Indoor and outdoor positioning method and device based on weight fusion and storage medium

Abstract

The application relates to the field of fusion positioning, and discloses a weight fusion indoor and outdoor positioning method, equipment and a storage medium. The method comprises the following steps: according to a preset inertial navigation algorithm, performing state prediction processing on a state vector to obtain a state prediction vector and a prediction confidence matrix; according to a preset factor analysis algorithm, generating satellite credibility factors, super-bandwidth credibility factors and IMU credibility factors; according to the satellite credibility factors, obtaining a satellite noise correction matrix, and according to the super-bandwidth credibility factors, obtaining a super-bandwidth noise correction matrix; performing Kalman fusion update processing on the state prediction vector and the prediction confidence matrix to obtain a state update vector and an update confidence matrix; and performing quality labeling on the update positioning data to generate fusion positioning data with quality labeling. In the embodiment of the application, the update positioning data approximates to the theoretical optimal accuracy under the current environment, and the positioning trajectory is seamlessly connected without abrupt breakpoints when crossing different technical coverage areas.

Description

Technical Field

[0001] This invention relates to the field of fusion positioning, and more particularly to a weighted fusion method, device and storage medium for indoor and outdoor positioning. Background Technology

[0002] With the rapid development of IoT, smart manufacturing, and unmanned systems technologies, the demand for high-precision, high-reliability location services is becoming increasingly urgent. Currently, the industry mainly uses a combination of Global Navigation Satellite Systems (GNSS, such as GPS / BeiDou) and Ultra-Wideband (UWB) technology to achieve indoor and outdoor positioning. However, existing fusion solutions still have the following shortcomings: 1. Single source dependency cannot achieve seamless coverage. In open outdoor scenarios, GNSS alone can achieve meter-level positioning; indoors, it switches to UWB to achieve centimeter-level positioning. However, this method will encounter signal coverage blind spots or alternating strong and weak signals in the boundary areas of the two types of scenarios (such as doorways, windowsills, tree-lined roads, and under overpasses), resulting in positioning interruptions or drastic jumps, which cannot meet the needs of continuous navigation.

[0003] 2. Simple switching combinations result in poor stability in the transition zone. Some solutions employ a hard handover strategy of "choosing between two modes": GNSS is used when the GNSS signal is good, and UWB is switched when the signal is lost. Because GNSS signals are susceptible to obstruction and multipath interference in the transition zone, the system frequently switches between the two modes, causing positioning results to jitter and resulting in a poor user experience.

[0004] 3. Loosely coupled fusion results in information loss and easy error coupling. More advanced fusion methods typically employ a loosely coupled architecture, fusing the position and velocity calculated by GNSS with those calculated by UWB through a weighted average or Kalman filter. This post-processing approach has significant drawbacks. It discards raw sensor observations such as pseudorange, carrier phase, and UWB time difference of arrival, which could provide richer constraints. Differences in sampling time and coordinate systems between different sensors make precise alignment difficult with loose coupling, resulting in a weak fusion foundation. If a sensor outputs an incorrect position due to multipath effects or non-line-of-sight obstruction, loose coupling cannot effectively identify and remove outliers, leading to contaminated fusion results – a classic case of "garbage in, garbage out."

[0005] In summary, existing positioning technologies struggle to achieve continuous and stable positioning data output under seamless indoor-outdoor transitions, necessitating a new technology to address these current challenges. Summary of the Invention

[0006] The main objective of this invention is to solve the technical problem that existing positioning technologies are unable to achieve continuous and stable positioning data output under seamless indoor and outdoor conditions.

[0007] The first aspect of this invention provides a weighted fusion indoor / outdoor positioning method, the weighted fusion indoor / outdoor positioning method comprising: Read the state vector, the angular velocity of the preset IMU, and the force measurement value of the preset IMU, wherein the state vector includes: positioning data; Based on the preset inertial navigation algorithm, the angular velocity, and the specific force measurement value, the state vector is subjected to state prediction processing to obtain the state prediction vector and the prediction confidence matrix; Based on the preset factor analysis algorithm, confidence levels are calculated for preset satellite signals, preset ultra-wideband signals, and preset IMU signals to generate satellite confidence factors, ultra-wideband confidence factors, and IMU confidence factors. Read satellite observations and ultra-wideband observations; Based on the satellite confidence factor, the preset satellite observation noise matrix is ​​subjected to factor correction processing to obtain the satellite noise correction matrix, and based on the ultra-bandwidth confidence factor, the preset ultra-bandwidth observation noise matrix is ​​subjected to factor correction processing to obtain the ultra-bandwidth noise correction matrix. Based on the satellite noise correction matrix, the ultra-wideband noise correction matrix, the satellite observations, and the ultra-wideband observations, Kalman fusion update processing is performed on the state prediction vector and the prediction confidence matrix to obtain the state update vector and the update confidence matrix, wherein the state update vector includes: updated positioning data; Based on the IMU reliability factor, the updated positioning data is quality-labeled to generate quality-labeled fused positioning data.

[0008] Optionally, in a first implementation of the first aspect of the present invention, the step of performing state prediction processing on the state vector according to a preset inertial navigation algorithm, the angular velocity, and the specific force measurement value to obtain a state prediction vector and a prediction confidence matrix includes:

[0009] Among them, U k X is a vector composed of angular velocity and specific force measurements. k-1 Let P be the state vector, f() be the prediction function, and P be the prediction function. k-1 Let F be the initial state covariance matrix. k X is the state transition Jacobian matrix, Q is the process noise covariance matrix, and X is the process noise covariance matrix. k|k-1 P is the state prediction vector. k|k-1 This is the prediction confidence matrix, where k is the index value.

[0010] Optionally, in a second implementation of the first aspect of the present invention, the step of calculating the confidence levels of a preset satellite signal, a preset ultra-wideband signal, and a preset IMU signal according to a preset factor analysis algorithm to generate a satellite confidence factor, an ultra-wideband confidence factor, and an IMU confidence factor includes:

[0011] Among them, F gnss N is the satellite credibility factor. norm C is the satellite number normalization factor. norm G is the carrier-to-noise ratio normalization factor. norm N is the geometric precision inverse normalization factor, w1, w2, and w3 are the satellite factor weighting coefficients, and N is the weighting factor. sat N represents the number of visible satellites. thresh To effectively calculate the threshold, M represents the total number of satellites, and CNR... i Let CNR be the carrier-to-noise ratio of the i-th satellite. max CNR is the maximum effective empirical threshold. min GDOP is the minimum effective empirical threshold, and GDOP is the current geometric precision factor. thresh This is the geometric empirical threshold.

[0012] Optionally, in a third implementation of the first aspect of the present invention, the step of calculating the confidence levels of a preset satellite signal, a preset ultra-wideband signal, and a preset IMU signal according to a preset factor analysis algorithm to generate a satellite confidence factor, an ultra-wideband confidence factor, and an IMU confidence factor further includes:

[0013] Among them, F uwb As the ultra-bandwidth confidence factor, A norm S is the base station number normalization factor. norm R is the signal strength normalization factor. norm λ1, λ2, and λ3 are the normalization factors for the ranging residuals, and N is the weighting coefficient of the superbandwidth factor. anchor The number of visible base stations, RSSI j Let RSSI be the signal strength of the j-th base station. max RSSI is the maximum empirical threshold for the signal. min d is the minimum empirical threshold for the signal. j meas Let d be the measured distance of the j-th base station. j pred This is the distance predicted for the j-th base station at the previous time step.

[0014] Optionally, in a fourth implementation of the first aspect of the present invention, the step of calculating the confidence levels of a preset satellite signal, a preset ultra-wideband signal, and a preset IMU signal according to a preset factor analysis algorithm to generate a satellite confidence factor, an ultra-wideband confidence factor, and an IMU confidence factor further includes:

[0015] Among them, F imu Here, γ is the IMU reliability factor, t is the mixing coefficient, β is the attenuation coefficient, and V is the IMU reliability factor. norm Here, is the motion intensity normalization factor, 'a' is the accelerometer reading, and 'b' is the motion intensity normalization factor. a To estimate the accelerometer zero bias, a thresh This is a high dynamic threshold.

[0016] Optionally, in a fifth implementation of the first aspect of the present invention, the steps of performing factor correction processing on a preset satellite observation noise matrix according to the satellite confidence factor to obtain a satellite noise correction matrix, and performing factor correction processing on a preset ultra-wideband observation noise matrix according to the ultra-wideband confidence factor to obtain an ultra-wideband noise correction matrix include: R gnss，k =R gnss，base / F gnss ; R uwb，k =R uwb，base / F uwb ; Among them, R gnss，base R is the satellite observation noise matrix. gnss，k For the satellite noise correction matrix, F gnss R is the satellite credibility factor. uwb，base R is the ultra-bandwidth observation noise matrix. uwb，k To obtain the ultra-bandwidth noise correction matrix, F uwb This is the ultra-wideband reliability factor.

[0017] Optionally, in a sixth implementation of the first aspect of the present invention, the step of performing Kalman fusion update processing on the state prediction vector and the prediction confidence matrix based on the satellite noise correction matrix, the ultra-wideband noise correction matrix, the satellite observations, and the ultra-wideband observations to obtain the state update vector and the update confidence matrix includes:

[0018] Among them, K k For the Kalman update matrix, P k|k-1 To predict the confidence matrix, H k To observe the Jacobian matrix, R kX is a diagonal matrix composed of the ultra-bandwidth noise correction matrix and the satellite noise correction matrix. k|k-1 Let Z be the state prediction vector, and Z be the combined vector of satellite observations and ultra-wideband observations, h(X) k|k-1 X represents the prediction distance corresponding to the state prediction vector. k For the state update vector, P k To update the confidence matrix, I is the identity matrix.

[0019] Optionally, in the seventh implementation of the first aspect of the present invention, the step of performing quality labeling on the updated positioning data based on the IMU trust factor to generate quality-labeled fused positioning data includes: Determine whether the IMU trust factor is less than a preset anomaly threshold; When the value is less than a preset abnormal threshold, the updated positioning data is marked as low confidence, and fused positioning data with low confidence is generated. If the value is not less than a preset abnormal threshold, then it is determined whether the IMU trust factor is greater than a preset health threshold. When the value exceeds a preset health threshold, the updated location data is marked as high confidence, and fused location data with high confidence is generated.

[0020] A second aspect of the present invention provides a weighted fusion indoor and outdoor positioning device, comprising: a memory and at least one processor, wherein the memory stores instructions, and the memory and the at least one processor are interconnected via a line; the at least one processor invokes the instructions in the memory to cause the weighted fusion indoor and outdoor positioning device to perform the weighted fusion indoor and outdoor positioning method described above.

[0021] A third aspect of the present invention provides a computer-readable storage medium storing instructions that, when executed on a computer, cause the computer to perform the aforementioned weighted fusion indoor and outdoor positioning method.

[0022] In this embodiment of the invention, a preliminary state vector is predicted from the inertial data of the IMU to obtain a state prediction vector. Confidence calculations are performed on satellite signals and ultra-wideband signals. Based on the confidence level, the highest weight is dynamically assigned to the most reliable sensor while suppressing interference from unreliable sources. Multi-sensor positioning data fusion calculations are then performed to generate updated positioning data. Furthermore, the updated positioning data is quality-labeled through confidence analysis of the IMU's inertial data to generate fused positioning data. The fusion process is adjusted according to the smooth transition weights of the multi-source signal quality factors, achieving updated positioning data that approximates the theoretical optimal accuracy under the current environment. The positioning trajectory seamlessly connects without abrupt breaks when crossing different technology coverage areas, solving the technical problem of existing positioning technologies struggling to achieve continuous and stable positioning data output under seamless indoor-outdoor transitions. Attached Figure Description

[0023] Figure 1 This is a schematic diagram of an embodiment of the weighted fusion indoor and outdoor positioning method in this invention; Figure 2 This is a schematic diagram of a specific embodiment of the 107 steps of the weighted fusion indoor and outdoor positioning method in this invention. Figure 3 This is a schematic diagram of an embodiment of an indoor and outdoor positioning device with weighted fusion according to the present invention. Detailed Implementation

[0024] This invention provides a weighted fusion method, device, and storage medium for indoor and outdoor positioning.

[0025] The embodiments of the present invention will now be described in more detail with reference to the accompanying drawings. While some embodiments of the present invention are shown in the drawings, it should be understood that the present invention can be implemented in various forms and should not be construed as limited to the embodiments set forth herein. Rather, these embodiments are provided to provide a more thorough and complete understanding of the present disclosure. It should be understood that the accompanying drawings and embodiments are for illustrative purposes only and are not intended to limit the scope of protection of the present invention.

[0026] In the description of the embodiments disclosed in this invention, the term "comprising" and similar terms should be understood as open-ended inclusion, i.e., "including but not limited to". The term "based on" should be understood as "at least partially based on". The term "one embodiment" or "the embodiment" should be understood as "at least one embodiment". The terms "first", "second", etc., may refer to different or the same objects. Other explicit and implicit definitions may also be included below.

[0027] For ease of understanding, the specific process of the embodiments of the present invention is described below. Please refer to [link / reference]. Figure 1One embodiment of the weighted fusion indoor / outdoor positioning method in this invention includes: 101. Read the state vector, the angular velocity of the preset IMU, and the force measurement value of the preset IMU, wherein the state vector includes: positioning data; In this embodiment, the state vector X=[p,v,q,ba,bg] of the IMU sensor is read. T Where p is the positioning data, v is the measured velocity, q is the attitude (four data points), ba is the accelerometer zero bias, and bg is the gyroscope zero bias. The zero bias is estimated because the IMU has temperature drift and factory error. Without calibration, the data will become increasingly off-target.

[0028] Furthermore, the angular velocity and specific force measurement values ​​of the IMU are read simultaneously. The specific force measurement value = carrier acceleration - gravitational acceleration. The accelerometer cannot sense gravity and can only collect the supporting force or motion acceleration.

[0029] 102. Based on the preset inertial navigation algorithm, the angular velocity, and the specific force measurement value, perform state prediction processing on the state vector to obtain a state prediction vector and a prediction confidence matrix; In this embodiment, the angular velocity and specific force measurements are combined into U. k Measurement vector, where U k =[angular velocity, force measurement], U k The measurement vector and state vector are imported into a preset inertial navigation algorithm for state prediction processing to obtain the state prediction vector and the prediction confidence matrix indicating the prediction confidence of the state prediction vector.

[0030] Specifically, step 102 includes the following specific implementation methods:

[0031] Among them, U k X is a vector composed of angular velocity and specific force measurements. k-1 Let P be the state vector, f() be the prediction function, and P be the prediction function. k-1 Let F be the initial state covariance matrix. k X is the state transition Jacobian matrix, Q is the process noise covariance matrix, and X is the process noise covariance matrix. k|k-1 P is the state prediction vector. k|k-1 This is the prediction confidence matrix, where k is the index value.

[0032] It should be noted that the optimal estimate is generated within an adaptive extended Kalman filter framework, and the confidence factor is directly used to dynamically adjust the key parameters of the filter to achieve intelligent fusion. Q, the process noise covariance matrix, is usually set to a fixed value.

[0033] 103. Based on the preset factor analysis algorithm, perform confidence calculations on preset satellite signals, preset ultra-wideband signals, and preset IMU signals to generate satellite confidence factor, ultra-wideband confidence factor, and IMU confidence factor; In this embodiment, a fixedly deployed master clock (which may be a high-precision rubidium clock, a BeiDou time synchronization module, etc.) sends precise synchronization signals to slave clocks in all UWB base stations and mobile terminals via wired or wireless means. The clocks of all devices are synchronized to nanosecond to microsecond precision, providing a unified time scale for all subsequent measurements. The precise geographic coordinates of the UWB base stations, obtained through surveying or differential BeiDou, are entered into the system as a spatial reference.

[0034] The satellite module receives satellite signals, calculates its own raw observation data (such as pseudorange and carrier phase) and rough position, and adds a high-precision timestamp.

[0035] The UWB module communicates bidirectionally with multiple base stations within its field of view, measures the radio flight time, calculates the distance, and each distance measurement is accompanied by a precise timestamp.

[0036] The IMU sensor continuously outputs raw data of triaxial acceleration and angular velocity at a frequency of hundreds of hertz, and its sampling time is also accurately recorded.

[0037] All timestamped heterogeneous data is sent in real time to the central processing unit in the terminal via an internal bus (such as CAN, SPI) or local area network. The algorithm calculates and extracts key factors that reflect the reliability of each sensor in real time.

[0038] The algorithm interpolates or aligns all data to the same point in time and transforms all location information to a unified map coordinate system. Factor analysis algorithms calculate satellite reliability factors based on raw satellite signal acquisition data; factor analysis algorithms calculate satellite reliability factors based on raw ultra-wideband signal acquisition data; and factor analysis algorithms calculate IMU reliability factors based on raw IMU signal acquisition data.

[0039] Specifically, step 103 includes the following specific implementation methods:

[0040] Among them, F gnss N is the satellite credibility factor. norm C is the satellite number normalization factor. norm G is the carrier-to-noise ratio normalization factor. norm N is the geometric precision inverse normalization factor, w1, w2, and w3 are the satellite factor weighting coefficients, and N is the weighting factor. sat N represents the number of visible satellites. thresh To effectively calculate the threshold, M represents the total number of satellites, and CNR...i Let CNR be the carrier-to-noise ratio of the i-th satellite. max CNR is the maximum effective empirical threshold. min GDOP is the minimum effective empirical threshold, and GDOP is the current geometric precision factor. thresh This is the geometric empirical threshold.

[0041] It should be noted that N thresh For effective calculation, the threshold is usually set to 5. The more satellites there are, the higher the satellite reliability factor F becomes. gnss The higher the number of satellites and the worse the geometry, the lower the satellite reliability factor F. gnss The lower the credibility, the better. max CNR is the maximum effective empirical threshold. min This is the minimum effective empirical threshold, for example, set to 50dB-Hz and 35dB-Hz respectively. This value reflects the average signal strength. GDOP thresh The geometric empirical threshold can be set to 3. The smaller the GDOP, the better the geometric configuration. The geometric accuracy inverse normalization factor G... norm The higher.

[0042] Furthermore, step 103 also includes the following specific implementation methods:

[0043] Among them, F uwb As the ultra-bandwidth confidence factor, A norm S is the base station number normalization factor. norm R is the signal strength normalization factor. norm λ1, λ2, and λ3 are the normalization factors for the ranging residuals, and N is the weighting coefficient of the superbandwidth factor. anchor The number of visible base stations, RSSI j Let RSSI be the signal strength of the j-th base station. max RSSI is the maximum empirical threshold for the signal. min d is the minimum empirical threshold for the signal. j meas Let d be the measured distance of the j-th base station. j pred This is the distance predicted for the j-th base station at the previous time step.

[0044] It should be noted that N anchor To obtain a good 3D positioning result, at least four base stations are required, representing the number of visible base stations. norm R is a signal strength normalization factor that reflects the average signal quality. norm The normalization factor for the ranging residuals indicates the possible existence of non-line-of-sight errors; therefore, 1-R can be used. norm Indicates credibility.

[0045] Furthermore, step 103 also includes the following specific implementation methods:

[0046] Among them, F imu Here, γ is the IMU reliability factor, t is the mixing coefficient, β is the attenuation coefficient, and V is the IMU reliability factor. norm Here, is the motion intensity normalization factor, 'a' is the accelerometer reading, and 'b' is the motion intensity normalization factor. a To estimate the accelerometer zero bias, a thresh This is a high dynamic threshold.

[0047] It should be noted that the reliability of the IMU is mainly related to the intensity and duration of motion, and the calculation focuses more on short-term dynamics. exp(-β*t) is the time decay term, where t is the time since the last receipt of a high-reliability external signal (BeiDou / UWB), and β is the decay coefficient simulating the characteristics of inertial navigation error accumulation over time. exp(-β*t) decays from 1 to 0 over time. V norm The value of the motion intensity normalization factor decreases during violent vehicle maneuvers because the scaling factor and nonlinear error of the IMU increase at this time.

[0048] 104. Read satellite observations and ultra-wideband observations; In this embodiment, the observation equation is z = h(x) + v h Where z = [satellite observations, UWB observations], h(x) is the satellite pseudorange and UWB ranging, and v h The observation noise represents the error in sensor measurements.

[0049] 105. Based on the satellite confidence factor, perform factor correction processing on the preset satellite observation noise matrix to obtain the satellite noise correction matrix, and based on the ultra-bandwidth confidence factor, perform factor correction processing on the preset ultra-bandwidth observation noise matrix to obtain the ultra-bandwidth noise correction matrix. In this embodiment, the preset satellite observation noise matrix is ​​first modified based on the satellite confidence factor to obtain the satellite noise correction matrix. When the satellite signal is good in the open sky, the satellite signal can be trusted. However, when the satellite signal is poor and there is multipath interference, the satellite signal cannot be trusted.

[0050] Based on the ultra-wideband reliability factor, the preset ultra-wideband observation noise matrix is ​​modified by factor correction to obtain the ultra-wideband noise correction matrix. When the UWB base station is in direct line of sight without obstruction, the ultra-wideband signal can be trusted. However, when the UWB signal is obstructed, the ultra-wideband signal cannot be trusted.

[0051] Specifically, step 105 includes the following specific implementation methods: R gnss，k =R gnss，base / F gnss ; R uwb，k =R uwb，base / F uwb ; Among them, R gnss，base R is the satellite observation noise matrix. gnss，k For the satellite noise correction matrix, F gnss R is the satellite credibility factor. uwb，base R is the ultra-bandwidth observation noise matrix. uwb，k To obtain the ultra-bandwidth noise correction matrix, F uwb This is the ultra-wideband reliability factor.

[0052] It should be noted that R base It is the baseline noise matrix of each sensor under ideal conditions. The confidence factor F is used as the denominator. The higher the value of F, the smaller the corresponding R, which means that the filter has a higher confidence weight for the observation.

[0053] 106. Based on the satellite noise correction matrix, the ultra-wideband noise correction matrix, the satellite observations, and the ultra-wideband observations, perform Kalman fusion update processing on the state prediction vector and the prediction confidence matrix to obtain a state update vector and an update confidence matrix, wherein the state update vector includes: updated positioning data; In this embodiment, the raw observations from satellites and UWB, rather than their individually calculated positions, are used as the observation updates, and the influence of these observations on the filter is adjusted according to real-time weights. The filter ultimately outputs an optimal estimate, namely the fused high-precision position, velocity, and attitude. When the satellite or UWB signal is of high quality, its high-confidence result will back-calibrate the IMU's zero-bias error, preparing for a possible pure inertial navigation phase through error compensation.

[0054] The satellite noise correction matrix, ultra-wideband noise correction matrix, satellite observations, and ultra-wideband observations are substituted into a pre-defined Kalman framework for prediction and updating. The state prediction vector and prediction confidence matrix are then subjected to Kalman fusion update processing to obtain the state update vector and update confidence matrix.

[0055] Specifically, the 106 steps include the following specific implementation methods:

[0056] Among them, K k For the Kalman update matrix, P k|k-1 To predict the confidence matrix, H k To observe the Jacobian matrix, R kX is a diagonal matrix composed of the ultra-bandwidth noise correction matrix and the satellite noise correction matrix. k|k-1 Let Z be the state prediction vector, and Z be the combined vector of satellite observations and ultra-wideband observations, h(X) k|k-1 X represents the prediction distance corresponding to the state prediction vector. k For the state update vector, P k To update the confidence matrix, I is the identity matrix.

[0057] It should be noted that the Kalman gain K k It is based on P k|k-1 To predict the confidence matrix and R k Adjust the diagonal matrix. In R k When the value of the diagonal matrix is ​​small, the Kalman gain K... k The larger the value, the greater the correction, and the more the weights are biased towards the combined vector Z of the original observations from the reference satellite and UWB. In R... k When the diagonal matrix is ​​large, the Kalman gain K... k If the value becomes smaller, the correction will be weaker, and the weights will be more inclined to refer to the IMU's state prediction vector X. k|k-1 In P k When the updated confidence matrix is ​​large, the Kalman gain K... k If the value increases, the correction will be stronger, and the weights will be more biased towards the combined vector Z of the original observations from the reference satellite and UWB.

[0058] 107. Based on the IMU reliability factor, perform quality labeling on the updated positioning data to generate quality-labeled fused positioning data.

[0059] In this embodiment, the IMU credibility factor F is calculated. imu As an indicator of system health, it is used for quality labeling and risk control of output data. When the IMU credibility factor F... imu If the location data remains below a certain danger threshold (e.g., 0.15) and no highly reliable external signal is received for a certain period of time, the updated location data is marked with an "external assistance reset required" warning, indicating to the user that the location reliability is about to be lost, and fused location data with quality labels is generated.

[0060] Parallel methods can be used for error boundary prediction: based on the state covariance matrix P k The diagonal element corresponding to the middle position, and the IMU confidence factor F imu The system calculates and outputs a conservative estimate of the "positioning error circle radius" in real time to assess the degree of attenuation, which is then used by upper-layer applications for risk assessment.

[0061] For details, please refer to Figure 2 , Figure 2This is a schematic diagram of a specific embodiment of the weighted fusion indoor and outdoor positioning method in this invention, which includes the following specific implementation methods in steps 107: 1071. Determine whether the IMU reliability factor is less than a preset anomaly threshold; 1072. When the value is less than a preset abnormal threshold, the updated positioning data is marked as low confidence, and fused positioning data with low confidence is generated. 1073. When the value is not less than a preset abnormal threshold, determine whether the IMU trust factor is greater than a preset health threshold. 1074. When the value is greater than the preset health threshold, the updated positioning data is marked as high confidence, and fused positioning data with high confidence is generated.

[0062] In steps 1071-1074, the IMU reliability factor F is first analyzed. imu If the value is less than a preset abnormal threshold of 0.3, the updated positioning data is marked as low confidence, and fused positioning data with low confidence is generated.

[0063] If the abnormal threshold is not less than 0.3, then the IMU confidence factor F is determined. imu If the value exceeds a preset health threshold of 0.7, the updated location data is marked as high confidence, and fused location data with high confidence labels is generated.

[0064] In a further processing method, if the IMU reliability factor F imu >0.8 but F gnss and F uwb All are low, marked as medium confidence, mainly relying on inertial estimation.

[0065] In this embodiment of the invention, a preliminary state vector is predicted from the inertial data of the IMU to obtain a state prediction vector. Confidence calculations are performed on satellite signals and ultra-wideband signals. Based on the confidence level, the highest weight is dynamically assigned to the most reliable sensor while suppressing interference from unreliable sources. Multi-sensor positioning data fusion calculations are then performed to generate updated positioning data. Furthermore, the updated positioning data is quality-labeled through confidence analysis of the IMU's inertial data to generate fused positioning data. The fusion process is adjusted according to the smooth transition weights of the multi-source signal quality factors, achieving updated positioning data that approximates the theoretical optimal accuracy under the current environment. The positioning trajectory seamlessly connects without abrupt breaks when crossing different technology coverage areas, solving the technical problem of existing positioning technologies struggling to achieve continuous and stable positioning data output under seamless indoor-outdoor transitions.

[0066] Figure 3This is a schematic diagram of the structure of a weighted fusion indoor / outdoor positioning device 300 provided in an embodiment of the present invention. The weighted fusion indoor / outdoor positioning device 300 can vary significantly due to different configurations or performance. It may include one or more central processing units (CPUs) 310 and memory 320, and one or more storage media 330 storing application programs 333 or data 332. The memory 320 and storage media 330 can be temporary or persistent storage. The program stored in the storage media 330 may include one or more modules (not shown in the diagram), each module may include a series of instruction operations on the weighted fusion indoor / outdoor positioning device 300. Furthermore, the processor 310 may be configured to communicate with the storage media 330 and execute the series of instruction operations in the storage media 330 on the weighted fusion indoor / outdoor positioning device 300.

[0067] The weighted fusion-based indoor / outdoor positioning device 300 may also include one or more power supplies 340, one or more wired or wireless network interfaces 350, one or more input / output interfaces 360, and / or one or more operating systems 331, such as Windows Server, Mac OS X, Unix, Linux, Free BSD, etc. Those skilled in the art will understand that... Figure 3 The illustrated structure of a weighted fusion-based indoor and outdoor positioning device does not constitute a limitation on weighted fusion-based indoor and outdoor positioning devices, which may include more or fewer components than illustrated, or combine certain components, or have different component arrangements.

[0068] The present invention also provides a computer-readable storage medium, which can be a non-volatile computer-readable storage medium or a volatile computer-readable storage medium, wherein the computer-readable storage medium stores instructions that, when the instructions are executed on a computer, cause the computer to perform the steps of the weighted fusion indoor and outdoor positioning method.

[0069] In the context of this disclosure, a machine-readable medium can be a tangible medium that may contain or store a program for use by or in conjunction with an instruction execution system, apparatus, or device. A machine-readable medium can be a machine-readable signal medium or a machine-readable storage medium. A machine-readable medium can be, but is not limited to, electronic, magnetic, optical, electromagnetic, infrared, or semiconductor systems, apparatus, or devices, or any suitable combination of the foregoing. More specific examples of machine-readable storage media include electrical connections based on one or more wires, portable computer disks, hard disks, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), optical fiber, portable compact disk read-only memory (CD-ROM), optical storage devices, magnetic storage devices, or any suitable combination of the foregoing.

[0070] Furthermore, although the operations are described in a specific order, this should be understood as requiring that such operations be performed in the specific order shown or in sequential order, or requiring that all illustrated operations be performed to achieve the desired result. In certain environments, multitasking and parallel processing may be advantageous. Similarly, although several specific implementation details are included in the above discussion, these should not be construed as limiting the scope of this disclosure. Certain features described in the context of individual embodiments may also be implemented in combination in a single implementation. Conversely, various features described in the context of a single implementation may also be implemented individually or in any suitable sub-combination in multiple implementations.

[0071] Although the subject matter has been described using language specific to structural features and / or methodological logic, it should be understood that the subject matter defined in the appended claims is not necessarily limited to the specific features or actions described above. Rather, the specific features and actions described above are merely illustrative examples of implementing the claims.

Claims

1. A multi-data source fusion positioning method, characterized in that, Including the following steps: Read the state vector, the angular velocity of the preset IMU, and the force measurement value of the preset IMU, wherein the state vector includes: positioning data; Based on the preset inertial navigation algorithm, the angular velocity, and the specific force measurement value, the state vector is subjected to state prediction processing to obtain the state prediction vector and the prediction confidence matrix; Based on the preset factor analysis algorithm, confidence levels are calculated for preset satellite signals, preset ultra-wideband signals, and preset IMU signals to generate satellite confidence factors, ultra-wideband confidence factors, and IMU confidence factors. Read satellite observations and ultra-wideband observations; Based on the satellite confidence factor, the preset satellite observation noise matrix is ​​subjected to factor correction processing to obtain the satellite noise correction matrix, and based on the ultra-bandwidth confidence factor, the preset ultra-bandwidth observation noise matrix is ​​subjected to factor correction processing to obtain the ultra-bandwidth noise correction matrix. Based on the satellite noise correction matrix, the ultra-wideband noise correction matrix, the satellite observations, and the ultra-wideband observations, Kalman fusion update processing is performed on the state prediction vector and the prediction confidence matrix to obtain the state update vector and the update confidence matrix, wherein the state update vector includes: updated positioning data; Based on the IMU reliability factor, the updated positioning data is quality-labeled to generate quality-labeled fused positioning data.

2. The multi-data source fusion positioning method according to claim 1, characterized in that, The step of performing state prediction processing on the state vector based on a preset inertial navigation algorithm, the angular velocity, and the specific force measurement value to obtain a state prediction vector and a prediction confidence matrix includes: Among them, U k X is a vector composed of angular velocity and specific force measurements. k-1 Let P be the state vector, f() be the prediction function, and P be the prediction function. k-1 Let F be the initial state covariance matrix. k X is the state transition Jacobian matrix, Q is the process noise covariance matrix, and X is the process noise covariance matrix. k|k-1 P is the state prediction vector. k|k-1 This is the prediction confidence matrix, where k is the index value.

3. The indoor and outdoor positioning method based on weighted fusion according to claim 1, characterized in that, The step of calculating the confidence levels of preset satellite signals, preset ultra-wideband signals, and preset IMU signals according to a preset factor analysis algorithm, and generating satellite confidence factors, ultra-wideband confidence factors, and IMU confidence factors, includes: Among them, F gnss N is the satellite credibility factor. norm C is the satellite number normalization factor. norm G is the carrier-to-noise ratio normalization factor. norm N is the geometric precision inverse normalization factor, w1, w2, and w3 are the satellite factor weighting coefficients, and N is the weighting factor. sat N represents the number of visible satellites. thresh To effectively calculate the threshold, M represents the total number of satellites, and CNR... i Let CNR be the carrier-to-noise ratio of the i-th satellite. max CNR is the maximum effective empirical threshold. min GDOP is the minimum effective empirical threshold, and GDOP is the current geometric precision factor. thresh This is the geometric empirical threshold.

4. The indoor and outdoor positioning method based on weighted fusion according to claim 1, characterized in that, The step of calculating the confidence levels of preset satellite signals, preset ultra-wideband signals, and preset IMU signals according to a preset factor analysis algorithm to generate satellite confidence factors, ultra-wideband confidence factors, and IMU confidence factors further includes: Among them, F uwb As the ultra-bandwidth confidence factor, A norm S is the base station number normalization factor. norm R is the signal strength normalization factor. norm λ1, λ2, and λ3 are the normalization factors for the ranging residuals, and N is the weighting coefficient of the superbandwidth factor. anchor The number of visible base stations, RSSI j Let RSSI be the signal strength of the j-th base station. max RSSI is the maximum empirical threshold for the signal. min d is the minimum empirical threshold for the signal. j meas Let d be the measured distance of the j-th base station. j pred This is the distance predicted for the j-th base station at the previous time step.

5. The indoor and outdoor positioning method based on weighted fusion according to claim 1, characterized in that, The step of calculating the confidence levels of preset satellite signals, preset ultra-wideband signals, and preset IMU signals according to a preset factor analysis algorithm to generate satellite confidence factors, ultra-wideband confidence factors, and IMU confidence factors further includes: Among them, F imu Here, γ is the IMU reliability factor, t is the mixing coefficient, β is the attenuation coefficient, and V is the IMU reliability factor. norm Here, is the motion intensity normalization factor, 'a' is the accelerometer reading, and 'b' is the motion intensity normalization factor. a To estimate the accelerometer zero bias, a thresh This is a high dynamic threshold.

6. The indoor and outdoor positioning method based on weighted fusion according to claim 1, characterized in that, The steps of performing factor correction processing on the preset satellite observation noise matrix according to the satellite confidence factor to obtain the satellite noise correction matrix, and performing factor correction processing on the preset ultra-wideband observation noise matrix according to the ultra-wideband confidence factor to obtain the ultra-wideband noise correction matrix include: R gnss，k =R gnss，base / F gnss ; R uwb，k =R uwb，base / F uwb ; Among them, R gnss，base R is the satellite observation noise matrix. gnss，k For the satellite noise correction matrix, F gnss R is the satellite credibility factor. uwb，base R is the ultra-bandwidth observation noise matrix. uwb，k To obtain the ultra-bandwidth noise correction matrix, F uwb This is the ultra-wideband reliability factor.

7. The indoor and outdoor positioning method based on weighted fusion according to claim 1, characterized in that, The step of performing Kalman fusion update processing on the state prediction vector and the prediction confidence matrix based on the satellite noise correction matrix, the ultra-bandwidth noise correction matrix, the satellite observations, and the ultra-bandwidth observations to obtain the state update vector and the update confidence matrix includes: Among them, K k For the Kalman update matrix, P k|k-1 To predict the confidence matrix, H k To observe the Jacobian matrix, R k X is a diagonal matrix composed of the ultra-bandwidth noise correction matrix and the satellite noise correction matrix. k|k-1 Let Z be the state prediction vector, and Z be the combined vector of satellite observations and ultra-wideband observations, h(X) k|k-1 X represents the prediction distance corresponding to the state prediction vector. k For the state update vector, P k To update the confidence matrix, I is the identity matrix.

8. The indoor and outdoor positioning method based on weighted fusion according to claim 1, characterized in that, The step of quality labeling the updated positioning data based on the IMU reliability factor to generate quality-labeled fused positioning data includes: Determine whether the IMU trust factor is less than a preset anomaly threshold; When the value is less than a preset abnormal threshold, the updated positioning data is marked as low confidence, and fused positioning data with low confidence is generated. If the value is not less than a preset abnormal threshold, then it is determined whether the IMU trust factor is greater than a preset health threshold. When the value exceeds a preset health threshold, the updated location data is marked as high confidence, and fused location data with high confidence is generated.

9. A weighted fusion indoor and outdoor positioning device, characterized in that, The weighted fusion indoor and outdoor positioning device includes: a memory and at least one processor, wherein the memory stores instructions, and the memory and the at least one processor are interconnected via a line; The at least one processor invokes the instructions in the memory to cause the weighted indoor / outdoor positioning device to perform the weighted indoor / outdoor positioning method as described in any one of claims 1-8.

10. A computer-readable storage medium storing a computer program thereon, characterized in that, When the computer program is executed by the processor, it implements the weighted fusion indoor and outdoor positioning method as described in any one of claims 1-8.

Images