Position prediction method and device for unmanned aerial vehicle based on beidou differential positioning

Abstract

This invention relates to a method and apparatus for predicting the position of an unmanned aerial vehicle (UAV) based on BeiDou differential positioning. The method includes: performing differential processing on BeiDou satellite data to iteratively update the baseline estimate to obtain an optimal estimate; analyzing the BeiDou satellite data to obtain the coordinate information and error correction value of the nearest reference station; calculating the UAV's three-dimensional coordinate information based on the optimal estimate, the coordinate information of the nearest reference station, and the error correction value; generating a training set based on historical data of the three-dimensional coordinate information; inputting the training set into a pre-constructed neural network model for iterative training to obtain a three-dimensional position prediction model for the UAV. This invention utilizes a neural network to predict the three-dimensional coordinate position of the UAV within a future period, which can verify the accuracy of the UAV's real-time position coordinates and improve the reliability of the device's positioning. It can also monitor the UAV's operational status when the BeiDou satellite signal is briefly lost, preventing loss of control due to position loss.

Description

Technical Field

[0001] This invention belongs to the field of unmanned aerial vehicle (UAV) technology, specifically relating to a method and device for predicting the position of an unmanned aerial vehicle based on BeiDou differential positioning. Background Technology

[0002] In recent years, with the increasing maturity of drone technology, it has played an increasingly important role in industrial and agricultural production as well as the military. Drones have advantages such as flexible flight, good stealth, strong survivability, low cost, ease of use, and strong adaptability, enabling them to adapt to various harsh environments and complete various tasks. Positioning and navigation of drones are crucial for the successful completion of various missions.

[0003] The advancement of UAV positioning technology has gone through the following stages: Early UAVs used PCM radio command uplinks for positioning and control within line-of-sight, and utilized pre-programmed flight plans to automatically climb to their operating altitude along a predetermined flight path. Flight altitude, airspeed, flight time, and range were all controlled by the program. As the complexity of UAV missions increased, real-time positioning and flight path control became necessary. Therefore, GPS systems were required on UAVs, leading to further improvements in autonomous navigation capabilities. In semi-autonomous navigation, operators only needed to issue flight path requests to the UAV autopilot, eliminating the need for continuous uplink control.

[0004] In related technologies, with the improvement and commercial application of my country's independently developed BeiDou Navigation Satellite System, the application of UAV positioning and control systems based on BeiDou navigation is becoming increasingly widespread. UAVs are equipped with various sensor systems, including the BeiDou positioning system, to transmit data via real-time uplink and downlink and to communicate with various ground relay platforms in real time, enabling UAVs to achieve swarm operations and beyond-line-of-sight positioning capabilities. Advances in BeiDou positioning technology have improved the positioning accuracy of UAVs during missions, shortened uplink control time, increased transmission bandwidth, extended "silent" time for UAVs, and enhanced the reliability and anti-interference capabilities of the navigation system.

[0005] Currently, UAV positioning generally employs visual methods, inertial navigation, and GPS positioning. Among these, visual methods have low depth calculation accuracy and are susceptible to the effects of fog, dust, natural light, and heat radiation. While inertial navigation methods are more resistant to interference from the external environment, their errors have a cumulative effect, requiring timely compensation and correction. GPS positioning only provides timing and positioning functions and cannot meet the real-time monitoring needs of UAVs during flight. Therefore, existing UAV positioning methods cannot satisfy their diverse requirements. Summary of the Invention

[0006] In view of this, the purpose of this invention is to overcome the shortcomings of the prior art and provide a method and device for predicting the position of unmanned aerial vehicles based on Beidou differential positioning, so as to solve the problem that the existing UAV position positioning methods cannot meet its various needs.

[0007] To achieve the above objectives, the present invention adopts the following technical solution: a method for predicting the position of an unmanned aerial vehicle based on BeiDou differential positioning, comprising:

[0008] Receive BeiDou satellite data sent by the BeiDou reference station;

[0009] The coordinates and error correction values ​​of the nearest reference station are obtained by analyzing the BeiDou satellite data.

[0010] The BeiDou satellite data is subjected to differential processing to iteratively update the baseline estimate and obtain the optimal estimate.

[0011] The three-dimensional coordinate information of the UAV is calculated based on the optimal estimate, the coordinate information of the nearest reference station, and the error correction value;

[0012] A training set is generated based on historical data of the three-dimensional coordinate information. The training set is then input into a pre-built neural network model for iterative training to obtain a three-dimensional position prediction model for the UAV.

[0013] Furthermore, the differential processing of BeiDou satellite data to iteratively update the baseline estimate and obtain the optimal estimate includes:

[0014] Based on the BeiDou satellite data, pseudorange observations and carrier phase observations were obtained;

[0015] Based on the pseudorange and carrier phase observations, baseline calculation is performed according to the pre-constructed double-difference observation model to obtain the differential error equation;

[0016] The Adam algorithm is used to solve the difference error equation and update the baseline estimate until the baseline estimate meets the preset conditions, at which point the update stops and the optimal estimate is obtained.

[0017] Furthermore, the step of using the Adam algorithm to solve the difference error equation and update the baseline estimate until the baseline estimate meets a preset condition, thereby obtaining the optimal estimate, includes:

[0018] Based on the difference error equation, the first-order gradient estimation expression and the second-order gradient estimation expression of the baseline estimate are obtained, and an iterative formula is established.

[0019] The iterative formula is used to correct the bias in the first and second gradients.

[0020] The baseline estimate is updated based on the deviation between the corrected first and second gradients;

[0021] Determine whether the baseline estimate meets the preset requirements. If it does, output the estimate as the optimal estimate; otherwise, continue updating.

[0022] Furthermore, the differential error equations include: pseudorange differential error equations and carrier phase differential error equations;

[0023] The pseudorange difference error equation is as follows:

[0024]

[0025] The carrier phase differential error equation is as follows

[0026]

[0027] Where W is the error term vector of the pseudorange difference, and A is the pseudorange difference coefficient matrix. Let H be the pseudorange difference vector, where H is a vector composed of measurable or computable constant terms, V is the error term vector of the carrier phase difference, and B is the coefficient matrix of the carrier phase difference. The carrier phase differential vector, Let L be the double-difference integer ambiguity of the carrier phase, L be the vector consisting of measurable or computable constant terms, and E be the identity matrix.

[0028] Furthermore, the iterative update stops when the baseline estimate meets the following preset requirements;

[0029]

[0030] in, Given the significance level α and the dimension m of the baseline error vector, determined through a chi-square distribution, the optimal estimate of the baseline vector is obtained.

[0031] Furthermore, based on the three-dimensional coordinates of the nearest reference station, the optimal estimate, and the error correction value, the three-dimensional coordinate information of the UAV is calculated in the following manner:

[0032]

[0033] Where, x r Let be the three-dimensional coordinate vector obtained from the nearest reference station, c be the error correction value, and x be the real-time three-dimensional coordinates of the unmanned aerial vehicle.

[0034] Furthermore, the pre-built neural network model includes a generator and a discriminator; the step of inputting the training set into the pre-built neural network model for iterative training to obtain the UAV 3D position prediction model includes:

[0035] Select n three-dimensional coordinate vectors {x} from the historical data of the three-dimensional coordinate information at epochs (t+1) to (t+n). t+1 ,x t+2 ,…,x t+n As a positive example, the generator takes k noise vectors as input and generates k negative examples, maximizes the loss function and updates the discriminator parameters;

[0036] With the parameters of the discriminator fixed, the BeiDou satellite data at epoch t is input into the generator to obtain the predicted three-dimensional coordinates from epoch (t+1) to (t+n). The predicted three-dimensional coordinates are then input into the discriminator, which outputs a truthiness parameter. The generator adjusts its parameters to make this truthiness parameter as large as possible, thus obtaining the generator parameters.

[0037] After network training is completed, the trained UAV 3D position prediction model is used to generate predicted 3D coordinates of the UAV for a future period of time based on the current BeiDou satellite data.

[0038] Furthermore, it also includes:

[0039] In each iteration, the parameters of the generator are fixed first, and only the parameters of the discriminator are updated.

[0040] This application provides a position prediction device for an unmanned aerial vehicle based on BeiDou differential positioning, comprising:

[0041] The system includes a satellite antenna, a communication module, a BeiDou receiver motherboard, and a data storage module. The satellite antenna is connected to the BeiDou receiver motherboard via the communication module, and the data storage module is also connected to the BeiDou receiver motherboard.

[0042] The satellite antenna is used to search for BeiDou satellite signals and receive BeiDou satellite data;

[0043] The communication module is used to transmit the BeiDou satellite data to the BeiDou receiver motherboard;

[0044] The Beidou receiver motherboard is used to perform differential processing on the Beidou satellite data to obtain the real-time three-dimensional coordinate information of the UAV and the predicted three-dimensional coordinate information for a future period of time.

[0045] The communication module is also used to receive information transmitted between the ground control station and the reference station, and to send the real-time three-dimensional coordinate information of the UAV and the three-dimensional coordinate prediction information for a period of time to the ground control station.

[0046] The data storage module is used to store BeiDou satellite data, UAV three-dimensional coordinate information, and UAV three-dimensional position prediction model.

[0047] Furthermore, it also includes:

[0048] A preamplifier is used to amplify the BeiDou satellite signals received by the satellite antenna.

[0049] The input terminal of the preamplifier is connected to the satellite antenna, and the output terminal of the preamplifier is connected to the main board of the Beidou receiver.

[0050] The beneficial effects that can be achieved by adopting the above technical solution in this invention include:

[0051] This invention provides a method and apparatus for predicting the position of an unmanned aerial vehicle (UAV) based on BeiDou differential positioning. First, BeiDou satellite data is acquired, then processed to obtain the UAV's three-dimensional coordinate information. Historical data of the three-dimensional coordinate information is input into a neural network for training, enabling the model to predict the UAV's three-dimensional coordinate position over a future period. This verifies the accuracy of the UAV's real-time position coordinates and further improves the reliability of the positioning device. When the BeiDou satellite signal is briefly lost, it can also be used to track the UAV's operational status, preventing loss of control due to position loss. Attached Figure Description

[0052] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0053] Figure 1 This is a schematic diagram illustrating the steps of the position prediction method for unmanned aerial vehicles based on BeiDou differential positioning according to the present invention.

[0054] Figure 2 A flowchart illustrating the baseline estimation provided by this invention;

[0055] Figure 3 A schematic diagram illustrating the iterative training process of the neural network model provided by this invention;

[0056] Figure 4 A schematic diagram of the position prediction device for an unmanned aerial vehicle using BeiDou differential positioning provided by the present invention;

[0057] Figure 5This is a schematic diagram of the structural environment for implementing the BeiDou differential positioning method for unmanned aerial vehicles according to the present invention. Detailed Implementation

[0058] To make the objectives, technical solutions, and advantages of this invention clearer, the technical solutions of this invention will be described in detail below. Obviously, the described embodiments are merely some embodiments of this invention, and not all embodiments. Based on the embodiments of this invention, all other implementation methods obtained by those skilled in the art without creative effort are within the scope of protection of this invention.

[0059] The following describes, with reference to the accompanying drawings, a specific method and apparatus for predicting the position of an unmanned aerial vehicle based on BeiDou differential positioning, as provided in an embodiment of this application.

[0060] like Figure 1 As shown in the embodiments of this application, the position prediction method for unmanned aerial vehicles based on BeiDou differential positioning includes:

[0061] S101 receives BeiDou satellite data sent by the BeiDou reference station;

[0062] S102, Analyze the BeiDou satellite data to obtain the coordinate information and error correction value of the nearest reference station;

[0063] It should be noted that obtaining the coordinate information of the reference station and determining the error correction value through BeiDou satellite data can be achieved using existing technologies, which will not be elaborated here.

[0064] S103, perform differential processing on the BeiDou satellite data to iteratively update the baseline estimate and obtain the optimal estimate;

[0065] In some embodiments, such as Figure 2 As shown, the differential processing of BeiDou satellite data to iteratively update the baseline estimate and obtain the optimal estimate includes:

[0066] Based on the BeiDou satellite data, pseudorange observations and carrier phase observations were obtained;

[0067] Based on the pseudorange and carrier phase observations, baseline calculation is performed according to the pre-constructed double-difference observation model to obtain the differential error equation;

[0068] The Adam algorithm is used to solve the difference error equation and update the baseline estimate until the baseline estimate meets the preset conditions, at which point the update stops and the optimal estimate is obtained.

[0069] It is understood that this application can employ various differential modes to fully eliminate or reduce errors generated during satellite signal propagation, such as inter-frequency differential, inter-satellite differential, and inter-epoch differential. By filtering BeiDou satellite signals through differential modes, the accuracy of BeiDou satellite data can be improved.

[0070] In some embodiments, the step of using the Adam algorithm to solve the difference error equation and update the baseline estimate until the baseline estimate meets a preset condition to stop updating, thereby obtaining the optimal estimate, includes:

[0071] Based on the difference error equation, the first-order gradient estimation expression and the second-order gradient estimation expression of the baseline estimate are obtained, and an iterative formula is established.

[0072] The iterative formula is used to correct the bias in the first and second gradients.

[0073] The baseline estimate is updated based on the deviation between the corrected first and second gradients;

[0074] Determine whether the baseline estimate meets the preset requirements. If it does, output the estimate as the optimal estimate; otherwise, continue updating.

[0075] Specifically, this application uses the Adam algorithm to obtain the optimal estimate of the baseline, where the error equation of the pseudorange difference is expressed in matrix form as follows:

[0076]

[0077] Where W is the error term vector of the pseudorange difference, and A is the pseudorange difference coefficient matrix. H is a pseudorange difference vector, where H is a vector consisting of measurable or computable constant terms.

[0078] Pseudorange difference is currently the most widely used type of difference. It involves observing all satellites at a base station, calculating the true distance from each satellite to the base station at each moment based on the precise coordinates of the base station and the coordinates of each satellite, comparing it with the measured pseudorange, obtaining the pseudorange correction, and transmitting it to the rover receiver to correct the measured pseudorange and improve positioning accuracy.

[0079] The error equation for carrier phase differential is shown below:

[0080]

[0081] Where V is the error term vector of the carrier phase difference, and B is the coefficient matrix of the carrier phase difference. The carrier phase differential vector, Let L be the double-difference integer ambiguity of the carrier phase, L be the vector consisting of measurable or computable constant terms, and E be the identity matrix.

[0082] Carrier phase differential is used to improve the accuracy of position data obtained from satellite-based positioning systems (Global Navigation Satellite System, GNSS). In addition to the information contained in the signal, carrier phase differential uses measurements of the phase of the carrier signal and relies on a single reference station or interpolated virtual station for real-time correction to provide positioning accuracy up to the centimeter level.

[0083] The baseline estimate is updated using the Adam algorithm. Iterative updates stop when the deviation of the baseline vector estimate meets the following condition.

[0084]

[0085] in, Given the significance level α and the dimension m of the baseline error vector, the baseline vector is determined using a chi-square distribution. The optimal estimate of the baseline vector is then obtained.

[0086] S104. Calculate the three-dimensional coordinate information of the UAV based on the optimal estimate, the coordinate information of the nearest reference station, and the error correction value.

[0087] Specifically, in this application, the three-dimensional coordinate vector obtained from the nearest reference station is x. r If the coordinate parameter correction value is c, then the real-time three-dimensional coordinates x of the unmanned aerial vehicle can be calculated using the following formula.

[0088]

[0089] S105, a training set is generated based on historical data of the three-dimensional coordinate information, and the training set is input into a pre-constructed neural network model for iterative training to obtain a three-dimensional position prediction model for the UAV.

[0090] Specifically, such as Figure 3 As shown, the pre-built neural network model described in this application includes a generator and a discriminator; the step of inputting the training set into the pre-built neural network model for iterative training to obtain a UAV 3D position prediction model includes:

[0091] In each iteration, the parameters of the generator are fixed first, and only the parameters of the discriminator are updated.

[0092] Select n three-dimensional coordinate vectors {x} from the historical data of the three-dimensional coordinate information at epochs (t+1) to (t+n). t+1 ,x t+2 ,…,x t+nAs a positive example, the generator takes k noise vectors as input and generates k negative examples, maximizes the loss function and updates the discriminator parameters;

[0093] With the parameters of the discriminator fixed, the BeiDou satellite data at epoch t is input into the generator to obtain the predicted three-dimensional coordinates from epoch (t+1) to (t+n). The predicted three-dimensional coordinates are then input into the discriminator, which outputs a truthiness parameter. The generator adjusts its parameters to make this truthiness parameter as large as possible, thus obtaining the generator parameters.

[0094] After network training is completed, the trained UAV 3D position prediction model is used to generate predicted 3D coordinates of the UAV for a future period of time based on the current BeiDou satellite data.

[0095] It should be noted that in this application, the generator network D and the discriminator network G are first initialized. Then, in each iteration, the parameters of the generator are fixed first, and only the parameters of the discriminator are updated. From the historical values ​​of the high-precision three-dimensional coordinates of the UAV obtained in this invention, n three-dimensional coordinate vectors {x} are selected from epochs (t+1) to (t+n). t+1 ,x t+2 ,…,x t+n As positive examples, m noise vectors are input into the generator to generate m negative examples, maximizing the loss function and updating the discriminator parameters. Next, the discriminator parameters are fixed, while the generator is changed. The BeiDou navigation data at epoch t is input into the generator to obtain the predicted 3D coordinates for epochs (t+1) to (t+n). These predicted 3D coordinates are then input into the discriminator, which provides a truthiness parameter. The generator adjusts its parameters to maximize this truthiness parameter. After the network training is complete, the trained generator can be used to generate predicted 3D positions of UAVs over a future period based on the current BeiDou satellite navigation data.

[0096] like Figure 4 As shown, this application embodiment provides a position prediction device for an unmanned aerial vehicle based on Beidou differential positioning, including: a satellite antenna 11, a communication module 12, a Beidou receiver motherboard 13 and a data storage module 14. The satellite antenna 11 is connected to the Beidou receiver motherboard 13 through the communication module 12, and the data storage module 14 is connected to the Beidou receiver motherboard 13.

[0097] The satellite antenna 11 is used to search for BeiDou satellite signals and receive BeiDou satellite data;

[0098] The communication module 12 is used to transmit the BeiDou satellite data to the BeiDou receiver motherboard;

[0099] The Beidou receiver motherboard 13 is used to perform differential processing on the Beidou satellite data to obtain the real-time three-dimensional coordinate information of the UAV and the predicted three-dimensional coordinate information for a future period of time.

[0100] The communication module 12 is also used to receive information transmitted between the ground control station and the reference station, and to send the real-time three-dimensional coordinate information of the UAV and the three-dimensional coordinate prediction information for a period of time to the ground control station.

[0101] The data storage module 14 is used to store BeiDou satellite data, UAV three-dimensional coordinate information, and UAV three-dimensional position prediction model.

[0102] In some embodiments, the position prediction device for unmanned aerial vehicles based on BeiDou differential positioning provided in this application further includes:

[0103] The preamplifier 15 is used to amplify the BeiDou satellite signals received by the satellite antenna.

[0104] The input terminal of the preamplifier 15 is connected to the satellite antenna 11, and the output terminal of the preamplifier 15 is connected to the Beidou receiver motherboard 13.

[0105] The satellite antenna 11 in this application is used to search for, track, and receive BeiDou satellite signals, and can simultaneously receive satellite signals in four frequency bands: B1I, B3I, B1C, and B2a. The preamplifier 15 is used to amplify the weak signals received by the satellite antenna 11. The communication module 12 consists of a BeiDou short message communication module, a 4G / 5G mobile network communication module, and a radio communication module. It is used to receive information transmitted from the ground control station and the reference station, and to return the real-time three-dimensional coordinates processed by the BeiDou receiver motherboard 13 to the remote control terminal and the ground control station. The BeiDou receiver motherboard 13 is used to parse and process the BeiDou satellite signals received by the antenna and convert them into BeiDou satellite data, calculate the real-time three-dimensional position coordinates and predicted coordinate values ​​of the UAV, and transmit the position information to the communication module. The data storage module is used to store the UAV's three-dimensional position data and BeiDou satellite navigation files, and to store trained network models, etc.

[0106] The position prediction device for unmanned aerial vehicles based on BeiDou differential positioning provided in this application also includes a power module (not shown in the figure) for providing power for the normal operation of the position prediction device for unmanned aerial vehicles based on BeiDou differential positioning.

[0107] The position prediction device for unmanned aerial vehicles based on BeiDou differential positioning provided in this application is also equipped with an electromagnetic shielding shell (not shown in the figure): used to resist external electromagnetic interference.

[0108] This application utilizes a position prediction device for an unmanned aerial vehicle based on BeiDou differential positioning to track BeiDou satellite signals and process and analyze the obtained satellite signals. For example... Figure 4 As shown, it mainly includes a satellite antenna 11, a preamplifier 15, an electromagnetic shielding shell, a communication module 12, a Beidou receiver motherboard 13, a data storage module 14, and a power module. The UAV position information obtained by the UAV positioning device is transmitted to the remote control terminal through the communication link.

[0109] This application provides a computer device, including a memory 1 and a processor 2, and may further include a network interface 3. The memory 1 stores a computer program. The memory 1 may include non-permanent memory in the form of computer-readable media, random access memory (RAM), and / or non-volatile memory, such as read-only memory (ROM) or flash RAM. The computer device stores an operating system 4. The memory is an example of a computer-readable medium. When the computer program is executed by the processor, it causes the processor to execute a position prediction method for an unmanned aerial vehicle based on BeiDou differential positioning. Figure 5 The structure shown is merely a block diagram of a portion of the structure related to the present application and does not constitute a limitation on the computer device to which the present application is applied. Specific computer devices may include more or fewer components than those shown in the figure, or combine certain components, or have different component arrangements.

[0110] In one embodiment, the position prediction method for unmanned aerial vehicles based on BeiDou differential positioning provided in this application can be implemented as a computer program, which can be implemented in, for example... Figure 5 It runs on the computer device shown.

[0111] In some embodiments, when the computer program is executed by the processor, it receives BeiDou satellite data sent by a BeiDou reference station; performs differential processing on the BeiDou satellite data to iteratively update the baseline estimate to obtain the optimal estimate; analyzes the BeiDou satellite data to obtain the coordinate information and error correction value of the nearest reference station; calculates the three-dimensional coordinate information of the UAV based on the optimal estimate, the coordinate information of the nearest reference station, and the error correction value; generates a training set based on historical data of the three-dimensional coordinate information; inputs the training set into a pre-built neural network model for iterative training to obtain a three-dimensional position prediction model for the UAV.

[0112] This application also provides a computer storage medium, examples of which include, but are not limited to, phase-change memory (PRAM), static random access memory (SRAM), dynamic random access memory (DRAM), other types of random access memory (RAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), flash memory or other memory technologies, CD-ROM, digital optical disc (DVD) or other optical storage, magnetic tape storage or other magnetic storage devices, or any other non-transfer medium, which can be used to store information that can be accessed by a computing device.

[0113] In some embodiments, the present invention also proposes a computer-readable storage medium storing a computer program, which, when executed by a processor, receives BeiDou satellite data sent by a BeiDou reference station; performs differential processing on the BeiDou satellite data to iteratively update the baseline estimate to obtain an optimal estimate; analyzes the BeiDou satellite data to obtain the coordinate information and error correction value of the nearest reference station; calculates the three-dimensional coordinate information of the UAV based on the optimal estimate, the coordinate information of the nearest reference station, and the error correction value; generates a training set based on historical data of the three-dimensional coordinate information; inputs the training set into a pre-built neural network model for iterative training to obtain a three-dimensional position prediction model for the UAV.

[0114] This application has the following beneficial effects:

[0115] This invention employs multiple differential modes to effectively eliminate or reduce errors generated during satellite signal propagation, thereby improving the positioning accuracy of unmanned aerial vehicles.

[0116] This invention integrates BeiDou short message communication, 4G / 5G mobile network communication, and radio communication, enriching the communication links, helping to achieve real-time location reporting, and enhancing communication reliability.

[0117] This invention utilizes a trained neural network model to predict the three-dimensional coordinates of an unmanned aerial vehicle (UAV) over a future period, which verifies the accuracy of the UAV's real-time position coordinates and further improves the reliability of the device's positioning. When the BeiDou satellite signal is briefly lost, it can also be used to track the UAV's operational status, preventing loss of control due to position loss.

[0118] It is understood that the method embodiments provided above correspond to the device embodiments described above, and the specific details can be referred to each other, which will not be repeated here.

[0119] Those skilled in the art will understand that embodiments of this application can be provided as methods, systems, or computer program products. Therefore, this application can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, this application can take the form of a computer program product implemented on one or more computer-usable storage media (including, but not limited to, disk storage and optical storage) containing computer-usable program code.

[0120] This application is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of this application. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, generate instructions for implementing the flowchart... Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0121] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to operate in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction methods implemented in a process. Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.

[0122] These computer program instructions may also be loaded onto a computer or other programmable data processing equipment to cause a series of operational steps to be performed on the computer or other programmable equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable equipment for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.

[0123] The above description is merely a specific embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the technical scope disclosed in the present invention should be included within the scope of protection of the present invention. Therefore, the scope of protection of the present invention should be determined by the scope of the claims.

Claims

1. A method for predicting the position of an unmanned aerial vehicle based on BeiDou differential positioning, characterized in that, include: Receive BeiDou satellite data sent by the BeiDou reference station; The coordinates and error correction values ​​of the nearest reference station are obtained by analyzing the BeiDou satellite data. The BeiDou satellite data is subjected to differential processing to iteratively update the baseline estimate and obtain the optimal estimate. The three-dimensional coordinate information of the unmanned aerial vehicle is calculated based on the optimal estimate, the coordinate information of the nearest reference station, and the error correction value. A training set is generated based on historical data of the three-dimensional coordinate information. The training set is then input into a pre-built neural network model for iterative training to obtain a three-dimensional position prediction model for the unmanned aerial vehicle. The step of performing differential processing on the BeiDou satellite data to iteratively update the baseline estimate and obtain the optimal estimate includes: Based on the BeiDou satellite data, pseudorange observations and carrier phase observations were obtained; Based on the pseudorange and carrier phase observations, baseline calculation is performed according to the pre-constructed double-difference observation model to obtain the differential error equation; The Adam algorithm is used to solve the difference error equation and update the baseline estimate until the baseline estimate meets the preset conditions, at which point the update stops and the optimal estimate is obtained. The process of solving the difference error equation using the Adam algorithm to update the baseline estimate until the baseline estimate meets a preset condition and then stopping the update to obtain the optimal estimate includes: Based on the difference error equation, the first-order gradient estimation expression and the second-order gradient estimation expression of the baseline estimate are obtained, and an iterative formula is established. The iterative formula is used to correct the bias in the first and second gradients. The baseline estimate is updated based on the deviation between the corrected first and second gradients; Determine whether the baseline estimate meets the preset requirements. If it does, output the estimate as the optimal estimate; otherwise, continue updating.

2. The method according to claim 1, characterized in that, The differential error equations include: pseudorange differential error equations and carrier phase differential error equations; The pseudorange difference error equation is as follows: The carrier phase differential error equation is as follows in, The error term vector of the pseudorange difference. This is the pseudo-range difference coefficient matrix. It is a pseudo-range difference vector. It is a vector composed of measurable or computable constant terms. Let be the error term vector of the carrier phase difference. This is the coefficient matrix of the carrier phase difference. The carrier phase differential vector, For the double-difference integer ambiguity of the carrier phase, A vector consisting of measurable or computable constant terms. It is an identity matrix.

3. The method according to claim 1, characterized in that, The iterative update stops when the baseline estimate meets the following preset requirements; in, Given a significance level Dimension of the baseline error vector The best estimate of the baseline vector is obtained by using the chi-square distribution. .

4. The method according to claim 3, characterized in that, The three-dimensional coordinate information of the unmanned aerial vehicle is calculated using the three-dimensional coordinates of the nearest reference station, the optimal estimate, and the error correction value in the following manner. in, The three-dimensional coordinate vector obtained from the nearest reference station. This is the error correction value. Real-time 3D coordinates of the unmanned aerial vehicle.

5. The method according to claim 1, characterized in that, The pre-built neural network model includes a generator and a discriminator; the step of inputting the training set into the pre-built neural network model for iterative training to obtain a three-dimensional position prediction model for the unmanned aerial vehicle includes: Select the first from the historical data of the three-dimensional coordinate information. to Historical Period Three-dimensional coordinate vectors As a positive example, using a generator input Generate a noise vector. Find a counterexample, maximize the loss function, and update the discriminator parameters; Fix the parameters of the discriminator, and The BeiDou satellite data input generator of the epoch obtained the predicted number of times. to The 3D coordinates of the epoch are input into the discriminator. The discriminator outputs a truth value parameter. The generator adjusts the parameters to make this truth value parameter as large as possible, thus obtaining the generator parameters. After network training is completed, the trained UAV 3D position prediction model is used to generate predicted 3D coordinates of the UAV for a future period of time based on the current BeiDou satellite data.

6. The method according to claim 5, characterized in that, Also includes: In each iteration, the parameters of the generator are fixed first, and only the parameters of the discriminator are updated.

7. A position prediction device for an unmanned aerial vehicle (UAV) based on BeiDou differential positioning, applied to the position prediction method for an UAV based on BeiDou differential positioning as described in claim 1, characterized in that, include: The system includes a satellite antenna, a communication module, a BeiDou receiver motherboard, and a data storage module. The satellite antenna is connected to the BeiDou receiver motherboard via the communication module, and the data storage module is also connected to the BeiDou receiver motherboard. The satellite antenna is used to search for BeiDou satellite signals and receive BeiDou satellite data; The communication module is used to transmit the BeiDou satellite data to the BeiDou receiver motherboard; The Beidou receiver motherboard is used to perform differential processing on the Beidou satellite data to obtain the real-time three-dimensional coordinate information of the unmanned aerial vehicle and the predicted three-dimensional coordinate information for a period of time in the future. The communication module is also used to receive information transmitted between the ground control station and the reference station, and to send the real-time three-dimensional coordinate information and the three-dimensional coordinate prediction information for a period of time to the ground control station. The data storage module is used to store BeiDou satellite data, the three-dimensional coordinate information of the unmanned aerial vehicle (UAV), and the three-dimensional position prediction model of the UAV.

8. The apparatus according to claim 7, characterized in that, Also includes: A preamplifier is used to amplify the BeiDou satellite signals received by the satellite antenna. The input terminal of the preamplifier is connected to the satellite antenna, and the output terminal of the preamplifier is connected to the main board of the Beidou receiver.

Images