A Redundant Integrated Navigation Method and System Based on Unmanned Aerial Vehicles

Abstract

This application provides a redundant integrated navigation method and system based on unmanned aerial vehicles (UAVs). Its core lies in a dual-layer redundancy mechanism of multi-sensor hardware and multi-fusion algorithm software, as well as a smooth switching mechanism for the multi-fusion processor. The method includes: acquiring sensor data and preprocessing it; performing fault diagnosis and performance evaluation on the preprocessed sensor data, comprehensively judging and executing sensor redundancy switching; performing multi-fusion algorithm software calculation and performance evaluation on the redundantly processed sensor data, comprehensively judging and executing fusion processor redundancy switching; and performing dynamic delay compensation and structured encapsulation and distribution on the redundantly processed fusion processor navigation state data. This system is used to implement the above method. This application overcomes the problems of high hardware resource consumption and insufficient reliability in traditional redundant integrated navigation systems, significantly improving the robustness and reliability of UAV integrated navigation systems.

Description

Technical Field

[0001] This application relates to the field of unmanned aerial vehicle (UAV) navigation technology, and more specifically, to a redundant integrated navigation method and system based on UAVs. Background Technology

[0002] The navigation system of an unmanned aerial vehicle (UAV) provides the necessary state data for its flight. The sensors used in this system mainly include inertial measurement units (IMUs), Global Positioning System (GNSS), barometers, and magnetic compasses. Inertial navigation systems directly measure the acceleration and angular velocity of the drone through the IMU, unaffected by external environmental interference. However, the integral characteristic of IMUs causes navigation errors to accumulate over time. For low-precision IMUs used in small UAVs, these errors accumulate even more rapidly. GNSS systems provide three-dimensional position and velocity information, which can be directly used for navigation and positioning, but they are more susceptible to external influences. Once satellite signals are interfered with, the measurement accuracy of GNSS sensors drops rapidly. Barometers measure atmospheric pressure and calculate the UAV's altitude using the conversion relationship with standard atmospheric pressure at sea level. Magnetic compasses measure the three-axis geomagnetic field strength of the environment in which the UAV is located to calculate its heading angle. However, both barometers and magnetic compasses can only provide single-state observation information. Based on the characteristics of the various sensors mentioned above, UAV navigation systems typically employ a redundant combined navigation method based on multi-sensor information fusion to obtain navigation state estimation information with more dimensions, higher accuracy, and greater stability than that obtained from a single sensor.

[0003] The performance of sensors directly impacts the state estimation performance of a navigation system. A malfunction in a single sensor can affect navigation accuracy and, in severe cases, lead to flight safety accidents. Engineering applications often employ redundant integrated navigation designs to mitigate the impact of a single sensor failure.

[0004] Traditional redundant integrated navigation systems include two schemes: sensor and processor redundancy design and sensor redundancy design.

[0005] The sensor and processor redundancy design scheme employs multiple independent hardware subsystems. Each subsystem contains independent and complete sensors, processors, and algorithm software, capable of independently outputting navigation status information such as attitude, velocity, and position. These subsystems act as hard backups for each other, ensuring the integrated navigation system can continue to operate normally even if some subsystems fail, thus enhancing system reliability. While this design scheme achieves good redundancy in the integrated navigation system, it suffers from drawbacks such as high hardware resource consumption, large size, and high cost.

[0006] Sensor redundancy design schemes only provide redundancy at the sensor hardware level, using a single processor and algorithm software. While this approach can solve navigation system hardware failures caused by sensor malfunctions at a lower cost, it cannot address navigation system software problems caused by algorithm malfunctions, resulting in insufficient system robustness and security. Summary of the Invention

[0007] The purpose of this application is to provide a redundant integrated navigation method and system based on unmanned aerial vehicles (UAVs). Through a dual-layer redundancy mechanism of multi-sensor hardware and multi-fusion algorithm software, as well as a smooth switching mechanism of multi-fusion, the robustness and reliability of UAV integrated navigation are significantly improved.

[0008] This application also provides a redundant integrated navigation method based on unmanned aerial vehicles (UAVs), including:

[0009] Based on the original sensor measurement data, the original measurement data is preprocessed, and numerical optimization and downsampling are performed to obtain the processed sensor data.

[0010] Based on the preprocessed sensor data, multi-sensor fault diagnosis and comprehensive performance evaluation are performed to determine whether redundancy switching is required.

[0011] Based on the sensor status observation data after switching, perform independent navigation calculation and comprehensive performance evaluation of the multi-fusion unit to determine whether redundant switching is required.

[0012] Based on the time-domain navigation state data with delay after switching fusion, time-domain delay compensation is performed to obtain the current time-domain navigation state data, which is then encapsulated and processed to publish navigation data.

[0013] Optionally, in the UAV-based redundant integrated navigation method described in this application, the raw sensor measurement data is preprocessed and subjected to numerical optimization and downsampling to obtain processed sensor data, specifically including:

[0014] Zero bias error compensation and scaling factor error compensation are performed on sensor data based on preset calibration parameters;

[0015] The sensor temperature characteristic curve was established through high and low temperature environment tests, and real-time online temperature compensation was performed based on the temperature characteristic curve.

[0016] Sensor data is synchronized in time by timestamp alignment, and abnormal data is identified and removed based on statistical methods.

[0017] Online adaptive filtering and smoothing of sensor data are performed based on Fast Fourier Transform (FFT) to obtain optimized sensor data;

[0018] The optimized sensor data is downsampled using the equal-interval averaging method to obtain the processing result.

[0019] Optionally, in the UAV-based redundant integrated navigation method described in this application, multi-sensor fault diagnosis is performed based on preprocessed sensor data to determine whether redundancy switching is required, specifically including:

[0020] The sensor's raw data is used to determine if the measured value exceeds the limit based on a preset threshold.

[0021] Identify data loss status through data continuity analysis;

[0022] Sensor jamming detection is achieved by combining time window numerical comparison;

[0023] The system integrates over-limit detection, data loss detection, and system crash detection to output fault diagnosis results.

[0024] If the diagnostic result is normal, sensor performance evaluation is performed. If the diagnostic result detects a fault response, sensor switching is performed. During sensor switching, the backup sensor data is connected to the fusion device data interface, and the faulty sensor is disconnected from the fusion device data interface.

[0025] Optionally, in the UAV-based redundant integrated navigation method described in this application, a multi-sensor comprehensive performance evaluation is performed based on the preprocessed sensor data to determine whether redundancy switching is required, specifically including:

[0026] FFT was used to perform spectral analysis on sensor data to identify abnormal vibration signals;

[0027] The output differences of redundant sensors are compared by multi-sensor cross-consistency verification, where redundant sensors refer to a set of multiple sensors of the same type.

[0028] The residuals are calculated using the aircraft's physical model, geographic information, and kinematic constraints to determine the degree of deviation between the data and the theoretical values.

[0029] Based on the above performance analysis results, the confidence weights of each sensor are dynamically updated, and it is determined whether to switch sensors.

[0030] Optionally, in the UAV-based redundant integrated navigation method described in this application, based on the sensor state observation data after switching, multi-fusion unit independent navigation calculation and comprehensive performance evaluation are performed to determine whether redundant switching is required, specifically including:

[0031] Based on the data after sensor switching, the multi-fusion unit uses the Error State Kalman Filter (ESKF) algorithm to generate multiple sets of navigation solution results;

[0032] Based on the navigation calculation results, the chi-square test and residual analysis methods are used, combined with the sensor noise level and historical reliability to calculate the comprehensive credibility score, which is then used as the evaluation result of the fusion unit.

[0033] Based on the fusion evaluation results, it is determined whether to perform a fusion switch. When performing a fusion switch, a low-pass filtering algorithm is used to suppress the transition during the switch. The fusion switch execution will cut out the navigation state data of the main fusion and connect the navigation state data of the backup fusion.

[0034] Optionally, in the UAV-based redundant integrated navigation method described in this application, time-domain delay compensation is performed based on the fusion delay time-domain navigation state data after switching to obtain the current time-domain navigation state data, which is then encapsulated and processed to publish navigation data, specifically including:

[0035] Online estimation of the delay error between observed data and predicted state based on the least squares method;

[0036] Based on delay error, navigation status data in the current time domain is generated by extrapolating real-time measurement data from the IMU.

[0037] The navigation state data synchronized in the time domain and the fusion state data are structured, encapsulated, and published.

[0038] Secondly, embodiments of this application provide a redundant integrated navigation system based on unmanned aerial vehicles (UAVs). The system includes a memory and a processor. The memory includes a program for a redundant integrated navigation method based on UAVs. When the program for the redundant integrated navigation method based on UAVs is executed by the processor, it implements the steps of the redundant integrated navigation method based on UAVs described above.

[0039] As can be seen from the above, the redundant integrated navigation method and system based on UAVs provided in this application preprocesses the original sensor measurement data, performs numerical optimization and downsampling to obtain processed sensor data; based on the preprocessed sensor data, performs multi-sensor fault diagnosis and comprehensive performance evaluation to determine whether redundancy switching is required; based on the sensor state observation data after switching, performs independent navigation calculation and comprehensive performance evaluation of multiple fusion units to determine whether redundancy switching is required; based on the delayed time-domain navigation state data of the fusion unit after switching, performs time-domain delay compensation to obtain the current time-domain navigation state data, encapsulates and processes it, and publishes the navigation data; through the dual redundancy mechanism of multi-sensor hardware and multi-fusion unit algorithm software, and the smooth switching mechanism of multiple fusion units, the robustness and reliability of UAV integrated navigation are significantly improved. Attached Figure Description

[0040] To more clearly illustrate the technical solution of this application, the accompanying drawings used in this application will be briefly introduced below. It should be understood that the following drawings only show some embodiments of this application and should not be regarded as a limitation of the scope. For those skilled in the art, other related drawings can be obtained from these drawings without creative effort.

[0041] Figure 1 A flowchart of the redundant integrated navigation method based on UAVs provided in this application;

[0042] Figure 2 A detailed flowchart of the redundant integrated navigation method based on unmanned aerial vehicles provided in this application;

[0043] Figure 3 A general block diagram of the redundant integrated navigation system based on unmanned aerial vehicles provided in this application;

[0044] Figure 4 A schematic diagram showing the composition and connection of the various subsystems of the redundant integrated navigation system based on unmanned aerial vehicles provided in this application. Detailed Implementation

[0045] The technical solutions of this application will now be clearly and completely described with reference to the accompanying drawings. Obviously, what is described is only a portion of the embodiments of this application, and not all of them. The components of this application described and shown in the accompanying drawings can generally be arranged and designed in various different configurations. Therefore, the following detailed description of this application provided in the accompanying drawings is not intended to limit the scope of the claimed application, but merely to illustrate selected embodiments of this application. All other embodiments obtained by those skilled in the art based on the embodiments of this application without inventive effort are within the scope of protection of this application.

[0046] It should be noted that similar reference numerals and letters in the following figures indicate similar items; therefore, once an item is defined in one figure, it does not need to be further defined and explained in subsequent figures. Furthermore, in the description of this application, the terms "first," "second," etc., are used only to distinguish descriptions and should not be construed as indicating or implying relative importance.

[0047] Please refer to Figures 1-2 . Figure 1 This is a flowchart of a redundant integrated navigation method based on unmanned aerial vehicles (UAVs) according to this application. Figure 2 This is a detailed flowchart of a redundant integrated navigation method based on unmanned aerial vehicles (UAVs) according to this application. This UAV-based redundant integrated navigation method is used in a terminal device and includes the following steps:

[0048] Based on the original sensor measurement data, the original measurement data is preprocessed, and numerical optimization and downsampling are performed to obtain the processed sensor data.

[0049] Based on the preprocessed sensor data, multi-sensor fault diagnosis and comprehensive performance evaluation are performed to determine whether redundancy switching is required.

[0050] Based on the sensor status observation data after switching, perform independent navigation calculation and comprehensive performance evaluation of the multi-fusion unit to determine whether redundant switching is required.

[0051] Based on the time-domain navigation state data with delay after switching fusion, time-domain delay compensation is performed to obtain the current time-domain navigation state data, which is then encapsulated and processed to publish navigation data.

[0052] It should be noted that, based on external sensor hardware, the data processing subsystem collects raw measurement data from redundant IMUs, GNSS, magnetic compasses, and barometers, and performs preliminary data processing (downsampling, filtering, etc.) on this data. The processed data is then transmitted as sensor data to the sensor redundancy subsystem. The sensor redundancy subsystem performs calculations, analysis, and judgment processing on the sensor data, converting it into state observation data, which is then transmitted to the fusion redundancy subsystem. The fusion redundancy subsystem receives the state observation data, performs data fusion algorithm processing on it, generates navigation state data, and transmits it to the data output subsystem. The data output subsystem performs time-domain delay compensation processing on the received navigation state data and then outputs it.

[0053] According to an embodiment of the present invention, based on the original sensor measurement data, the original measurement data is preprocessed, and numerical optimization and downsampling are performed to obtain processed sensor data, specifically including:

[0054] Zero bias error compensation and scaling factor error compensation are performed on sensor data based on preset calibration parameters;

[0055] The sensor temperature characteristic curve was established through high and low temperature environment tests, and real-time online temperature compensation was performed based on the temperature characteristic curve.

[0056] Sensor data is synchronized in time by timestamp alignment, and abnormal data is identified and removed based on statistical methods.

[0057] Online adaptive filtering and smoothing of sensor data are performed based on Fast Fourier Transform (FFT) to obtain optimized sensor data;

[0058] The optimized sensor data is downsampled using the equal-interval averaging method to obtain the processing result.

[0059] According to an embodiment of the present invention, based on preprocessed sensor data, multi-sensor fault diagnosis is performed to determine whether redundancy switching is required, specifically including:

[0060] The sensor's raw data is used to determine if the measured value exceeds the limit based on a preset threshold.

[0061] Identify data loss status through data continuity analysis;

[0062] Sensor jamming detection is achieved by combining time window numerical comparison;

[0063] The system integrates over-limit detection, data loss detection, and system crash detection to output fault diagnosis results.

[0064] If the diagnostic result is normal, perform sensor performance evaluation; if the diagnostic result indicates a fault response, switch the sensor.

[0065] It should be noted that the sensor's raw data is judged to exceed the limit based on a preset threshold, the data loss status is identified through data continuity analysis, and sensor jamming is detected by combining time window value comparison. Based on the above detection results, if the data is normal, it is output to the sensor evaluation module; if a fault response is detected, a sensor switching command is triggered.

[0066] According to an embodiment of the present invention, based on preprocessed sensor data, a comprehensive performance evaluation of multiple sensors is performed to determine whether redundancy switching is required, specifically including:

[0067] FFT was used to perform spectral analysis on sensor data to identify abnormal vibration signals;

[0068] The output differences of redundant sensors are compared by multi-sensor cross-consistency verification, where redundant sensors refer to a set of multiple sensors of the same type.

[0069] The residuals are calculated using the aircraft's physical model, geographic information, and kinematic constraints to determine the degree of deviation between the data and the theoretical values.

[0070] Based on the above performance analysis results, the confidence weights of each sensor are dynamically updated, and it is determined whether to switch sensors.

[0071] It should be noted that, through multi-sensor cross-consistency verification comparing the output differences of redundant sensors, and combining the aircraft physical model, geographic information, and kinematic constraints to calculate residuals, the degree of deviation between the data and theoretical values ​​is determined. Based on the above analysis results, the confidence weights of each sensor are dynamically updated and output to the sensor switching module. The sensor switching module performs optimal sensor combination decision-making: based on real-time confidence weights, it selects a primary sensor data group and an auxiliary sensor data group from the redundant sensor pool, consisting of two sets of IMU data, one set of magnetic compass data, one set of barometer data, and one set of GNSS data. The selected two sets of sensors are then output to the fusion redundancy subsystem. Through the above steps, the sensor redundancy subsystem ensures the normal operation of the two sets of sensor data input to the fusion redundancy subsystem through a three-level joint control mechanism of fault diagnosis-evaluation-switching.

[0072] According to an embodiment of the present invention, based on the sensor state observation data after switching, multi-fusion unit independent navigation calculation and comprehensive performance evaluation are performed to determine whether redundant switching is required, specifically including:

[0073] Based on the data after sensor switching, the multi-fusion unit uses the Error State Kalman Filter (ESKF) algorithm to generate multiple sets of navigation solution results;

[0074] Based on the navigation calculation results, the chi-square test and residual analysis methods are used, combined with the sensor noise level and historical reliability to calculate the comprehensive credibility score, which is then used as the evaluation result of the fusion unit.

[0075] Based on the merging evaluation results, it is determined whether to perform merging switching. When merging switching is performed, a low-pass filtering algorithm is used to suppress transitions during the switching process.

[0076] It should be noted that the fusion redundancy subsystem consists of a fusion module, a fusion evaluation module, and a fusion switching module. The reliability of state estimation is achieved through a dual-channel fault-tolerant fusion and dynamic switching mechanism. The fusion module adopts a parallel computing software architecture with dual independent fusion algorithms: the main fusion module and the backup fusion module receive data from the main sensor group and the auxiliary sensor group, respectively, output by the sensor redundancy subsystem. Based on the Error State Kalman Filter (ESKF) algorithm, a 16-dimensional error state vector is defined, including position error, velocity error, attitude error, angular velocity error, and acceleration error. Nominal state prediction, error covariance matrix update, and observation residual calculation are performed recursively. Based on observation data from the accelerometer, GNSS, and magnetic compass, error state injection and covariance reset are completed, generating dual-channel navigation solution results and outputting them to the fusion evaluation module. The fusion evaluation module performs multi-dimensional verification of navigation performance: it uses a chi-square test to determine the statistical significance of the solution residuals, evaluates the convergence of the filter through the residual covariance matrix, and calculates a comprehensive reliability score of the fusion result based on the dynamic confidence score of the fusion result, considering factors such as sensor noise level and historical reliability. The evaluation result is then transmitted to the fusion switching module. The fusion switching module performs seamless switching of navigation output: it compares the dual-channel solution results in real time based on the reliability score. When the performance of the primary fusion unit deteriorates, a low-pass filtering algorithm is used to weight the outputs of the primary and backup fusion units, suppressing state jumps during the switching process. Finally, the stabilized navigation state data is output to the data output subsystem. Through these steps, the fusion redundancy subsystem achieves accurate output of navigation state estimation under abnormal fusion algorithm scenarios through a three-level collaborative mechanism of dual-redundancy solution, dynamic evaluation, and smooth switching, significantly improving the robustness and reliability of the integrated navigation system.

[0077] According to an embodiment of the present invention, based on the time-domain navigation state data with delay after switching fusion, time-domain delay compensation is performed to obtain the current time-domain navigation state data, which is then encapsulated and processed to publish navigation data, specifically including:

[0078] Online estimation of the delay error between observed data and predicted state based on the least squares method;

[0079] Based on delay error, navigation status data in the current time domain is generated by extrapolating real-time measurement data from the IMU.

[0080] The navigation state data synchronized in the time domain and the fusion state data are structured, encapsulated, and published.

[0081] It should be noted that the data output subsystem consists of a time-domain delay compensation module and a data encapsulation module. Through dynamic compensation and structured encapsulation, it achieves high timeliness and completeness in the output of navigation data. The time-domain delay compensation module performs adaptive time-domain synchronization: it dynamically adjusts the storage length of the navigation state history buffer based on the current motion intensity of the aircraft, estimates the delay error between observed data and predicted state online using the least squares method, and generates the current time-domain navigation state information by extrapolating real-time IMU measurement data, eliminating the time lag caused by data transmission and processing links. The data encapsulation module performs multi-dimensional information integration: it structurally encapsulates the time-domain synchronized navigation state information and system operating state information. The navigation state information includes attitude, velocity, and position, while the system operating state information includes sensor health flags, fusion unit operating state information, and switching flags. The encapsulated data packet is output through a standard communication protocol. Through these steps, the data output subsystem ensures the real-time performance and resolvability of navigation data through dynamic delay compensation and structured data encapsulation technology, while providing a complete state feedback data chain.

[0082] like Figures 3-4 As shown, in a second aspect, embodiments of this application provide a redundant integrated navigation system based on unmanned aerial vehicles (UAVs). The system includes a memory and a processor. The memory includes a program for a redundant integrated navigation method based on UAVs. When the program for the redundant integrated navigation method based on UAVs is executed by the processor, it implements the following steps:

[0083] Based on the original sensor measurement data, the original measurement data is preprocessed, and numerical optimization and downsampling are performed to obtain the processed sensor data.

[0084] Based on the preprocessed sensor data, multi-sensor fault diagnosis and comprehensive performance evaluation are performed to determine whether redundancy switching is required.

[0085] Based on the sensor status observation data after switching, perform independent navigation calculation and comprehensive performance evaluation of the multi-fusion unit to determine whether redundant switching is required.

[0086] Based on the time-domain navigation state data with delay after switching fusion, time-domain delay compensation is performed to obtain the current time-domain navigation state data, which is then encapsulated and processed to publish navigation data.

[0087] It should be noted that, based on external sensor hardware, the data processing subsystem collects raw measurement data from redundant IMUs, GNSS, magnetic compasses, and barometers, and performs preliminary data processing (downsampling, filtering, etc.) on this data. The processed data is then transmitted as sensor data to the sensor redundancy subsystem. The sensor redundancy subsystem performs calculations, analysis, and judgment processing on the sensor data, converting it into state observation data, which is then transmitted to the fusion redundancy subsystem. The fusion redundancy subsystem receives the state observation data, performs data fusion algorithm processing on it, generates navigation state data, and transmits it to the data output subsystem. The data output subsystem performs time-domain delay compensation processing on the received navigation state data and then outputs it.

[0088] According to an embodiment of the present invention, based on the original sensor measurement data, the original measurement data is preprocessed, and numerical optimization and downsampling are performed to obtain processed sensor data, specifically including:

[0089] Zero bias error compensation and scaling factor error compensation are performed on sensor data based on preset calibration parameters;

[0090] The sensor temperature characteristic curve was established through high and low temperature environment tests, and real-time online temperature compensation was performed based on the temperature characteristic curve.

[0091] Sensor data is synchronized in time by timestamp alignment, and abnormal data is identified and removed based on statistical methods.

[0092] Online adaptive filtering and smoothing of sensor data are performed based on Fast Fourier Transform (FFT) to obtain optimized sensor data;

[0093] The optimized sensor data is downsampled using the equal-interval averaging method to obtain the processing result.

[0094] It should be noted that the data processing subsystem consists of a preprocessing module, a numerical optimization module, and a downsampling module, achieving precise optimization of sensor data through a multi-level data processing flow. The preprocessing module performs three levels of error compensation and outputs the processing results to the sensor redundancy subsystem. Through the above steps, the data processing subsystem achieves step-by-step correction of sensor errors and precise processing of data density through a hierarchical data processing mechanism.

[0095] According to an embodiment of the present invention, sensor fault diagnosis is performed based on the processing results to obtain diagnostic results, specifically including:

[0096] The sensor data is measured based on the processing results to obtain the measured values;

[0097] The measured value is compared with a set threshold to obtain the measurement deviation rate;

[0098] Determine whether the measurement deviation rate is greater than or equal to the set measurement deviation rate threshold;

[0099] If the measurement deviation rate is greater than or equal to the set threshold, FFT is used to perform spectrum analysis on the sensor data to identify abnormal vibration signals and obtain diagnostic results.

[0100] If the deviation rate is less than the set threshold, the sensor is considered to be fault-free.

[0101] It should be noted that the sensor's raw data is judged to exceed the measurement limit based on a preset threshold, the data loss status is identified through data continuity analysis, and sensor jamming is detected by combining time window value comparison. If the detection result is normal data, it is output to the sensor evaluation module. If a fault response is detected, a sensor switching command is triggered. The sensor fault diagnosis module belongs to the sensor redundancy subsystem and is used to realize sensor hardware fault identification.

[0102] According to an embodiment of the present invention, based on preprocessed sensor data, a comprehensive performance evaluation of multiple sensors is performed to determine whether redundancy switching is required, specifically including:

[0103] FFT was used to perform spectral analysis on sensor data to identify abnormal vibration signals;

[0104] The output differences of redundant sensors were compared through multi-sensor cross-consistency verification.

[0105] The residuals are calculated using the aircraft's physical model, geographic information, and kinematic constraints to determine the degree of deviation between the data and the theoretical values.

[0106] Based on the above performance analysis results, the confidence weights of each sensor are dynamically updated, and it is determined whether to switch sensors.

[0107] It should be noted that the sensor performance evaluation module belongs to the sensor redundancy subsystem. It compares the output differences of redundant sensors through multi-sensor cross-consistency verification, calculates residuals based on the aircraft physical model, geographic information, and kinematic constraints, determines the degree of deviation between the data and theoretical values, and dynamically updates the confidence weights of each sensor based on the above analysis results, outputting them to the sensor switching module. The sensor switching module, also part of the sensor redundancy subsystem, performs optimal sensor combination decisions: based on real-time confidence weights, it selects a primary sensor data group and an auxiliary sensor data group from the redundant sensor pool, consisting of two sets of IMU data, one set of magnetic compass data, one set of barometer data, and one set of GNSS data. The selected two sets of sensor data are then output to the fusion redundancy subsystem. Through these steps, the sensor redundancy subsystem ensures the normal operation of the two sets of sensor data input to the fusion redundancy subsystem via a three-level joint control mechanism of fault diagnosis, evaluation, and switching.

[0108] According to an embodiment of the present invention, based on the sensor state observation data after switching, multi-fusion unit independent navigation calculation and comprehensive performance evaluation are performed to determine whether redundant switching is required, specifically including:

[0109] Based on the data after sensor switching, the multi-fusion unit uses the Error State Kalman Filter (ESKF) algorithm to generate multiple sets of navigation solution results;

[0110] Based on the navigation calculation results, the chi-square test and residual analysis methods are used, combined with the sensor noise level and historical reliability to calculate the comprehensive credibility score, which is then used as the evaluation result of the fusion unit.

[0111] Based on the merging evaluation results, it is determined whether to perform merging switching. When merging switching is performed, a low-pass filtering algorithm is used to suppress transitions during the switching process.

[0112] It should be noted that the fusion redundancy subsystem consists of a fusion module, a fusion evaluation module, and a fusion switching module. The reliability of state estimation is achieved through a dual-channel fault-tolerant fusion and dynamic switching mechanism. The fusion module adopts a parallel computing software architecture with dual independent fusion algorithms: the main fusion module and the backup fusion module receive data from the main sensor group and the auxiliary sensor group, respectively, output by the sensor redundancy subsystem. Based on the Error State Kalman Filter (ESKF) algorithm, a 16-dimensional error state vector is defined, including position error, velocity error, attitude error, angular velocity error, and acceleration error. Nominal state prediction, error covariance matrix update, and observation residual calculation are performed recursively. Based on observation data from the accelerometer, GNSS, and magnetic compass, error state injection and covariance reset are completed, generating dual-channel navigation solution results and outputting them to the fusion evaluation module. The fusion evaluation module performs multi-dimensional verification of navigation performance: it uses a chi-square test to determine the statistical significance of the solution residuals, evaluates the convergence of the filter through the residual covariance matrix, and calculates a comprehensive reliability score of the fusion result based on the dynamic confidence score of the fusion result, considering factors such as sensor noise level and historical reliability. The evaluation result is then transmitted to the fusion switching module. The fusion switching module performs seamless switching of navigation output: it compares the dual-channel solution results in real time based on the reliability score. When the performance of the primary fusion unit deteriorates, a low-pass filtering algorithm is used to weight the outputs of the primary and backup fusion units, suppressing state jumps during the switching process. Finally, the stabilized navigation state data is output to the data output subsystem. Through these steps, the fusion redundancy subsystem achieves accurate output of navigation state estimation under abnormal fusion algorithm scenarios through a three-level collaborative mechanism of dual-redundancy solution, dynamic evaluation, and smooth switching, significantly improving the robustness and reliability of the integrated navigation system.

[0113] According to an embodiment of the present invention, based on the time-domain navigation state data with delay after switching fusion, time-domain delay compensation is performed to obtain the current time-domain navigation state data, which is then encapsulated and processed to publish navigation data, specifically including:

[0114] Online estimation of the delay error between observed data and predicted state based on the least squares method;

[0115] Based on delay error, navigation status data in the current time domain is generated by extrapolating real-time measurement data from the IMU.

[0116] The navigation state data synchronized in the time domain and the fusion state data are structured, encapsulated, and published.

[0117] It should be noted that the fusion unit status data includes status flag information and navigation solution information during the fusion unit's operation. The data output subsystem consists of a time-domain delay compensation module and a data encapsulation module. Through dynamic compensation and structured encapsulation, it achieves high timeliness and completeness in the output of navigation data. The time-domain delay compensation module performs adaptive time-domain synchronization: it dynamically adjusts the storage length of the navigation state history buffer based on the current motion intensity of the aircraft, estimates the delay error between the observed data and the predicted state online using the least squares method, and generates the current time-domain navigation state information by extrapolating real-time IMU measurement data, eliminating the time lag caused by data transmission and processing links. The data encapsulation module performs multi-dimensional information integration: it structures and encapsulates the time-domain synchronized navigation state information and system operating state information. The navigation state information includes attitude, velocity, and position, while the system operating state information includes sensor health flags, fusion unit operating state information, and switching flags. The encapsulated data packet is output through a standard communication protocol. Through these steps, the data output subsystem ensures the real-time performance and resolvability of navigation data through dynamic delay compensation and structured data encapsulation technology, while providing a complete state feedback data chain.

[0118] A third aspect of the present invention provides a computer-readable storage medium including a program for a redundant integrated navigation method based on an unmanned aerial vehicle (UAV), wherein when the program is executed by a processor, it implements the steps of the redundant integrated navigation method based on an UAV as described in any of the preceding claims.

[0119] This invention discloses a redundant integrated navigation method and system based on unmanned aerial vehicles (UAVs). The method involves preprocessing the raw sensor measurement data, performing numerical optimization and downsampling to obtain processed sensor data. Based on the preprocessed sensor data, multi-sensor fault diagnosis and comprehensive performance evaluation are performed to determine whether redundancy switching is necessary. Based on the sensor state observation data after switching, independent navigation calculations and comprehensive performance evaluations of multiple fusion units are performed to determine whether redundancy switching is necessary. Based on the delayed time-domain navigation state data of the fusion unit after switching, time-domain delay compensation is performed to obtain the current time-domain navigation state data, which is then encapsulated and published as navigation data. Through a dual redundancy mechanism of multi-sensor hardware and multi-fusion unit algorithm software, as well as a smooth switching mechanism for multiple fusion units, the robustness and reliability of UAV integrated navigation are significantly improved.

[0120] In the several embodiments provided in this application, it should be understood that the disclosed devices and methods can be implemented in other ways. The device embodiments described above are merely illustrative. For example, the division of units is only a logical functional division, and in actual implementation, there may be other division methods, such as: multiple units or components can be combined, or integrated into another system, or some features can be ignored or not executed. In addition, the coupling, direct coupling, or communication connection between the various components shown or discussed can be through some interfaces, and the indirect coupling or communication connection between devices or units can be electrical, mechanical, or other forms.

[0121] The units described above as separate components may or may not be physically separate. The components shown as units may or may not be physical units. They may be located in one place or distributed across multiple network units. Some or all of the units may be selected to achieve the purpose of this embodiment according to actual needs.

[0122] In addition, in the various embodiments of the present invention, each functional unit can be integrated into one processing unit, or each unit can be a separate unit, or two or more units can be integrated into one unit; the integrated unit can be implemented in hardware or in the form of hardware plus software functional units.

[0123] Those skilled in the art will understand that all or part of the steps of the above method embodiments can be implemented by hardware related to program instructions. The aforementioned program can be stored in a readable storage medium. When the program is executed, it performs the steps of the above method embodiments. The aforementioned storage medium includes various media capable of storing program code, such as mobile storage devices, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0124] Alternatively, if the integrated units of the present invention are implemented as software functional modules and sold or used as independent products, they can also be stored in a readable storage medium. Based on this understanding, the technical solutions of the embodiments of the present invention, or the parts that contribute to the prior art, can be embodied in the form of a software product. This software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the methods described in the various embodiments of the present invention. The aforementioned storage medium includes various media capable of storing program code, such as mobile storage devices, ROM, RAM, magnetic disks, or optical disks.

Claims

1. A redundancy combined navigation method based on unmanned aerial vehicle, characterized in that, The method comprises the following steps: Based on the sensor raw measurement data, the raw measurement data is preprocessed, and numerical optimization and downsampling are performed to obtain processed sensor data; Based on the preprocessed sensor data, multi-sensor fault diagnosis and comprehensive performance evaluation are performed to determine whether redundancy switching is needed; Based on the sensor state observation data after switching, multi-fusion independent navigation solution and comprehensive performance evaluation are performed to determine whether redundancy switching is needed; Based on the fusion delay time domain navigation state data after switching, time domain delay compensation is performed to obtain current time domain navigation state data, which is then packaged and processed to publish navigation data; The multi-sensor comprehensive performance evaluation specifically includes: FFT is used to analyze and identify abnormal vibration signals in the frequency spectrum of sensor data; The output difference of redundant sensors is compared through multi-sensor cross-consistency verification, and the redundant sensors refer to a set of multiple sensors of the same type; The residual error is calculated based on the aircraft physical model, geographic information and kinematic constraints to determine the deviation of the data from the theoretical value; The confidence weight of each sensor is dynamically updated based on the performance evaluation results, and it is determined whether to switch the sensor. 2.The UAV-based redundant integrated navigation method of claim 1, wherein, Based on the sensor raw measurement data, the raw measurement data is preprocessed, and numerical optimization and downsampling are performed to obtain processed sensor data, specifically including: Zero bias error compensation and scale factor error compensation are performed on the raw sensor data based on pre-set calibration parameters; A temperature characteristic curve of the sensor is established through high and low temperature environment test, and real-time online temperature compensation is performed based on the temperature characteristic curve; The original sensor data is time-synchronized through timestamp alignment, and abnormal data is identified and removed based on statistical methods; Optimized sensor data is obtained through online adaptive filtering and smoothing of the original sensor data based on fast Fourier transform (FFT); The optimized sensor data is downsampled based on the equal interval average method to obtain the processing result. 3.The UAV-based redundant integrated navigation method of claim 1, wherein, Based on the preprocessed sensor data, multi-sensor fault diagnosis and comprehensive performance evaluation are performed to determine whether redundancy switching is needed, wherein the multi-sensor fault diagnosis specifically includes: The preprocessed sensor data is measured based on a pre-set threshold value to determine whether it is out of range; Data loss state is identified through data continuity analysis; Sensor sticking detection is achieved by comparing the values in the time window; The fault diagnosis result is output by integrating the out-of-range detection, data loss detection and sticking detection; If the diagnosis result is normal, multi-sensor performance evaluation is performed, and if the diagnosis result is a fault response, sensor switching is performed. 4.The UAV-based redundant integrated navigation method of claim 1, wherein, Based on the sensor state observation data after switching, multi-fusion independent navigation solution and comprehensive performance evaluation are performed to determine whether redundancy switching is needed, specifically including: Based on the sensor switching data, multiple fusion devices use error state Kalman filter algorithm to generate multiple sets of navigation solution results; Based on the navigation solution results, chi-square test and residual analysis methods are used to calculate the comprehensive confidence score based on the sensor noise level and historical reliability, which is used as the fusion device evaluation result; Based on the fusion evaluation result, it is judged whether to perform fusion switch, and when the fusion switch is performed, a low-pass filtering algorithm is used to suppress the jump in the switching process.

5. The drone-based redundant integrated navigation method of claim 1, wherein, Based on the fusion delay time domain navigation state data after switching, time domain delay compensation is performed to obtain current time domain navigation state data, and encapsulation processing is performed thereon to publish navigation data, specifically including: Based on the least square method, the delay error between the observation data and the predicted state is estimated online; Based on the delay error, the current time domain navigation state data is generated by extrapolation through the IMU real-time measurement data; The time domain synchronized navigation state data and the fusion state data are structured and encapsulated and published.

6. An unmanned aerial vehicle based redundant integrated navigation system, characterized in that, The system comprises a memory and a processor, the memory comprising a program of a redundancy combined navigation method based on a UAV, and the processor executing the method of any one of claims 1-5.

Images