A fault-tolerant multi-source fusion navigation and positioning method based on multi-solution separation

Abstract

This invention discloses a fault-tolerant multi-source fusion navigation and positioning method based on multiple solution separation, belonging to the field of integrity monitoring technology. The method includes steps such as constructing a risk fault tree model, dividing the fault mode set, parallel filtering positioning, spatiotemporal synchronization, robust fusion, and protection level calculation to address multiple risk sources faced by multi-source fusion navigation systems in complex environments. This achieves rigorous monitoring and elimination of risk sources from different sensors, ultimately obtaining a fault-tolerant positioning solution. This invention integrates integrity prior information with dynamic risk monitoring capabilities, constructing a highly robust and flexibly adaptable navigation and positioning framework that can ensure the integrity of multi-source fusion systems and meet the high-precision, high-safety requirements of intelligent transportation, unmanned systems, and other scenarios.

Description

Technical Field

[0001] This invention belongs to the field of integrity monitoring technology, specifically relating to a fault-tolerant multi-source fusion navigation and positioning method based on multi-solution separation. Background Technology

[0002] Global Navigation Satellite System (GNSS) is currently the most widely used radio signal positioning and navigation technology. Its low cost, simple and reliable technology have made it a popular choice among users, making it an indispensable infrastructure in modern production and daily life. It is also widely used in autonomous systems closely related to life safety, such as autonomous driving and drones. However, GNSS navigation signals are susceptible to complex environmental factors such as urban obstruction, multipath effects, and ionospheric interference, resulting in significant limitations for single GNSS systems in practical applications. Therefore, current solutions typically employ a combined navigation system centered on GNSS / INS, integrating multi-source heterogeneous information from visual sensors and odometry to improve positioning accuracy, stability, and robustness.

[0003] While multi-sensor fusion enhances system performance, it also introduces more potential risk sources, leading to complex fault propagation paths and making it difficult to guarantee system integrity. How to effectively monitor and suppress potential faults in multi-sensor fusion systems while improving positioning accuracy has become a pressing technical challenge. Therefore, there is an urgent need to build a multi-sensor integrated positioning platform that is pluggable and supports integrity assurance. Summary of the Invention

[0004] To address the aforementioned technical issues, this invention provides a fault-tolerant multi-source fusion navigation and positioning method based on multi-solution separation. By providing integrity prior information at the system end and using user terminal positioning and integrity monitoring algorithms, it rigorously monitors and eliminates different sensor risk sources that threaten navigation and positioning. This constructs an integrated navigation and positioning framework with strong robustness and flexible adaptability, aiming to provide technical support for future intelligent transportation, unmanned systems, and other scenarios with extremely high requirements for safety and high precision.

[0005] To achieve the above objectives, the technical solution adopted by the present invention is as follows:

[0006] A fault-tolerant multi-source fusion navigation and positioning method based on multi-solution separation, the method comprising:

[0007] Step 1: Analyze the risk sources of GNSS and auxiliary sensors in complex environments, determine the prior probability of failure based on historical statistical data, construct a risk fault tree model, and allocate integrity risk and continuity risk.

[0008] Step 2: Based on the risk source type and prior probability of failure of the multi-source sensor, summarize and calculate the fault modes to be monitored, and generate a set of fault modes for which the location solution needs to be calculated later.

[0009] Step 3: Locate the fault mode set using parallel filtering methods, and design a parallel filter management strategy and a sensor dynamic plug-in / plug-out management strategy.

[0010] Step 4: Achieve hard synchronization of sensor time through PPS, and unify the spatial coordinate system using a rigid body transformation model to ensure spatiotemporal consistency;

[0011] Step 5: Robustly fuse the sensor data under each filter and output the localization solution, localization error covariance, and solution separation error covariance for each fault mode.

[0012] Step 6: Calculate the protection level of the integrated navigation system by combining the prior fault probability, integrity risk and continuity risk, positioning error covariance and deseparation error covariance.

[0013] Step 7: Based on the protection level and the location solution, perform fault-tolerant location estimation.

[0014] Furthermore, step 1 includes considering GNSS risk source types including occlusion, reflection, multipath effect, and cycle slip, as well as the fault behavior characteristics of inertial and auxiliary sensors including INS and visual odometry, and determining the prior probability of faults based on system operation scenarios and historical statistical data; by analyzing the causes and evolution paths of risk sources, establishing a risk source fault tree model and allocating it to continuity risk and integrity risk.

[0015] Furthermore, in step 2, the maximum number of satellites that can fail simultaneously is determined using the prior probability of GNSS satellite and constellation failures. At the same time, the failure situations of inertial and auxiliary sensors, including INS and visual odometry, are considered. Finally, the failure modes to be monitored in the multi-source fusion navigation and positioning system are determined, and a set of failure modes is generated, including a complete set of failure modes, a subset of failure modes, and a sub-subset of failure modes. The complete set of failure modes refers to all sensor observations corresponding to the failure-free mode. The subset of failure modes refers to the sensor observations corresponding to any one failure mode among the failure modes to be monitored. The sub-subset of failure modes refers to the sensor observations corresponding to any two failure modes among the failure modes to be monitored.

[0016] Furthermore, in step 3, the parallel filter management strategy includes using the main filter to calculate parameters that are insensitive to user location and sharing them with the remaining sub-filters, with each filter independently performing combined navigation and positioning estimation; the sensor plug-in / plug-out management strategy includes that when a sensor is online, the main / sub-filters perform parameter inheritance and bi-differentiation; when a sensor is offline, the monitoring of relevant fault modes is retained to ensure the ability to identify historical fault states.

[0017] Furthermore, step 4 includes hard synchronization using high-precision satellite pulse-per-second (PPS) signals, spatial calibration of observation data from each sensor in its own coordinate system using a rigid body transformation model, and unified conversion to a unified reference coordinate system for data fusion and filtering.

[0018] Furthermore, step 5 includes using an extended Kalman filter or an unscented Kalman filter robust fusion algorithm to perform positioning calculations on the observation data, and outputting the positioning solution, positioning error covariance, and solution separation error covariance for each set respectively; the positioning solution of the set under different fault modes is determined by whether there are sensors other than GNSS to adopt precise single-point positioning or loose combination navigation positioning methods.

[0019] Furthermore, step 6 includes using the positioning error covariance of all sets and the solution to separate the positioning error covariance, and combining the prior probabilities of all failure modes, based on the dynamic allocation of continuity risk and integrity risk on the whole set and subsets, and using the bisection method to calculate the protection level of the integrated navigation system in the current state, that is, the maximum tolerable error limit that the system positioning solution can achieve.

[0020] Furthermore, step 7 includes performing fault-tolerant location estimation using the protection level and the location solution results of all sets. The fault-tolerant location estimation is obtained by the union of the consistency check intervals of the location solutions under each fault mode, wherein the consistency check interval under each fault mode is considered as a weight for the location solution of each fault mode.

[0021] In a second aspect, the present invention provides an electronic device, comprising: one or more processors; and a memory for storing one or more programs; wherein, when the one or more programs are executed by the one or more processors, the one or more processors implement the aforementioned fault-tolerant multi-source fusion navigation and positioning method based on multiple decoupling.

[0022] Thirdly, the present invention provides a computer-readable storage medium having executable instructions stored thereon, which, when executed by a processor, enable the processor to implement the aforementioned fault-tolerant multi-source fusion navigation and positioning method based on multiple solution separation.

[0023] The beneficial effects of this invention are as follows:

[0024] This invention addresses the impact of multiple risk sources on the reliability of positioning estimation in complex environments for multi-source fusion navigation systems, proposing a highly flexible and robust multi-solution separation fault-tolerant navigation and positioning method. By constructing a multi-level fault mode system, a parallel filtering mechanism, and a dynamic sensor plug-and-play management strategy, elastic positioning of the multi-source fusion navigation system in complex environments is achieved. This method can shield the system output from the influence of risk sources under any single-point failure or multi-point common-mode failure, ensuring that the system still provides high-precision and high-reliability positioning results. Simultaneously, it supports dynamic sensor plug-and-play, possessing good scalability and adaptability, and can meet the needs of high-precision and high-safety application scenarios such as intelligent transportation and unmanned systems, providing reliable technical support for future intelligent and unmanned applications. Attached Figure Description

[0025] Figure 1 This is a flowchart of a fault-tolerant multi-source fusion navigation and positioning method based on multiple solution separation according to the present invention. Detailed Implementation

[0026] The present invention will be further described below with reference to the accompanying drawings and embodiments.

[0027] The core of this invention lies in the deep integration of an "integrity assurance mechanism" and a "pluggable sensor architecture." For different risk source types, it constructs a multi-level fault mode system and a dynamic estimation algorithm based on parallel filtering, achieving coordinated operation of sensor pluggable adaptation, real-time fault mode updates, and location of integrity constraints. The invention includes the following key technical steps: Specific steps include:

[0028] Step 1: Analyze the risk sources of GNSS and auxiliary sensors in complex urban environments. Determine the prior probability of failure based on historical statistical data, construct a corresponding fault tree model, and allocate integrity risk and continuity risk. Considering common GNSS risk source types in urban environments such as occlusion, reflection, multipath effect, and cycle slip, as well as the fault behavior characteristics of inertial and auxiliary sensors such as INS and visual odometry, establish a risk source fault tree model based on system operation scenarios and statistical data. The impact of risk sources on sensors is reflected in the measurement values ​​of each sensor. By analyzing the causes and evolution paths of risk sources, allocate continuity risk and integrity risk to them, providing prior risk input for subsequent confidence interval calculation of positioning solutions and protection level assessment.

[0029] Step 2: Based on the multiple risk source types and prior failure probabilities of the multi-source sensors summarized in Step 1, calculate the failure modes to be monitored and generate a set of failure modes. First, calculate the maximum number of GNSS satellites that can fail simultaneously, N (e.g., N=1, meaning at most one satellite can fail at the same time). Simultaneously consider failures of inertial and auxiliary sensors, including INS and visual odometry, to determine the failure modes to be monitored for the multi-source fusion navigation and positioning system (including the fault-free mode H0, the fault mode H...). i（i=1….k） The fault mode Hi can be any of the possible fault modes, such as multiple single-satellite faults or faults in other sensors like inertial navigation systems, totaling k. Secondly, a fault mode set is generated using the fault modes to be monitored, specifically including the complete set (containing all sensor observations corresponding to the fault-free mode H0) and a subset (excluding the fault modes to be monitored H0). i The fault mode set is divided into three parts: a set of sensor observations corresponding to any one fault mode, a subset (excluding sensor observations corresponding to any two fault modes in the fault mode to be monitored), and a subset. Finally, the fault mode set localization solution is defined as the whole set solution, the subset solution, and the subsubset solution, which are all obtained by localization estimation from the observations corresponding to the whole set, subset, and subsubset of the fault mode set.

[0030] Step 3 involves using parallel filters to locate all fault mode sets from Step 2, and designing corresponding management strategies for these parallel filters and a dynamic sensor plug-in / plug-out management strategy. The management strategy for the parallel filters involves using a main filter to calculate parameters insensitive to user location (such as satellite position) and sharing them with the remaining sub-filters. The main filter is used for positioning estimation based on observations corresponding to the entire fault mode set, while the remaining positioning estimation filters are sub-filters, reducing the number of state parameters that need to be repeatedly estimated during filtering. Each filter independently performs combined navigation positioning estimation. The dynamic sensor plug-in / plug-out management measure involves adjusting the fault mode set monitored by the algorithm and subsequently adjusting the filter set when the navigation system suddenly receives data from a new sensor (referred to as a "sensor online" event) or when a sensor in use suddenly becomes unavailable due to interference (referred to as a "sensor offline" event). Specifically, when a new sensor becomes online, each main / sub-filter splits into two filters, both inheriting the parameters of the original filters. When a sensor in use suddenly becomes unavailable, the impact of that unavailable sensor on most filters does not disappear. Therefore, it is necessary to consider the possibility that the sensor may have failed before it became unavailable, that is, the failure modes associated with the unavailable sensor still need to be monitored (that is, no adjustments should be made to the filter set after the sensor is taken offline).

[0031] Step 4: Sensor time hard synchronization is achieved through PPS. The rigid body transformation model is used to unify the spatial coordinate system and ensure the spatiotemporal consistency of observation data from various sensors during fusion. To ensure the spatiotemporal consistency of observations from different sources, this positioning platform uses high-precision satellite pulse-per-second (PPS) signals for hard synchronization, achieving time unification for GNSS, INS, odometers, and other devices. At the same time, a rigid body transformation model is used to achieve spatial calibration of observation data from each sensor in its own coordinate system, and then uniformly transforms the data to a unified reference coordinate system for data fusion and filtering, ensuring the geometric consistency of the fused solution.

[0032] Step 5: For each sensor that has undergone spatial and temporal synchronization under each filter in Step 4, robust fusion algorithms such as Extended Kalman Filter (EKF) or Unscented Kalman Filter (UKF) are used to calculate the positioning solution of the observation data in each subset. The positioning solution, positioning error covariance, and solution separation error covariance of each subset are output for fault-tolerant positioning estimation. The positioning solution of the subset under different fault modes needs to be determined by whether there are sensors other than GNSS to determine whether to use precise point positioning or loosely combined navigation positioning. Specifically, if a subset considers INS failure, the positioning of this subset requires the use of GNSS and odometer observations, so a loosely combined positioning method is used; if a subset considers both INS and odometer failure, the positioning of this subset only uses GNSS observations, so a precise point positioning method is used.

[0033] Step 6: Using the positioning error covariance of all sets calculated in Step 5 and the separation positioning error covariance, and combining the prior probabilities of all failure modes in Step 1, the optimal protection level of the integrated navigation system in the current state is calculated based on the dynamic allocation of continuity risk and integrity risk on the whole set and subsets. That is, the maximum tolerable error limit that the system positioning solution can achieve. This is obtained by using the bisection method based on the integrity risk and continuity risk constraints.

[0034] Step 7: Using the protection level obtained in Step 6 and combining it with the positioning results of all sets obtained in Step 5, fault-tolerant positioning estimation is performed. This fault-tolerant positioning estimation is obtained by taking the union of the consistency check intervals of the positioning solutions under each fault mode. The consistency check interval for each fault mode is considered a weight for the positioning solution under each fault mode, thereby eliminating potential fault influences and ensuring the integrity of the multi-source sensor integrated navigation positioning solution. This solution can shield the impact of risk sources on the system output, ensuring that the system can still provide a reliable and intact positioning output even under any single-point failure or multi-point common-mode failure.

[0035] In a second aspect, the present invention provides an electronic device, comprising: one or more processors; and a memory for storing one or more programs; wherein, when the one or more programs are executed by the one or more processors, the one or more processors implement the aforementioned fault-tolerant multi-source fusion navigation and positioning method based on multiple decoupling.

[0036] Thirdly, the present invention provides a computer-readable storage medium having executable instructions stored thereon, which, when executed by a processor, enable the processor to implement the aforementioned fault-tolerant multi-source fusion navigation and positioning method based on multiple solution separation.

[0037] The specific embodiments described above further illustrate the purpose, technical solution, and beneficial effects of the present invention. It should be understood that the above descriptions are merely specific embodiments of the present invention and are not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A fault-tolerant multi-source fusion navigation and positioning method based on multiple solution separation, characterized in that, The method includes: Step 1: Analyze the risk sources of GNSS and auxiliary sensors in complex environments, determine the prior probability of failure based on historical statistical data, construct a risk fault tree model, and allocate integrity risk and continuity risk. Step 2: Based on the risk source type and prior probability of failure of the multi-source sensor, summarize and calculate the fault modes to be monitored, and generate a set of fault modes for which the location solution needs to be calculated later. Step 3: Locate the fault mode set using parallel filtering methods, and design a parallel filter management strategy and a sensor dynamic plug-in / plug-out management strategy. Step 4: Achieve sensor time hard synchronization through high-precision satellite pulse-per-second (PPS) signals, and unify the spatial coordinate system using a rigid body transformation model to ensure spatiotemporal consistency; Step 5: Robustly fuse the sensor data under each filter and output the localization solution, localization error covariance, and solution separation error covariance for each fault mode. Step 6: Calculate the protection level of the integrated navigation system by combining the prior probability of failure, integrity risk and continuity risk, positioning error covariance and deseparation error covariance. Step 7: Based on the protection level and the location solution, perform fault-tolerant location estimation.

2. The fault-tolerant multi-source fusion navigation and positioning method based on multiple solution separation according to claim 1, characterized in that, Step 1 includes GNSS risk source types such as occlusion, reflection, multipath effect, and cycle slip, as well as fault behavior characteristics of inertial and auxiliary sensors including INS and visual odometry, and determines the prior probability of faults based on system operating scenarios and historical statistical data. By analyzing the causes and evolution paths of risk sources, a fault tree model of risk sources is established and its continuity risk and integrity risk are allocated.

3. The fault-tolerant multi-source fusion navigation and positioning method based on multiple solution separation according to claim 1, characterized in that, In step 2, the maximum number of satellites that can fail simultaneously is determined using the prior probability of GNSS satellite and constellation failures. At the same time, based on the failure status of inertial and auxiliary sensors such as INS and visual odometry, the failure modes to be monitored in the multi-source fusion navigation and positioning system are finally determined, and a set of failure modes is generated, including a complete set of failure modes, a subset of failure modes, and a sub-subset of failure modes. The complete set of failure modes refers to all sensor observations corresponding to the failure-free mode. The subset of failure modes refers to the sensor observations corresponding to any one failure mode among the failure modes to be monitored. The sub-subset of failure modes refers to the sensor observations corresponding to any two failure modes among the failure modes to be monitored.

4. The fault-tolerant multi-source fusion navigation and positioning method based on multiple solution separation according to claim 1, characterized in that, In step 3, the parallel filter management strategy includes using the main filter to calculate parameters that are insensitive to user location and sharing them with the remaining sub-filters, with each filter independently performing combined navigation and positioning estimation; the sensor dynamic plug-in / plug-out management strategy includes that when a sensor is online, the main / sub-filters inherit and bi-differentiate parameters; when a sensor is offline, monitoring of relevant fault modes is retained to ensure the ability to identify historical fault states.

5. The fault-tolerant multi-source fusion navigation and positioning method based on multiple solution separation according to claim 1, characterized in that, Step 4 includes hard synchronization using high-precision satellite pulse-per-second (PPS) signals, spatial calibration of observation data from each sensor in its own coordinate system using a rigid body transformation model, and unified conversion to a unified reference coordinate system for data fusion and filtering.

6. The fault-tolerant multi-source fusion navigation and positioning method based on multiple solution separation according to claim 1, characterized in that, Step 6 includes using the positioning error covariance of all sets and the solution to separate the positioning error covariance, and combining the prior probabilities of all failure modes. Based on the dynamic allocation of continuity risk and integrity risk on the whole set and subsets, the bisection method is used to calculate the protection level of the integrated navigation system in the current state, that is, the maximum tolerable error limit that the system positioning solution can achieve.

7. The fault-tolerant multi-source fusion navigation and positioning method based on multiple solution separation according to claim 1, characterized in that, Step 7 includes performing fault-tolerant location estimation using the protection level and the location solution results of all sets. The fault-tolerant location estimation is obtained by the union of the consistency check intervals of the location solutions under each fault mode, wherein the consistency check interval under each fault mode is used to determine the weight of the location solution for each fault mode.

8. An electronic device, characterized in that, include: One or more processors; Memory, used to store one or more programs; When one or more programs are executed by the one or more processors, the one or more processors implement the fault-tolerant multi-source fusion navigation and positioning method based on multiple solution separation as described in any one of claims 1-7.

9. A computer-readable storage medium, characterized in that, It stores executable instructions that, when executed by a processor, enable the processor to implement the fault-tolerant multi-source fusion navigation and positioning method based on multiple solution separation as described in any one of claims 1-7.

Images