High-reliability intelligent navigation method based on granular computing and uncertainty analysis

Abstract

The invention discloses a high-reliability intelligent navigation method based on granular calculation and uncertainty analysis, which belongs to the technical field of navigation and positioning, and comprises the following steps: uniformly abstracting original data from multi-source sensors such as satellite positioning, vision, radio frequency identification and the like into information granules containing a core navigation state, uncertainty description and mathematical representation; constructing a fusion model formed by fuzzy reasoning and a rough set based on the information grains, and calculating certainty and possibility evaluation of a navigation state; a specific index is introduced, and a sensor combination is optimized through iteration, so that dynamic self-adaptive sensor selection aiming at the current environment is realized; and finally, outputting a navigation decision according to a fusion result of the selected combination. According to the method, the weight of the sensor can be adaptively adjusted under complex working conditions that the GNSS is affected by electromagnetic interference or vision is affected by severe weather, the navigation robustness and continuity are improved, and the method is suitable for key application scenes such as power substation inspection.

Description

Technical Field

[0001] This application relates to the field of navigation and positioning technology, and in particular to a highly reliable intelligent navigation method based on granular computing and uncertainty analysis. Background Technology

[0002] Currently, substation inspection robots commonly employ a variety of navigation technologies, including GNSS (such as BeiDou), lidar point cloud, visual navigation, and RFID. However, most systems still rely on a single sensor source or simple coupling, such as GNSS+IMU or lidar+vision combinations. While these methods can achieve high-precision positioning and path planning in ideal environments, their stability and reliability are severely lacking in the complex electromagnetic environment and harsh weather conditions of substations.

[0003] First, traditional sensor fusion algorithms are mostly based on Kalman filtering or its extensions (EKF, UKF). These algorithms assume that the observation noise follows a Gaussian distribution and has fixed statistical properties. However, in substation scenarios: GNSS signals are susceptible to high-voltage electromagnetic interference and obstruction, resulting in significant multipath errors; lidar generates nonlinear scattering noise in fog and freezing rain environments; and visual feature points are unstable under strong reflective light or low illumination conditions. These anomalous data all violate the linear Gaussian assumption, causing the fusion results to drift or lose lock. Traditional algorithms struggle to handle this ambiguity and non-Gaussian uncertainty, thus the information fusion layer exhibits a "coarsening" characteristic.

[0004] Secondly, existing systems mostly employ fixed-weight models or preset logical rules, lacking dynamic evaluation of the reliability of multi-source navigation data. When the accuracy of the primary navigation source (such as GNSS) drops sharply due to interference, the system cannot automatically reduce its weight or switch auxiliary sources (such as vision or radar). This rigid fusion mechanism makes the navigation system lack self-adjustment capabilities in complex environments, prone to cascading failures, and seriously affects the system's robustness.

[0005] Finally, the high-humidity mountainous environments of southern China, exemplified by Guizhou, are frequently characterized by freezing rain, fog, high humidity, and strong electromagnetic interference. These factors cause distortion or loss of observations from navigation sensors, while existing systems generally lack targeted modeling and compensation mechanisms. As a result, robots experience path drift, misjudgment of obstacles, and frequent task interruptions during inspections, significantly reducing their reliability.

[0006] Due to the aforementioned issues, current inspection robots have the following limitations: low reliability, navigation is easily interfered with and fails; high maintenance costs, requiring frequent manual intervention and recalibration; and insufficient intelligence, making it difficult to achieve fully autonomous inspection tasks.

[0007] Therefore, there is an urgent need for a new navigation technology system that can operate stably in complex environments and has adaptive and generalization capabilities. Summary of the Invention

[0008] This invention addresses the technical problems of existing substation inspection robots in complex electromagnetic and harsh weather environments, such as positioning errors, inspection interruptions, and even equipment collisions caused by high uncertainty in sensor data and poor reliability of fusion navigation. It proposes a highly reliable intelligent navigation method based on granular computing and uncertainty analysis. By establishing a granular representation model of multi-source information, fine-grained modeling and reliability assessment of sensor observation data are performed. Furthermore, by combining fuzzy rough set theory, the weights of each sensor source are dynamically optimized, achieving intelligent adaptive adjustment of the fusion strategy. Thus, even under all operating conditions including strong interference, obstruction, freezing rain, and fog, high-precision, continuous, and robust autonomous navigation can still be achieved, effectively avoiding problems such as positioning errors, inspection interruptions, and equipment collisions, significantly improving the safety and operational reliability of the inspection robot.

[0009] To solve the technical problem, the technical solution of the present invention is as follows:

[0010] A highly reliable intelligent navigation method based on granular computation and uncertainty analysis, the method comprising:

[0011] Acquire multi-source navigation sensor data and abstract the multi-source navigation sensor data into a unified information granule containing navigation state information and its uncertainty description;

[0012] Based on the uncertainty description contained in the information granules, a fuzzy inference model is constructed to characterize the degree to which the navigation state meets the target navigation conditions. Uncertainty fusion is performed on the multi-source navigation sensor data to obtain the deterministic and probability assessment results of the navigation state.

[0013] Based on the determination and probability assessment results, the sensor set participating in navigation decision-making is dynamically evaluated and selected to generate a sensor combination adapted to the current environmental state.

[0014] Based on the information granule fusion results corresponding to the sensor combination, the navigation decision result is output.

[0015] Furthermore, the information granules include: core information for characterizing the most likely navigation state, uncertainty information for characterizing the reliability of the core information, and a mathematical representation for describing the possible range of values ​​for the navigation state; wherein, the uncertainty information is used to reflect the reliability of the corresponding navigation sensor under the current environmental conditions and serves as a weighting basis in the subsequent fusion and decision-making process.

[0016] Furthermore, the uncertainty fusion process is based on a fuzzy relation model, which represents whether the navigation state meets the target navigation conditions as a fuzzy concept, and calculates the deterministic evaluation result and the probability evaluation result corresponding to the fuzzy concept; wherein, the deterministic evaluation result is characterized by the lower approximation of the fuzzy rough set, and the probability evaluation result is characterized by the upper approximation of the fuzzy rough set.

[0017] Furthermore, the process of dynamically evaluating and selecting the sensor set is based on a specificity index used to measure the ability of different sensor combinations to distinguish navigation decisions. The specificity index is used to reflect the degree of distinction of the current sensor combination over different navigation decision results. The larger the specificity index, the stronger the ability of the sensor combination to distinguish the navigation state. Sensors that improve the specificity index are gradually added to the current sensor set through an iterative approach, and the iteration is terminated when the specificity index no longer improves significantly.

[0018] Furthermore, the multi-source navigation sensor includes: a satellite positioning sensor, a visual sensor, and a radio frequency identification sensor;

[0019] The uncertainty information of the information particles constructed by the satellite positioning sensor is characterized based on positioning accuracy parameters and signal quality parameters;

[0020] The uncertainty information of the information particles constructed by the aforementioned visual sensor is obtained through fuzzy reasoning, based on a combination of environmental factors and visual feature information.

[0021] The uncertainty information of the information granules constructed by the radio frequency identification sensor is characterized using evidence theory to describe the credibility of the navigation state in different regions.

[0022] Furthermore, the uncertainty information is autonomously calculated by the corresponding navigation sensor based on its own operating status and environmental conditions.

[0023] A highly reliable intelligent navigation system based on granular computing and uncertainty analysis, characterized in that the system is used to execute any of the methods described above, the system comprising:

[0024] The data acquisition and granulation module is used to acquire multi-source navigation sensor data and abstract the multi-source navigation sensor data into information granules that contain navigation state information and its uncertainty description.

[0025] The uncertainty reasoning and fusion module is used to construct a fuzzy reasoning model based on the uncertainty description contained in the information granules to characterize the degree to which the navigation state meets the target navigation conditions, and to perform uncertainty fusion on the multi-source navigation sensor data to obtain the deterministic and probability assessment results of the navigation state.

[0026] The sensor set dynamic optimization module is used to dynamically evaluate and select the sensor set participating in navigation decision-making based on the deterministic and probabilities assessment results, and generate a sensor combination that is adapted to the current environmental state.

[0027] The decision fusion and navigation output module is used to output navigation decision results based on the information granule fusion results corresponding to the sensor combination.

[0028] A computer device includes a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the program to implement a highly reliable intelligent navigation method based on granular computing and uncertainty analysis as described above.

[0029] A computer-readable storage medium storing a computer program that, when executed by a processor, implements any of the above-described methods for highly reliable intelligent navigation based on granular computing and uncertainty analysis.

[0030] This application has the following advantages:

[0031] (1) This application takes granular computing as its core idea and performs fine-grained modeling and uncertainty quantification on data from various sensors (such as lidar, vision, inertial navigation, GPS, etc.). By constructing an information granular hierarchical structure, the system can not only extract features and evaluate the credibility of data from each type of sensor source, but also achieve robust processing of noise and abnormal data.

[0032] In harsh environments such as strong electromagnetic interference, obstruction, rain, and fog, traditional navigation systems are prone to positioning drift, data loss, or signal lock-off. Granular computing models, however, introduce fuzzy weighting and coarse boundary analysis mechanisms during the information fusion stage. This dynamically adjusts the influence factors of each navigation source, maximizing the retention of highly reliable data and achieving high-precision, continuous, and secure fusion navigation.

[0033] Through this mechanism, the robot can maintain stable positioning in complex substation scenarios, avoid colliding with equipment or deviating from the inspection route, and significantly improve the reliability of task execution.

[0034] (2) The built-in GNFRS-Nav (Granular and Neural Fuzzy Rough Set Navigation) algorithm integrates the ideas of neural fuzzy reasoning and rough set simplification, and can dynamically evaluate the credibility of each navigation source according to the real-time environment.

[0035] In different scenarios (such as indoor / outdoor switching, obstruction, signal attenuation, etc.), the algorithm can automatically identify the current optimal sensor combination and achieve adaptive switching between sensor selection and navigation strategy. For example, when GNSS signals are interfered with, the system will automatically increase the weight of visual odometry and inertial measurement unit to maintain navigation continuity; while in areas with sparse visual features, it will re-weight and fuse GNSS and radar data.

[0036] This intelligent adaptive mechanism allows the system to adjust the navigation mode without human intervention, ensuring a smooth, stable, and efficient navigation process.

[0037] (3) One of the core advantages of this application is the universality and scalability of its model. Since the granular computing and fuzzy rough set model are independent of specific sensor types and robot platforms, the system does not need to be reconstructed extensively for different models or configurations.

[0038] By defining a unified information granular structure and a standardized fusion interface, this framework can easily connect to navigation modules from different brands (such as IMUs, radars, or cameras from different manufacturers) and quickly achieve adaptation.

[0039] This means that the solution is versatile across platforms and scenarios, and can be widely applied to inspection robot systems in various substations, power transmission channels, and other complex industrial environments. It not only reduces the cost of system upgrades and maintenance, but also lays a solid technical foundation for future scenarios such as autonomous inspection and collaborative navigation.

[0040] In summary, the high-reliability navigation method based on granular computing and fuzzy rough sets can achieve robust data fusion, intelligent adaptation at the decision layer, and generalized scalability at the system layer in complex substation environments.

[0041] This technology breaks through the limitations of traditional filtering models and provides a unified, intelligent, and evolvable navigation solution for the next generation of intelligent inspection robots, which has significant engineering application value and theoretical innovation significance. Attached Figure Description

[0042] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0043] Figure 1 This application provides a technical roadmap for a highly reliable intelligent navigation method based on granular computing and uncertainty analysis in its embodiments.

[0044] Figure 2 A flowchart of the GNFRS-Nav dynamic sensor optimization algorithm provided for embodiments of this application. Detailed Implementation

[0045] The technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, and not all embodiments. Based on the embodiments of this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.

[0046] Example 1:

[0047] The core of this embodiment lies in constructing a navigation decision-making framework that integrates information granulation, uncertainty assessment, and adaptive fusion. Its system architecture and workflow are as follows: Figure 1 As shown.

[0048] Step 1: Granular processing of multi-source navigation information;

[0049] Abstracting raw data from different sensors into information particles with a unified description Each information particle is a triple:

[0050] (1)

[0051] in, It is the core vector of the information granule, representing the most likely navigation state (such as position, attitude). It is the uncertainty vector of the information granules, which quantifies the credibility of the core vector. It is a mathematical representation function of information particles (such as intervals, fuzzy sets, probability distributions) used to describe the possibility space of states.

[0052] GNSS positioning beads (G-Granule):

[0053] (2)

[0054] Among them, the core Three-dimensional coordinates provided for GNSS. Uncertainty. , , , , These are empirical coefficients; HDOP measures horizontal positioning accuracy, while VDOP measures vertical positioning accuracy. The signal strength is used to describe the quality of a signal, thus... Indicates uncertainty in the horizontal direction. Characterizes uncertainty in the vertical direction. Function Defined as an interval vector:

[0055] (3)

[0056] This visually illustrates the possible error range of GNSS positioning.

[0057] GNSS positioning particles not only provide location coordinates but also a self-evaluation via HDOP / SNR and VDOP / SNR. This self-evaluation (uncertainty) is crucial, as it tells subsequent fusion algorithms: at this current moment, my credibility is as follows; please determine the weight given to this credibility in the final decision.

[0058] When GNSS signals are affected by the complex electromagnetic environment of a substation, the SNR will decrease significantly, leading to uncertainty in the calculation. and The weight of GNSS data increases dramatically. Upon receiving this information, the fusion algorithm automatically reduces the weight of GNSS data, relying more on sensors such as vision and RFID that are immune to electromagnetic interference, thus ensuring the overall reliability of the navigation system.

[0059] Visual localization granules (V-Granules):

[0060] (4)

[0061] core The relative position is calculated through visual feature matching. Uncertainty. It is a scalar representing the confidence level of visual positioning, calculated through a fuzzy inference system:

[0062] (5)

[0063] Fog (haze), Light (illuminance), and Features (number of feature points) are the input fuzzy variables. Through a pre-defined fuzzy rule base, such as IF Fog IS Heavy THEN Credibility IS Low, inference is performed, ultimately defuzzifying the data into a precise value between 0 and 1. Characterization function It can be regarded as a A fuzzy set centered on a certain element, whose membership degree is determined by... Decide.

[0064] Visual positioning particles not only provide location coordinates, but more importantly, they use an intelligent fuzzy inference system (FIS) to conduct a comprehensive self-evaluation of their reliability in the current environment (i.e., their credibility). This self-evaluation is like it's telling the fusion algorithm: "Please note that when I saw this location, the environmental conditions were like this (fog, low light, few feature points), so my confidence in this result is only..." That's a lot. Please take my current situation into full consideration when deciding on the final position.

[0065] RFID positioning granules (R-Granule):

[0066] (6)

[0067] core The known absolute coordinates of the RFID tag; uncertainty. It is the basic probability assignment in evidence theory (DS theory). It is not a single value, but a mapping: ,in It is a framework for identifying all possible locations. For example, This indicates that 80% of the evidence supports the use of robots in the region. Internal. Characterization function For the reason Define the trust function and the likelihood function.

[0068] Core position of RFID positioning particles It provides known, fixed absolute coordinates (i.e., the tag's installation location). However, its true wisdom lies in its self-evaluation. Unlike GNSS, which reports a numerical error, or vision, which provides an overall level of confidence, it acts as a rigorous witness, offering a testimony to the fusion algorithm with clearly defined boundaries of uncertainty. This testimony states: I detected the tag, but I cannot be 100% certain that the robot is at the tag's precise location. I have... I am certain that the robot is within area A, and at the same time, I acknowledge that... It is possible that it could be anywhere within the framework.

[0069] Step 2: Uncertainty fusion based on grouping function and fuzzy rough set;

[0070] Construct a grouping function and vague negation The fuzzy rough set model is used to calculate the fuzzy concept that the robot is on the correct path for each navigation particle. The lower and upper approximations.

[0071] Lower approximation The degree to which a robot is certain to be on the correct path is defined as follows:

[0072] (7)

[0073] ,variable In the domain Traversal of the middle, The universe of discourse comprising all navigation states, This is a logical AND operation, and the minimum value operation is performed.

[0074] Lower approximation The degree to which a robot is likely on the correct path is defined as follows:

[0075] (8)

[0076] in, It is a fuzzy relation matrix describing the dependency between sensors and navigation state. This is achieved by calculating each information granular pair. The upper and lower approximations can quantify their certainty and probability in decision-making.

[0077] Step 3: Optimization of dynamic sensors based on specific indicators;

[0078] Introducing specific indicators Dynamically evaluate the current sensor combination The importance of navigation decisions.

[0079] (9)

[0080] in, Representing the Navigation decisions (such as going straight, turning, or stopping). Based on the current sensor set The calculated upper approximation. Specificity index. The larger the value, the stronger the ability of the current sensor combination to distinguish navigation states, and the more important the combination is; where |U| represents the total number of states in the universe of discourse U (or W).

[0081] The GNFRS-Nav dynamic optimization algorithm is designed, and its process is as follows: Figure 2 As shown:

[0082] Initialize sensor set ;

[0083] Loop through all available sensors:

[0084] Calculate the new set after adding this sensor. Specificity ;

[0085] Select the sensor that maximizes the specificity gain and add it to the set. .

[0086] Stop iterating when specificity no longer improves significantly.

[0087] Output the current optimal sensor combination And use its fusion result as the final navigation basis;

[0088] For example, when a robot is inspecting a main transformer area, the GNSS signal is subject to strong electromagnetic interference, resulting in uncertainty in its G-particles. A sharp increase. The visual system, due to slight condensation, affects the reliability of V-particles. It drops to 0.7. At this point, the GNFRS-Nav algorithm calculates that, in the current state, the set... It has the highest specificity. The fusion center mainly makes navigation decisions based on V-particles and R-particles, successfully avoiding GNSS failure areas and maintaining the continuity of inspections.

[0089] It is understood that, in terms of uncertainty handling, this application introduces information granules to explicitly model and quantify the uncertainty of each navigation source; while previous technologies have insufficient uncertainty handling or typically assume it to be Gaussian noise. This makes the present invention more tolerant to anomalies and disturbances, and the navigation more robust.

[0090] Secondly, regarding the fusion strategy, this application employs an adaptive fusion and sensor optimization strategy based on fuzzy rough sets and specific indicators; while previous technologies mostly used fixed weights or simple switching methods. Therefore, this invention can intelligently rely on more reliable navigation sources, resulting in more rational decision-making.

[0091] Third, in terms of environmental adaptability, this application utilizes dynamic sensor optimization to adapt in real time to complex operating conditions such as freezing rain, fog, and electromagnetic interference; while previous technologies relied on pre-programmed logic, resulting in poor flexibility. This makes the present invention highly adaptable to the unique climate and electromagnetic environment of Guizhou region.

[0092] Fourth, regarding system robustness, this application, thanks to its multi-source complementary characteristics, ensures reliable navigation even if some sensors fail temporarily; whereas previous technologies suffered from single-point failures that could lead to system paralysis. Therefore, this application significantly improves the all-weather, all-terrain completion rate of inspection tasks.

[0093] Fifth, regarding the granularity of information representation and fusion, as well as the intelligence of decision-making logic, this application uses a unified information granular structure to perform fusion at the information granular level, preserving the original uncertainty and simulating the human thought process of comprehensive judgment and reasonable doubt based on fuzzy reasoning and evidence theory. Previous technologies, however, mostly performed fusion at the level of raw data or point estimation, resulting in significant information loss, and their decisions were based on deterministic thresholds or simple logical judgments. This makes the fusion process of this invention more consistent with human cognition, provides a richer decision-making basis, and yields more reliable results. Simultaneously, the system behavior is more intelligent and smoother, avoiding the decision jumps or oscillations at threshold boundaries inherent in traditional methods.

[0094] Example 2:

[0095] This invention provides a highly reliable intelligent navigation system based on granular computation and uncertainty analysis. This system can be used to implement the aforementioned highly reliable intelligent navigation method based on granular computation and uncertainty analysis, specifically including:

[0096] The data acquisition and granulation module is used to acquire multi-source navigation sensor data and abstract the multi-source navigation sensor data into information granules that contain navigation state information and its uncertainty description.

[0097] The uncertainty reasoning and fusion module is used to construct a fuzzy reasoning model based on the uncertainty description contained in the information granules to characterize the degree to which the navigation state meets the target navigation conditions, and to perform uncertainty fusion on the multi-source navigation sensor data to obtain the deterministic and probability assessment results of the navigation state.

[0098] The sensor set dynamic optimization module is used to dynamically evaluate and select the sensor set participating in navigation decision-making based on the deterministic and probabilities assessment results, and generate a sensor combination that is adapted to the current environmental state.

[0099] The decision fusion and navigation output module is used to output navigation decision results based on the information granule fusion results corresponding to the sensor combination.

[0100] Example 3:

[0101] This embodiment provides a terminal device, which includes a processor and a memory. The memory stores a computer program, which includes program instructions. The processor executes the program instructions stored in the computer storage medium. The processor may be a Central Processing Unit (CPU), or other general-purpose processors, digital signal processors (DSPs), application-specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. It is the computing and control core of the terminal, suitable for implementing one or more instructions, specifically suitable for loading and executing one or more instructions to achieve a corresponding method flow or function. The processor described in this embodiment can be used in the operation of a highly reliable intelligent navigation method based on granular computing and uncertainty analysis, including the following steps:

[0102] Acquire multi-source navigation sensor data and abstract the multi-source navigation sensor data into a unified information granule containing navigation state information and its uncertainty description;

[0103] Based on the uncertainty description contained in the information granules, a fuzzy inference model is constructed to characterize the degree to which the navigation state meets the target navigation conditions. Uncertainty fusion is performed on the multi-source navigation sensor data to obtain the deterministic and probability assessment results of the navigation state.

[0104] Based on the determination and probability assessment results, the sensor set participating in navigation decision-making is dynamically evaluated and selected to generate a sensor combination adapted to the current environmental state.

[0105] Based on the information granule fusion results corresponding to the sensor combination, the navigation decision result is output.

[0106] Example 4:

[0107] This embodiment provides a storage medium, specifically a computer-readable storage medium (Memory), which is a memory device in a terminal device used to store programs and data. It is understood that the computer-readable storage medium here can include both the built-in storage medium in the terminal device and extended storage media supported by the terminal device. The computer-readable storage medium provides storage space that stores the terminal's operating system. Furthermore, this storage space also stores one or more instructions suitable for loading and execution by a processor. These instructions can be one or more computer programs (including program code). It should be noted that the computer-readable storage medium here can be high-speed RAM or non-volatile memory, such as at least one disk storage device.

[0108] One or more instructions stored in a computer-readable storage medium can be loaded and executed by a processor to implement the corresponding steps of the high-reliability intelligent navigation method based on granular computing and uncertainty analysis in the above embodiments; one or more instructions in the computer-readable storage medium are loaded and executed by the processor in the following steps:

[0109] Acquire multi-source navigation sensor data and abstract the multi-source navigation sensor data into a unified information granule containing navigation state information and its uncertainty description;

[0110] Based on the uncertainty description contained in the information granules, a fuzzy inference model is constructed to characterize the degree to which the navigation state meets the target navigation conditions. Uncertainty fusion is performed on the multi-source navigation sensor data to obtain the deterministic and probability assessment results of the navigation state.

[0111] Based on the determination and probability assessment results, the sensor set participating in navigation decision-making is dynamically evaluated and selected to generate a sensor combination adapted to the current environmental state.

[0112] Based on the information granule fusion results corresponding to the sensor combination, the navigation decision result is output.

[0113] Those skilled in the art will understand that embodiments of the present invention can be provided as methods, systems, or computer program products. Therefore, the present invention can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, the present invention can take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0114] This invention is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of the invention. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, generate instructions for implementing the flowchart illustrations and / or block diagrams. Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0115] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.

[0116] These computer program instructions may also be loaded onto a computer or other programmable data processing equipment to cause a series of operational steps to be performed on the computer or other programmable equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable equipment for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.

[0117] Although preferred embodiments of this application have been described, those skilled in the art, upon learning the basic inventive concept, can make other changes and modifications to these embodiments. Therefore, the appended claims are intended to be interpreted as including the preferred embodiments as well as all changes and modifications falling within the scope of this application.

[0118] Obviously, those skilled in the art can make various modifications and variations to this application without departing from the spirit and scope of this application. Therefore, if such modifications and variations fall within the scope of the claims of this application and their equivalents, this application also intends to include such modifications and variations.

Claims

1. A highly reliable intelligent navigation method based on granular computation and uncertainty analysis, characterized in that, The method includes: Acquire multi-source navigation sensor data and abstract the multi-source navigation sensor data into a unified information granule containing navigation state information and its uncertainty description; Based on the uncertainty description contained in the information granules, a fuzzy inference model is constructed to characterize the degree to which the navigation state meets the target navigation conditions. Uncertainty fusion is performed on the multi-source navigation sensor data to obtain the deterministic and probability assessment results of the navigation state. Based on the determination and probability assessment results, the sensor set participating in navigation decision-making is dynamically evaluated and selected to generate a sensor combination adapted to the current environmental state. Based on the information granule fusion results corresponding to the sensor combination, the navigation decision result is output.

2. The highly reliable intelligent navigation method based on granular computation and uncertainty analysis according to claim 1, characterized in that, The information granules include: core information for characterizing the most likely navigation state, uncertainty information for characterizing the reliability of the core information, and a mathematical representation for describing the possible range of values ​​for the navigation state; wherein, the uncertainty information is used to reflect the reliability of the corresponding navigation sensor under the current environmental conditions and serves as a weighting basis in the subsequent fusion and decision-making process.

3. The highly reliable intelligent navigation method based on granular computation and uncertainty analysis according to claim 1, characterized in that, The uncertainty fusion process is based on a fuzzy relation model, which represents whether the navigation state meets the target navigation conditions as a fuzzy concept, and calculates the deterministic evaluation result and the probability evaluation result corresponding to the fuzzy concept; wherein, the deterministic evaluation result is characterized by the lower approximation of the fuzzy rough set, and the probability evaluation result is characterized by the upper approximation of the fuzzy rough set.

4. The highly reliable intelligent navigation method based on granular computation and uncertainty analysis according to claim 1, characterized in that, The process of dynamically evaluating and selecting sensor sets is based on a specificity index used to measure the ability of different sensor combinations to distinguish navigation decisions; the specificity index is used to reflect the degree of distinction of the current sensor combination over different navigation decision results, and the larger the specificity index, the stronger the ability of the sensor combination to distinguish navigation states. Sensors that improve the specificity index are gradually added to the current sensor set through an iterative process, and the iteration is terminated when the specificity index no longer improves significantly.

5. The highly reliable intelligent navigation method based on granular computation and uncertainty analysis according to claim 1, characterized in that, The multi-source navigation sensor includes: a satellite positioning sensor, a visual sensor, and a radio frequency identification sensor; The uncertainty information of the information particles constructed by the satellite positioning sensor is characterized based on positioning accuracy parameters and signal quality parameters; The uncertainty information of the information particles constructed by the aforementioned visual sensor is obtained by fuzzy reasoning, which integrates environmental factors and visual feature information. The uncertainty information of the information granules constructed by the radio frequency identification sensor is characterized using evidence theory to describe the credibility of the navigation state in different regions.

6. The highly reliable intelligent navigation method based on granular computation and uncertainty analysis according to claim 2, characterized in that, The uncertainty information is obtained autonomously by the corresponding navigation sensor based on its own working status and environmental conditions.

7. A highly reliable intelligent navigation system based on granular computing and uncertainty analysis, characterized in that, The system is used to perform the method according to any one of claims 1-6, the system comprising: The data acquisition and granulation module is used to acquire multi-source navigation sensor data and abstract the multi-source navigation sensor data into information granules that contain navigation state information and its uncertainty description. The uncertainty reasoning and fusion module is used to construct a fuzzy reasoning model based on the uncertainty description contained in the information granules to characterize the degree to which the navigation state meets the target navigation conditions, and to perform uncertainty fusion on the multi-source navigation sensor data to obtain the deterministic and probability assessment results of the navigation state. The sensor set dynamic optimization module is used to dynamically evaluate and select the sensor set participating in navigation decision-making based on the deterministic and probabilities assessment results, and generate a sensor combination that is adapted to the current environmental state. The decision fusion and navigation output module is used to output navigation decision results based on the information granule fusion results corresponding to the sensor combination.

8. A computer device, characterized in that, The system includes a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the program, it implements a highly reliable intelligent navigation method based on granular computing and uncertainty analysis as described in any one of claims 1 to 6.

9. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores a computer program that, when executed by a processor, implements a highly reliable intelligent navigation method based on granular computing and uncertainty analysis as described in any one of claims 1 to 6.

Images