High-precision real-time positioning system and method for positioning personnel based on multi-mode fusion

Abstract

The invention discloses a high-precision real-time positioning system and method for positioning personnel based on multi-mode fusion. The system comprises a multi-source data acquisition module, an intelligent data processing module, a fusion positioning model construction module and a positioning result application module, provides visual monitoring, trajectory analysis and scene customization functions, and supports flexible expansion through modular design. According to the method, satellite positioning, Bluetooth beacon and inertial navigation technologies are integrated, the weight is dynamically adjusted in combination with an adaptive weighted fusion algorithm, and a fusion positioning model supports outdoor (satellite dominant + EKF calibration, the precision being 2-5 meters) and indoor (beacon dominant + fingerprint matching, the precision being lt); according to the invention, intelligent switching between three modes (inertial navigation + ZUPT correction) and signal loss (1 meter) is realized, data noise is optimized by using Kalman filtering, and high-precision real-time positioning in indoor and outdoor complex scenes is realized. According to the invention, the problems of insufficient precision, poor environmental adaptability and weak expansibility of a single positioning technology are solved, and the comprehensive performance of the positioning system is significantly improved.

Description

Technical Field

[0001] This invention relates to the field of intelligent positioning technology, specifically to a high-precision real-time positioning system and method for personnel based on multi-mode fusion positioning. Background Art

[0002] The development of personnel positioning systems is a complex engineering project integrating technologies from multiple fields, involving high-precision positioning technology, data processing algorithms, and other aspects. Current technical specifications and processes are not merely development constraints, but rather a complete system built through standardized design documents and rigorous testing procedures to ensure the reliability, stability, and high accuracy of the positioning system. An effective personnel positioning solution must possess three core elements: simplicity and efficiency in technical implementation, facilitating rapid system deployment and operation; emphasis on visualization in information presentation, using a graphical interface to intuitively display personnel dynamics; and a focus on accuracy in data processing to meet the stringent positioning precision requirements of different scenarios. High-precision real-time personnel positioning systems based on multi-mode fusion positioning, while possessing the technical characteristics of traditional positioning systems, exhibit distinct industry features.

[0003] Existing technologies have the following drawbacks: First, their positioning accuracy and adaptability are insufficient, and a single technology cannot meet the high-precision positioning needs of different environments. Second, their system scalability and compatibility are poor, and it is difficult for different positioning technologies to work together, making it difficult to adapt to the positioning needs of multiple scenarios and large-scale personnel. Third, there is an imbalance between cost and efficiency, with high-precision technologies having high deployment costs and low-cost technologies having poor positioning efficiency and accuracy. Fourth, their real-time performance and reliability are lacking, with data transmission and processing delays leading to untimely updates of location information and poor positioning stability in complex environments, failing to meet the requirements of emergency management scenarios.

[0004] Chinese patent document CN115639573A discloses a "multi-composite high-precision positioning fusion communication method," which uses two or more high-precision positioning methods to fuse tracking methods for real-time positioning of objects in space. The high-precision positioning methods employ complementary fusion of BeiDou differential positioning technology, Bluetooth AOA positioning technology, and multi-axis sensor positioning technology. The effective positioning outputs of the above methods are fused using a projection weight allocation method. A fusion formula is used to achieve the fusion of positioning data from multiple technologies. This method uses the displacement direction output by the multi-axis sensor as a standard, calculating the degree of overlap between the displacement direction output by other positioning methods and this standard. The higher the degree of overlap, the larger the corresponding weight allocation. This algorithm has low complexity and can achieve rapid data fusion and real-time positioning. After locating the object, the positioning data and attitude information are sent to a gateway via a long-distance radio ad hoc network using the terminal node attached to the object, achieving centralized management of the object's positioning and attitude information. However, the weight allocation method in the above technical solution is singular, lacks adaptability, and has limited positioning accuracy; furthermore, its real-time performance and reliability are insufficient. Summary of the Invention

[0005] This invention proposes a high-precision real-time positioning system and method for personnel based on multi-mode fusion positioning. Through the collaboration of multi-mode fusion architecture and intelligent algorithms, it achieves accurate, real-time, and efficient personnel positioning in complex scenarios. It solves the problems of insufficient accuracy, poor environmental adaptability, and weak scalability of single positioning technologies (such as GPS and Bluetooth) in complex environments (indoor, underground, and electromagnetic interference areas). It also overcomes the technical difficulties of insufficient real-time and reliability of existing positioning technologies, which cannot meet the needs of emergency management and other scenarios, and significantly improves the overall performance of the positioning system.

[0006] The invention also aims to integrate satellite positioning, Bluetooth beacon, and inertial navigation technologies, using an adaptive weighted fusion algorithm decision tree model to dynamically allocate the weights of each technology (based on environmental characteristics, signal quality, and historical data). The fusion positioning model supports intelligent switching between three modes: outdoor (satellite-dominated + EKF calibration, accuracy 2-5 meters), indoor (beacon-dominated + fingerprint matching, accuracy <1 meter), and signal loss (inertial navigation + ZUPT correction). Kalman filtering is used to optimize data noise, achieving high-precision real-time positioning in complex indoor and outdoor scenarios. This technology overcomes existing technological bottlenecks, reduces technological dependence, optimizes cost and efficiency, improves real-time performance and reliability, meets diverse personnel positioning needs across multiple scenarios, and enhances the overall performance of personnel positioning systems. It can be widely applied in various fields such as industrial production (e.g., personnel positioning management in factories and mines), security monitoring (e.g., oil depots, LNG facilities), smart venues (e.g., personnel navigation and management in large shopping malls and stadiums), and emergency rescue (e.g., personnel search and rescue in disaster scenarios such as fires and earthquakes), aiming to improve the accuracy, real-time performance, and reliability of personnel positioning, thereby enhancing management efficiency and safety in various industries.

[0007] To achieve the above objectives, this invention proposes a high-precision real-time personnel positioning system based on multi-mode fusion positioning, comprising: The multi-source data acquisition module collects data from various hardware devices in real time, enabling comprehensive data acquisition. The intelligent data processing module and the data cleaning and format unification unit perform preliminary data processing, and use an adaptive weighted fusion algorithm to perform in-depth processing of multi-source data and dynamically allocate the weights of each positioning technology. The integrated positioning model construction module builds an adaptive positioning model based on environmental characteristics to achieve accurate positioning. The positioning result application module provides real-time monitoring and trajectory query functions, and also provides open interface scenario customization applications; the system optimizes the weighted fused data through the Kalman filter algorithm, and outputs positioning results with an error of less than 1 meter.

[0008] Multi-source data acquisition is a fundamental component of the entire positioning system, aiming to obtain rich and accurate information related to personnel location. This invention utilizes multiple hardware devices working collaboratively, with each device leveraging its own strengths to achieve comprehensive data acquisition.

[0009] Preferably, the data collected by the multi-source data acquisition module covers key information such as the location coordinates, movement speed, and direction of travel of personnel, and is classified and stored according to the device type to provide a data source for subsequent data processing. The module includes: a satellite positioning unit, which adopts GPS or Beidou module, and the data acquisition frequency can be flexibly adjusted between 1-10Hz. Wireless beacon units, including Bluetooth or Wi-Fi beacons, are used for indoor assisted positioning; The inertial navigation unit measures human motion data using gyroscopes and accelerometers and initiates independent positioning when the signal is lost.

[0010] Satellite positioning modules (such as GPS / BeiDou), with their wide-area coverage, can provide approximate location information for people in open outdoor areas, providing a basic framework for positioning. Wireless beacons (such as RFID and WiFi) play an important auxiliary positioning role indoors or in areas with weak satellite signals. Their low power consumption, flexible deployment (Bluetooth positioning), or ability to achieve large-area coverage by relying on existing network infrastructure (Wi-Fi positioning) compensate for the shortcomings of satellite positioning in these areas. Inertial navigation devices become crucial in extreme situations such as signal loss. They use gyroscopes and accelerometers to measure the angular velocity and acceleration of a person's movement, and record the person's motion information through integration calculations to ensure the continuity of positioning. These hardware devices are stable and reliable, and the data acquisition frequency can be flexibly adjusted between 1-10Hz to meet the real-time requirements of different scenarios. The collected data covers key information such as the person's location coordinates, movement speed, and direction of travel, and is classified and stored according to device type, providing a rich and orderly data source for subsequent data processing.

[0011] The intelligent data processing module is the core of improving positioning accuracy. It uses innovative algorithms to perform in-depth processing on multi-source heterogeneous data, effectively eliminating noise errors and improving data quality.

[0012] Preferably, the intelligent data processing module includes a data cleaning and format unification unit and an adaptive weighted fusion algorithm unit. The data cleaning and format unification unit uses a data cleaning algorithm to accurately identify and remove abnormal data by setting reasonable thresholds and data change range rules, ensuring data accuracy. It also unifies the data format collected by the multi-source data acquisition module into a latitude and longitude combined with a timestamp, facilitating subsequent fusion processing and laying the foundation for the effective operation of the subsequent algorithm. The adaptive weighted fusion algorithm unit dynamically allocates the weights of each positioning technology based on environmental characteristics, signal quality, and historical data.

[0013] Preferably, the fusion positioning model construction module constructs an adaptive positioning model based on the results of the intelligent data processing module and combined with the characteristics of different environments to achieve accurate positioning; the adaptive positioning model includes an outdoor positioning model and an indoor positioning model, as well as a signal loss processing model; the signal loss processing model uses an inertial navigation unit and a zero-velocity correction algorithm to suppress error accumulation.

[0014] Preferably, the outdoor positioning model primarily uses satellite positioning data, combined with an extended Kalman filter (EKF) algorithm to calibrate inertial navigation data, achieving a positioning accuracy of 2–5 meters. When satellite signals are partially interfered with, inertial navigation and wireless beacon data serve as supplementary data, and the positioning results are calibrated using the EKF algorithm. The EKF algorithm transforms the nonlinear problem of fusing satellite positioning, inertial navigation, and wireless beacon data into a linear problem, ensuring the accuracy and stability of positioning in complex outdoor environments and effectively addressing situations such as satellite signal obstruction and interference.

[0015] Preferably, the indoor positioning model primarily uses wireless beacon data, combined with signal fingerprint matching technology, achieving a positioning accuracy of less than 1 meter. Simultaneously, it incorporates an indoor signal propagation model and signal fingerprint matching technology. By comparing real-time collected signal characteristics with a pre-stored signal fingerprint database, the positioning effect is further optimized, reducing errors caused by signal interference. For example, in indoor environments such as shopping malls, signal fingerprint matching technology can accurately determine the location of individuals within a store, meeting the requirements for high-precision indoor positioning.

[0016] When both satellite and Bluetooth signals are lost, the signal loss handling model initiates independent positioning for the inertial navigation device. To suppress the accumulation of inertial navigation errors, a Zero Velocity Correction (ZUPT) algorithm is employed. By monitoring accelerometer and gyroscope data, it determines whether the person is stationary; if so, the position and velocity of the inertial navigation system are corrected. During positioning, the system continuously monitors signal recovery. Once the signal is restored, it immediately switches back to multi-mode fusion positioning mode and uses the recovered signal to quickly calibrate the accumulated inertial navigation errors, ensuring continuous positioning and uninterrupted positioning even in complex environments.

[0017] Preferably, the positioning result application module transforms positioning data into intuitive and visual information to meet the application needs of different scenarios and improve management efficiency and security in various industries, including: The web-based visualization platform supports real-time display of personnel location on electronic maps and electronic fence alarm functions. Once personnel enter a preset danger zone (such as a high-risk work area in a factory or a restricted area in a venue), the system automatically triggers an alarm mechanism to promptly notify management personnel to take measures to ensure personnel safety and achieve real-time control of personnel location and hazard warning. The historical trajectory playback function is used for inspection assessment or accident analysis. Managers can analyze the trajectory to evaluate the work performance of inspection personnel, or after an accident, trace the movement path of personnel to find the cause of the accident and potential problems, which can help management decisions and safety improvements. The open interface supports secondary development, adapting to the customized needs of industrial plants and large venues; it provides customized interfaces and functions for different application scenarios; in factory scenarios, it focuses on monitoring personnel activities in high-risk areas and displays the distance between personnel and dangerous equipment in real time; in stadium scenarios, it optimizes personnel evacuation route planning according to the needs of events or activities, improving emergency response capabilities; the visualization platform supports access from multiple terminals such as PCs and mobile phones, and provides open interfaces to facilitate secondary development, meet the personalized needs of different users, and realize the flexible application of the positioning system.

[0018] This invention also proposes a high-precision real-time positioning method for personnel based on multi-mode fusion positioning, applied to the aforementioned high-precision real-time positioning system for personnel based on multi-mode fusion positioning, comprising the following steps: S1: The multi-source data acquisition module collects multi-source positioning data in real time; S2: Clean and standardize the format of the collected data, remove outliers and convert it into a latitude and longitude combined with timestamp format, and use an adaptive weighted fusion algorithm to dynamically allocate weights and fuse the data; S3: The Kalman filter algorithm optimizes and fuses the data, outputting positioning results with an error of less than 1 meter; S4: The fusion positioning model construction module switches to an adaptive positioning model based on environmental characteristics and updates personnel location in real time; S5: The visualization platform displays the location results and provides trajectory query and scene customization functions.

[0019] Preferably, the adaptive weighted fusion algorithm dynamically allocates weights through the following steps: Environmental characteristic analysis uses barometric and magnetic field sensors to determine whether the current environment is indoors or outdoors. Signal quality assessment, real-time monitoring of satellite signal signal-to-noise ratio and Bluetooth signal strength standard deviation; Historical data mining uses cluster analysis to determine the location distribution patterns of people in similar environments; Weight calculation and allocation: The weight values ​​of each positioning technology are output based on the decision tree algorithm.

[0020] Environmental feature analysis, utilizing data collected by sensors and other auxiliary equipment, combined with location data, and employing pre-built environmental classification models, accurately determines whether the current location environment is indoors or outdoors, and the complexity of the environment. For example, changes in air pressure measured by a barometric pressure sensor determine whether one is indoors, and a magnetic field sensor detects the presence of numerous metallic interfering objects indoors. This environmental feature information provides an important basis for weight allocation.

[0021] Signal quality assessment involves real-time monitoring of signal quality metrics for various positioning technologies, such as the signal-to-noise ratio (SNR) of satellite positioning signals and the standard deviation of Bluetooth signal strength (a measure of signal stability). Better signal quality results in higher weighting of the corresponding data during the fusion process, ensuring that more reliable data plays a greater role in positioning.

[0022] Historical data mining delves into historical location data to analyze the distribution patterns of people's locations under similar environmental and signal conditions. For example, cluster analysis can identify areas where people frequently appear in specific indoor environments, providing a reference for the weighting of current location data and making the weighting more consistent with reality.

[0023] Weight calculation and allocation take the aforementioned environmental characteristics, signal quality, and historical data as input, and use a decision tree algorithm from machine learning for training and analysis, outputting the weights of each positioning technology data. For example, in an indoor environment with a stable Bluetooth signal, the decision tree algorithm will increase the weight of Bluetooth data; in an open outdoor area with good satellite signal, the weight of satellite positioning data will be increased. By dynamically adjusting the weights in this way, the fused data is made closer to the true location, significantly improving positioning accuracy.

[0024] Kalman filtering optimization is performed on the weighted fused data, which still contains some noise. The classic Kalman filtering algorithm is then used to further optimize the data. Kalman filtering establishes system state equations and observation equations, iteratively predicting and correcting positioning data to effectively eliminate noise errors and output smoother, more accurate positioning data. This provides high-quality data support for subsequent positioning models and ensures the reliability of the positioning process.

[0025] Preferably, the specific implementation method of the dynamic weight allocation is as follows: In an indoor environment where the Bluetooth signal is stable, the decision tree algorithm increases the weight of Bluetooth data. Increase the weight of satellite positioning data when satellite signals are good in open outdoor environments; When both satellite and Bluetooth signals are lost, the system relies entirely on inertial navigation data and combines it with a zero-velocity correction algorithm to suppress errors.

[0026] Compared with the prior art, the beneficial effects achieved by the present invention are as follows: 1. Ultra-high precision positioning: Through innovative multi-mode fusion and advanced data processing algorithms, it achieves high-precision positioning both indoors and outdoors; for example, in complex indoor environments, it can achieve centimeter-level positioning accuracy, effectively solving the problem of insufficient accuracy of single technologies in complex environments, and meeting the stringent requirements of high-precision positioning in industries such as industrial and security. 2. Superior environmental adaptability: Multiple technologies work together intelligently to adapt to different scenarios; in areas with weak satellite signals, technologies such as Bluetooth and inertial navigation automatically enhance positioning capabilities, while in open outdoor areas, satellite positioning dominates the positioning process; it can adapt to complex and diverse environments such as industrial plants, large stadiums, and underground spaces. 3. Superior real-time performance and reliability: Real-time acquisition and processing of multi-source data, rapid updates of personnel location information; Multi-technology redundancy design combined with intelligent switching mechanism, greatly reducing the impact of single technical failures, ensuring reliable operation of the system in complex environments, and providing timely and accurate location data for emergency response and other scenarios; 4. High flexibility and scalability: The modular design concept runs through the entire system, allowing for flexible addition, removal, and replacement of positioning technology modules according to scenarios and needs; as new technologies develop and application scenarios expand, it is easy to integrate new positioning technologies or functional modules to meet ever-changing positioning requirements; 5. Optimize cost and efficiency: Balance positioning accuracy and cost through multi-technology fusion and intelligent algorithms; in areas where accuracy requirements are not high, use low-cost technologies such as Bluetooth, and in areas with high accuracy requirements, use a combination of high-precision positioning technologies; at the same time, efficient data fusion and processing algorithms improve positioning efficiency, reduce resource consumption, and achieve a good balance between performance and cost. Attached Figure Description

[0027] Figure 1 This is a block diagram of a high-precision real-time positioning system for personnel based on multi-mode fusion positioning, provided in Embodiment 1 of the present invention.

[0028] Figure 2 This is a flowchart of a high-precision real-time positioning method for personnel based on multi-mode fusion positioning provided in Embodiment 2 of the present invention. DETAILED DESCRIPTION

[0029] To make the objectives, technical solutions, and advantages of the present invention clearer, the technical solutions of the present invention will be further described in detail below through embodiments and in conjunction with the accompanying drawings. It should be understood that the specific embodiments described herein are only one preferred embodiment of the present invention and are only used to explain the technical solutions of the present invention. They do not limit the scope of protection of the present invention. All other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0030] First, let's explain the technical terms involved in this invention: Inertial navigation: a navigation technology that uses gyroscopes and accelerometers to measure the angular velocity and acceleration of a vehicle, and determines its position and attitude through integration calculations; Kalman filter algorithm: A mathematical algorithm that iteratively calculates the optimal estimate. In personnel positioning, it is used to fuse multi-source sensor data (such as GPS, inertial navigation, and beacons) to dynamically eliminate noise errors and output a smoother and more accurate position trajectory in real time. Multi-mode fusion positioning: It integrates multiple positioning technologies such as satellite positioning, Bluetooth positioning, and inertial navigation to leverage the advantages of each technology and improve positioning accuracy and environmental adaptability.

[0031] The technical solution of this invention revolves around achieving high-precision real-time positioning of personnel in complex scenarios, mainly comprising four key parts: multi-source data acquisition, intelligent data processing, fusion positioning model construction, and application of positioning results. The following will be explained in conjunction with the principle block diagram (…). Figure 1 ),flow chart( Figure 2 Each part is explained in detail.

[0032] Example 1: Figure 1 This paper demonstrates the core architecture of a high-precision real-time personnel positioning system based on multi-modal fusion positioning. The system consists of four main modules, which collaborate and interact with each other through layered mechanisms to achieve high-precision real-time positioning in complex scenarios. The main components of the system include: The multi-source data acquisition module collects data from various hardware devices in real time, enabling comprehensive data acquisition. The intelligent data processing module and the data cleaning and format unification unit perform preliminary data processing, and use an adaptive weighted fusion algorithm to perform in-depth processing of multi-source data and dynamically allocate the weights of each positioning technology. The integrated positioning model construction module builds an adaptive positioning model based on environmental characteristics to achieve accurate positioning. The positioning result application module provides real-time monitoring and trajectory query functions, and also provides open interface scenario customization applications; the system optimizes the weighted fused data through the Kalman filter algorithm, and outputs positioning results with an error of less than 1 meter.

[0033] The following provides a detailed explanation of the functions and collaborative relationships of each module and its submodules: 1) Multi-source data acquisition module This module forms the foundation of the entire positioning system, designed to acquire rich and accurate information related to personnel location. It is responsible for real-time data collection from various hardware devices, covering all indoor and outdoor scenarios. Its sub-modules include: The satellite positioning unit, using GPS or BeiDou modules, provides wide-area coverage positioning data in open outdoor areas, with a positioning accuracy of approximately 2-5 meters and a data acquisition frequency of 1-10Hz, supporting dynamic adjustment to adapt to different real-time requirements. Wireless beacon units, based on Bluetooth or Wi-Fi beacon technology, provide auxiliary positioning in indoor or satellite signal-weak areas (such as factories and underground spaces); Bluetooth beacons are flexible in deployment, while Wi-Fi beacons can rely on existing network infrastructure, and the combination of the two can achieve low-cost indoor positioning; The inertial navigation unit measures the angular velocity and acceleration of the person using gyroscopes and accelerometers, and calculates the trajectory through integration. When satellite and wireless signals are completely lost (such as in tunnels or enclosed spaces), the unit initiates independent positioning to ensure positioning continuity and suppresses error accumulation through the Zero Velocity Correction (ZUPT) algorithm.

[0034] The multi-source data acquisition module utilizes a variety of hardware devices working collaboratively, each leveraging its own strengths to achieve comprehensive data acquisition. These hardware devices are stable and reliable, with data acquisition frequencies flexibly adjustable between 1-10Hz to meet the real-time requirements of different scenarios. The acquired data covers key information such as personnel location coordinates, movement speed, and direction of travel, and is categorized and stored according to device type, providing a rich and organized data source for subsequent data processing. Its collaborative logic is as follows: the satellite positioning unit and the wireless beacon unit work complementaryly based on environmental characteristics, while the inertial navigation unit serves as a redundant backup. For example, in indoor scenarios, the wireless beacon unit dominates data acquisition, while the satellite positioning unit only provides auxiliary reference; when the signal is lost, the inertial navigation unit takes over the positioning task.

[0035] 2) Intelligent Data Processing Module This module is the core layer of the system, responsible for cleaning, fusing, and optimizing multi-source heterogeneous data, effectively eliminating noise and errors, and improving data quality. It includes the following sub-modules: The data cleaning unit is responsible for identifying and removing outliers from the collected data, which are often affected by various environmental factors. This unit uses threshold rules (such as coordinate jump range and signal strength standard deviation) to identify and remove outliers (such as GPS signal drift and Bluetooth signal fluctuation) to ensure the reliability of the input data. The format unification unit unifies heterogeneous formats such as latitude and longitude data from satellite positioning, RSSI (signal strength) data from Bluetooth beacons, and acceleration data from inertial navigation into a standardized "latitude and longitude + timestamp" format, providing a consistent foundation for subsequent fusion processing; The adaptive weighted fusion algorithm processing unit is the core of the technological innovation. It achieves optimal fusion of multi-source data through dynamic weight allocation.

[0036] The specific workflow of the above-mentioned adaptive weighted fusion algorithm processing unit includes: Environmental feature analysis, utilizing data collected by sensors and other auxiliary equipment combined with location data, employs a pre-built environmental classification model to accurately determine whether the current location environment is indoors or outdoors, and the complexity of the environment. For example, a barometric pressure sensor can be used to determine the indoor / outdoor environment, while a magnetic field sensor can detect metallic interference, providing a scenario-based basis for weight allocation. Signal quality assessment involves real-time monitoring of signal quality indicators for various positioning technologies, such as the signal-to-noise ratio (SNR) of satellite signals and the standard deviation of Bluetooth signal strength (a measure of signal stability). Higher signal quality results in greater weight for the corresponding data during the fusion process. Historical data mining delves into historical location data, analyzing the distribution patterns of people's locations under similar environmental and signal conditions to optimize current weight allocation strategies. For example, cluster analysis can be used to extract frequently occurring location areas in specific indoor environments, providing a reference for the weight allocation of current location data and making the weight allocation more consistent with reality. The weight calculation and allocation process takes the aforementioned environmental features, signal quality, and historical data as input, and uses a decision tree algorithm from machine learning for training and analysis, outputting the weights of each positioning technology data. Based on the decision tree algorithm, the weight values ​​of each data source are dynamically output (e.g., Bluetooth accounts for 70% and inertial navigation accounts for 30% in indoor scenes), and a Kalman filter algorithm is used to further eliminate noise, outputting high-precision positioning results with an error of less than 1 meter. Kalman filtering, by establishing system state equations and observation equations, iteratively predicts and corrects positioning data, effectively eliminating noise errors and outputting smoother, more accurate location data, providing high-quality data support for subsequent positioning models and ensuring the reliability of positioning.

[0037] 3) Fusion positioning model construction module This module constructs an adaptive positioning model based on the results of intelligent data processing and the characteristics of different environments to achieve accurate positioning. It includes three sub-models: Outdoor Positioning Model: In outdoor environments, satellite positioning data dominates. An Extended Kalman Filter (EKF) algorithm is used to calibrate inertial navigation data, addressing positioning errors caused by partial satellite signal obstruction, resulting in stable positioning accuracy of 2-5 meters. When satellite signals are partially interfered with, inertial navigation and radio beacon data serve as supplementary data, with the EKF algorithm used to calibrate the positioning results. The EKF algorithm transforms the nonlinear problem of fusing satellite positioning, inertial navigation, and radio beacon data into a linear problem, ensuring accuracy and stability in complex outdoor environments and effectively addressing satellite signal obstruction and interference. Indoor Positioning Model: In indoor environments, Bluetooth and other wireless beacon data become the primary positioning basis. Combined with sensor data and inertial navigation data, a weighted fusion algorithm improves positioning accuracy to within 1 meter. Simultaneously, an indoor signal propagation model and signal fingerprint matching technology are introduced. By comparing real-time collected signal characteristics with a pre-stored signal fingerprint database, the positioning effect is further optimized, reducing errors caused by signal interference. In indoor environments such as shopping malls, the combination of signal fingerprint matching technology further improves positioning accuracy to within 1 meter, meeting the needs of precise navigation at the store level. Signal Loss Handling Model: When both satellite and wireless signals are interrupted, the system relies entirely on the inertial navigation unit (INS), using the ZUPT algorithm to correct position and velocity errors in real time while stationary, ensuring short-term positioning continuity. By monitoring accelerometer and gyroscope data, the system determines whether the person is stationary; if so, it corrects the position and velocity of the inertial navigation system. During positioning, the system automatically triggers model switching based on real-time environmental assessments (such as sudden changes in air pressure or a sharp drop in signal strength). The system continuously monitors signal recovery; once the signal is restored, it immediately switches back to multi-mode fusion positioning and uses the recovered signal to quickly calibrate accumulated errors in the inertial navigation system, ensuring positioning continuity and uninterrupted positioning even in complex environments. For example, when a person moves from outdoors to indoors, the system seamlessly switches from satellite-dominated mode to beacon-dominated mode.

[0038] 4) Location Result Application Module This module transforms high-precision positioning data into visualization and scenario-based services to meet the application needs of different scenarios and improve management efficiency and security across industries. Specific functions include: Visualization Platform: A web-based electronic map displays personnel locations in real time, supporting multi-device access (PC, mobile phone), and intuitively presents personnel distribution and movement trends through heat maps, trajectory lines, and other formats; it supports querying and replaying historical personnel movement trajectories, providing strong evidence for inspection assessment and accident analysis; managers can analyze trajectories to evaluate the work performance of inspection personnel, or after an accident, trace personnel movement paths to identify the cause of the accident and potential problems, assisting in management decisions and safety improvements; Alarm Unit: Pre-set electronic fences (such as hazardous areas in factories and restricted areas in venues). Once personnel cross the boundary or linger for an extended period, the system automatically triggers an audible and visual alarm and pushes a notification to management personnel, improving safety management efficiency. If personnel enter a pre-set hazardous area (such as a high-risk work area in a factory or a restricted area in a venue), the system automatically triggers an alarm mechanism, promptly notifying management personnel to take measures to ensure personnel safety and achieve real-time control of personnel location and hazard warning. Customized application units for specific scenarios: These units provide open API interfaces and customized interfaces and functions for different application scenarios. They support secondary development by users according to different scenario requirements, meeting the personalized needs of different users and enabling flexible application of the positioning system. In factory scenarios, they focus on monitoring personnel activities in high-risk areas and displaying the distance between personnel and dangerous equipment in real time. In stadium scenarios, they optimize personnel evacuation route planning according to the needs of events or activities, improving emergency response capabilities.

[0039] The high-precision real-time positioning system for personnel based on multi-mode fusion positioning provided by this invention offers a standardized solution for personnel positioning in multiple scenarios through multi-mode fusion architecture and intelligent algorithm optimization, hierarchical architecture and intelligent collaboration mechanism. It significantly improves the overall performance and application value of the positioning system and achieves the following breakthroughs: High precision, with indoor and outdoor positioning errors of less than 1 meter and 5 meters respectively, meeting the demanding requirements of industrial, security and other scenarios; It is highly adaptable, dynamically switching positioning models to cover extreme scenarios such as satellite signal blind spots and complex electromagnetic environments; High reliability, redundant design and ZUPT algorithm ensure that positioning is not interrupted when the signal is interrupted, supporting time-sensitive tasks such as emergency rescue; Low-cost expansion and modular design allow for the addition or removal of hardware devices as needed, balancing performance and cost.

[0040] Example 2: Figure 2 This paper demonstrates the complete workflow of a high-precision real-time personnel positioning method based on multi-modal fusion positioning, covering the entire chain of operations from multi-source data acquisition to final positioning result output. The flowchart, through a hierarchical and progressive approach and dynamic feedback mechanism, achieves accurate positioning and intelligent decision-making in complex scenarios. The steps include: S1: The multi-source data acquisition module collects multi-source positioning data in real time; S2: Clean and standardize the format of the collected data, remove outliers and convert it into a latitude and longitude combined with timestamp format, and use an adaptive weighted fusion algorithm to dynamically allocate weights and fuse the data; S3: The Kalman filter algorithm optimizes and fuses the data, outputting positioning results with an error of less than 1 meter; S4: The fusion positioning model construction module switches to an adaptive positioning model based on environmental characteristics and updates personnel location in real time; S5: The visualization platform displays the location results and provides trajectory query and scene customization functions.

[0041] The following is a detailed explanation of each stage in the process sequence: After system initialization, the system enters the "multi-source data acquisition" phase, where it collaboratively collects three types of data through multiple devices: Satellite positioning data, using GPS or BeiDou modules to obtain latitude and longitude information of personnel, is suitable for open outdoor areas, with a data acquisition frequency of 1-10Hz; Wireless beacon data, received via Bluetooth or Wi-Fi beacon, including signal strength (RSSI) and time of arrival (ToA), is used for indoor positioning; Inertial navigation data is obtained by measuring the angular velocity and acceleration of the person using gyroscopes and accelerometers, and then calculating the trajectory through integration.

[0042] In the process of equipment division of labor and collaboration, the system dynamically allocates primary and secondary positioning devices according to the current environment; for example, satellite positioning is the primary method in outdoor scenarios, while wireless beacons are used in indoor scenarios, and inertial navigation is relied upon when the signal is lost.

[0043] After data acquisition is completed, the system enters the data availability judgment node: if the data is complete and the signal strength meets the preset threshold (e.g., satellite signal-to-noise ratio > 20dB, Bluetooth signal strength > -70dBm), it is marked as "available" and enters the "classification and storage" step, storing it to the database according to device type and time sequence; if the data is missing or the signal is abnormal (e.g., satellite signal is blocked, Bluetooth beacon is offline), the system triggers an alarm and attempts to re-acquire data or switch to a backup device.

[0044] After data storage, the system performs anomaly checks on the collected data: Abnormal data is identified by setting dynamic threshold rules, such as coordinate jump detection (if the difference in latitude and longitude between two consecutive frames exceeds 10 meters (outdoors) or 1 meter (indoors), it is determined to be GPS drift or signal interference); signal strength fluctuation detection (if the standard deviation of Bluetooth signal strength exceeds 5 dBm, it is considered environmental interference). After abnormal data is removed, the remaining data enters the "data cleaning" step, where interpolation algorithms (such as linear interpolation) are used to fill in missing values ​​to ensure data continuity. The cleaned heterogeneous data (such as satellite latitude and longitude, Bluetooth RSSI, and inertial navigation acceleration) is converted into a standardized "latitude and longitude + timestamp" unified format for easy subsequent fusion processing.

[0045] After data standardization, the process enters the adaptive weighted fusion algorithm weight allocation stage, with the core steps as follows: Environmental characteristic analysis utilizes barometric pressure sensors to determine indoor and outdoor environments (indoor barometric pressure fluctuations are relatively small), combined with magnetic field sensors to detect interference from metal structures (such as the impact of factory equipment on Bluetooth signals); for example, if the rate of change of barometric pressure is <0.1 hPa / s and the magnetic field strength is >50 μT, it is determined to be a complex indoor environment; Signal quality assessment involves real-time monitoring of satellite signal-to-noise ratio (SNR) and Bluetooth signal strength standard deviation to quantify signal reliability. If the satellite SNR > 25 dB and the Bluetooth standard deviation < 3 dBm, the signal quality is rated as "excellent". Historical data mining extracts patterns in the location distribution of people in similar scenarios through cluster analysis (such as the K-means algorithm). For example, in a shopping mall, people often congregate at entrances / exits or in popular store areas. If the current environment matches historical scenarios by more than 80%, historical weighting strategies are prioritized. Weight calculation and allocation are based on a decision tree algorithm that dynamically outputs weight values. One specific implementation method is as follows: Outdoor open environment: satellite weighting 70%, inertial navigation 30%; In complex indoor environments: Bluetooth weighting 60%, inertial navigation 40%; When the signal is lost: the inertial navigation weight is 100%, and the error is corrected by combining the ZUPT algorithm.

[0046] After weight allocation, the process enters the "Kalman filter optimization" step: the Kalman filter iteratively predicts and corrects the position through the state equation (motion model) and the observation equation (fusion data) to eliminate random noise; for example, when people are moving quickly, the filter prioritizes trusting inertial navigation data; when stationary, it relies on beacon data.

[0047] The system constructs a positioning model based on the environmental assessment results ("current environmental assessment") and dynamically switches between the following models: an outdoor positioning model, primarily based on satellite data, combined with an extended Kalman filter (EKF) algorithm to calibrate inertial navigation errors. When satellite signals are partially blocked (e.g., between tall buildings), the EKF linearly fuses satellite and inertial navigation data, stabilizing the positioning accuracy at 2-5 meters. The indoor positioning model uses Bluetooth beacon data as its core, combined with signal fingerprint matching technology. The system pre-stores signal feature databases for each area (e.g., the RSSI distribution of shop A1 in a shopping mall), compares the current signal fingerprint in real time, and achieves positioning accuracy within 1 meter. The signal loss handling model relies entirely on the inertial navigation unit, suppressing error accumulation through "inertial navigation + ZUPT correction." The ZUPT algorithm monitors accelerometer data in real time; if a stationary state is detected (acceleration standard deviation < 0.1 m / s²), the system will detect the signal loss. 2 If the speed error is reset, the position offset will be corrected.

[0048] The optimized location data enters the "Update Location Information" step, and then outputs the application results through the following functional modules: the electronic map displays the real-time location, the web-based visualization platform dynamically marks the location of people on the electronic map, and supports 2D / 3D view switching; the heat map function can display densely populated areas to help managers optimize resource allocation. Track query and analysis: The system stores historical track data and supports multi-dimensional queries by time, region, and personnel ID; the track playback function can be used for accident debriefing, such as analyzing missed areas during factory inspections. The system triggers an alarm mechanism with preset electronic fences (such as hazardous areas in chemical plants or restricted areas in stadiums). If personnel cross the boundary or stay for an extended period, the system triggers an audible and visual alarm and pushes a notification to the management terminal. The alarm strategy is customizable, such as increasing sensitivity by 20% in night mode. The system also provides a scene-customizable interface with a RESTful API interface to support integration with third-party systems. For example, in a mining scenario, a gas concentration sensor can be connected to achieve "location + environmental safety" linked monitoring.

[0049] The above process incorporates multiple feedback loops to ensure continuous system optimization: In the process of adjusting the data acquisition frequency, if the judgment of "whether historical data exists" is "no" (such as in a newly deployed area), the system will automatically increase the data acquisition frequency (such as from 1Hz to 5Hz) to accelerate the construction of the initial database. In the adaptive update stage of model parameters, the algorithm parameters are dynamically adjusted according to the real-time positioning error (such as Kalman filter residuals); for example, if the residuals continue to increase, the satellite weights are reduced and the proportion of inertial navigation is increased. In the user behavior learning phase, machine learning is used to analyze personnel movement patterns (such as inspection route preferences) to optimize trajectory prediction algorithms and improve positioning response speed.

[0050] The process ends with an "End" node, but in actual operation, the system continuously executes in a loop. Abnormal situation handling mechanisms include: communication interruption recovery and hardware fault redundancy; if wireless beacon communication is interrupted, the system switches to local caching mode, temporarily storing data for batch uploading after recovery; in case of satellite module failure, the backup BeiDou module is automatically activated; if the Bluetooth beacon is damaged, it switches to Wi-Fi positioning.

[0051] Figure 2 Through meticulous step design and a closed-loop feedback mechanism, the flowchart enables the efficient operation of a multi-mode fusion positioning system. Its core value lies in: full-scene coverage, seamlessly connecting complex indoor and outdoor environments from data collection to result application; intelligent decision-making, with dynamic weight allocation and model switching improving positioning accuracy and reliability; and flexible expansion, with modular design supporting rapid integration of new technologies to adapt to future positioning needs. This process not only provides standardized solutions for industries such as industry, security, and emergency rescue, but also lays a solid foundation for subsequent algorithm optimization and hardware upgrades.

[0052] In summary, this invention discloses a high-precision real-time positioning system and method for personnel based on multi-mode fusion positioning, belonging to the field of intelligent positioning technology. The system includes a multi-source data acquisition module, an intelligent data processing module, a fusion positioning model construction module, and a positioning result application module, providing visualization monitoring, trajectory analysis, and scene customization functions. Its modular design supports flexible expansion. The method integrates satellite positioning, Bluetooth beacon, and inertial navigation technologies, dynamically allocating the weights of each technology (based on environmental characteristics, signal quality, and historical data) using an adaptive weighted fusion algorithm, and optimizing data noise using Kalman filtering to achieve high-precision real-time positioning in complex indoor and outdoor scenarios. Specifically, the adaptive weighted fusion algorithm dynamically adjusts the weights through a decision tree model, and the fusion positioning model supports intelligent switching between three modes: outdoor (satellite-dominated + EKF calibration, accuracy 2-5 meters), indoor (beacon-dominated + fingerprint matching, accuracy <1 meter), and signal loss (inertial navigation + ZUPT correction). This invention achieves accurate, real-time, and efficient personnel positioning in complex scenarios through multi-mode fusion and intelligent algorithm collaboration. It solves the problems of insufficient accuracy, poor environmental adaptability, and weak scalability of single positioning technologies, significantly improving the overall performance of the positioning system and demonstrating significant technological innovation and application value.

[0053] The above description is only a preferred embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. It should be noted that for those skilled in the art, any changes or substitutions that can be easily conceived without departing from the technical principles of the present invention should be included within the scope of protection of the present invention.

Claims

1. A high-precision real-time positioning system for personnel based on multi-mode fusion positioning, characterized in that, include: The multi-source data acquisition module collects data from various hardware devices in real time, enabling comprehensive data acquisition. The intelligent data processing module and the data cleaning and format unification unit perform preliminary data processing, and use an adaptive weighted fusion algorithm to perform in-depth processing of multi-source data and dynamically allocate the weights of each positioning technology. The integrated positioning model construction module builds an adaptive positioning model based on environmental characteristics to achieve accurate positioning. The location results application module provides real-time monitoring and trajectory query functions, and also provides open interface scenarios for customized applications; The system optimizes the weighted and fused data using a Kalman filter algorithm, outputting a positioning result with an error of less than 1 meter.

2. The high-precision real-time positioning system for personnel based on multi-mode fusion positioning according to claim 1, characterized in that, The multi-source data acquisition module collects key information including personnel location coordinates, movement speed, and direction of travel, and stores it according to device type to provide a data source for subsequent data processing. The module includes: The satellite positioning unit uses a GPS or BeiDou module, and the data acquisition frequency can be flexibly adjusted between 1-10Hz. Wireless beacon units, including Bluetooth or Wi-Fi beacons, are used for indoor assisted positioning; The inertial navigation unit measures human motion data using gyroscopes and accelerometers, and initiates independent positioning when the signal is lost.

3. A high-precision real-time positioning system for personnel based on multi-mode fusion positioning according to claim 1 or 2, characterized in that, The intelligent data processing module includes a data cleaning and format unification unit and an adaptive weighted fusion algorithm unit. The data cleaning and format unification unit uses a data cleaning algorithm to accurately identify and remove abnormal data by setting reasonable thresholds and data change range rules, and unifies the data format collected by the multi-source data acquisition module into a latitude and longitude combined with timestamp. The adaptive weighted fusion algorithm unit dynamically allocates the weights of each positioning technology based on environmental characteristics, signal quality and historical data.

4. A high-precision real-time positioning system for personnel based on multi-mode fusion positioning according to claim 1, characterized in that, The fusion positioning model construction module constructs an adaptive positioning model based on the results of the intelligent data processing module and combined with the characteristics of different environments to achieve accurate positioning. The adaptive positioning model includes an outdoor positioning model and an indoor positioning model, as well as a signal loss processing model. The signal loss processing model uses an inertial navigation unit and a zero-speed correction algorithm to suppress error accumulation.

5. A high-precision real-time positioning system for personnel based on multi-mode fusion positioning according to claim 4, characterized in that, The outdoor positioning model primarily uses satellite positioning data, combined with an extended Kalman filter (EKF) algorithm to calibrate inertial navigation data, achieving a positioning accuracy of 2–5 meters. When satellite signals are partially interfered with, inertial navigation and wireless beacon data serve as supplementary data, and the positioning results are calibrated using the EKF algorithm. The EKF algorithm transforms the nonlinear problem of fusing satellite positioning, inertial navigation, and wireless beacon data into a linear problem for processing.

6. A high-precision real-time positioning system for personnel based on multi-mode fusion positioning according to claim 4, characterized in that, The indoor positioning model primarily uses wireless beacon data, combined with signal fingerprint matching technology, achieving a positioning accuracy of less than 1 meter. Simultaneously, it incorporates an indoor signal propagation model and signal fingerprint matching technology, further optimizing the positioning effect and reducing signal interference errors by comparing real-time collected signal characteristics with a pre-stored signal fingerprint database.

7. A high-precision real-time positioning system for personnel based on multi-mode fusion positioning according to claim 1, characterized in that, The location result application module transforms the location data into intuitive and visual information, including: A web-based visualization platform that supports real-time display of personnel locations on electronic maps and electronic fence alarm functions; The historical trajectory playback function is used for inspection assessment or accident analysis; The open interface supports secondary development and can be adapted to the customized needs of industrial plants and large venues.

8. A high-precision real-time positioning method for personnel based on multi-mode fusion positioning, applied to the high-precision real-time positioning system for personnel based on multi-mode fusion positioning as described in any one of claims 1 to 7, characterized in that, The following steps are involved: S1: The multi-source data acquisition module collects multi-source positioning data in real time; S2: Clean and standardize the format of the collected data, remove outliers and convert it into a latitude and longitude combined with timestamp format, and use an adaptive weighted fusion algorithm to dynamically allocate weights and fuse the data; S3: The Kalman filter algorithm optimizes and fuses the data, outputting positioning results with an error of less than 1 meter; S4: The fusion positioning model construction module switches to an adaptive positioning model based on environmental characteristics and updates personnel location in real time; S5: The visualization platform displays the location results and provides trajectory query and scene customization functions.

9. A high-precision real-time positioning method for personnel based on multi-mode fusion positioning according to claim 8, characterized in that, The adaptive weighted fusion algorithm dynamically allocates weights through the following steps: Environmental characteristic analysis uses barometric and magnetic field sensors to determine whether the current environment is indoors or outdoors. Signal quality assessment, real-time monitoring of satellite signal signal-to-noise ratio and Bluetooth signal strength standard deviation; Historical data mining uses cluster analysis to determine the location distribution patterns of people in similar environments; Weight calculation and allocation: The weight values ​​of each positioning technology are output based on the decision tree algorithm.

10. A high-precision real-time positioning method for personnel based on multi-mode fusion positioning according to claim 8 or 9, characterized in that, The specific implementation method of the dynamic weight allocation is as follows: In an indoor environment where the Bluetooth signal is stable, the decision tree algorithm increases the weight of Bluetooth data. Increase the weight of satellite positioning data when satellite signals are good in open outdoor environments; When both satellite and Bluetooth signals are lost, the system relies entirely on inertial navigation data and combines it with a zero-velocity correction algorithm to suppress errors.

Images