Indoor and outdoor integrated positioning and navigation method, device, equipment and storage medium

Abstract

The invention relates to an indoor and outdoor integrated positioning and navigation method, device and equipment and a storage medium. The method comprises the following steps: dynamically performing indoor and outdoor scene recognition through multiple sensors, dynamically adjusting the weight according to the real-time credibility of each sensor and a satellite signal, and fusing and determining an outdoor scene, an indoor scene or an indoor and outdoor junction scene where a user is currently located; a corresponding fusion positioning mode is activated in a self-adaptive mode according to the recognized scene category, and a federated Kalman filtering model with an INS error equation as a state variable and RTK and UWB positioning information as observed quantities is constructed; and calculating a space precision factor of the RTK and a horizontal precision factor of the UWB in real time to serve as dynamic indexes for measuring the current positioning precision of each sub-positioning system, dynamically calculating an information distribution coefficient of each sub-filter according to the dynamic indexes, fusing local state estimation of each sub-filter, obtaining a global optimal state estimation value, and feeding the global optimal state estimation value back to the INS to correct an accumulated error.

Description

Technical Field

[0001] This application relates to the field of satellite positioning technology, and in particular to a positioning and navigation method, apparatus, device and storage medium that integrates indoor and outdoor positioning. Background Technology

[0002] With the accelerating construction of modern smart cities and the rapid development of IoT technology, people's demand for positioning and navigation services has expanded from single outdoor environments to all indoor and outdoor scenarios. In outdoor scenarios, the Global Navigation Satellite System (GNSS) can provide all-weather, global location services and has been widely used in vehicle navigation, personnel tracking, and other fields. However, in indoor or near-indoor scenarios such as underground parking lots, bridges and tunnels, and large shopping malls, satellite signals are easily blocked and interfered with, making GNSS positioning ineffective.

[0003] Currently, Chinese patent application CN110234072A proposes an indoor positioning method based on the fusion of Wi-Fi fingerprint and pedestrian trajectory estimation. This method first collects Wi-Fi signal strength data at various indoor reference points in an offline phase to construct a fingerprint map database. In the online positioning phase, the real-time collected Wi-Fi signal strength is matched with the fingerprint database to obtain a preliminary location estimate. Simultaneously, pedestrian trajectory estimation is performed using the inertial sensor built into the smartphone. A Kalman filter is then used to fuse the Wi-Fi positioning results with the PDR (pedestrian trajectory prediction) track to improve the continuity and stability of the positioning. This method alleviates, to some extent, the problem of pure Wi-Fi positioning being susceptible to signal fluctuations and also compensates for the long-term cumulative drift of PDR.

[0004] The aforementioned technologies have limited application scenarios and low accuracy in indoor / outdoor environments, signal-blocked environments, or special weather conditions, making it difficult to provide reliable preliminary information for fusion positioning algorithms. Secondly, the weight allocation of Wi-Fi positioning and PDR during fusion positioning typically uses fixed values, failing to fully consider the real-time positioning quality changes of different positioning systems in complex environments. When the signal quality of a certain subsystem deteriorates, the fixed-weight fusion method will introduce errors into the final positioning result.

[0005] Based on this, this application provides an indoor-outdoor integrated positioning and navigation method, apparatus, device, and storage medium. Summary of the Invention

[0006] To improve the simplicity of the scenario, the recognition accuracy is low at indoor and outdoor boundaries, in signal-blocked environments, or under special weather conditions, making it difficult to provide reliable preliminary information for the fusion positioning algorithm. Secondly, the weight allocation of Wi-Fi positioning and PDR in the fusion positioning process usually adopts a fixed value, which fails to fully consider the real-time positioning quality changes of different positioning systems in complex environments. When the signal quality of a certain subsystem deteriorates, the fixed weight fusion method will introduce errors into the final positioning result. This application provides an indoor and outdoor fusion positioning and navigation method, device, equipment, and storage medium.

[0007] Firstly, this application provides an indoor-outdoor integrated positioning and navigation method, which adopts the following technical solution: including: Real-time acquisition of data from optical sensors, temperature sensors, satellite signals, UWB ranging, and inertial measurement units; acquisition of multi-source data; and time alignment and coordinate system unification processing of the multi-source data. The system independently makes preliminary judgments about indoor and outdoor scenarios based on light sensor data, temperature sensor data, and satellite signal data. It dynamically adjusts the voting weight of each sensor in the scenario judgment according to the real-time reliability of each sensor data, and merges the preliminary judgment results of each sensor through a weighted voting mechanism to determine the scenario category of the user. The scenario categories include outdoor scenarios, indoor scenarios, and indoor-outdoor boundary scenarios. Based on the scene category, the corresponding fusion positioning mode is adaptively selected; if it is an outdoor scene, the RTK and INS tightly combined positioning mode is activated; if it is an indoor scene, the UWB and INS tightly combined positioning mode is activated; if it is an indoor-outdoor scene, the RTK, UWB and INS full-source fusion positioning mode is activated. Construct a system state model with the INS error equation as the state variable; construct corresponding system observation models according to the selected fusion positioning mode; calculate the spatial accuracy factor of RTK positioning and the horizontal accuracy factor of UWB positioning in real time, and use them as dynamic indicators to measure the current positioning accuracy of each sub-positioning system. Based on the spatial accuracy factor and the horizontal accuracy factor, the information allocation coefficients of each sub-filter in the federated Kalman filter are dynamically calculated. The information allocation coefficients are positively correlated with the positioning accuracy of the corresponding sub-positioning system. The information allocation coefficients are injected into the federated Kalman filter to adaptively fuse the local state estimates of each sub-filter to obtain the global optimal state estimate. The global optimal state estimate is used to correct the cumulative error of the INS in real time, and the corrected fused positioning result is output.

[0008] Preferably, the step of dynamically adjusting the voting weight of each sensor in scene determination based on the real-time reliability of each sensor data, and fusing the preliminary determination results of each sensor through a weighted voting mechanism to determine the scene category in which the user is currently located, includes: Day and night are distinguished based on sunrise and sunset times. During the day, a dynamic threshold method is used, with the threshold adjusted linearly based on the time period after sunrise and before sunset and the air humidity. If the light intensity is higher than the threshold, it is considered outdoor; otherwise, it is considered indoor. At night, a fixed threshold is used, with light intensity higher than the threshold considered indoor; otherwise, it is considered outdoor. Based on the current date, determine whether it is a generalized summer or a generalized winter. Construct short-term and long-term average temperatures, calculate the absolute value of the difference, and compare it with a preset threshold: In summer, a positive difference indicates a switch from indoor to outdoor, and a negative difference indicates the opposite; the decision logic is reversed in winter; if the difference is less than the threshold, retain the result of the previous moment. If the number of visible satellites is not less than the first satellite number threshold and the signal-to-noise ratio is not less than the first signal-to-noise ratio threshold, it is judged as outdoor; if the number of visible satellites is not greater than the second satellite number threshold and the signal-to-noise ratio is not greater than the second signal-to-noise ratio threshold, it is judged as indoor; if the number of visible satellites and the signal-to-noise ratio are in the middle, the signal-to-noise ratio is used as the main criterion for judgment. If the signal-to-noise ratio is greater than the second signal-to-noise ratio threshold, it is judged as outdoor, otherwise it is judged as indoor. Based on the real-time reliability of each sensor's data, its weight in the weighted voting mechanism is dynamically adjusted; the preliminary judgment result of each sensor is multiplied by the dynamic weight, and the total number of votes for indoor and outdoor scenarios is accumulated respectively. The scenario with the highest number of votes is the user's current location.

[0009] Preferably, the real-time calculation of the spatial accuracy factor of RTK positioning and the horizontal accuracy factor of UWB positioning, and the use of these as dynamic indicators to measure the current positioning accuracy of each sub-positioning system, includes: Based on the error equation of the inertial navigation system, a fifteen-dimensional system state variable is established, including the three-dimensional platform misalignment angle, three-dimensional velocity error, three-dimensional position error, three-dimensional gyroscope zero bias, and three-dimensional accelerometer zero bias. Based on the system state variable, a system state equation is constructed to describe the evolution of the system state over time. Based on the current activated fusion positioning mode, corresponding observation equations are constructed respectively: If the RTK and INS tightly combined positioning mode is activated, the observation model of the first sub-filter is constructed, and the difference between the position and velocity calculated by RTK and the position and velocity output by INS is used as the observation vector to establish the observation relationship between the two; If the UWB and INS tightly combined positioning mode is activated, the observation model of the second sub-filter is constructed, and the difference between the position calculated by UWB and the position output by INS is used as the observation vector to establish the observation relationship between the two; If the RTK, UWB and INS full-source fusion positioning mode is activated, the observation models of the first sub-filter and the second sub-filter are constructed simultaneously, so that the two sub-filters run in parallel. The system collects satellite geometric distribution information in the RTK positioning system in real time and calculates the spatial accuracy factor to quantify the current error amplification of RTK positioning. At the same time, the concept of accuracy factor is extended to UWB indoor positioning. The system collects the geometric distribution information of each base station in the UWB positioning system in real time and calculates the horizontal accuracy factor to quantify the current error amplification of UWB positioning. The spatial accuracy factor and the horizontal accuracy factor are used as dynamic indicators to measure the current positioning accuracy of each sub-positioning system.

[0010] Preferably, the calculation of the level precision factor includes: Real-time acquisition of satellite signal data currently involved in the calculation in the RTK positioning system, including the azimuth angle, elevation angle and pseudorange observation value of each satellite relative to the receiver; Based on the azimuth and elevation angles of each satellite, the direction cosine from the receiver position to each satellite is calculated, and the geometric observation matrix of the RTK positioning system is constructed. The number of rows in the geometric observation matrix corresponds to the number of currently visible satellites, and the number of columns corresponds to the three-dimensional spatial coordinates and receiver clock error parameters. Based on the geometric observation matrix, calculate its cofactor matrix; extract the diagonal elements of the corresponding three-dimensional spatial coordinates in the cofactor matrix, calculate the sum of squares and roots, and obtain the spatial precision factor at the current time. Real-time acquisition of data from each base station currently participating in the calculation in the UWB positioning system, including the azimuth angle, elevation angle, and ranging value of each base station relative to the positioning tag; The concept of accuracy factor is extended to indoor UWB positioning. Based on the azimuth and elevation angles of each base station, the direction cosine from the location of the positioning tag to each base station is calculated to construct the geometric observation matrix of the UWB positioning system. The number of rows in the geometric observation matrix corresponds to the number of currently available base stations, and the number of columns corresponds to the horizontal two-dimensional coordinates and ranging error parameters. Based on the geometric observation matrix of the UWB positioning system, calculate its cofactor matrix; extract the diagonal elements corresponding to the horizontal two-dimensional coordinates in the cofactor matrix, calculate the sum of squares and roots, and obtain the horizontal precision factor at the current time.

[0011] Preferably, the global optimal state estimation includes: The spatial accuracy factor and the horizontal accuracy factor are compared with the preset first accuracy threshold and the second accuracy threshold, respectively. Based on the preset judgment rules, it is determined whether the RTK positioning system and the UWB positioning system are currently available. Based on the availability status and accuracy factor value of each sub-positioning system, the information allocation coefficients of each sub-filter in the federated Kalman filter are dynamically determined. The information allocation coefficients are injected into the first sub-filter and the second sub-filter respectively. The system noise covariance matrix and state error covariance matrix of the sub-filter are redistributed according to the information allocation coefficients to complete the initialization of the sub-filter. Each sub-filter is controlled to independently perform time updates and observation updates based on the system state equation and the corresponding observation equation, thereby obtaining its own local state estimate and local state error covariance matrix. Based on the local state estimates and local state error covariance matrices of each sub-filter received by the main filter, and combined with the information allocation coefficients, the local estimates of each sub-filter are weighted and fused to obtain the global optimal state estimate and global state error covariance matrix. The global optimal state estimate is fed back to the INS to correct the INS's accumulated error. At the same time, the global optimal state estimate and the global state error covariance matrix are redistributed to each sub-filter according to the information allocation coefficients as the initial values ​​for the next filtering cycle.

[0012] Preferably, the step of dynamically determining the information allocation coefficients of each sub-filter in the federated Kalman filter based on the availability status and accuracy factor value of each sub-positioning system includes: When both the RTK positioning system and the UWB positioning system are determined to be available, a dynamic allocation strategy based on the relative size of the accuracy factor is adopted. When only the RTK positioning system is available and the UWB positioning system is unavailable, the information allocation coefficient of the RTK corresponding sub-filter is set to the maximum value, and the information allocation coefficient of the UWB corresponding sub-filter is set to zero. When only the UWB positioning system is available and the RTK positioning system is unavailable, the information allocation coefficient of the UWB corresponding sub-filter is set to the maximum value, and the information allocation coefficient of the RTK corresponding sub-filter is set to zero. When both the RTK and UWB positioning systems are unavailable, the information allocation coefficients of both sub-filters are set to zero, and independent positioning is performed solely by INS.

[0013] Preferably, when both the RTK positioning system and the UWB positioning system are determined to be available, a dynamic allocation strategy based on the relative magnitude of the accuracy factor is adopted, including: Read the spatial accuracy factor output by the RTK positioning system and the horizontal accuracy factor output by the UWB positioning system at the current moment, and ensure that both values ​​are valid positive numbers; The values ​​of the spatial precision factor and the horizontal precision factor are summed to obtain the sum of precision factors, which is used as the reference denominator for subsequent normalization processing. Based on the ratio of the spatial precision factor to the sum of the precision factors, the initial information allocation coefficient of the first sub-filter is calculated. The initial information allocation coefficient is negatively correlated with the magnitude of the spatial precision factor. The proportion of the spatial precision factor to the sum of the precision factors is calculated, and the initial information allocation coefficient of the first sub-filter is obtained by subtracting the proportion from the maximum value. Based on the ratio of the horizontal precision factor to the sum of the precision factors, the initial information allocation coefficient of the second sub-filter is calculated. The initial information allocation coefficient is negatively correlated with the magnitude of the horizontal precision factor. The proportion of the horizontal precision factor to the sum of the precision factors is calculated, and the corresponding proportion is subtracted from the maximum value to obtain the initial information allocation coefficient of the second sub-filter. The initial information allocation coefficients of the first sub-filter and the second sub-filter are summed and verified to ensure that the sum of the two is equal to the preset maximum allocation coefficient value; The initial information allocation coefficients of the verified first and second sub-filters are used as the final information allocation coefficients.

[0014] Secondly, this application discloses an indoor / outdoor integrated positioning and navigation device, which adopts the following technical solution, including: The data acquisition module is used to acquire real-time data from optical sensors, temperature sensors, satellite signals, UWB ranging, and inertial measurement units, obtain multi-source data, and perform time alignment and coordinate system unification processing on the multi-source data. The data discrimination module is used to independently make preliminary judgments on indoor and outdoor scenes based on light sensor data, temperature sensor data, and satellite signal data respectively. According to the real-time reliability of each sensor data, the voting weight of each sensor in scene judgment is dynamically adjusted, and the preliminary judgment results of each sensor are integrated through a weighted voting mechanism to determine the scene category where the user is currently located. The scene categories include outdoor scene, indoor scene, and indoor-outdoor boundary scene. The positioning selection module is used to adaptively select the corresponding fusion positioning mode according to the scene category; if it is an outdoor scene, the RTK and INS tightly combined positioning mode is activated; if it is an indoor scene, the UWB and INS tightly combined positioning mode is activated; if it is an indoor-outdoor scene, the RTK, UWB and INS full-source fusion positioning mode is activated. The positioning accuracy module is used to construct a system state model with the INS error equation as the state variable; construct corresponding system observation models according to the selected fusion positioning mode; calculate the spatial accuracy factor of RTK positioning and the horizontal accuracy factor of UWB positioning in real time, and use them as dynamic indicators to measure the current positioning accuracy of each sub-positioning system. The positioning correction module is used to dynamically calculate the information allocation coefficients of each sub-filter in the federated Kalman filter based on the spatial accuracy factor and the horizontal accuracy factor. The information allocation coefficients are positively correlated with the positioning accuracy of the corresponding sub-positioning system. The information allocation coefficients are injected into the federated Kalman filter to adaptively fuse the local state estimates of each sub-filter to obtain the global optimal state estimate. The global optimal state estimate is used to correct the cumulative error of the INS in real time, and the corrected fused positioning result is output.

[0015] Thirdly, this application also provides a control device, the device comprising: It includes a memory and a processor, wherein the memory stores a computer program that can be loaded by the processor and executed as described above for indoor-outdoor integrated positioning and navigation.

[0016] Fourthly, this application also provides a computer-readable storage medium storing a computer program that can be loaded by a processor and executed as described above for the indoor-outdoor integrated positioning and navigation method.

[0017] In summary, this application uses real-time acquisition of optical sensor, temperature sensor, and BeiDou satellite signal data to make independent preliminary scene judgments based on dynamic thresholds, temperature change trends, and satellite signal-to-noise ratio. Voting weights are dynamically adjusted based on the real-time reliability of each sensor's data, and a weighted voting mechanism is used to ultimately determine whether the user is currently in an outdoor, indoor, or indoor-outdoor scenario, providing reliable preliminary information for subsequent adaptive positioning. Based on this, an adaptive fusion positioning framework based on federated Kalman filtering is constructed: a 15-dimensional system state model is established based on the inertial navigation system error equation; RTK and INS tight combination, UWB and INS tight combination, or all three-source fusion positioning modes are dynamically activated based on scene recognition results, and corresponding observation models are constructed for each; simultaneously, the concept of accuracy factor is extended to indoor UWB positioning, and the spatial accuracy factor of RTK and the horizontal accuracy factor of UWB are calculated in real time, serving as dynamic indicators to measure the current positioning accuracy of each sub-positioning system. During the fusion process, the information allocation coefficients of each sub-filter are dynamically calculated based on the aforementioned accuracy factor, ensuring a positive correlation between the allocation coefficients and positioning accuracy. An adaptive fusion of the local state estimates of each sub-filter is then performed using a federated Kalman filter to obtain the globally optimal state estimate, which is then fed back to the inertial navigation system to correct accumulated errors. This dynamic weighted voting mechanism improves the accuracy and robustness of indoor and outdoor scene recognition. The adaptive weight allocation based on the accuracy factor overcomes the limitations of traditional fixed-weight fusion, maintaining sub-meter positioning accuracy at indoor and outdoor boundaries, in signal-obstructed areas, and in complex geometric layouts. This significantly enhances the environmental adaptability and continuous stability of the positioning system. Attached Figure Description

[0018] Figure 1 This is a flowchart illustrating a positioning and navigation method that integrates indoor and outdoor environments.

[0019] Figure 2 This is a structural block diagram of a positioning and navigation device that integrates indoor and outdoor environments. Detailed Implementation

[0020] The following combination Figure 1 - Figure 2 This application will be described in further detail.

[0021] The execution control entity of this application is a positioning and navigation system, which includes three parts: a positioning terminal, a cloud platform, and a web monitoring platform.

[0022] The positioning terminal includes: an RTK positioning module for receiving BeiDou satellite signals, calculating position and velocity information in outdoor scenarios, and outputting a spatial precision factor (PDOP) value; a UWB positioning module, comprising multiple positioning base stations deployed in indoor scenarios and a positioning tag, for calculating position information in indoor scenarios and outputting a horizontal precision factor (HDOP) value; an INS inertial navigation module, integrated into the RTK positioning module, for outputting continuous position, velocity, and attitude information; a light sensor module for real-time acquisition of ambient light intensity data; a temperature and humidity sensor module for real-time acquisition of ambient temperature and humidity data; an information processing module for fusing multi-source sensor data to calculate a high-precision fused positioning result; a 5G communication module for real-time uploading of the fused positioning result and environmental data collected by each sensor to a cloud platform; and a power supply module for providing stable power to all modules of the positioning terminal.

[0023] The cloud platform includes: a cloud database, using MySQL, for storing location data, environmental data, and user information data uploaded by the positioning terminals; and a cloud server for deploying backend services, receiving and processing data uploaded by the positioning terminals, and providing data access interfaces. The web monitoring platform includes: a backend service module, developed based on the Spring Boot framework and running on a cloud server, responsible for interacting with the cloud database and providing data query and processing interfaces; a frontend display module, developed based on the Vue framework, using a browser / server architecture to provide users with a visual interactive interface through a browser; an outdoor positioning display submodule, integrating the Baidu Maps API to display the user's location information and movement trajectory in outdoor scenes in real time; an indoor positioning display submodule, embedding an indoor floor plan to display the user's location information and movement trajectory in indoor scenes in real time; an environmental data display submodule to display environmental data collected by light sensors and temperature and humidity sensors and their change curves in real time in chart form; and an access control submodule, based on the Shiro security framework, to perform user authentication and access authorization.

[0024] Reference Figure 1 The embodiments of this application include at least steps S10 to S50.

[0025] S10 acquires real-time data from optical sensors, temperature sensors, satellite signals, UWB ranging, and inertial measurement units, obtaining multi-source data and performing time alignment and coordinate system unification processing on the multi-source data.

[0026] S20 independently makes preliminary judgments on indoor and outdoor scenarios based on light sensor data, temperature sensor data, and satellite signal data respectively; dynamically adjusts the voting weight of each sensor in the scenario judgment according to the real-time reliability of each sensor data, and merges the preliminary judgment results of each sensor through a weighted voting mechanism to determine the current scenario category of the user. The scenario categories include outdoor scenarios, indoor scenarios, and indoor-outdoor boundary scenarios.

[0027] S30 adaptively selects the corresponding fusion positioning mode according to the scene category; if it is an outdoor scene, it activates the RTK+INS tight combination positioning mode; if it is an indoor scene, it activates the UWB+INS tight combination positioning mode; if it is an indoor-outdoor scene, it activates the RTK+UWB+INS full-source fusion positioning mode.

[0028] S40. Construct a system state model with the INS error equation as the state variable; construct corresponding system observation models according to the selected fusion positioning mode; calculate the spatial accuracy factor of RTK positioning and the horizontal accuracy factor of UWB positioning in real time, and use them as dynamic indicators to measure the current positioning accuracy of each sub-positioning system.

[0029] S50 dynamically calculates the information allocation coefficients of each sub-filter in the federated Kalman filter based on the spatial accuracy factor and the horizontal accuracy factor. The information allocation coefficients are positively correlated with the positioning accuracy of the corresponding sub-positioning system. The information allocation coefficients are injected into the federated Kalman filter to adaptively fuse the local state estimates of each sub-filter to obtain the global optimal state estimate. The global optimal state estimate is used to correct the cumulative error of the INS in real time, and the corrected fused positioning result is output.

[0030] Specifically, through multi-sensor data acquisition and preprocessing, multi-source data from optical sensors, temperature sensors, BeiDou satellite signals, UWB ranging, and inertial measurement units are acquired and time-aligned and coordinate unified. Based on this, an indoor / outdoor scene recognition mechanism based on dynamic weighted voting is adopted. After each sensor independently makes an initial scene judgment, the voting weights are dynamically adjusted according to real-time reliability. Weighted voting fusion determines whether the user is currently in an outdoor, indoor, or indoor / outdoor boundary scene. Based on the identified scene category, the corresponding fusion positioning mode is adaptively activated: outdoor activation uses a tight combination of RTK and INS; indoor activation uses a tight combination of UWB and INS; and indoor / outdoor boundary activation uses a full-source fusion of all three. Subsequently, a system state model with the INS error equation as the state variable is constructed, and a corresponding observation model is built according to the selected mode. Simultaneously, the spatial accuracy factor of RTK and the horizontal accuracy factor of UWB are calculated in real time as dynamic indicators to measure the current positioning accuracy of each sub-positioning system. Based on the aforementioned accuracy factor, the information allocation coefficients of each sub-filter in the federated Kalman filter are dynamically calculated, ensuring a positive correlation between the allocation coefficients and positioning accuracy. Then, the local state estimates of each sub-filter are adaptively fused to obtain the globally optimal state estimate, which is then fed back to the INS to correct accumulated errors. This technical solution achieves seamless indoor / outdoor switching and high-precision continuous positioning through scene perception and accuracy factor-driven adaptive fusion.

[0031] In some embodiments, step S20 specifically includes the following steps: calculating the local sunrise and sunset times based on the current time, and combining the real-time collected light intensity values, using a dynamic threshold method to make a preliminary judgment on indoor and outdoor scenes: if the current time is daytime, the light intensity is compared with a dynamically adjusted light threshold; if the light intensity is higher than the threshold, it is initially judged as outdoor, otherwise as indoor; if the current time is nighttime, the light intensity is compared with a fixed nighttime threshold; if the light intensity is higher than the fixed nighttime threshold, it is initially judged as indoor, otherwise as outdoor; the dynamically adjusted light threshold is linearly adjusted based on the time period after sunrise and before sunset and the air humidity; determining whether it is a generalized summer or a generalized winter based on the current date, constructing a short-term temperature average and a long-term temperature average, calculating the absolute value of the difference between the two, and comparing it with a preset temperature change threshold: In generalized summer, if the temperature difference is greater than a threshold and the difference is positive, the initial decision is to switch from indoors to outdoors; if the difference is negative, the initial decision is to switch from outdoors to indoors. In generalized winter, the decision logic is the opposite of that in generalized summer. If the temperature difference is less than a threshold, the scene decision result of the previous moment is maintained. The initial decision of indoor / outdoor scene is made based on the real-time collected number of visible satellites and the signal-to-noise ratio (SNR): if the number of visible satellites is greater than or equal to the first satellite number threshold and the SNR is greater than or equal to the first SNR threshold, the initial decision is outdoor; if the number of visible satellites is less than or equal to the second satellite number threshold and the SNR is less than or equal to the second SNR threshold, the initial decision is indoors; if it is in an intermediate state, the SNR is used as the main decision criterion: if the SNR is greater than the second SNR threshold, the initial decision is outdoor; otherwise, it is indoors. Based on the real-time reliability of each sensor's data, its weight in the weighted voting mechanism is dynamically adjusted: For light sensors, when the light intensity exceeds a preset strong light threshold, its voting weight is increased to the first upper limit; when it is within half an hour after sunrise or half an hour before sunset, its weight is reset to zero; for satellite signals, when the number of visible satellites is greater than the first satellite number threshold and the SNR is greater than the first signal-to-noise ratio threshold, or when the number of visible satellites is less than the second satellite number threshold and the SNR is less than the second signal-to-noise ratio threshold, its voting weight is increased to the second upper limit; when the SNR is in the middle range, its weight is linearly adjusted according to the SNR value; the weight of the temperature sensor is dynamically determined by subtracting the weights of the light sensor and satellite signals from the total weight; the preliminary judgment results of each sensor are combined with the dynamically adjusted voting weights to calculate the total number of votes for indoor scenes and the total number of votes for outdoor scenes respectively, and the scene category with the higher number of votes is the final determined scene where the user is currently located.

[0032] Specifically, the system uses a light sensor combined with sunrise and sunset times and air humidity to dynamically determine thresholds, a temperature sensor to detect temperature change trends based on a generalized seasonal division, and BeiDou satellites to assess signal quality based on the number of visible satellites and signal-to-noise ratio. These three sensors independently complete the initial scene judgment. Subsequently, the voting weights are dynamically adjusted based on the real-time reliability of the data from each sensor: the weight of the light sensor is increased when the light intensity is extremely high, and it is set to zero at the sunrise and sunset boundary; its weight is increased when the satellite signal is extremely good or poor, and it is adjusted linearly in intermediate states; the weight of the temperature sensor is dynamically determined by the remaining weights. Finally, the judgment results of each sensor are multiplied by the dynamic weights, and the total number of votes for indoor and outdoor scenes is accumulated, with the scene with the highest number of votes being the final scene category. This mechanism overcomes the recognition blind spots of a single sensor during day-night transitions and signal transition zones through dynamic weight adjustment, significantly improving the accuracy and robustness of scene recognition at indoor and outdoor boundaries.

[0033] In some embodiments, step S40 specifically includes the following steps: based on the error equation of the inertial navigation system, establish fifteen-dimensional system state variables including the three-dimensional platform misalignment angle, three-dimensional velocity error, three-dimensional position error, three-dimensional gyroscope zero bias, and three-dimensional accelerometer zero bias; construct a system state equation based on the system state variables to describe the evolution of the system state over time; the formula of the system state equation is as follows: ; in, The system state vector includes the three-dimensional platform misalignment angle, three-dimensional velocity error, three-dimensional position error, three-dimensional gyroscope zero bias, and three-dimensional accelerometer zero bias. Let be the state transition matrix, which describes the linear transformation relationship of the state from time t to the next time step; This is the noise transition matrix, which maps the system noise to the state space; The system excitation noise sequence is usually assumed to be zero-mean Gaussian white noise, and its covariance matrix reflects the uncertainty of the system model.

[0034] Based on the current activated fusion positioning mode, corresponding observation equations are constructed respectively: If the RTK and INS tightly combined positioning mode is activated, the observation model of the first sub-filter is constructed, and the difference between the position and velocity calculated by RTK and the position and velocity output by INS is used as the observation vector to establish the observation relationship between the two; If the UWB and INS tightly combined positioning mode is activated, the observation model of the second sub-filter is constructed, and the difference between the position calculated by UWB and the position output by INS is used as the observation vector to establish the observation relationship between the two; If the RTK, UWB and INS full-source fusion positioning mode is activated, the observation models of the first sub-filter and the second sub-filter are constructed simultaneously, so that the two sub-filters run in parallel. The system collects satellite geometric distribution information in the RTK positioning system in real time and calculates the spatial accuracy factor to quantify the current error amplification of RTK positioning. At the same time, the concept of accuracy factor is extended to UWB indoor positioning. The system collects the geometric distribution information of each base station in the UWB positioning system in real time and calculates the horizontal accuracy factor to quantify the current error amplification of UWB positioning. The spatial accuracy factor and the horizontal accuracy factor are used as dynamic indicators to measure the current positioning accuracy of each sub-positioning system.

[0035] Specifically, firstly, based on the INS error equation, a 15-dimensional system state model is established, incorporating misalignment angle, velocity error, position error, and inertial device zero bias, to describe the evolution of positioning errors. Secondly, according to the currently active fusion positioning mode, a corresponding observation model is dynamically constructed: in RTK+INS mode, position and velocity differences are used as observation vectors; in UWB+INS mode, position difference is used as the observation vector; and in full-source fusion mode, two sub-filters run in parallel. Finally, the concept of accuracy factor in satellite positioning is extended to UWB positioning, and the spatial accuracy factor of RTK and the horizontal accuracy factor of UWB are calculated in real time as dynamic indicators to quantify the current error amplification of each sub-positioning system. This step, by establishing a unified state-space model and multi-source observation equations, and introducing the accuracy factor as a real-time reliability metric, lays the model foundation for subsequent dynamic adjustment of fusion weights based on signal quality.

[0036] Furthermore, in some embodiments, considering that real-time quantitative evaluation of RTK and UWB positioning accuracy is achieved through the construction and calculation of the geometric observation matrix, the corresponding processing steps are as follows: Real-time acquisition of satellite signal data currently involved in the calculation within the RTK positioning system, including the azimuth angle, elevation angle, and pseudorange observation values ​​of each satellite relative to the receiver; calculation of the direction cosine from the receiver position to each satellite based on the azimuth and elevation angles of each satellite, constructing the geometric observation matrix of the RTK positioning system, where the number of rows in the geometric observation matrix corresponds to the number of currently visible satellites, and the number of columns corresponds to the spatial three-dimensional coordinates and the receiver clock bias parameters; calculation of its cofactor matrix based on the geometric observation matrix; extraction of the diagonal elements corresponding to the spatial three-dimensional coordinates from the cofactor matrix, and calculation of... The spatial accuracy factor at the current moment is obtained by summing the squares and roots. Real-time data from each base station participating in the calculation within the UWB positioning system is collected, including the azimuth, elevation, and ranging values ​​of each base station relative to the positioning tag. The concept of accuracy factor is extended to indoor UWB positioning. Based on the azimuth and elevation angles of each base station, the direction cosine from the positioning tag location to each base station is calculated, constructing the geometric observation matrix of the UWB positioning system. The number of rows in the geometric observation matrix corresponds to the number of currently available base stations, and the number of columns corresponds to the horizontal two-dimensional coordinates and ranging error parameters. Based on the geometric observation matrix of the UWB positioning system, its cofactor matrix is ​​calculated. The diagonal elements corresponding to the horizontal two-dimensional coordinates in the cofactor matrix are extracted, and the sum of the squares and roots are calculated to obtain the horizontal accuracy factor at the current moment.

[0037] When calculating the accuracy factor, the geometric observation matrix is ​​first constructed based on the observation equations of the positioning system. For RTK positioning, the linearized observation equation for the i-th satellite can be expressed as: ; in, The deviation between the receiver's actual and approximate positions in the X direction (eastward) is expressed in meters, and is the parameter to be estimated. The deviation between the receiver's actual and approximate positions in the Y direction (north direction) is expressed in meters, and is the parameter to be estimated. This represents the deviation between the receiver's actual position and its approximate position in the Z direction (upward direction). For pseudo-distance residuals; Let be the direction cosine of the receiver pointing to the i-th satellite, derived from the azimuth angle. and pitch angle Calculated; Right now ; c represents the speed of light and corresponds to the receiver clock bias parameter. Combining the equations for all visible satellites yields the geometric observation matrix. The spatial precision factor PDOP is defined as follows: ; in, Cofactor matrix The j-th diagonal element corresponds to the variance factor of the three-dimensional spatial coordinates. For UWB positioning, a similar method is used, with the geometric observation matrix... The horizontal precision factor HDOP, composed of the direction cosines of each base station, is calculated as follows: ; in, , The variance factor corresponds to the horizontal two-dimensional coordinates. Using the above formula, the real-time spatial geometric distribution of satellites and base stations is quantified into specific accuracy indicators. This provides a numerical basis for dynamically adjusting the fusion weights based on the current reliability of each sub-positioning system, ensuring the effectiveness of the adaptive fusion algorithm.

[0038] Specifically, for RTK positioning systems, the azimuth, elevation, and pseudorange values ​​of each satellite are first collected. The direction cosine is then calculated to construct a geometric observation matrix, where the number of rows corresponds to the number of visible satellites, and the number of columns corresponds to the 3D coordinates and clock error parameters. Based on this matrix, a cofactor matrix is ​​calculated, and the diagonal elements corresponding to the 3D coordinates are extracted and their square roots are calculated to obtain the spatial accuracy factor for the current moment. For UWB positioning systems, the concept of accuracy factor is extended to indoor scenarios. The azimuth, elevation, and ranging values ​​of each base station are collected, and a geometric observation matrix is ​​constructed with horizontal 2D coordinates and ranging error as parameters. Similarly, the horizontal accuracy factor is calculated using the cofactor matrix. Thus, through mathematical modeling, the real-time spatial geometric distribution of satellites and base stations is quantified into specific accuracy indicators. This provides accurate numerical data for dynamically adjusting the fusion weights based on the current reliability of each sub-positioning system, ensuring the effectiveness of the adaptive fusion algorithm.

[0039] In some embodiments, step S50 specifically includes the following steps: comparing the spatial precision factor and the horizontal precision factor with a preset first precision threshold and a second precision threshold respectively; if the spatial precision factor is less than or equal to the first precision threshold, then the RTK positioning system is currently available; if the spatial precision factor is greater than the second precision threshold, then the RTK positioning system is currently unavailable; if the horizontal precision factor is less than or equal to the first precision threshold, then the UWB positioning system is currently available; if the horizontal precision factor is greater than the second precision threshold, then the UWB positioning system is currently unavailable; dynamically determining the information allocation coefficients of each sub-filter in the federated Kalman filter according to the availability status and precision factor values ​​of each sub-positioning system; injecting the information allocation coefficients into the first sub-filter and the second sub-filter respectively, and redistributing the system noise covariance matrix and state error covariance matrix of the sub-filter according to the information allocation coefficients to complete the initialization of the sub-filter; Each sub-filter independently performs time updates and observation updates based on the system state equation and the corresponding observation equation, obtaining its own local state estimate and local state error covariance matrix. Based on the local state estimates and local state error covariance matrices output by each sub-filter received by the main filter, and combined with the information allocation coefficient, the local estimates of each sub-filter are weighted and fused to obtain the global optimal state estimate and global state error covariance matrix. The global optimal state estimate is fed back to the INS to correct the accumulated error of the INS. At the same time, the global optimal state estimate and global state error covariance matrix are redistributed to each sub-filter according to the information allocation coefficient as the initial values ​​for the next filtering cycle.

[0040] Specifically, the real-time calculated spatial accuracy factor and horizontal accuracy factor are compared with preset thresholds to determine the current availability status of the RTK and UWB sub-positioning systems. Based on the availability status and accuracy factor values, the information allocation coefficients of each sub-filter are dynamically determined and injected into the sub-filters for initialization. Subsequently, each sub-filter independently performs time updates and observation updates based on the system state equation and the corresponding observation equation to obtain local state estimates. After receiving the local estimates, the main filter performs weighted fusion with the dynamically allocated coefficients to obtain the globally optimal state estimate. Finally, the global estimate is fed back to the INS to correct accumulated errors, and the sub-filters are reset with the same allocation coefficients as the initial values ​​for the next cycle. This mechanism isolates faulty subsystems through threshold judgment, achieves optimal fusion of multi-source information through dynamic weights, and ensures continuous and stable filtering through feedback reset, significantly improving the output accuracy of the positioning system in complex environments.

[0041] Furthermore, in some embodiments, considering the dynamic allocation of information coefficients based on the availability status of the sub-positioning system to achieve adaptive fusion, the corresponding processing steps are as follows: When both the RTK positioning system and the UWB positioning system are determined to be available, a dynamic allocation strategy based on the relative magnitude of the accuracy factor is adopted; when only the RTK positioning system is available and the UWB positioning system is unavailable, the information allocation coefficient of the corresponding sub-filter of RTK is set to the maximum value, and the information allocation coefficient of the corresponding sub-filter of UWB is set to zero; when only the UWB positioning system is available and the RTK positioning system is unavailable, the information allocation coefficient of the corresponding sub-filter of UWB is set to the maximum value, and the information allocation coefficient of the corresponding sub-filter of RTK is set to zero; when both the RTK positioning system and the UWB positioning system are unavailable, the information allocation coefficients of both sub-filters are set to zero, and independent positioning is performed solely by INS.

[0042] Specifically, based on the accuracy factor threshold judgment result, four cases are handled: when both systems are available, weights are dynamically allocated according to the relative magnitude of the accuracy factor, with higher accuracy resulting in larger allocation coefficients; when only RTK is available, its coefficient is set to the maximum value, and the UWB coefficient is set to zero; when only UWB is available, the opposite is true; when neither system is available, the coefficients of both sub-filters are set to zero, and positioning relies solely on INS. This mechanism achieves fault isolation through availability judgment, ensures the dominance of high-quality signal sources through dynamic weight allocation, avoids low-quality observation data from polluting the fusion results, significantly improves the system's robustness and adaptability in complex environments, and ensures that the output positioning accuracy is always dominated by the most reliable subsystem.

[0043] Furthermore, in some embodiments, considering the calculation of information allocation coefficients under dual-system availability, the corresponding processing steps are as follows: Read the spatial precision factor output by the RTK positioning system and the horizontal precision factor output by the UWB positioning system at the current moment, ensuring that both values ​​are valid positive numbers; accumulate the values ​​of the spatial precision factor and the horizontal precision factor to obtain the precision factor sum, which serves as the reference denominator for subsequent normalization processing; based on the ratio of the spatial precision factor to the precision factor sum, calculate the initial information allocation coefficient of the first sub-filter. The initial information allocation coefficient is negatively correlated with the magnitude of the spatial precision factor. Calculate the proportion of the spatial precision factor to the precision factor sum, using the maximum... Subtracting this ratio from the value yields the initial information allocation coefficients of the first sub-filter. Based on the ratio of the horizontal precision factor to the sum of precision factors, the initial information allocation coefficients of the second sub-filter are calculated. The initial information allocation coefficients are negatively correlated with the magnitude of the horizontal precision factor. The proportion of the horizontal precision factor to the sum of precision factors is calculated, and the corresponding proportion is subtracted from the maximum value to obtain the initial information allocation coefficients of the second sub-filter. The initial information allocation coefficients of the first and second sub-filters are summed and verified to ensure that the sum equals the preset maximum allocation coefficient value. The verified initial information allocation coefficients of the first and second sub-filters are used as the final information allocation coefficients.

[0044] The formula for calculating the information allocation coefficient is as follows: When both RTK and UWB are available, assuming the maximum allocation coefficient is 1 (i.e., the sum is 1), then the information allocation coefficient of the first sub-filter (RTK+INS) is... Information allocation coefficients of the second sub-filter (UWB+INS) They are respectively: ; in, This represents the spatial accuracy factor of the RTK positioning system. This represents the horizontal accuracy factor of the UWB positioning system. As the formula shows, when a certain accuracy factor is small, the corresponding allocation coefficient is large, meaning that subsystems with higher positioning accuracy receive higher fusion weights. The principle of information conservation is ensured through summation verification.

[0045] Specifically, the spatial precision factor of RTK and the horizontal precision factor of UWB are read and verified to be valid positive numbers. The sum of these two precision factors is used as a normalization benchmark. The proportion of each precision factor to the sum is calculated, and this proportion is subtracted from the preset maximum allocation coefficient to obtain the initial information allocation coefficient for the corresponding sub-filter. This ensures that the coefficients are negatively correlated with the magnitude of the precision factor; that is, the higher the precision, the smaller the factor value, and the larger the allocation coefficient. Finally, the two initial coefficients are summed and verified to ensure that their sum equals the maximum allocation coefficient value. After satisfying the information conservation principle, the verified coefficients are used as the final allocation coefficients. This mechanism, through normalization and negative correlation calculation, accurately maps the spatial geometric distribution quality to fusion weights, ensuring that when both systems are available, the subsystem with higher positioning accuracy receives a larger fusion weight, thereby achieving optimal weighted fusion of multi-source information.

[0046] The implementation principle of the indoor-outdoor fusion positioning and navigation method in this application is as follows: Real-time acquisition of data from optical sensors, temperature sensors, and BeiDou satellite signals is used. Independent scene preliminary judgments are made based on dynamic thresholds, temperature change trends, and satellite signal-to-noise ratio. Voting weights are dynamically adjusted according to the real-time reliability of each sensor's data. A weighted voting mechanism ultimately determines whether the user is currently in an outdoor scene, an indoor scene, or an indoor-outdoor boundary scene, providing reliable preliminary information for subsequent adaptive positioning. Based on this, an adaptive fusion positioning framework based on federated Kalman filtering is constructed: a 15-dimensional system state model is established based on the inertial navigation system error equation. The positioning modes of RTK and INS tight combination, UWB and INS tight combination, or all three are dynamically activated according to the scene recognition results, and corresponding observation models are constructed for each. Simultaneously, the concept of accuracy factor is extended to indoor UWB positioning, and the spatial accuracy factor of RTK and the horizontal accuracy factor of UWB are calculated in real time, serving as dynamic indicators to measure the current positioning accuracy of each sub-positioning system. During the fusion process, the information allocation coefficients of each sub-filter are dynamically calculated based on the aforementioned accuracy factor, ensuring a positive correlation between the allocation coefficients and positioning accuracy. An adaptive fusion of the local state estimates of each sub-filter is then performed using a federated Kalman filter to obtain the globally optimal state estimate, which is then fed back to the inertial navigation system to correct accumulated errors. This dynamic weighted voting mechanism improves the accuracy and robustness of indoor and outdoor scene recognition. The adaptive weight allocation based on the accuracy factor overcomes the limitations of traditional fixed-weight fusion, maintaining sub-meter positioning accuracy at indoor and outdoor boundaries, in signal-obstructed areas, and in complex geometric layouts. This significantly enhances the environmental adaptability and continuous stability of the positioning system.

[0047] Figure 1 This is a flowchart illustrating an indoor / outdoor integrated positioning and navigation method in one embodiment. It should be understood that, although... Figure 1The steps in the flowchart are shown sequentially as indicated by the arrows, but these steps are not necessarily executed in the order indicated by the arrows; unless explicitly stated otherwise, there is no strict order requirement for the execution of these steps, and they can be executed in other orders; and Figure 1 At least some of the steps in the process may include multiple sub-steps or multiple stages. These sub-steps or stages are not necessarily completed at the same time, but can be executed at different times. The execution order of these sub-steps or stages is not necessarily sequential, but can be executed in turn or alternately with other steps or at least some of the sub-steps or stages of other steps.

[0048] Based on the same technical concept, referring to Figure 2 This application also provides an indoor / outdoor integrated positioning and navigation device, which adopts the following technical solution: the device includes: The data acquisition module is used to acquire real-time data from optical sensors, temperature sensors, satellite signals, UWB ranging, and inertial measurement units, obtaining multi-source data and performing time alignment and coordinate system unification processing on the multi-source data. The data discrimination module is used to independently make preliminary judgments on indoor and outdoor scenes based on light sensor data, temperature sensor data, and satellite signal data respectively. According to the real-time reliability of each sensor data, the voting weight of each sensor in scene judgment is dynamically adjusted, and the preliminary judgment results of each sensor are integrated through a weighted voting mechanism to determine the scene category where the user is currently located. The scene categories include outdoor scene, indoor scene, and indoor-outdoor boundary scene. The positioning selection module is used to adaptively select the corresponding fusion positioning mode according to the scene category; if it is an outdoor scene, the RTK+INS tight combination positioning mode is activated; if it is an indoor scene, the UWB+INS tight combination positioning mode is activated; if it is an indoor-outdoor scene, the RTK+UWB+INS full-source fusion positioning mode is activated. The positioning accuracy module is used to construct a system state model with the INS error equation as the state variable; construct corresponding system observation models according to the selected fusion positioning mode; calculate the spatial accuracy factor of RTK positioning and the horizontal accuracy factor of UWB positioning in real time, and use them as dynamic indicators to measure the current positioning accuracy of each sub-positioning system. The positioning correction module is used to dynamically calculate the information allocation coefficients of each sub-filter in the federated Kalman filter based on the spatial accuracy factor and the horizontal accuracy factor. The information allocation coefficients are positively correlated with the positioning accuracy of the corresponding sub-positioning system. The information allocation coefficients are injected into the federated Kalman filter to adaptively fuse the local state estimates of each sub-filter to obtain the global optimal state estimate. The global optimal state estimate is used to correct the cumulative error of the INS in real time, and the corrected fused positioning result is output.

[0049] In some embodiments, the data discrimination module is specifically used to calculate the local sunrise and sunset times based on the current time, and, in conjunction with real-time collected light intensity values, to make a preliminary judgment on indoor and outdoor scenes using a dynamic threshold method: If the current time is daytime, the light intensity is compared with a dynamically adjusted light threshold. If the light intensity is higher than the threshold, it is initially determined to be outdoors; otherwise, it is indoors. If the current time is nighttime, the light intensity is compared with a fixed nighttime threshold. If the light intensity is higher than the fixed nighttime threshold, it is initially determined to be indoors; otherwise, it is outdoors. The dynamically adjusted light threshold is linearly adjusted based on the time period after sunrise and before sunset, as well as the air humidity. Determine whether it is a generalized summer or a generalized winter based on the current date, construct short-term and long-term temperature averages, calculate the absolute value of the difference between the two, and compare it with a preset temperature change threshold: In generalized summer, if the temperature difference is greater than the threshold and the difference is positive, the initial decision is to switch from indoors to outdoors; if the difference is negative, the initial decision is to switch from outdoors to indoors. In generalized winter, the decision logic is the opposite of that in generalized summer. If the temperature difference is less than the threshold, the scene decision result of the previous moment is maintained. The indoor / outdoor scene is initially determined based on the real-time data of the number of visible satellites and the signal-to-noise ratio (SNR): if the number of visible satellites is greater than or equal to the first satellite count threshold and the SNR is greater than or equal to the first SNR threshold, the scene is initially determined to be outdoors; if the number of visible satellites is less than or equal to the second satellite count threshold and the SNR is less than or equal to the second SNR threshold, the scene is initially determined to be indoors; if the scene is in an intermediate state, the SNR is used as the primary criterion: if the SNR is greater than the second SNR threshold, the scene is initially determined to be outdoors; otherwise, it is indoors. Based on the real-time reliability of each sensor's data, its weight in the weighted voting mechanism is dynamically adjusted: For light sensors, when the light intensity exceeds a preset strong light threshold, its voting weight is increased to the first upper limit; when it is within half an hour after sunrise or half an hour before sunset, its weight is reset to zero; for satellite signals, when the number of visible satellites is greater than the first satellite number threshold and the SNR is greater than the first signal-to-noise ratio threshold, or when the number of visible satellites is less than the second satellite number threshold and the SNR is less than the second signal-to-noise ratio threshold, its voting weight is increased to the second upper limit; when the SNR is in the middle range, its weight is linearly adjusted according to the SNR value; the weight of the temperature sensor is dynamically determined by subtracting the weights of the light sensor and satellite signals from the total weight. The preliminary judgment results of each sensor are combined with the dynamically adjusted voting weights to calculate the total number of votes for indoor scenes and the total number of votes for outdoor scenes respectively. The two are compared, and the scene category with the higher number of votes is the final determined scene where the user is currently located.

[0050] In some embodiments, the positioning accuracy module is specifically used to establish fifteen-dimensional system state variables based on the error equation of the inertial navigation system, including the three-dimensional platform misalignment angle, three-dimensional velocity error, three-dimensional position error, three-dimensional gyroscope zero bias, and three-dimensional accelerometer zero bias; and to construct system state equations based on the system state variables to describe the evolution of the system state over time. Based on the current activated fusion positioning mode, corresponding observation equations are constructed respectively: If the RTK and INS tightly combined positioning mode is activated, the observation model of the first sub-filter is constructed, and the difference between the position and velocity calculated by RTK and the position and velocity output by INS is used as the observation vector to establish the observation relationship between the two; If the UWB and INS tightly combined positioning mode is activated, the observation model of the second sub-filter is constructed, and the difference between the position calculated by UWB and the position output by INS is used as the observation vector to establish the observation relationship between the two; If the RTK, UWB and INS full-source fusion positioning mode is activated, the observation models of the first sub-filter and the second sub-filter are constructed simultaneously, so that the two sub-filters run in parallel. The system collects satellite geometric distribution information in the RTK positioning system in real time and calculates the spatial accuracy factor to quantify the current error amplification of RTK positioning. At the same time, the concept of accuracy factor is extended to UWB indoor positioning. The system collects the geometric distribution information of each base station in the UWB positioning system in real time and calculates the horizontal accuracy factor to quantify the current error amplification of UWB positioning. The spatial accuracy factor and the horizontal accuracy factor are used as dynamic indicators to measure the current positioning accuracy of each sub-positioning system.

[0051] In some embodiments, the positioning accuracy module is specifically used to collect satellite signal data currently being calculated in the RTK positioning system in real time, including the azimuth angle, elevation angle and pseudorange observation value of each satellite relative to the receiver; Based on the azimuth and elevation angles of each satellite, the direction cosine from the receiver position to each satellite is calculated, and the geometric observation matrix of the RTK positioning system is constructed. The number of rows in the geometric observation matrix corresponds to the number of currently visible satellites, and the number of columns corresponds to the three-dimensional spatial coordinates and receiver clock error parameters. Based on the geometric observation matrix, calculate its cofactor matrix; extract the diagonal elements of the corresponding three-dimensional spatial coordinates in the cofactor matrix, calculate the sum of squares and roots, and obtain the spatial precision factor at the current time. Real-time acquisition of data from each base station currently participating in the calculation in the UWB positioning system, including the azimuth angle, elevation angle, and ranging value of each base station relative to the positioning tag; The concept of accuracy factor is extended to indoor UWB positioning. Based on the azimuth and elevation angles of each base station, the direction cosine from the location of the positioning tag to each base station is calculated to construct the geometric observation matrix of the UWB positioning system. The number of rows in the geometric observation matrix corresponds to the number of currently available base stations, and the number of columns corresponds to the horizontal two-dimensional coordinates and ranging error parameters. Based on the geometric observation matrix of the UWB positioning system, calculate its cofactor matrix; extract the diagonal elements corresponding to the horizontal two-dimensional coordinates in the cofactor matrix, calculate the sum of squares and roots, and obtain the horizontal precision factor at the current time.

[0052] In some embodiments, the positioning correction module is specifically used to compare the spatial accuracy factor and the horizontal accuracy factor with a preset first accuracy threshold and a second accuracy threshold, respectively. If the spatial accuracy factor is less than or equal to the first accuracy threshold, the RTK positioning system is determined to be currently available; if the spatial accuracy factor is greater than the second accuracy threshold, the RTK positioning system is determined to be currently unavailable; if the horizontal accuracy factor is less than or equal to the first accuracy threshold, the UWB positioning system is determined to be currently available; if the horizontal accuracy factor is greater than the second accuracy threshold, the UWB positioning system is determined to be currently unavailable. Based on the availability status and accuracy factor value of each sub-positioning system, the information allocation coefficients of each sub-filter in the federated Kalman filter are dynamically determined. The information allocation coefficients are injected into the first sub-filter and the second sub-filter respectively. The system noise covariance matrix and state error covariance matrix of the sub-filter are redistributed according to the information allocation coefficients to complete the initialization of the sub-filter. Each sub-filter is controlled to independently perform time updates and observation updates based on the system state equation and the corresponding observation equation, thereby obtaining its own local state estimate and local state error covariance matrix. Based on the local state estimates and local state error covariance matrix of each sub-filter received by the main filter, and combined with the information allocation coefficient, the local estimates of each sub-filter are weighted and fused to obtain the global optimal state estimate and global state error covariance matrix. The global optimal state estimate is fed back to the INS to correct the INS's accumulated error. At the same time, the global optimal state estimate and the global state error covariance matrix are redistributed to each sub-filter according to the information allocation coefficients, serving as the initial values ​​for the next filtering cycle.

[0053] In some embodiments, the positioning correction module is specifically used to employ a dynamic allocation strategy based on the relative magnitude of the accuracy factor when both the RTK positioning system and the UWB positioning system are determined to be available. When only the RTK positioning system is available and the UWB positioning system is unavailable, the information allocation coefficient of the RTK corresponding sub-filter is set to the maximum value, and the information allocation coefficient of the UWB corresponding sub-filter is set to zero. When only the UWB positioning system is available and the RTK positioning system is unavailable, the information allocation coefficient of the corresponding UWB sub-filter is set to the maximum value, and the information allocation coefficient of the corresponding RTK sub-filter is set to zero. When both the RTK and UWB positioning systems are unavailable, the information allocation coefficients of both sub-filters are set to zero, and independent positioning is then achieved solely by relying on the INS.

[0054] In some embodiments, the positioning correction module is specifically used to read the spatial accuracy factor output by the RTK positioning system and the horizontal accuracy factor output by the UWB positioning system at the current moment, ensuring that both values ​​are valid positive numbers. The values ​​of the spatial precision factor and the horizontal precision factor are summed to obtain the sum of precision factors, which is used as the reference denominator for subsequent normalization processing. Based on the ratio of the spatial precision factor to the sum of precision factors, the initial information allocation coefficient of the first sub-filter is calculated. The initial information allocation coefficient is negatively correlated with the magnitude of the spatial precision factor. The proportion of the spatial precision factor to the sum of precision factors is calculated, and the initial information allocation coefficient of the first sub-filter is obtained by subtracting the proportion from the maximum value. Based on the ratio of the horizontal precision factor to the sum of precision factors, the initial information allocation coefficient of the second sub-filter is calculated. The initial information allocation coefficient is negatively correlated with the magnitude of the horizontal precision factor. The proportion of the horizontal precision factor to the sum of precision factors is calculated, and the corresponding proportion is subtracted from the maximum value to obtain the initial information allocation coefficient of the second sub-filter. The initial information allocation coefficients of the first sub-filter and the second sub-filter are summed and verified to ensure that the sum of the two is equal to the preset maximum allocation coefficient value; The initial information allocation coefficients of the verified first and second sub-filters are used as the final information allocation coefficients.

[0055] This application also discloses a control device.

[0056] Specifically, the control device includes a memory and a processor, the memory storing a computer program that can be loaded by the processor and executed to perform the aforementioned indoor-outdoor integrated positioning and navigation method.

[0057] This application also discloses a computer-readable storage medium.

[0058] Specifically, the computer-readable storage medium stores a computer program that can be loaded by a processor and executed as described above for indoor-outdoor integrated positioning and navigation. The computer-readable storage medium includes, for example, various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0059] The above are all preferred embodiments of this application and are not intended to limit the scope of protection of this application. Therefore, all equivalent changes made in accordance with the structure, shape and principle of this application should be covered within the scope of protection of this application.

Claims

1. A positioning and navigation method integrating indoor and outdoor environments, characterized in that, include: Real-time acquisition of data from optical sensors, temperature sensors, satellite signals, UWB ranging, and inertial measurement units; acquisition of multi-source data; and time alignment and coordinate system unification processing of the multi-source data. The system independently makes preliminary judgments about indoor and outdoor scenarios based on light sensor data, temperature sensor data, and satellite signal data. It dynamically adjusts the voting weight of each sensor in the scenario judgment according to the real-time reliability of each sensor data, and merges the preliminary judgment results of each sensor through a weighted voting mechanism to determine the scenario category of the user. The scenario categories include outdoor scenarios, indoor scenarios, and indoor-outdoor boundary scenarios. Based on the scene category, the corresponding fusion positioning mode is adaptively selected; If it is an outdoor scene, then activate the RTK and INS tightly coupled positioning mode; If it is an indoor scene, the UWB and INS tightly combined positioning mode is activated; if it is an indoor-outdoor scene, the RTK, UWB and INS full-source fusion positioning mode is activated. Construct a system state model with the INS error equation as the state variable; construct corresponding system observation models according to the selected fusion positioning mode; calculate the spatial accuracy factor of RTK positioning and the horizontal accuracy factor of UWB positioning in real time, and use them as dynamic indicators to measure the current positioning accuracy of each sub-positioning system. Based on the spatial accuracy factor and the horizontal accuracy factor, the information allocation coefficients of each sub-filter in the federated Kalman filter are dynamically calculated. The information allocation coefficients are positively correlated with the positioning accuracy of the corresponding sub-positioning system. The information allocation coefficients are injected into the federated Kalman filter to adaptively fuse the local state estimates of each sub-filter to obtain the global optimal state estimate. The global optimal state estimate is used to correct the cumulative error of the INS in real time, and the corrected fused positioning result is output.

2. The indoor-outdoor integrated positioning and navigation method according to claim 1, characterized in that, The process involves dynamically adjusting the voting weight of each sensor in scene determination based on the real-time reliability of the sensor data, and fusing the preliminary determination results of each sensor through a weighted voting mechanism to determine the current scene category of the user. This includes: Day and night are distinguished based on sunrise and sunset times. During the day, a dynamic threshold method is used, with the threshold adjusted linearly based on the time period after sunrise and before sunset and the air humidity. If the light intensity is higher than the threshold, it is considered outdoor; otherwise, it is considered indoor. At night, a fixed threshold is used, with light intensity higher than the threshold considered indoor; otherwise, it is considered outdoor. Based on the current date, determine whether it is a generalized summer or a generalized winter. Construct short-term and long-term average temperatures, calculate the absolute value of the difference, and compare it with a preset threshold: In summer, a positive difference indicates a switch from indoor to outdoor, and a negative difference indicates the opposite; the decision logic is reversed in winter; if the difference is less than the threshold, retain the result of the previous moment. If the number of visible satellites is not less than the first satellite number threshold and the signal-to-noise ratio is not less than the first signal-to-noise ratio threshold, it is judged as outdoor; if the number of visible satellites is not greater than the second satellite number threshold and the signal-to-noise ratio is not greater than the second signal-to-noise ratio threshold, it is judged as indoor; if the number of visible satellites and the signal-to-noise ratio are in the middle, the signal-to-noise ratio is used as the main criterion for judgment. If the signal-to-noise ratio is greater than the second signal-to-noise ratio threshold, it is judged as outdoor, otherwise it is judged as indoor. Based on the real-time reliability of each sensor's data, its weight in the weighted voting mechanism is dynamically adjusted; the preliminary judgment result of each sensor is multiplied by the dynamic weight, and the total number of votes for indoor and outdoor scenarios is accumulated respectively. The scenario with the highest number of votes is the user's current location.

3. The indoor-outdoor integrated positioning and navigation method according to claim 2, characterized in that, The real-time calculation of the spatial accuracy factor of RTK positioning and the horizontal accuracy factor of UWB positioning, and their use as dynamic indicators to measure the current positioning accuracy of each sub-positioning system, includes: Based on the error equation of the inertial navigation system, a fifteen-dimensional system state variable is established, including the three-dimensional platform misalignment angle, three-dimensional velocity error, three-dimensional position error, three-dimensional gyroscope zero bias, and three-dimensional accelerometer zero bias. Based on the system state variable, a system state equation is constructed to describe the evolution of the system state over time. Based on the current activated fusion positioning mode, corresponding observation equations are constructed respectively: If the RTK and INS tightly combined positioning mode is activated, the observation model of the first sub-filter is constructed, and the difference between the position and velocity calculated by RTK and the position and velocity output by INS is used as the observation vector to establish the observation relationship between the two; If the UWB and INS tightly combined positioning mode is activated, the observation model of the second sub-filter is constructed, and the difference between the position calculated by UWB and the position output by INS is used as the observation vector to establish the observation relationship between the two; If the RTK, UWB and INS full-source fusion positioning mode is activated, the observation models of the first sub-filter and the second sub-filter are constructed simultaneously, so that the two sub-filters run in parallel. The system collects satellite geometric distribution information in the RTK positioning system in real time and calculates the spatial accuracy factor to quantify the current error amplification of RTK positioning. At the same time, the concept of accuracy factor is extended to UWB indoor positioning. The system collects the geometric distribution information of each base station in the UWB positioning system in real time and calculates the horizontal accuracy factor to quantify the current error amplification of UWB positioning. The spatial accuracy factor and the horizontal accuracy factor are used as dynamic indicators to measure the current positioning accuracy of each sub-positioning system.

4. The indoor-outdoor integrated positioning and navigation method according to claim 3, characterized in that, The calculation level precision factor includes: Real-time acquisition of satellite signal data currently involved in the calculation in the RTK positioning system, including the azimuth angle, elevation angle and pseudorange observation value of each satellite relative to the receiver; Based on the azimuth and elevation angles of each satellite, the direction cosine from the receiver position to each satellite is calculated, and the geometric observation matrix of the RTK positioning system is constructed. The number of rows in the geometric observation matrix corresponds to the number of currently visible satellites, and the number of columns corresponds to the three-dimensional spatial coordinates and receiver clock error parameters. Based on the geometric observation matrix, calculate its cofactor matrix; extract the diagonal elements of the corresponding three-dimensional spatial coordinates in the cofactor matrix, calculate the sum of squares and roots, and obtain the spatial precision factor at the current time. Real-time acquisition of data from each base station currently participating in the calculation in the UWB positioning system, including the azimuth angle, elevation angle, and ranging value of each base station relative to the positioning tag; The concept of accuracy factor is extended to indoor UWB positioning. Based on the azimuth and elevation angles of each base station, the direction cosine from the location of the positioning tag to each base station is calculated to construct the geometric observation matrix of the UWB positioning system. The number of rows in the geometric observation matrix corresponds to the number of currently available base stations, and the number of columns corresponds to the horizontal two-dimensional coordinates and ranging error parameters. Based on the geometric observation matrix of the UWB positioning system, calculate its cofactor matrix; extract the diagonal elements corresponding to the horizontal two-dimensional coordinates in the cofactor matrix, calculate the sum of squares and roots, and obtain the horizontal precision factor at the current time.

5. The indoor-outdoor integrated positioning and navigation method according to claim 3, characterized in that, The global optimal state estimation includes: The spatial accuracy factor and the horizontal accuracy factor are compared with the preset first accuracy threshold and the second accuracy threshold, respectively. Based on the preset judgment rules, it is determined whether the RTK positioning system and the UWB positioning system are currently available. Based on the availability status and accuracy factor value of each sub-positioning system, the information allocation coefficients of each sub-filter in the federated Kalman filter are dynamically determined. The information allocation coefficients are injected into the first sub-filter and the second sub-filter respectively. The system noise covariance matrix and state error covariance matrix of the sub-filter are redistributed according to the information allocation coefficients to complete the initialization of the sub-filter. Each sub-filter is controlled to independently perform time updates and observation updates based on the system state equation and the corresponding observation equation, thereby obtaining its own local state estimate and local state error covariance matrix. Based on the local state estimates and local state error covariance matrices of each sub-filter received by the main filter, and combined with the information allocation coefficients, the local estimates of each sub-filter are weighted and fused to obtain the global optimal state estimate and global state error covariance matrix. The global optimal state estimate is fed back to the INS to correct the INS's accumulated error. At the same time, the global optimal state estimate and the global state error covariance matrix are redistributed to each sub-filter according to the information allocation coefficients as the initial values ​​for the next filtering cycle.

6. The indoor-outdoor integrated positioning and navigation method according to claim 5, characterized in that, The step of dynamically determining the information allocation coefficients of each sub-filter in the federated Kalman filter based on the availability status and accuracy factor value of each sub-positioning system includes: When both the RTK positioning system and the UWB positioning system are determined to be available, a dynamic allocation strategy based on the relative size of the accuracy factor is adopted. When only the RTK positioning system is available and the UWB positioning system is unavailable, the information allocation coefficient of the RTK corresponding sub-filter is set to the maximum value, and the information allocation coefficient of the UWB corresponding sub-filter is set to zero. When only the UWB positioning system is available and the RTK positioning system is unavailable, the information allocation coefficient of the UWB corresponding sub-filter is set to the maximum value, and the information allocation coefficient of the RTK corresponding sub-filter is set to zero. When both the RTK and UWB positioning systems are unavailable, the information allocation coefficients of both sub-filters are set to zero, and independent positioning is performed solely by INS.

7. The indoor-outdoor integrated positioning and navigation method according to claim 6, characterized in that, When both the RTK positioning system and the UWB positioning system are determined to be available, a dynamic allocation strategy based on the relative magnitude of the accuracy factor is adopted, including: Read the spatial accuracy factor output by the RTK positioning system and the horizontal accuracy factor output by the UWB positioning system at the current moment, and ensure that both values ​​are valid positive numbers; The values ​​of the spatial precision factor and the horizontal precision factor are summed to obtain the sum of precision factors, which is used as the reference denominator for subsequent normalization processing. Based on the ratio of the spatial precision factor to the sum of the precision factors, the initial information allocation coefficient of the first sub-filter is calculated. The initial information allocation coefficient is negatively correlated with the magnitude of the spatial precision factor. The proportion of the spatial precision factor to the sum of the precision factors is calculated, and the initial information allocation coefficient of the first sub-filter is obtained by subtracting the proportion from the maximum value. Based on the ratio of the horizontal precision factor to the sum of the precision factors, the initial information allocation coefficient of the second sub-filter is calculated. The initial information allocation coefficient is negatively correlated with the magnitude of the horizontal precision factor. The proportion of the horizontal precision factor to the sum of the precision factors is calculated, and the corresponding proportion is subtracted from the maximum value to obtain the initial information allocation coefficient of the second sub-filter. The initial information allocation coefficients of the first sub-filter and the second sub-filter are summed and verified to ensure that the sum of the two is equal to the preset maximum allocation coefficient value; The initial information allocation coefficients of the verified first and second sub-filters are used as the final information allocation coefficients.

8. A positioning and navigation device integrating indoor and outdoor environments, characterized in that, The device includes: The data acquisition module is used to acquire real-time data from optical sensors, temperature sensors, satellite signals, UWB ranging, and inertial measurement units, obtain multi-source data, and perform time alignment and coordinate system unification processing on the multi-source data. The data discrimination module is used to independently make preliminary judgments on indoor and outdoor scenes based on light sensor data, temperature sensor data, and satellite signal data respectively. According to the real-time reliability of each sensor data, the voting weight of each sensor in scene judgment is dynamically adjusted, and the preliminary judgment results of each sensor are integrated through a weighted voting mechanism to determine the scene category where the user is currently located. The scene categories include outdoor scene, indoor scene, and indoor-outdoor boundary scene. The positioning selection module is used to adaptively select the corresponding fusion positioning mode according to the scene category; if it is an outdoor scene, the RTK and INS tightly combined positioning mode is activated; if it is an indoor scene, the UWB and INS tightly combined positioning mode is activated; if it is an indoor-outdoor scene, the RTK, UWB and INS full-source fusion positioning mode is activated. The positioning accuracy module is used to construct a system state model with the INS error equation as the state variable; construct corresponding system observation models according to the selected fusion positioning mode; calculate the spatial accuracy factor of RTK positioning and the horizontal accuracy factor of UWB positioning in real time, and use them as dynamic indicators to measure the current positioning accuracy of each sub-positioning system. The positioning correction module is used to dynamically calculate the information allocation coefficients of each sub-filter in the federated Kalman filter based on the spatial accuracy factor and the horizontal accuracy factor. The information allocation coefficients are positively correlated with the positioning accuracy of the corresponding sub-positioning system. The information allocation coefficients are injected into the federated Kalman filter to adaptively fuse the local state estimates of each sub-filter to obtain the global optimal state estimate. The global optimal state estimate is used to correct the cumulative error of the INS in real time, and the corrected fused positioning result is output.

9. A control device, characterized in that, The device includes: A memory and a processor, wherein the memory stores a computer program that can be loaded by the processor and executed as described in any one of claims 1 to 7.

10. A computer-readable storage medium, characterized in that, The computer program is stored that can be loaded by a processor and executed as described in any one of claims 1 to 7.

Images