocessor or other data processor, the invention described herein includes these and other different types of non-transitory computer-readable storage media. When programmed according to the methods and techniques of the present invention, the present invention also includes the computer itself.

[0117] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and not to limit them. Although the present invention has been described in detail with reference to the preferred embodiments, those of ordinary skill in the art should understand that the technical solutions of the present invention can be modified or equivalently replaced without departing from the spirit and scope of the present technical solution, and they should all be covered by the scope of the claims of the present invention.

Claims

1. A high-precision positioning and orientation method for a single Beidou satellite signal receiver based on deep learning, characterized in that: The method includes the following steps: Step S1: Obtain the raw observation values of the reference station and the raw observation value data of the rover station; Step S2: Preprocess the obtained data; Step S3: Construct a mixed-frequency single-difference model, and use this model to fix the ambiguity to complete baseline solution; Step S4: Achieve high-precision positioning and high-precision orientation.

2. The high-precision positioning and orientation method of a single BeiDou satellite signal receiver based on deep learning according to claim 1, characterized in that: In the said Step S2, it includes preprocessing the data by using a quality control technology based on deep learning. The quality control technology based on deep learning includes cycle slip detection based on the recurrent neural network (RNN) and chi-square detection based on the convolutional neural network (CNN); The RNN-based cycle slip detection includes: Extract the residual, signal strength, and Doppler frequency change from the raw data as the input of the recurrent neural network (RNN); use the labeled data to train the model, adjust the model parameters to minimize the prediction error, and use cross-validation technology to improve the generalization ability of the model; use the trained model to predict the new observation data and identify cycle slip events; for the identified cycle slip events, take corrective, elimination, or downweighting measures to ensure the accuracy of the positioning result; The CNN-based chi-square detection includes: Use the convolutional neural network (CNN) and use the labeled data to train the model. The training data includes data that conforms to the expected distribution and data that does not conform to the expected distribution, so as to effectively distinguish the two types of data; eliminate and downweight the data that does not conform to the chi-square distribution.

3. The high-precision positioning and orientation method of a single Beidou satellite signal receiver based on deep learning according to claim 2, characterized in that: After preprocessing the data by using the quality control technology based on deep learning, it also includes: constructing single-difference observation equations and double-difference observation equations, then performing extended Kalman filtering. After the filtering is completed, fix the wide-lane ambiguity. If the fixation is successful, then fix the narrow-lane ambiguity. If the fixation fails, enter the next epoch.

4. The high-precision positioning and orientation method of a single Beidou satellite signal receiver based on deep learning according to claim 1, characterized in that: The construction and solution method of the mixed-frequency single-difference model is as follows: Step S31: Use the data of single-frequency satellites, co-frequency satellites, and multi-frequency satellites to construct a mixed-frequency single-difference model; Step S32: Obtain the sequential filtering float solution and the float ambiguity; Step S33: Perform posterior quality control. If the verification fails, eliminate the observation values; Step S34: If the verification passes, perform full-set and subset ambiguity fixation.

5. The high-precision positioning and orientation method of a single Beidou satellite signal receiver based on deep learning according to claim 4, characterized in that: In the said Step S31, the said data includes carrier phase observation values, pseudorange observation values, and broadcast ephemeris data.

6. The high-precision positioning and orientation system of a single Beidou satellite signal receiver based on deep learning according to claim 4, characterized in that: In the said Step S32, the sequential filtering uses sequential filtering methods such as Kalman filtering to filter the single-difference model to obtain the float solution and the float ambiguity.

7. The high-precision positioning and orientation system for a single Beidou satellite signal receiver based on deep learning according to claim 4, characterized in that: In the said Step S34, Full-set ambiguity fixation is to fix the single-difference ambiguities of all satellites and use an algorithm to fix the float ambiguity to an integer; Subset ambiguity fixation is to first fix the ambiguities of some satellites and gradually expand to full-set ambiguity fixation.

8. The high-precision positioning and orientation method of a single Beidou satellite signal receiver based on deep learning according to claim 4, characterized in that: When further processing the result data output by the mixed-frequency single-difference model, a mixed-frequency differential ambiguity fixation method combining partial ambiguities and various combinations of wide and narrow lanes is adopted to comprehensively improve the fixation effect. The specific steps of the fixation method are as follows: Determine whether it is multi-frequency fixation or single-frequency fixation. If it is multi-frequency fixation, then adopt Method 1 below. If it is single-frequency fixation, then adopt Method 2 below; Method 1: First, perform wide-lane ambiguity fixing. If the fixing is successful, output the corresponding result data for partial wide-lane ambiguity fixing. If it fails, end. If partial wide-lane ambiguity fixing is successfully completed, output the corresponding result data for narrow-lane ambiguity fixing. If it fails, end. If narrow-lane ambiguity fixing is successfully completed, output the corresponding result data for partial narrow-lane ambiguity fixing. If it fails, end. If partial narrow-lane ambiguity fixing is successfully completed, the ambiguity fixing and positioning and orientation gain are achieved. If not, end. Method 2: First, perform non-combination ambiguity fixing. If the fixing is successful, output the corresponding result data for partial non-combination ambiguity fixing. If it fails, end. If partial non-combination ambiguity fixing is successfully completed, the ambiguity fixing and positioning and orientation gain are achieved. If not, end.

9. The high-precision positioning and orientation system of a single Beidou satellite signal receiver based on deep learning according to claim 1, characterized in that: The system includes an ARM subsystem, a radio frequency circuit unit, a power supply unit, an interface unit, a clock unit, a WIFI unit, and a Beidou satellite signal receiving antenna; The Beidou satellite signal received by the antenna first passes through a low-noise amplifier and a power divider, and enters the radio frequency circuit unit through a band-pass filter. The radio frequency circuit unit digitizes the signal and sends it to the ARM subsystem. The high-precision positioning and orientation method of a single Beidou satellite signal receiver based on deep learning as described in any one of claims 1 to 8 is implemented by executing the computer program deployed on the operation module in the ARM subsystem, and real-time high-precision positioning and orientation calculation is completed.

Images