Reliability improvement and anti-degradation control method and system of unmanned aerial vehicle navigation system

Abstract

The invention belongs to the technical field of unmanned aerial vehicle flight control, and particularly relates to an unmanned aerial vehicle navigation system reliability improvement and anti-degradation control method and system, and the method comprises the steps: constructing a multi-mode GNSS positioning device, and synchronously outputting RTK position information, an RTK fixed state and SPP position information; constructing a two-channel extended Kalman filter (EKF), fusing the position information of the RTK and the position information of the SPP, and correspondingly obtaining a first routine and a second routine of position estimation; acquiring the position deviation estimation of the EKF at the current moment by using a state monitor so as to calculate the current position deviation compensation amount for correcting a position estimation second routine; performing two-dimensional reliability quality evaluation on the SPP fusion channel to obtain a two-dimensional detection mark; and based on the RTK fixed state and the two-dimensional detection mark, selecting a position estimation first routine or a corrected position estimation second routine as position information finally output by the unmanned aerial vehicle navigation system, and realizing smooth switching between the RTK positioning mode and the SPP positioning mode.

Description

Technical Field

[0001] The present invention belongs to the technical field of unmanned aerial vehicle (UAV) flight control, and particularly relates to a method and system for improving the reliability and anti-degradation control of a UAV navigation system. Background Art

[0002] In the industrial background of the large-scale development of the low-altitude economy, the UAV flight control system has put forward higher requirements for navigation and positioning accuracy. In particular, the accuracy attenuation phenomenon of positioning signals and the system reliability problem have become technical bottlenecks that urgently need to be solved for the large-scale application of UAVs in key fields such as logistics distribution, emergency rescue, and low-altitude inspection.

[0003] The existing technology mainly adopts a fusion navigation system based on Kalman filtering, which realizes positioning by fusing the integration prediction of inertial measurement unit (IMU) data and the external correction of a GNSS module with a single positioning output mode. However, during actual flight, the GNSS positioning accuracy has a decisive impact on the fusion accuracy of UAVs. When encountering network signal attenuation, the GNSS module with a single positioning output mode will degrade from centimeter-level accuracy real-time kinematic (RTK) positioning to meter-level accuracy pseudorange single point positioning (SPP), resulting in a serious position drift phenomenon in the navigation system and affecting the operation accuracy. In addition, for solutions to this problem, technical means such as adding external vision sensors or lidar SLAM positioning systems are usually adopted. This method effectively enhances the robustness of the system by increasing the redundancy at the system hardware level, and thus can better cope with the negative impacts brought by the decline in GNSS positioning accuracy.

[0004] For example, Chinese patent document CN118483728A discloses a multi-mode positioning device, method, and system for UAVs under GNSS interference conditions. By verifying the reliability of GNSS satellite signals in the environment where the UAV is located, if the reliability meets the preset standard, the UAV positioning data is obtained through GNSS joint positioning by combining GNSS satellite signals and UAV images. If the reliability does not meet the preset standard, the images of the scene where the UAV is located are collected to obtain the UAV positioning data through visual positioning; and during UAV positioning and navigation, the reliability of GNSS satellites in the environment where the UAV is located is continuously monitored, and the switching between GNSS joint positioning and visual positioning is carried out according to the reliability of GNSS satellites to achieve continuous navigation and positioning during the UAV mission execution process.

[0005] Chinese patent document CN116753937A discloses a real-time mapping fusion method based on UAV laser radar and visual SLAM, including the following steps: Step 1, inertial navigation model system fusion algorithm calculates the flight Euler angle coordinate (x, y, z) system; Step 2, GNSS positioning and real-time image splicing to generate 2D map; Step 3, real-time image generates 2D map and visual attitude estimation; Step 4, GPU accelerated stereo reconstruction to visual attitude estimation to generate 2.5D digital elevation model. The present invention greatly reduces the probability of UAV harming ground personnel, and can even improve the accuracy of landing at a designated airport location.

[0006] However, the GNSS module using a single positioning output mode can only output one positioning result at any time, either real-time kinematic differential (RTK) positioning data or pseudo-range single point positioning (SPP) data. When the GNSS module degenerates from the RTK positioning state to the SPP positioning state, since the GNSS module needs to reinitialize the state vector and state covariance matrix, and cannot inherit the historical state information of the SPP, the SPP positioning result needs to go through a convergence process and cannot immediately reach the optimal positioning accuracy. During this period, significant positioning deviations will occur. This sharp drop in positioning accuracy will have a serious impact on the UAV navigation system: if the navigation system integrates the position information with a large deviation in this segment, it will cause the UAV to have a significant position drift phenomenon; if the navigation system chooses to discard this segment of positioning data, it will cause the interruption of external positioning information, causing the system to lose an important position reference source.

[0007] The above-mentioned solution of adding external sensors (such as visual SLAM, lidar point cloud matching, etc.) can build a multi-level hardware redundant system, but the solution faces many systematic challenges in practical applications: First, the introduction of core components such as visual sensors and lidars has significantly increased the hardware cost, which has greatly increased the overall cost of the system; secondly, the calibration process of the multi-sensor system not only involves complex spatial parameter alignment, but also needs to consider the timing synchronization problem between sensors, and its complexity grows exponentially with the number of sensors; thirdly, the spatiotemporal synchronization problem of multi-source heterogeneous data is particularly prominent, requiring precise timestamp alignment and spatial coordinate unification, which poses a severe test to the real-time performance of the system; finally, the parallel processing and fusion calculation of multi-sensor data will lead to a geometric increase in the computing power demand of the airborne computing platform, which not only increases power consumption, but also puts higher requirements on the computing architecture of the processor. Summary of the invention

[0008] The present invention aims to solve the position deviation problem caused by the degradation of GNSS positioning accuracy of UAVs, and provide a method for improving the reliability and anti-degradation control of UAV navigation systems, so as to effectively improve the positioning accuracy and reliability of UAVs in weak network signal environments without significantly increasing hardware costs and system complexity.

[0009] The present invention also discloses a system loaded with a method for improving the reliability and anti-degradation control of an unmanned aerial vehicle (UAV) navigation system.

[0010] The detailed technical solution of the present invention is as follows: A method for improving the reliability and anti-degradation control of a UAV navigation system, the method comprising: S1. Construct a multi-modal GNSS positioning device that outputs in parallel with a dual positioning mode by combining real-time kinematic (RTK) differential positioning and single point positioning (SPP) pseudorange, and the multi-modal GNSS positioning device synchronously outputs RTK position information in the navigation coordinate system , RTK fixed status and SPP position information in the navigation coordinate system ; S2. Construct a dual-channel extended Kalman filter (EKF) including an RTK fusion channel and an SPP fusion channel, and use the dual-channel extended Kalman filter (EKF) to fuse the RTK position information and SPP position information respectively to obtain a first position estimation routine and a second position estimation routine ; S3. Use a state monitor to continuously obtain the dual-channel position deviation estimation of the dual-channel extended Kalman filter (EKF) at the current moment to calculate the current position deviation compensation amount ; ; S4. Construct a dynamic deviation compensation model for the SPP fusion channel, and use the current position deviation compensation amount to correct the second position estimation routine to obtain a target second position estimation routine corresponding to the SPP fusion channel ; S5. Perform a two-dimensional reliability quality assessment on the SPP fusion channel to obtain its two-dimensional detection flag, and the two-dimensional detection flag includes an SPP position measurement residual detection flag and an SPP fusion channel state variance detection flag ; S6. Perform GNSS positioning mode degradation control, that is, based on the RTK fixed status and the two-dimensional detection flag, select the first position estimation routine or the target second position estimation routine as the position information finally output by the UAV navigation system , to achieve smooth switching between real-time kinematic positioning RTK and single-point positioning SPP

[0011] According to a preferred embodiment of the present invention, in S1, the RTK position information in the navigation coordinate system is , the RTK fixed state is , the SPP position information in the navigation coordinate system is ; wherein, , , respectively represent the coordinate values of the drone on the N-axis, E-axis, and D-axis under real-time kinematic positioning RTK; , , respectively represent the coordinate values of the drone on the N-axis, E-axis, and D-axis under single-point positioning SPP; T represents matrix transpose; The origin of the navigation coordinate system is the initial position when the drone takes off, and the N, E, and D axes are orthogonal; wherein, the N-axis points to the due north direction; the E-axis points to the east and is perpendicular to the N-axis; the D-axis is perpendicular to the plane where the N-axis and E-axis are located and the direction is downward.

[0012] According to a preferred embodiment of the present invention, S2 includes using the Kalman filtering method to fuse the RTK position information output by the multi-modal GNSS positioning device , correcting the first routine inertial navigation position predicted by the drone according to the inertial measurement unit IMU , estimating the first routine fusion positioning information with centimeter-level high precision, denoted as the first routine position estimation ; wherein, the first routine inertial navigation position is , the first routine position estimation is ; and, using the Kalman filtering method to fuse the SPP position information output by the multi-modal GNSS positioning device , correcting the second routine inertial navigation position predicted by the drone according to the inertial measurement unit IMU , estimating the second routine fusion positioning information with meter-level accuracy, denoted as the second routine position estimation ; wherein, the second routine inertial navigation position is , the second routine position estimation is .

[0013] According to a preferred embodiment of the present invention, in S3, the state monitor estimates the first routine based on the position and the second position estimation routine Calculate the current time Dual-channel position deviation estimation ,Right now: (1); For the current moment Dual-channel position deviation estimation Perform low-pass filtering to calculate the current position deviation compensation ,Right now: (2); In formula (2): represents the filtering parameters, which are calculated from the position estimation update interval and time constant of the SPP fusion channel; Indicates the current moment, Indicates the position deviation compensation amount at the previous moment; And, determine the RTK fixed state ,like , that is, when the RTK state is a fixed solution, the current position deviation compensation is updated based on the above equations (1) and (2): ; When the real-time dynamic differential positioning RTK fails, the dynamic update of the dual-channel position deviation is stopped and the current position deviation compensation is locked. .

[0014] According to the preferred embodiment of the present invention, in said S4, the current position deviation compensation amount is Dynamic injection position estimation second routine Perform position deviation compensation, the compensation formula is: (3); In formula (3): For the current moment The smoothing gain coefficient decays over time to reduce the step response during switching; its recursive formula is expressed as: (4); In formula (4): is the smoothing gain coefficient of the previous moment; Represents the position estimation update interval of the SPP fusion channel; represents the time constant, and the compensation intensity is dynamically adjusted through an exponential decay function, where the time constant , which is used to ensure that the compensation amount decays smoothly with the RTK failure time.

[0015] Preferably according to the present invention, in step S5, obtaining the SPP position measurement residual detection flag , including: Based on the SPP position information output by the multi-modal GNSS positioning device and the second-routine inertial navigation position predicted by the inertial measurement unit IMU , calculating the second-routine position information value, that is, the position residual : (5); Based on the position residual calculating the second-routine elevation position normalized residual : (6); In formula (6): represents the variance of the second-routine elevation position information; and, calculating the second-routine horizontal position normalized residual : (7); In formula (7): , respectively represent the second-routine N component and E component position information variances; Based on a preset threshold for residual detection, that is, judging the second-routine elevation position normalized residual and the second-routine horizontal position normalized residual , obtaining the SPP position measurement residual detection flag : (8); In formula (8): , respectively represent the horizontal and elevation position normalized residual detection thresholds.

[0016] Preferably according to the present invention, in step S5, obtaining the SPP fusion channel state variance detection flag , including: Extracting the diagonal elements of the position estimation state covariance matrix, and calculating the sum of the SPP fusion channel position estimation variances: (9); In formula (9): represents the sum of the SPP fusion channel position estimation variances; , , respectively represent N component, E component andD The position state variance of the component Among them, the position estimation state covariance matrix is expressed as: (10); In formula (10): is a vector composed of diagonal elements, that is ; Perform state variance detection based on a preset threshold to obtain the SPP fusion channel state variance detection flag : (11); In formula (11): represents the position state variance detection threshold.

[0017] According to the preference of the present invention, the S6 includes the switching of the first routine of position estimation to the second routine, that is: Initial state: The current position of the UAV navigation system using the first routine of position estimation output by the RTK fusion channel is used as the final output position information ; RTK fixed state detection: If , that is, RTK is in a fixed state, then continue to output: ; If , that is, RTK is in a non-fixed state, enter the SPP fusion channel reliability evaluation; SPP channel reliability evaluation: If the SPP position measurement residual detection flag , that is, the measurement residual detection passes, and the SPP fusion channel state variance detection flag , that is, the state variance detection passes, then switch to the second routine, that is, output: ; If any detection fails, keep using the first routine, that is, output .

[0018] According to the preference of the present invention, the S6 also includes the switching of the second routine of position estimation to the first routine, that is: Initial state: The current position of the UAV navigation system using the second routine of target position estimation output by the SPP fusion channel and compensated and corrected is used as the final output position information ; RTK fixed state detection: If , that is, RTK is in a non-fixed state, then continue to output: ; If , that is, the RTK is in a fixed state, then switch back to the first routine, that is, output: .

[0019] In another aspect of the present invention, a system for implementing a method for improving the reliability and anti-degradation control of a UAV navigation system is provided. The system includes: A multi-modal GNSS positioning device, constructed based on real-time kinematic differential positioning (RTK) and pseudo-range single-point positioning (SPP), for synchronously outputting RTK position information in the navigation coordinate system , the RTK fixed state and SPP position information in the navigation coordinate system ; A dual-channel extended Kalman filter (EKF), including an RTK fusion channel and an SPP fusion channel, for respectively fusing the RTK position information and the SPP position information to correspondingly obtain a first routine of position estimation and a second routine of position estimation ; A state monitor, for real-time obtaining the dual-channel position deviation estimation of the dual-channel extended Kalman filter (EKF) at the current moment to calculate the current position deviation compensation amount ; and, performing GNSS positioning mode degradation control, that is, based on the RTK fixed state and the two-dimensional detection flag to select the first routine of position estimation or the second routine of target position estimation as the position information finally output by the UAV navigation system to realize the smooth switching between real-time kinematic differential positioning (RTK) and pseudo-range single-point positioning (SPP); A dynamic deviation compensation module, constructing a dynamic deviation compensation model for the SPP fusion channel, for using the current position deviation compensation amount to correct the second routine of position estimation to obtain a second routine of target position estimation corresponding to the SPP fusion channel ; A reliability quality evaluation module, for performing two-dimensional reliability quality evaluation on the SPP fusion channel to obtain its two-dimensional detection flag, and the two-dimensional detection flag includes an SPP position measurement residual detection flag and an SPP fusion channel state variance detection flag and an SPP fusion channel state variance detection flag .

[0020] Compared with the prior art, the beneficial effects of the present invention are: (1)A method for improving the reliability and anti-degradation control of an unmanned aerial vehicle (UAV) navigation system provided by the present invention significantly improves the performance and reliability of the UAV navigation system through the collaborative optimization of core technologies such as a multi-modal GNSS positioning device, a dual-channel extended Kalman filter (EKF) architecture, a dynamic deviation compensation model, and a two-dimensional reliability detection mechanism.

[0021] (2)The multi-modal GNSS positioning device in the present invention realizes the parallel processing and synchronous output of the RTK and SPP dual modes, ensuring high-precision positioning and robustness in complex environments; the dual-channel EKF architecture effectively suppresses the position drift during RTK failure and realizes seamless mode switching through independently operating RTK and SPP fusion channels, combined with a real-time deviation compensation and smooth switching mechanism; the state monitor and the two-dimensional reliability detection mechanism significantly reduce the probability of mis-switching through the real-time evaluation of measurement residuals and state variances, ensuring the stability of the system in weak signal environments.

[0022] (3)The present invention introduces a dynamic deviation compensation model, which dynamically adjusts the compensation intensity through an exponential decay function, further optimizing the transition process from RTK to SPP and improving the positioning accuracy and continuity.

[0023] (4)The overall design of the present invention ensures high precision and high reliability while avoiding the high costs and high complexities of traditional redundancy schemes, significantly improving the adaptability and operation efficiency of UAVs in weak network signal environments.

[0024] (5)The present invention provides an innovative solution for UAV navigation in the low-altitude economy field, with broad application prospects and high application value. BRIEF DESCRIPTION OF THE DRAWINGS

[0025] Figure 1 is a flowchart of the method for improving the reliability and anti-degradation control of the UAV navigation system described in the present invention.

[0026] Figure 2 is an execution block diagram of the method for improving the reliability and anti-degradation control of the UAV navigation system described in the present invention. DETAILED DESCRIPTION OF THE INVENTION

[0027] The present invention will be further described below in conjunction with the drawings and embodiments.

[0028] It should be noted that the following detailed description is exemplary and is intended to provide further explanation of the present invention. Unless otherwise specified, all technical and scientific terms used herein have the same meaning as commonly understood by those of ordinary skill in the technical field to which the present invention belongs.

[0029] It should be noted that the terms used herein are only for describing specific embodiments and are not intended to limit the exemplary embodiments according to the present invention. As used herein, unless the context clearly indicates otherwise, the singular forms are also intended to include the plural forms. In addition, it should be understood that when the terms "comprising" and / or "including" are used in this specification, they specify the presence of features, steps, operations, devices, components, and / or combinations thereof.

[0030] In the case of no conflict, the embodiments in the present invention and the features in the embodiments can be combined with each other.

[0031] The present invention provides a method for improving the reliability of a navigation and positioning system based on multi-state fusion and anti-drift control when the GNSS positioning mode degrades to solve the problem of position offset generated when the GNSS positioning accuracy of an unmanned aerial vehicle (UAV) degrades. It independently and parallelly estimates the RTK and SPP position information through a multi-modal GNSS positioning device and synchronously outputs the dual data streams to construct a dual-channel position information fusion architecture for the UAV navigation system. Moreover, this method introduces a dynamic deviation compensation mechanism to achieve seamless switching between the RTK and SPP positioning modes, effectively suppressing the position drift caused by the degradation of positioning accuracy. At the same time, the safety and stability of mode switching are ensured through a two-dimensional reliability detection method, ultimately achieving high-precision and high-reliability positioning of the UAV in a complex environment and meeting the stringent requirements of the navigation system in the low-altitude economy field.

[0032] The following further describes the method and system for improving the reliability of the UAV navigation system and anti-degradation control of the present invention in combination with specific embodiments.

[0033] Embodiment 1 Refer Figure 1 and Figure 2 , this embodiment provides a method for improving the reliability of a UAV navigation system and anti-degradation control, and the method includes: S1. Construct a multi-modal GNSS positioning device with parallel output of dual positioning modes by combining real-time kinematic (RTK) positioning and single-point positioning with pseudorange (SPP). The multi-modal GNSS positioning device synchronously outputs the RTK position information in the navigation coordinate system , RTK fixed state and the SPP position information in the navigation coordinate system .

[0034] In this embodiment, the traditional GNSS positioning device with a single positioning mode output is improved to a multi-modal GNSS positioning device with parallel output of dual positioning modes, that is, a parallel processing architecture is constructed, enabling the UAV navigation system to simultaneously run RTK and SPP, and synchronously output the calculation results of the dual positioning modes.

[0035] The multi-modal GNSS positioning device synchronously outputs RTK position information in the navigation coordinate system , RTK fixed status and SPP position information in the navigation coordinate system . Among them, the origin of the navigation coordinate system is the initial position when the UAV takes off; the three axes of north (N), east (E), and down (D) are orthogonal; among them, the N axis points to the due north direction; the E axis points to the east and is perpendicular to the N axis; the D axis is perpendicular to the plane where the N axis and the E axis are located, and the direction is downward.

[0036] The RTK position information in the navigation coordinate system described above is , the RTK fixed status is , the SPP position information in the navigation coordinate system is ; among them, , , respectively represent the coordinate values of the UAV on the N axis, E axis, and D axis under real-time kinematic differential positioning RTK; , , respectively represent the coordinate values of the UAV on the N axis, E axis, and D axis under pseudorange point positioning SPP; T represents matrix transpose.

[0037] Based on the above, by constructing a multi-modal GNSS positioning device with a parallel processing architecture, adopting a multi-threaded solution engine and a hardware-level time synchronization technology, the independent parallel operation of real-time kinematic differential positioning RTK based on carrier phase observations and single-point positioning SPP based on pseudorange observations is realized, supporting the synchronous estimation and data output of the RTK and SPP dual positioning modes, ensuring the time alignment accuracy of the dual data streams, and providing a high-precision and highly reliable positioning data basis for the UAV navigation system.

[0038] Furthermore, the multi-modal GNSS positioning device is connected to the UAV flight control through a UART serial port for data communication.

[0039] S2. Construct a dual-channel extended Kalman filter EKF including an RTK fusion channel and an SPP fusion channel, and use the dual-channel extended Kalman filter EKF to fuse the RTK position information and the SPP position information respectively to obtain a first routine of position estimation and a second routine of position estimation .

[0040] Specifically, the RTK position information output by the multi-modal GNSS positioning device is fused using the Kalman filtering method , and the first routine inertial navigation position predicted by the drone according to the inertial measurement unit (IMU) is corrected , and the first routine fusion positioning information with centimeter-level high precision is estimated, denoted as the first routine position estimation . Among them, the first routine inertial navigation position is , and the first routine position estimation is .

[0041] In addition, the SPP position information output by the multi-modal GNSS positioning device is fused using the Kalman filtering method , and the second routine inertial navigation position predicted by the drone according to the inertial measurement unit (IMU) is corrected , and the second routine fusion positioning information with meter-level accuracy is estimated, denoted as the second routine position estimation . Among them, the second routine inertial navigation position is , and the second routine position estimation is .

[0042] Based on the above, by constructing a dual-channel extended Kalman filter (EKF) architecture, the independent operation and collaborative optimization of the RTK fusion channel and the SPP fusion channel are respectively realized. Among them, the RTK fusion channel adopts a high-precision data fusion algorithm to achieve centimeter-level positioning accuracy; the SPP fusion channel ensures stable positioning in complex environments through a robust data fusion algorithm

[0043] On the other hand, the prediction of the routine inertial navigation position by the inertial measurement unit (IMU) above is a conventional operation without improvement in the present invention, so it will not be elaborated in detail in the present invention

[0044] S3. Use the state monitor to continuously obtain the dual-channel position deviation estimation of the dual-channel extended Kalman filter (EKF) at the current moment to calculate the current position deviation compensation amount . .

[0045] In this step, the state monitor calculates the dual-channel position deviation estimation at the current moment based on the first routine position estimation and the second routine position estimation , that is : (1); Subsequently, at the current moment Dual-channel position deviation estimation Perform low-pass filtering to calculate the current position deviation compensation amount , that is: (2); In formula (2): represents the filtering parameter, which is calculated from the position estimation update interval and time constant of the SPP fusion channel; represents the current moment, represents the position deviation compensation amount at the previous moment.

[0046] Based on the above, this step further includes: judging the RTK fixed state , if , that is, when the RTK state is a fixed solution, execute the above steps to update the current position deviation compensation amount ; otherwise, when it is detected that the real-time kinematic differential positioning RTK fails, stop the dynamic update of the dual-channel position deviation and lock the current position deviation compensation amount .

[0047] The above state monitor operates as an independent module, is decoupled from the dual-channel EKF routine design, and calculates the position deviation compensation amount between the RTK fusion channel and the SPP fusion channel in real time for subsequent error correction of the SPP channel. At the same time, the state monitor executes the degradation control strategy of the GNSS positioning mode in the subsequent steps to achieve seamless switching between the RTK and SPP positioning modes and suppress the position drift caused by RTK failure. This design significantly improves the positioning accuracy and system robustness of the UAV in complex environments.

[0048] On the other hand, the low-pass filtering technology is used to smooth the high-frequency noise between the two channels, output a stable and reliable deviation compensation amount, and ensure the compensation accuracy. When it is detected that the RTK fails, the system automatically locks the current deviation compensation amount, stops the dynamic update, realizes RTK failure protection and smooth transition, effectively suppresses the position drift, and improves the navigation stability of the UAV in the GNSS positioning degradation scenario.

[0049] S4. Construct a dynamic deviation compensation model for the SPP fusion channel, and use the current position deviation compensation amount to correct the second position estimation routine to obtain the corresponding target second position estimation routine of the SPP fusion channel .

[0050] That is, dynamically inject the current position deviation compensation amount into the second position estimation routine for position deviation compensation, and the compensation formula is: (3); In formula (3): is the current moment is the smoothing gain coefficient, which decays with time to reduce the step response during switching; its recurrence formula is expressed as: (4); In formula (4): is the smoothing gain coefficient at the previous moment; represents the position estimation update interval of the SPP fusion channel; represents the time constant, and the compensation intensity is dynamically adjusted through an exponential decay function, where the time constant , ensuring that the compensation amount decays smoothly with the RTK failure time.

[0051] The above dynamic deviation compensation model provides strong compensation at the initial stage of RTK failure to suppress position jumps, and then gradually reduces the compensation intensity to achieve seamless switching from RTK to SPP, significantly improving the navigation accuracy and stability of the UAV when GNSS positioning degrades.

[0052] S5. Perform a two-dimensional reliability quality assessment on the SPP fusion channel to obtain its two-dimensional detection flag, and the two-dimensional detection flag includes the SPP position measurement residual detection flag and the SPP fusion channel state variance detection flag .

[0053] In this step, the specific process of obtaining the SPP position measurement residual detection flag is as follows: Based on the SPP position information output by the multi-modal GNSS positioning device and the second routine inertial navigation position predicted by the inertial measurement unit IMU , calculate the second routine position information value, that is, the position residual : (5); Furthermore, calculate the second routine elevation position normalized residual : (6); In formula (6): represents the second routine elevation position information variance; And, calculate the second routine horizontal position normalized residual : (7); In formula (7): , respectively represent the second routine N component andE Variance of component position information.

[0054] Perform residual detection based on a preset threshold, that is, judge the normalized residual of the elevation position of the second routine and the normalized residual of the horizontal position of the second routine , and obtain the SPP position measurement residual detection flag : (8); In formula (8): , respectively represent the normalized residual detection thresholds for horizontal and elevation positions.

[0055] And, the specific process of obtaining the SPP fusion channel state variance detection flag is as follows: Extract the diagonal elements of the position estimation state covariance matrix, and calculate the sum of the SPP fusion channel position estimation variances: (9); In formula (9): represents the sum of the SPP fusion channel position estimation variances; , , respectively represent N component, E component and D the position state variances of the components; Among them, the position estimation state covariance matrix is expressed as: (10); In formula (10): is a vector composed of diagonal elements, that is .

[0056] Based on the above, through a two-dimensional evaluation mechanism of position measurement residual detection and state variance detection, the positioning quality of the SPP channel is monitored in real time. Among them, the measurement residual detection is based on the calculation of the normalized residual, and the : (11); In formula (11): represents the position state variance detection threshold.

[0057] Based on the above, through a two-dimensional evaluation mechanism of position measurement residual detection and state variance detection, the positioning quality of the SPP channel is monitored in real time. Among them, the measurement residual detection is based on the calculation of the normalized residual, and the multiple standard deviations ( )(As a threshold, it conforms to the Gaussian distribution characteristics; the state variance detection evaluates the positioning stability through the diagonal elements of the covariance matrix. This mechanism effectively reduces the probability of mis-switching and significantly improves the navigation reliability of the UAV in complex environments.)

[0058] S6. Execute GNSS positioning mode degradation control, that is, based on the RTK fixed state and the two-dimensional detection flag, select the first routine of position estimation or the second routine of target position estimation as the position information finally output by the UAV navigation system , realizing the smooth switching between real-time kinematic positioning RTK and single point positioning with pseudorange SPP.)

[0059] This step specifically includes: the switching from the first routine of position estimation to the second routine, that is: (1) Initial state: The UAV navigation system currently uses the position estimated by the first routine output by the RTK fusion channel as the position information finally output .

[0060] (2) RTK fixed state detection: If , that is, RTK is in a fixed state, then continue to output: ; If , that is, RTK is in a non-fixed state, enter the reliability evaluation of the SPP fusion channel.)

[0061] (3) SPP channel reliability evaluation: If the SPP position measurement residual detection flag (the measurement residual detection passes) and the SPP fusion channel state variance detection flag (the state variance detection passes), then switch to the second routine (SPP fusion channel), that is, output: .

[0062] If any of the detections fails, keep using the first routine, that is, output .

[0063] This step specifically also includes: the switching from the second routine of position estimation to the first routine, that is: (1) Initial state: The UAV navigation system currently uses the position of the second routine of target position estimation output by the SPP fusion channel and compensated and corrected as the position information finally output .

[0064] (2) RTK fixed state detection: If , that is, when RTK is in an unfixed state, continue to output: ; If , that is, when RTK is in a fixed state, switch back to the first routine, that is, output: .

[0065] Based on the above, through the RTK fixed state and the two-dimensional detection flag, seamless switching between the first routine and the second routine of position estimation is realized, and the optimal position information of the UAV navigation system is dynamically selected to ensure high-reliable navigation can still be maintained when RTK fails, while avoiding mis-switching.

[0066] Embodiment 2, This embodiment provides a system for implementing a method for improving the reliability and anti-degradation control of a UAV navigation system. The system includes: A multi-modal GNSS positioning device, constructed based on real-time kinematic differential positioning RTK and pseudorange single-point positioning SPP, for synchronously outputting RTK position information in the navigation coordinate system , RTK fixed state and SPP position information in the navigation coordinate system ; A dual-channel extended Kalman filter EKF, including an RTK fusion channel and an SPP fusion channel, for respectively fusing RTK position information and SPP position information to correspondingly obtain the first routine of position estimation and the second routine of position estimation ; A state monitor, for real-time obtaining the dual-channel position deviation estimation of the dual-channel extended Kalman filter EKF at the current moment to calculate the current position deviation compensation amount ; and, performing GNSS positioning mode degradation control, that is, based on the RTK fixed state and the two-dimensional detection flag to select the first routine of position estimation or the target second routine of position estimation as the position information finally output by the UAV navigation system to realize the smooth switching between real-time kinematic differential positioning RTK and pseudorange single-point positioning SPP; A dynamic deviation compensation module, constructing a dynamic deviation compensation model for the SPP fusion channel, for using the current position deviation compensation amount to correct the second routine of position estimation to obtain the target second routine of position estimation corresponding to the SPP fusion channel ; ; A reliability quality assessment module is used to perform a two-dimensional reliability quality assessment on the SPP fusion channel to obtain its two-dimensional detection flag, where the two-dimensional detection flag includes an SPP position measurement residual detection flag and an SPP fusion channel state variance detection flag .

[0067] Obviously, the above embodiments of the present invention are merely examples for clearly illustrating the technical solutions of the present invention, rather than limitations on the specific implementation manners of the present invention. Any modifications, equivalent replacements, and improvements made within the spirit and principles of the claims of the present invention shall be included within the protection scope of the claims of the present invention.

Claims

1. A method for improving the reliability and anti-degradation control of an unmanned aerial vehicle navigation system, characterized in that: The method comprises: S1. Combining real-time dynamic differential positioning RTK and pseudo-range single point positioning SPP to build a multi-modal GNSS positioning device with dual positioning mode parallel output, the multi-modal GNSS positioning device synchronously outputs RTK position information in the navigation coordinate system ,RTK fixed status and the SPP position information in the navigation coordinate system ; S2, construct a dual-channel extended Kalman filter EKF including an RTK fusion channel and an SPP fusion channel, and use the dual-channel extended Kalman filter EKF to fuse the RTK position information respectively and SPP location information , to obtain the first routine of position estimation and the second position estimation routine ; S3, using the state monitor to obtain the dual-channel extended Kalman filter EKF in real time at the current moment Dual-channel position deviation estimation , to calculate the current position deviation compensation ; S4, construct a dynamic deviation compensation model for the SPP fusion channel, and use the current position deviation compensation amount The second routine for estimating the position Correction is performed to obtain the target position estimation second routine corresponding to the SPP fusion channel ; S5. Perform a dual-dimensional reliability quality assessment on the SPP fusion channel to obtain a dual-dimensional detection mark, wherein the dual-dimensional detection mark includes an SPP position measurement residual detection mark. and SPP fusion channel state variance detection flag ; S6, executing GNSS positioning mode degradation control, that is, based on the RTK fixed state and the dual-dimensional detection flag selects the first position estimation routine Or target position estimation second routine As the final output of the UAV navigation system , realizing smooth switching between real-time dynamic differential positioning RTK and pseudo-range single point positioning SPP.

2. The method for improving the reliability and anti-degradation control of the UAV navigation system according to claim 1 is characterized in that: In S1, the RTK position information in the navigation coordinate system for , RTK fixed state for , SPP position information in the navigation coordinate system for ;in, , , They respectively represent the coordinate values ​​of the UAV on the N-axis, E-axis, and D-axis under real-time dynamic differential positioning RTK; , , They represent the coordinate values ​​of the UAV on the N-axis, E-axis, and D-axis under pseudo-range single point positioning SPP; T represents matrix transposition; The origin of the navigation coordinate system is the initial position of the drone when it takes off, and the three axes N, E, and D are orthogonal; the N axis points to the north; the E axis points to the east and is perpendicular to the N axis; the D axis is perpendicular to the plane where the N axis and the E axis are located, and is directed downward.

3. The method for improving the reliability and anti-degradation control of the UAV navigation system according to claim 2 is characterized in that: S2 includes fusing the RTK position information output by the multi-modal GNSS positioning device using a Kalman filter method. , correct the first routine inertial navigation position predicted by the UAV’s IMU , the first routine fusion positioning information with high accuracy at the centimeter level is estimated, which is recorded as the first routine of position estimation ; Among them, the first routine inertial navigation position for , the first routine of position estimation for ; And, the Kalman filter method is used to fuse the SPP position information output by the multi-modal GNSS positioning device , correct the second routine inertial navigation position predicted by the drone based on the inertial measurement unit IMU , the second routine of estimating the meter-level accuracy is integrated with the positioning information, which is recorded as the second routine of position estimation ; Among them, the second routine inertial navigation position for , the second position estimation routine for .

4. The method for improving the reliability and anti-degradation control of the unmanned aerial vehicle navigation system according to claim 1 is characterized in that: In S3, the state monitor estimates the first routine based on the position and the second position estimation routine Calculate the current time Dual-channel position deviation estimation ,Right now: （1）； For the current moment Dual-channel position deviation estimation Perform low-pass filtering to calculate the current position deviation compensation ,Right now: （2）； In formula (2): represents the filtering parameters, which are calculated from the position estimation update interval and time constant of the SPP fusion channel; Indicates the current moment, Indicates the position deviation compensation amount at the previous moment; And, determine the RTK fixed state ,like , that is, when the RTK state is a fixed solution, the current position deviation compensation is updated based on the above equations (1) and (2): ; When the real-time dynamic differential positioning RTK fails, the dynamic update of the dual-channel position deviation is stopped and the current position deviation compensation is locked. .

5. The method for improving the reliability and anti-degradation control of the unmanned aerial vehicle navigation system according to claim 1, characterized in that: In S4, the current position deviation compensation amount Dynamic injection position estimation second routine Perform position deviation compensation, the compensation formula is: （3）； In formula (3): For the current moment A smooth gain factor that decays over time to reduce the step response during switching; Its recursive formula is expressed as: （4）； In formula (4): is the smoothing gain coefficient of the previous moment; Represents the position estimation update interval of the SPP fusion channel; represents the time constant, and the compensation intensity is dynamically adjusted through an exponential decay function, where the time constant , which is used to ensure that the compensation amount decays smoothly with the RTK failure time.

6. The method for improving the reliability and anti-degradation control of the unmanned aerial vehicle navigation system according to claim 3 is characterized in that: In S5, the SPP position measurement residual detection flag is obtained. ,include: SPP position information based on the output of multi-modal GNSS positioning device and the second routine inertial navigation position predicted by the inertial measurement unit IMU , calculate the second routine position information value, that is, the position residual : （5）； Based on the position residual Calculate the second routine elevation position standardized residual : （6）； In formula (6): represents the variance of the elevation position information of the second routine; And, calculate the second routine horizontal position standardized residual : （7）； In formula (7): , Respectively represent the second routine N Quantity and E Component position information variance; Residual detection is performed based on the preset threshold, that is, the standardized residual of the second routine elevation position is determined and the second routine horizontally positions the standardized residuals , get the SPP position measurement residual detection flag : （8）； In formula (8): , Represent the horizontal and elevation position standardized residual detection thresholds respectively.

7. The method for improving the reliability and anti-degradation control of the UAV navigation system according to claim 6, characterized in that: In S5, the SPP fusion channel state variance detection flag is obtained. ,include: Extract the diagonal elements of the position estimation state covariance matrix and calculate the SPP fusion channel position estimation variance sum: （9）； In formula (9): represents the sum of variances of SPP fusion channel position estimates; , , Respectively N Quantity, E Quantity and D The position state variance of the component; Among them, the position estimation state covariance matrix is ​​expressed as: （10）； In formula (10): is a vector composed of diagonal elements, that is ; Perform state variance detection based on the preset threshold to obtain the SPP fusion channel state variance detection flag : （11）； In formula (11): Represents the position state variance detection threshold.

8. The method for improving the reliability and anti-degradation control of the unmanned aerial vehicle navigation system according to claim 7, characterized in that: The S6 includes switching from the first position estimation routine to the second routine, that is: Initial state: The first routine of the position estimation output by the UAV navigation system using the RTK fusion channel The location of the final output ; RTK fixed status detection: like , that is, RTK is in a fixed state, then continue to output: ; like , that is, RTK is in a non-fixed state and enters the SPP fusion channel reliability assessment; SPP channel reliability assessment: If the SPP position measurement residual detection flag , that is, the measurement residual detection passes, and the SPP fusion channel state variance detection flag , that is, the state variance test passes, then switch to the second routine, that is, output: ; If any of the tests fails, the first routine is used, that is, the output .

9. The method for improving the reliability and anti-degradation control of the unmanned aerial vehicle navigation system according to claim 7, characterized in that: The S6 also includes switching from the second position estimation routine to the first routine, that is: Initial state: The UAV navigation system currently uses the SPP fusion channel output and the compensated target position estimation routine 2 The location of the final output ; RTK fixed status detection: like , that is, RTK is in a non-fixed state, then continue to output: ; like , that is, RTK is in a fixed state, then switch back to the first routine, that is, output: .

10. A system for realizing a method for improving the reliability and anti-degradation control of a drone navigation system, characterized in that: The system comprises: Multi-modal GNSS positioning device, based on real-time dynamic differential positioning RTK and pseudo-range single point positioning SPP, is used to synchronously output RTK position information in the navigation coordinate system ,RTK fixed status and the SPP position information in the navigation coordinate system ; Dual-channel extended Kalman filter EKF, including RTK fusion channel and SPP fusion channel, used to fuse RTK position information separately and SPP location information , to obtain the first routine of position estimation and the second position estimation routine ; State monitor, used to obtain the dual-channel extended Kalman filter EKF at the current moment in real time Dual-channel position deviation estimation , to calculate the current position deviation compensation ; and, perform GNSS positioning mode degradation control, that is, based on the RTK fixed state and dual-dimensional detection landmark selection position estimation first routine Or target position estimation second routine As the final output of the UAV navigation system , realize smooth switching between real-time dynamic differential positioning RTK and pseudo-range single point positioning SPP; Dynamic deviation compensation module, which builds a dynamic deviation compensation model with SPP fusion channel to use the current position deviation compensation amount Second routine for position estimation Correction is performed to obtain the target position estimation second routine corresponding to the SPP fusion channel ; A reliability quality assessment module is used to perform a two-dimensional reliability quality assessment on the SPP fusion channel to obtain its two-dimensional detection mark, which includes an SPP position measurement residual detection mark. and SPP fusion channel state variance detection flag .

Images