Evaluation-based multi-source fusion positioning method and positioning system

Abstract

The embodiment of the invention provides an evaluation-based multi-source fusion positioning method and positioning system, which are applied to navigation services. According to the method, UWB positioning data, GNNS positioning data and Bluetooth RSSI positioning data are fused at the same time, the current signal quality of each positioning signal source is determined through a preset data quality evaluation strategy, the current positioning mode is determined based on the evaluation result of the current signal quality, different positioning modes have different preset corresponding positioning calculation fusion strategies, and the positioning calculation fusion is more accurate. Therefore, the positioning source and the fusion strategy can be intelligently, simply and adaptively selected to realize positioning.

Description

Technical Field

[0001] This invention belongs to the field of mobile positioning technology, and in particular to a positioning method and system based on multi-source fusion evaluation. Background Technology

[0002] With the increasing demands for positioning accuracy and stability in industries such as manufacturing, healthcare, mining, power, and elderly care, traditional single positioning technologies can no longer meet the needs of indoor and outdoor positioning scenarios. Multi-source data fusion for positioning has become an inevitable trend. For example, application number CN2022116325493 discloses a multi-source fusion positioning method and device for navigation services. This positioning device uses BDS, UWB, and IMU technologies to achieve indoor and outdoor positioning. While it can effectively handle different scenarios, when selecting from multiple positioning sources, it relies on the positioning requirements and environmental parameters of the target object, increasing the power consumption of the positioning system and failing to intelligently and adaptively select the specific positioning source.

[0003] Therefore, it is necessary to improve the existing technology to overcome the aforementioned defects. Summary of the Invention

[0004] Therefore, the present invention aims to solve the technical problem of enabling a positioning system to intelligently and adaptively select positioning sources and fusion strategies to achieve positioning when multiple positioning sources are fused.

[0005] To address the aforementioned technical problems, a multi-source fusion positioning method based on evaluation is proposed and applied to positioning and navigation services. The method includes: acquiring terminal data from multiple positioning sources in a positioning system, including a UWB positioning module, a Bluetooth RSSI positioning module, and a GNSS satellite positioning module; preprocessing the terminal data to obtain valid UWB, RSSI, and GNSS datasets respectively; extracting UWB feature vectors from the valid UWB dataset and evaluating the current data quality of the UWB positioning module; extracting RSSI feature vectors from the valid RSSI dataset and evaluating the current data quality of the Bluetooth RSSI positioning module; extracting GNNS feature vectors from the valid GNSS dataset and evaluating the current data quality of the GNSS satellite positioning module; based on the evaluated current data quality, selecting and determining the current mode according to a preset mode switching logic threshold; and dynamically adjusting the terminal data weights of the multiple positioning sources based on the selected current mode to perform positioning fusion and obtain the final positioning coordinates.

[0006] In one embodiment, the specific calculation steps for "extracting UWB feature vectors from the valid UWB dataset and evaluating the current data quality of the UWB positioning module" are as follows:

[0007]

[0008] Q UWB =w1·Q range +w2·Q angle (Formula 1)

[0009] Among them, Q UWB Q is the current data quality metric for the UWB positioning module. range d is the quality assessment index for ranging. meas d represents the current measured distance of the UWB positioning module. pred Q is the distance calculated based on the predicted location, where δ is the error tolerance factor; angle θ is the quality evaluation index for angle measurement. t θ is the current measured angle of the UWB positioning module. t-1 Let w1 be the angle of the previous frame, σ be the jump tolerance factor, and w1 + w2 = 1.

[0010] In one embodiment, if the positioning system includes multiple UWB positioning modules, the step of "extracting UWB feature vectors from the valid UWB dataset and evaluating the current data quality of the UWB positioning modules" further includes the step of...

[0011]

[0012] Among them, Q i w is the current data quality metric for the i-th UWB positioning module. i Let w be the weight corresponding to the i-th UWB positioning module. i =1 / range i .

[0013] In one embodiment, the specific calculation steps for "extracting RSSI feature vectors from the valid RSSI dataset and evaluating the current data quality of the Bluetooth RSSI positioning module" are as follows:

[0014]

[0015] Q rssi =w1·Q sign +w2·Q stab (Formula 2)

[0016] Q rssi Q is the current data quality metric for the RSSI positioning module. sign Q is a metric for evaluating signal strength quality. stab σ is a stability quality assessment index. rssi Let w1 be the standard deviation of the sliding window, α be the tolerance standard deviation, and w1 + w2 = 1.

[0017] In one embodiment, if the positioning system includes multiple Bluetooth RSSI positioning modules, the step of "extracting RSSI feature vectors from the valid RSSI dataset and evaluating the current data quality of the Bluetooth RSSI positioning modules" further includes the step of...

[0018]

[0019] Among them, Q i w is the quality metric for the i-th Bluetooth device. i Let be the weight corresponding to the i-th Bluetooth device.

[0020] In one embodiment, the specific calculation steps for "extracting GNNS feature vectors from the valid GNNS dataset and evaluating the current data quality of the GNNS satellite positioning module" are as follows:

[0021]

[0022] Q gnss =w1·Q cnr +w2·Q nstat +w3·Q hdop (Formula 3)

[0023] Among them, Q gnss Q is the current data quality metric for the GNNS positioning module. cnr Q is a carrier-to-noise ratio quality assessment metric. nstat Q is a metric for evaluating the quantity and quality of satellites. hdop σ is the quality assessment index for HDOP, and w1+w2+w3=1.

[0024] In one embodiment, the step of "selecting and determining the current mode based on the results of the current data quality assessment and according to a preset mode switching logic threshold" includes performing logical judgments in the following order until the current mode is determined:

[0025] If UWB Q angle ≥θ uwb_angel If the number is less than 1, then enter downgrade mode A;

[0026] If UWB Q range ≥θ uwb_range If the quantity is ≥1 but <3, then enter downgrade mode B;

[0027] If UWB Q range ≥θ uwb_range If the number is less than 1, then UWB RSSI distance estimation is enabled, and the system enters degraded mode C.

[0028] If no UWB ranging data is available, then enter degraded mode D;

[0029] If BLE Q ble <θ ble If so, BLE is determined to be an invalid auxiliary function;

[0030] If GNSS Q gnss <θ gnss If so, then GNSS is determined to be an invalid aid.

[0031] One embodiment, "dynamically adjusting the terminal data weights of each positioning source based on the selected current mode to perform positioning fusion and obtain the final positioning coordinates," includes using an extended Kalman filter algorithm to perform matrix fusion of positioning results from the UWB positioning module, the Bluetooth RSSI positioning module, and the GNSS positioning module.

[0032] In one embodiment, the matrix fusion includes the following steps: extrapolating the current state using a state transition model; constructing observation equations based on available data sources to calculate residual and Jacobian matrices; introducing confidence levels and dynamically adjusting the observation noise covariance matrix; if the observation confidence level of a certain data source is low, its observation error covariance is set to a larger value to reduce its weight in the fusion; if the positioning system is in degraded mode, only the currently available observations are used to update the state.

[0033] This invention also provides a multi-source fusion positioning system based on evaluation. The multi-source fusion positioning system includes: multiple positioning sources, including a UWB positioning module, a Bluetooth RSSI positioning module, and a GNSS satellite positioning module; a data preprocessing and quality evaluation module, used to preprocess the terminal data to obtain valid UWB, RSSI, and GNSS datasets respectively, extract UWB feature vectors from the valid UWB dataset and evaluate the current data quality of the UWB positioning module, extract RSSI feature vectors from the valid RSSI dataset and evaluate the current data quality of the Bluetooth RSSI positioning module, and extract GNNS feature vectors from the valid GNSS dataset and evaluate the current data quality of the GNNS satellite positioning module; a degradation decision module, used to select and determine the current mode based on the evaluation of the current data quality and a preset mode switching logic threshold; and a positioning calculation module, used to dynamically adjust the terminal data weights of each positioning source based on the current mode obtained by the degradation decision module to perform positioning fusion and obtain the final positioning coordinates.

[0034] The technical solution provided by this invention has the following advantages:

[0035] The positioning system and method provided in this invention evaluate the signal quality of each positioning signal source through data quality assessment, and determine the current positioning mode based on the evaluation results of the signal quality. Different positioning modes have preset corresponding different positioning calculation strategies, thereby enabling intelligent and adaptive selection of positioning sources and fusion strategies to achieve positioning. Attached Figure Description

[0036] To more clearly illustrate the specific embodiments of the present invention or the technical solutions in the prior art, the drawings used in the description of the specific embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of the present invention. For those skilled in the art, other drawings can be obtained from these drawings without creative effort.

[0037] Figure 1 A flowchart illustrating a multi-source fusion-based localization method based on evaluation, provided in an embodiment of the present invention;

[0038] Figure 2 for Figure 1 An example embodiment provides a schematic diagram of a localization process based on multi-source fusion and evaluation.

[0039] Figure 3 This is a schematic diagram of a multi-source fusion-based positioning system provided in an embodiment of the present invention. Detailed Implementation

[0040] The technical solution of the present invention will now be clearly and completely described with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of the present invention. The present invention will be described in detail below with reference to the accompanying drawings and embodiments. It should be noted that, unless otherwise specified, the embodiments and features in the embodiments of the present invention can be combined with each other.

[0041] It should be noted that the terms "first," "second," etc., in the specification, claims, and drawings of this invention are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence.

[0042] In this invention, unless otherwise stated, the examples described below are merely specific examples and are not intended to limit the embodiments of the invention to the specific steps, values, conditions, data, order, etc. Those skilled in the art can utilize the concept of this invention to construct more embodiments not mentioned herein by reading this specification.

[0043] In this document, the term "embodiment" means that a particular feature, structure, or characteristic described in connection with an embodiment may be included in at least one embodiment of this application. The appearance of this phrase in various places throughout the specification does not necessarily refer to the same embodiment, nor is it a separate or alternative embodiment mutually exclusive with other embodiments. It will be explicitly and implicitly understood by those skilled in the art that the embodiments described herein can be combined with other embodiments.

[0044] Here, we first explain some terms used in the embodiments of this invention. Ultra-wideband (UWB) technology is a wireless carrier communication technology that does not use sinusoidal carriers but instead uses nanosecond-level non-sinusoidal narrow pulses to transmit data, thus occupying a wide spectrum. UWB technology has advantages such as low system complexity, low transmitted signal power spectral density, insensitivity to channel fading, low interception capability, and high positioning accuracy, making it particularly suitable for high-speed wireless access in dense multipath environments such as indoor spaces. Time-of-flight (TOF) technology can be broadly understood as a technology that measures the time it takes for an object, particle, or wave to travel a certain distance in a fixed medium (where the medium / distance / time are known or measurable), thereby further understanding certain properties of ions or media. The ToF ranging method belongs to bidirectional ranging technology, which mainly uses the time it takes for a signal to travel back and forth between two asynchronous transceivers (or reflected surfaces) to measure the distance between nodes. Angle-of-Arrival (AOA) ranging: This is a typical ranging-based localization algorithm. It uses hardware to sense the direction of arrival of the transmitting node's signal, calculates the relative azimuth or angle between the base station and the tag (terminal), and then uses triangulation or other methods to calculate the location of the unknown node. AOA-based localization is a common self-localization algorithm for wireless sensor network nodes, with low communication overhead.

[0045] Global Navigation Satellite System (GNSS) is a space-based radio navigation system that provides all-weather, high-precision positioning, navigation, and timing services to various users worldwide through multiple satellites. The core principle of GNSS is to calculate three-dimensional position by utilizing the relationship between satellite signal propagation time and distance, through multi-satellite geometric intersection. Current GNSS technology can provide positioning accuracy up to centimeter-level, meeting civilian requirements. However, in urban environments, the presence of tall buildings and other obstructions causes reflections of satellite radio waves, leading to reduced measurement accuracy. This multipath effect remains unresolved. In heavily obstructed environments such as garages, the number of visible satellites decreases, necessitating the combination of GNSS with other positioning methods to achieve accurate positioning.

[0046] Bluetooth RSSI (Received Signal Strength Indicator) positioning technology is based on Bluetooth Low Energy (BLE) technology. It estimates the distance between devices by measuring signal strength attenuation, thereby achieving indoor or short-range positioning. Bluetooth RSSI transmits signals through Bluetooth beacons at known locations. The receiving end measures the RSSI value, calculates the distance to the beacon, and then determines its own location through the geometric intersection of multiple beacons. The core advantages of this technology are low hardware cost (mobile phones and Bluetooth beacons can both be used as nodes) and flexible deployment. The main disadvantage is the need to overcome the interference of environmental noise on RSSI.

[0047] This embodiment provides an evaluation-based multi-source fusion positioning method for positioning and navigation services. This method simultaneously fuses UWB positioning data, GNNS positioning data, and Bluetooth RSSI positioning data. It determines the current signal quality of each positioning signal source through a preset data quality evaluation strategy, and determines the current positioning mode based on the evaluation result. Different positioning modes have preset corresponding different positioning calculation and fusion strategies, thereby enabling intelligent and simple adaptive selection of positioning sources and fusion strategies to achieve positioning.

[0048] like Figure 1 and Figure 2 As shown, the multi-source fusion-based localization method based on evaluation includes the following steps.

[0049] Step S1: Acquire terminal data from multiple positioning sources in the positioning system. These multiple positioning sources include a UWB positioning module, a Bluetooth RSSI positioning module, and a GNSS satellite positioning module. The UWB positioning module includes a UWB angle measurement unit and a UWB ranging unit. The UWB angle measurement unit is used to sense the direction of arrival of the transmitting node signal and calculate the relative azimuth or angle between the receiving base station and the tag (terminal). The UWB ranging unit is used to obtain the round-trip flight time between two asynchronous transceivers (or reflective surfaces).

[0050] Step S2: Preprocess the terminal data and extract the corresponding feature vectors. Specifically, preprocess the terminal data to obtain UWB effective datasets, RSSI effective datasets, and GNSS effective datasets. Extract UWB feature vectors from the UWB effective datasets and evaluate the current data quality of the UWB positioning module; extract RSSI feature vectors from the RSSI effective datasets and evaluate the current data quality of the Bluetooth RSSI positioning module; extract GNNS feature vectors from the GNNS effective datasets and evaluate the current data quality of the GNNS satellite positioning module.

[0051] The phrase "data preprocessing of the terminal data" includes the following steps.

[0052] Step S21: Remove noise and invalid data from the terminal data. Noise and invalid data mainly refer to missing values, outliers, etc., in the data transmitted from the three positioning signal sources.

[0053] Step S22: Synchronize the terminal data of the UWB positioning module, the Bluetooth RSSI positioning module, and the GNSS satellite positioning module according to the timestamp to ensure that each data point is the observation value at the same time.

[0054] In one embodiment, UWB feature vectors are extracted from the valid UWB dataset, and the current data quality of the UWB positioning module is evaluated. The specific calculation steps are shown in Formula 1:

[0055]

[0056] Q UWB =w1·Q range +w2·Q angle (Formula 1)

[0057] Among them, Q UWB Q is the current data quality metric for the UWB positioning module. range d is the quality assessment index for ranging. meas d represents the current measured distance of the UWB positioning module. predQ is the distance calculated based on the predicted location, where δ is the error tolerance factor; angle θ is the quality evaluation index for angle measurement. t θ is the current measured angle of the UWB positioning module. t-1 The angle is from the previous frame, σ is the jump tolerance factor, and w1 + w2 = 1. The UWB feature vector is d. meas and θ t .

[0058] δ is a constant, such as 1m, and its specific value depends on the positioning system. σ is a constant, such as 5°, and its specific value depends on the positioning system. w1 and w2 are both constants, which can be 0.6 and 0.4 respectively, and their specific values ​​depend on the positioning system. After this normalization process, the current data quality index value is between [0,1], and the closer it is to 1, the more reliable it is.

[0059] In one embodiment, if the positioning system includes multiple UWB positioning modules, the step of "extracting UWB feature vectors from the valid UWB dataset and evaluating the current data quality of the UWB positioning modules" further includes the step of...

[0060]

[0061] Among them, Q i w is the current data quality metric for the i-th UWB positioning module. i w represents the weight corresponding to the i-th UWB positioning module. i =1 / range i .

[0062] In one embodiment, the specific calculation steps for "extracting RSSI feature vectors from the valid RSSI dataset and evaluating the current data quality of the Bluetooth RSSI positioning module" are as follows:

[0063]

[0064] Q rssi =w1·Q sign +w2·Q stab (Formula 2)

[0065] Q rssi Q is the current data quality metric for the RSSI positioning module. sign Q is a metric for evaluating signal strength quality. stab σ is a stability quality assessment index. rssi Let w1 be the standard deviation of the sliding window, α be the tolerance standard deviation, and w1 + w2 = 1.

[0066] Furthermore, using normalization, the RSSI range is set, Q signTypically, it ranges from -100 to -40 dBm. Current data quality index values ​​are between [0,1], with values ​​closer to 1 indicating higher reliability. α is a constant, such as 4 dBm, but the specific value depends on the positioning system. w1 and w2 are both constants, which can be 0.7 and 0.3 respectively, with specific values ​​depending on the positioning system.

[0067] If the positioning system includes multiple Bluetooth RSSI positioning modules, the step of "extracting RSSI feature vectors from the valid RSSI dataset and evaluating the current data quality of the Bluetooth RSSI positioning modules" further includes the step of...

[0068]

[0069] Among them, Q i w is the quality metric for the i-th Bluetooth device. i Let w be the weight corresponding to the i-th Bluetooth device. i You can press Q sign Weighting depends on the positioning system.

[0070] In one embodiment, the specific calculation steps for "extracting GNNS feature vectors from the valid GNNS dataset and evaluating the current data quality of the GNNS satellite positioning module" are as follows:

[0071]

[0072] Q gnss =w1·Q cnr +w2·Q nstat +w3·Q hdop (Formula 3)

[0073] Among them, Q tatal Q represents the current data quality metrics for the GNNS positioning module. cnr As a carrier-to-noise ratio (CNR) quality assessment metric, normalization is used in this embodiment to set the CNR range to 20 to 45. cnr The value is between [0,1], and the closer it is to 1, the more reliable it is. Q nstat As a metric for evaluating the quantity and quality of satellites, this embodiment uses normalization to set the quantity range to 4 to 12. Q nstat The value is between [0,1], and the closer it is to 1, the more reliable it is. Q hdop This is the HDOP quality evaluation index. σ is the tolerance factor, a constant. In this embodiment, σ is 2, but other values ​​can be selected depending on the positioning system. w1+w2+w3=1, where w1, w2, and w3 are all constants, which can be 0.4, 0.2, and 0.4 respectively. The specific values ​​depend on the positioning system.

[0074] Step S3: Based on the current data quality assessment results, determine the current mode selection according to the preset mode switching logic threshold.

[0075] In one specific embodiment, logical judgments are performed in the following order until the current mode is determined:

[0076] If UWB Q angle ≥θ uwb_angel If the number is less than 1, the system enters degraded mode A. In degraded mode A, the positioning system is dominated by the UWB positioning module for ranging, with the Bluetooth RSSI positioning module and GNSS positioning module as auxiliary.

[0077] If UWB Q range ≥θ uwb_range If the number is ≥1 but <3, then it enters downgrade mode B; in downgrade mode B, the positioning system retains the effective ranging information of UWB and integrates the positioning results of the Bluetooth RSSI positioning module and the GNSS positioning module.

[0078] If UWB Q range ≥θ uwb_range If the number is less than 1, then UWB RSSI ranging is enabled and the system enters degraded mode C. In degraded mode C, the positioning system enables the Bluetooth RSSI positioning module and integrates the positioning results assisted by the Bluetooth RSSI positioning module and the GNSS positioning module.

[0079] If there is no UWB ranging data, it enters degraded mode D; in degraded mode D, the positioning system integrates the positioning results assisted by the BLE RSSI positioning module and the GNSS positioning module.

[0080] If BLE Q ble <θ ble If so, BLE is determined to be an invalid auxiliary function.

[0081] If GNSS Q gnss <θ gnss If so, then GNSS is determined to be an invalid aid.

[0082] Step S4: Based on the selected current mode, dynamically adjust the terminal data weights of each positioning source to perform positioning fusion and obtain the final positioning coordinates. Specifically, an extended Kalman filter algorithm is used to perform matrix fusion on the positioning results from the UWB positioning module, the Bluetooth RSSI positioning module, and the GNSS positioning module.

[0083] In this embodiment, a 2D coordinate system is used as an example for explanation, and the target state vector is:

[0084]

[0085] Where x k y k Represents position coordinates, v x,k x y,k Represents the velocity component.

[0086] State transition model:

[0087] X k =F·X k-1 +W k

[0088] State transition matrix F:

[0089]

[0090] Δt is the sampling period, W k For process noise, satisfy W k ~N(0,Q).

[0091] The observation vector is set according to the actual data source:

[0092] Among them, UWB angle measurement:

[0093] Z angle =arctan2(yy i ,xx i )+v angle

[0094] UWB ranging:

[0095]

[0096] UWB / BLE RSSI distance estimation:

[0097]

[0098] in:

[0099] RSSI0: RSSI at a reference distance d0, typically d0 = 1 meter;

[0100] GNSS:

[0101]

[0102] Furthermore, a Kalman filter algorithm is used to perform matrix fusion on the positioning results from the UWB positioning module, the Bluetooth RSSI positioning module, and the GNSS positioning module. The specific steps include:

[0103] S41: Extrapolate the current state using a state transition model;

[0104] S42: Construct observation equations based on available data sources to calculate residuals and Jacobian matrices; introduce confidence level Q. i And dynamically adjust the observation noise covariance matrix R k ;

[0105] S43: If the observation confidence of a certain data source is low, its observation error covariance is set to a larger value to reduce its weight in the fusion.

[0106] S44: If the positioning system is in degraded mode, update the state only using the currently available observations.

[0107] The fusion method in this embodiment of the invention is very simple, which greatly simplifies the hardware computation of fusion coefficients. However, these fusion coefficients are all based on the positioning pattern matching after the quality judgment of each data source in the early stage. Therefore, the positioning accuracy is not degraded due to the simplification of fusion parameters.

[0108] Each module in the above positioning method can be implemented entirely or partially through software, hardware, or a combination thereof. These modules can be embedded in or independent of the processor in a computer device, or stored in the memory of a computer device as software, so that the processor can call and execute the operations corresponding to each module.

[0109] like Figure 3 As shown, this embodiment of the invention also provides a multi-source fusion positioning system based on evaluation. The multi-source fusion positioning system 10 includes multiple positioning sources, a data preprocessing and quality evaluation module 20, a degradation decision module 30, and a positioning calculation module 40.

[0110] Multiple positioning sources include a UWB positioning module 11, a Bluetooth RSSI positioning module 12, and a GNSS satellite positioning module 13. A data preprocessing and quality assessment module 20 is used to preprocess the terminal data to obtain valid UWB, RSSI, and GNSS datasets respectively. It extracts UWB feature vectors from the valid UWB dataset and assesses the current data quality of the UWB positioning module; extracts RSSI feature vectors from the valid RSSI dataset and assesses the current data quality of the Bluetooth RSSI positioning module; and extracts GNNS feature vectors from the valid GNSS dataset and assesses the current data quality of the GNNS satellite positioning module. A degradation decision module 30 uses the assessed current data quality result and a preset mode switching logic threshold to determine the current mode selection. A positioning calculation module 40 dynamically adjusts the terminal data weights of each positioning source based on the current mode obtained by the degradation decision module to perform positioning fusion and obtain the final positioning coordinates.

[0111] The specific positioning method of the multi-source fusion positioning system 10 is the same as the multi-fusion positioning method described above in this application, and will not be repeated here.

[0112] Obviously, the embodiments described above are merely some, not all, embodiments of the present invention. Based on the embodiments of the present invention, those skilled in the art can make other variations or modifications without creative effort, and all such variations or modifications should fall within the scope of protection of the present invention.

Claims

1. A multi-source fusion-based positioning method based on evaluation, applied to positioning and navigation services, characterized in that, The system acquires terminal data from multiple positioning sources in a positioning system, including a UWB positioning module, a Bluetooth RSSI positioning module, and a GNSS satellite positioning module. The terminal data is preprocessed to obtain UWB effective dataset, RSSI effective dataset and GNSS effective dataset respectively. UWB feature vectors are extracted from the UWB effective dataset and the current data quality of the UWB positioning module is evaluated. RSSI feature vectors are extracted from the RSSI effective dataset and the current data quality of the Bluetooth RSSI positioning module is evaluated. GNNS feature vectors are extracted from the GNNS effective dataset and the current data quality of the GNNS satellite positioning module is evaluated. Based on the current data quality assessment results, the current mode is selected and determined according to the preset mode switching logic threshold. Based on the selected current mode, the terminal data weights of the multiple positioning sources are dynamically adjusted to perform positioning fusion and obtain the final positioning coordinates.

2. The localization method based on multi-source fusion evaluation according to claim 1, characterized in that, The specific calculation steps for "extracting UWB feature vectors from the valid UWB dataset and evaluating the current data quality of the UWB positioning module" are as follows: Q UWB =w1·Q range +w2·Q angle (Formula 1) Among them, Q UWB Q is the current data quality metric for the UWB positioning module. range d is the quality assessment index for ranging. meas d represents the current measured distance of the UWB positioning module. pred Q is the distance calculated based on the predicted location, where δ is the error tolerance factor; angle θ is the quality evaluation index for angle measurement. t θ is the current measured angle of the UWB positioning module. t-1 Let w1 be the angle of the previous frame, σ be the jump tolerance factor, and w1 + w2 = 1.

3. The multi-source fusion-based localization method according to claim 2, characterized in that, If the positioning system includes multiple UWB positioning modules, the step of "extracting UWB feature vectors from the valid UWB dataset and evaluating the current data quality of the UWB positioning modules" further includes the step of... Among them, Q i w is the current data quality metric for the i-th UWB positioning module. i w represents the weight corresponding to the i-th UWB positioning module. i =1 / range i .

4. The localization method based on multi-source fusion evaluation according to claim 1, characterized in that, The specific calculation steps for "extracting RSSI feature vectors from the valid RSSI dataset and evaluating the current data quality of the Bluetooth RSSI positioning module" are as follows: Q rssi =w1·Q sign +w2·Q stab (Formula 2) Among them, Q rssi Q is the current data quality metric for the RSSI positioning module. sign Q is a metric for evaluating signal strength quality. stab σ is a stability quality assessment index. rssi Let w1 be the standard deviation of the sliding window, α be the tolerance standard deviation, and w1 + w2 = 1.

5. The multi-source fusion-based localization method according to claim 4, characterized in that, If the positioning system includes multiple Bluetooth RSSI positioning modules, the step of "extracting RSSI feature vectors from the valid RSSI dataset and evaluating the current data quality of the Bluetooth RSSI positioning modules" further includes the step of... Among them, Q i w is the quality metric for the i-th Bluetooth device. i Let be the weight corresponding to the i-th Bluetooth device.

6. The localization method based on multi-source fusion evaluation according to claim 1, characterized in that, The specific calculation steps for "extracting GNNS feature vectors from the valid GNNS dataset and evaluating the current data quality of the GNNS satellite positioning module" are as follows: Q gnss = w1 Q cnr +w2·Q nstat +w3·Q hdop (Formula 3) Among them, Q gnss Q is the current data quality metric for the GNNS positioning module. cnr Q is a carrier-to-noise ratio quality assessment metric. nstat Q is a metric for evaluating the quantity and quality of satellites. hdop σ is the quality assessment index for HDOP, and w1+w2+w3=1.

7. A multi-source fusion-based localization method according to any one of claims 1 to 6, characterized in that, The phrase "based on the results of the current data quality assessment, selecting and determining the current mode according to a preset mode switching logic threshold" includes performing logical judgments in the following order until the current mode is determined: If UWB Q angle ≥θ uwb_angel If the number is less than 1, then enter downgrade mode A; If UWB Q range ≥θ uwb_range If the quantity is ≥1 but <3, then enter downgrade mode B; If UWB Q range ≥θ uwb_range If the number is less than 1, then UWB RSSI distance estimation is enabled, and the system enters degraded mode C. If no UWB ranging data is available, then enter degraded mode D; If BLE Q ble <θ ble If so, BLE is determined to be an invalid auxiliary function; If GNSS Q gnss <θ gnss If so, then GNSS is determined to be an invalid aid.

8. The localization method based on multi-source fusion evaluation according to claim 7, characterized in that, "Based on the selected current mode, dynamically adjust the terminal data weights of each positioning source to perform positioning fusion and obtain the final positioning coordinates" includes using the extended Kalman filter algorithm to perform matrix fusion of positioning results from the UWB positioning module, the Bluetooth RSSI positioning module, and the GNSS positioning module.

9. The localization method based on multi-source fusion evaluation according to claim 8, characterized in that, The matrix fusion includes the following steps: The current state is extrapolated using a state transition model; Observation equations are constructed based on available data sources to calculate residuals and Jacobian matrices; confidence levels are introduced and the observation noise covariance matrix is ​​dynamically adjusted. If the confidence level of a certain data source observation is low, its observation error covariance is set to a larger value to reduce its weight in the fusion; if the positioning system is in degraded mode, only the currently available observations are used for state updates.

10. A multi-source fusion-based positioning system, characterized in that, The system includes: Multiple positioning sources, including a UWB positioning module, a Bluetooth RSSI positioning module, and a GNSS satellite positioning module; The data preprocessing and quality assessment module is used to preprocess the terminal data to obtain UWB effective dataset, RSSI effective dataset and GNSS effective dataset respectively, extract UWB feature vectors from the UWB effective dataset and evaluate the current data quality of the UWB positioning module, extract RSSI feature vectors from the RSSI effective dataset and evaluate the current data quality of the Bluetooth RSSI positioning module, and extract GNNS feature vectors from the GNNS effective dataset and evaluate the current data quality of the GNNS satellite positioning module. The degradation decision module is used to evaluate the current data quality and determine the current mode based on the preset mode switching logic threshold. The positioning calculation module is used to dynamically adjust the terminal data weights of each positioning source based on the current mode obtained by the degradation decision module in order to perform positioning fusion and obtain the final positioning coordinates.

Images