A blind museum precise positioning and obstacle avoidance guide system based on UWB
The blind tour guide system, which integrates UWB base station networks and multiple sensors, combined with improved A* algorithms and multimodal human-computer interaction, solves the shortcomings of traditional blind tour guide systems in terms of navigation accuracy and environmental perception. It achieves centimeter-level precise navigation and personalized services, enhancing the autonomy and safety of blind users in museums.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- ZHEJIANG UNIV OF SCI & TECH
- Filing Date
- 2026-03-06
- Publication Date
- 2026-06-12
AI Technical Summary
Traditional blind navigation systems suffer from low navigation accuracy, weak environmental perception, inability to avoid obstacles in real time, and insufficient personalized services. Furthermore, existing indoor positioning technologies are insufficient to meet centimeter-level accuracy requirements, and cannot effectively address systemic challenges in real-world scenarios such as multi-user concurrency, group collaboration, and emergency assistance.
It adopts UWB base station network and multi-sensor fusion, combined with improved A* algorithm and multimodal human-computer interaction, to achieve centimeter-level positioning and obstacle avoidance. It integrates UWB module, inertial measurement unit, millimeter-wave radar and multimodal human-computer interaction device, supports personalized path planning, group collaboration and emergency assistance, and has environmental self-learning and map calibration functions.
It enables blind users to navigate museums with centimeter-level precision, provides personalized and safe guided tours, enhances user autonomy and experience, supports stable positioning and emergency assistance in multi-user scenarios, and ensures the reliability and maintainability of the system.
Smart Images
Figure CN122192303A_ABST
Abstract
Description
Technical Field
[0001] This invention relates to the field of intelligent navigation technology, specifically to a UWB-based precise positioning and obstacle avoidance navigation system for blind museums. Background Technology
[0002] Intelligent navigation refers to a systematic technology that comprehensively utilizes multi-source sensor data, high-precision positioning technology, real-time environmental perception and cognitive algorithms, combined with users' personalized needs and behavioral patterns, to dynamically plan and guide users to reach their destination from the starting point along the optimal or safest path.
[0003] Traditional guided tours for the blind mainly rely on audio guides, guide dogs, or human companions. These methods suffer from limitations such as low navigation accuracy, weak environmental perception, inability to avoid obstacles in real time, and insufficient personalized services. In recent years, indoor positioning technologies such as Bluetooth, Wi-Fi, and RFID have been attempted for guided tours, but they generally suffer from problems such as low positioning accuracy, susceptibility to signal interference, and severe multipath effects. These limitations make it difficult to meet the centimeter-level accuracy safety navigation needs of the blind in complex indoor environments. At the same time, existing guided tour systems mostly focus on one-way information delivery and lack the ability to learn and adapt to user behavior. They also fail to effectively address the systemic challenges in real-world scenarios such as multi-user concurrency, group collaboration, and emergency assistance. Summary of the Invention
[0004] The purpose of this invention is to provide a UWB-based system for precise positioning and obstacle avoidance in museums for the blind, in order to solve the problems mentioned in the background art.
[0005] To solve the above-mentioned technical problems, the present invention provides the following technical solution: a UWB-based accurate positioning and obstacle avoidance guide system for a museum for the blind, comprising: a cloud management platform, a UWB positioning base station network, and a mobile guide terminal; The cloud management platform stores a high-precision 3D map containing obstacle information and a path planning engine. The UWB positioning base station network consists of multiple UWB anchor points deployed in spatial density, used to interact with the terminal to perform positioning and ranging. The mobile terminal integrates a UWB module, an inertial measurement unit, a millimeter-wave radar, and a multi-mode human-computer interaction device. The UWB module is used to achieve centimeter-level positioning; the inertial measurement unit is used to maintain short-term positioning when the signal is blocked; the millimeter-wave radar is used to detect moving obstacles; the multi-mode human-computer interaction device includes a wristband-type haptic feedback unit that provides directional encoded vibration, bone conduction headphones that play navigation voice, and a smart guide cane with integrated proximity sensors.
[0006] Preferably, the mobile navigation terminal also integrates a lightweight visual processing module. This module extracts edge contour features from the images captured by the terminal camera and matches the extracted contours with local environmental feature templates issued by the cloud management platform. This is used to assist in identifying large glass or mirror-like fixed obstacles that are difficult to reflect effectively by UWB signals. The identification results are used as supplementary information to be input into the extended Kalman filter algorithm and obstacle avoidance decision module.
[0007] Preferably, the wristband-type haptic feedback unit in the multi-mode human-computer interaction device adopts a ring array vibration motor, which can encode and generate various haptic commands for indicating forward movement, left turn, right turn, stop, and fine-tuning direction by controlling the vibration timing and mode of motors at different positions; the handle of the smart guide cane has a built-in haptic warning device, which will generate a vibration with an intensity inversely proportional to the distance to the obstacle when the near-field emergency layer senses an obstacle; the navigation voice and exhibit explanation information broadcast by the bone conduction headphones are processed by directional sound field to create a spatial sense of virtual sound source to assist users in judging direction.
[0008] Preferably, the path planning algorithm engine of the cloud management platform adopts an improved A* algorithm framework. When calculating the path cost, it not only considers the geometric distance, but also introduces the real-time dynamic obstacle prediction cost, the user's historical walking preference cost, and the path turning frequency cost. The algorithm outputs a path with width generated on the basis of a safe area, and provides guidance and correction through gradually increasing tactile or voice prompts when the user deviates from the center line of the output path.
[0009] Preferably, the system has context-adaptive capabilities. The mobile navigation terminal has a built-in user behavior learning module that can continuously record and analyze the user's average walking speed, turning radius, and response sensitivity parameters to various prompts, and form a personalized user model. The personalized user model is uploaded to a cloud management platform for personalized adjustment of path planning strategies, advance warning prompts, and the intensity and frequency of various feedbacks in the multi-modal human-computer interaction device.
[0010] Preferably, the UWB positioning base station network adopts frequency division multiplexing and dynamic anchor point selection mechanism to cope with multi-user scenarios; each mobile navigation terminal's UWB communication module is assigned a specific working frequency band or time slot; the system monitors the signal quality of each UWB anchor point in real time, and dynamically selects the three to four anchor points with the strongest signals for each terminal to form the optimal positioning set, which is used to perform high-precision fusion positioning calculation, thereby reducing signal interference between multiple terminals and optimizing network load.
[0011] Preferably, the system includes a group guided tour collaborative mode. When multiple mobile guided tour terminals are set up for the same group, the cloud management platform plans a common tour route for the entire group and calculates the relative positions of each member in real time. The system provides members with prompts to maintain formation through the human-computer interaction devices of each terminal. When a member deviates from the predetermined formation or the safe distance is too large, a warning message is sent to the member and the leader's terminal at the same time.
[0012] Preferably, the mobile guide terminal is equipped with a one-click help module. When triggered, the terminal will immediately send a help data packet containing its precise real-time location, current sensor status, and environmental recordings from a few seconds before and after to the cloud management platform via a wireless network. The cloud management platform will push the help information and the terminal's real-time movement trajectory to the staff monitoring terminal and establish a voice call connection between the staff terminal and the help terminal.
[0013] Preferably, the system supports self-learning of the museum environment and map calibration. After initial deployment or exhibition changes, staff can carry mobile guide terminals in learning mode along a preset route. The system automatically collects UWB base station signal characteristics, inertial data, and key landmarks identified by other sensors, and combines them with architectural drawings to semi-automatically generate or update the high-precision 3D map. During daily operation, the system continuously monitors the stability of fixed anchor point signals and automatically calibrates for minor offsets caused by environmental changes.
[0014] Compared with the prior art, the beneficial effects achieved by the present invention are: First, this invention achieves centimeter-level anti-interference positioning by fusing UWB base station networks with multiple sensors, overcoming the challenges of signal obstruction and special reflective surfaces. The improved A* algorithm incorporates dynamic obstacle prediction and user preferences to plan personalized safety paths. Furthermore, it utilizes multimodal interactions such as tactile coding, proportional vibration warnings, and directional sound fields to provide precise guidance and risk warnings in an intuitive and private manner, significantly enhancing the autonomy and experience of blind users walking independently and safely in museums.
[0015] Secondly, this invention achieves end-to-end personalized adaptive optimization through a user behavior learning module, making the service fit individual needs. The environmental self-learning and map calibration functions support rapid deployment and updates, daily automatic calibration ensures long-term reliability, group collaborative guidance ensures orderly and safe group activities, and one-click help establishes a rapid emergency channel. These features make the system outstanding in terms of availability, maintainability, and handling of complex scenarios such as multiple users and environmental changes, providing a complete and sustainable barrier-free guidance solution. Attached Figure Description
[0016] Figure 1 This is a system architecture diagram of the present invention; Figure 2This is a flowchart of the system operation of the present invention. Detailed Implementation
[0017] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments 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.
[0018] This invention provides the following technical solutions: Example Please see Figure 1 and Figure 2 A UWB-based precision positioning and obstacle avoidance navigation system for museums for the blind includes: a cloud management platform, a UWB positioning base station network, and a mobile navigation terminal. The cloud management platform stores a high-precision 3D map containing obstacle information and a path planning engine; The UWB positioning base station network consists of multiple UWB anchor points deployed in spatial density, used to interact with terminals to perform positioning and ranging. The mobile terminal integrates a UWB module, an inertial measurement unit, a millimeter-wave radar, and a multi-mode human-computer interaction device. The UWB module is used to achieve centimeter-level positioning; the inertial measurement unit is used to maintain short-term positioning when the signal is blocked; the millimeter-wave radar is used to detect moving obstacles; the multi-mode human-computer interaction device includes a wristband-type haptic feedback unit that provides directional encoded vibration, bone conduction headphones that play navigation voice, and a smart guide cane with integrated proximity sensors.
[0019] The mobile navigation terminal also integrates a lightweight visual processing module. This module extracts edge contour features from the images captured by the terminal's camera and matches the extracted contours with local environmental feature templates sent by the cloud management platform. This helps to identify large glass or mirror-like fixed obstacles that are difficult to reflect effectively by UWB signals. The identification results are used as supplementary information to input the extended Kalman filter algorithm and obstacle avoidance decision module.
[0020] Among them, the Extended Kalman Filter (EKF) algorithm is an optimal recursive estimation algorithm suitable for nonlinear system state estimation. It estimates the system's state variables with minimum mean square error through two steps: prediction and update. In this application, the algorithm is applied to the multi-sensor data fusion positioning process of a mobile navigation terminal: centimeter-level ranging data provided by the UWB module is used as the main observation value, and acceleration and angular velocity data output by the inertial measurement unit are used as the basis for system state prediction. At the same time, the contour features and position information of special obstacles such as glass and mirrors identified by the lightweight vision processing module are used as supplementary observation values. By establishing the system state equation and observation equation, the algorithm fuses the data from these three types of heterogeneous sensors in real time, effectively suppressing the error and blind zone of a single sensor. Even when the UWB signal is blocked or encounters a special reflective surface, it can still maintain stable and high-precision position and attitude estimation, and output the optimized state estimate to the obstacle avoidance decision module, providing a reliable position basis for safe navigation.
[0021] The wristband-type haptic feedback unit in the multi-mode human-computer interaction device uses a ring array vibration motor, which can encode and generate various haptic commands for indicating forward movement, left turn, right turn, stop, and fine-tuning direction by controlling the vibration timing and mode of motors in different positions. The handle of the smart guide cane has a built-in haptic warning device. When the near-field emergency layer senses an obstacle, the warning device will generate a vibration with an intensity inversely proportional to the distance from the obstacle. The navigation voice and exhibit explanation information broadcast by the bone conduction headphones are processed by directional sound field to create a sense of spatial orientation of virtual sound sources to assist users in judging direction.
[0022] By encoding multidimensional tactile commands through a ring-array vibration motor, the system provides blind users with intuitive, private, and noise-free directional guidance, effectively overcoming the shortcomings of traditional voice prompts that are easily masked in noisy environments. Secondly, the smart guide cane's vibration warnings, inversely proportional to the distance to obstacles, provide tiered alerts for emergencies, allowing users to intuitively perceive the approach of danger through vibration intensity and thus react promptly and accurately. Finally, the bone conduction headphones' directional sound field processing creates a spatial sense of virtual sound sources at the auditory level, forming sensory synergy with the wristband's tactile commands, enhancing the user's directional perception and spatial awareness. This multimodal interaction mode improves the reliability, timeliness, and intuitiveness of navigation information transmission, reducing the cognitive load and walking risks for blind users in complex museum environments.
[0023] The cloud management platform's path planning algorithm engine adopts an improved A* algorithm framework. When calculating path costs, it not only considers geometric distance but also incorporates real-time dynamic obstacle prediction costs, user historical walking preference costs, and path turning frequency costs. The algorithm outputs a path with width generated based on a safe area and provides guided correction through increasingly stronger tactile or voice prompts when the user deviates from the center line of the output path.
[0024] The A* algorithm is a heuristic search algorithm for finding the shortest path from the starting point to the destination in a static environment. It guides the search direction through an evaluation function f(n) = g(n) + h(n), where g(n) represents the actual cost from the starting point to the current node n, and h(n) is the estimated cost from the current node to the target node. In this application, the algorithm framework has been significantly improved based on the classic A* algorithm. The calculation of the path cost function g(n) includes not only the basic geometric distance cost, but also the real-time dynamic obstacle prediction cost, the user's historical walking preference cost, and the path turning frequency cost. At the same time, the algorithm output is not a traditional single-line path, but a path corridor with a reasonable width is generated within the safe passage area of a high-precision 3D map. When the user detects a deviation from the center line of this corridor through fusion positioning, the system will provide non-intrusive guidance and correction through gradual changes in wristband vibration intensity or voice prompt frequency. This ensures the global path optimality while taking into account the flexibility of real-time obstacle avoidance, the personalization of user habits, and the safety and comfort of blind people walking.
[0025] The system has the ability to adapt to different situations. The mobile navigation terminal has a built-in user behavior learning module that can continuously record and analyze the user's average walking speed, turning range, and response sensitivity parameters to various prompts, and form a personalized user model. The personalized user model is uploaded to the cloud management platform for personalized adjustment of path planning strategies, advance warning prompts, and the intensity and frequency of various feedbacks in the multi-modal human-computer interaction device.
[0026] By continuously learning and quantifying users' individual behavioral characteristics such as walking speed, turning habits, and responsiveness to prompts, the system can build accurate personalized user models and dynamically optimize core navigation strategies accordingly. The system can prioritize routes that match the user's walking rhythm and turning preferences, improving walking smoothness and comfort. In terms of warning prompts, the system can personalize the warning lead time based on the user's reaction speed, avoiding interference caused by premature warnings and preventing risks caused by late warnings. In terms of human-computer interaction, the system can adaptively adjust the intensity of tactile feedback and the frequency of voice prompts to match the user's sensory sensitivity, thereby significantly reducing cognitive load, improving navigation efficiency and user experience. Ultimately, the system can truly adapt to the unique needs and behavioral patterns of different blind users, providing considerate, natural, and efficient personalized navigation services.
[0027] The UWB positioning base station network adopts frequency division multiplexing and dynamic anchor point selection mechanisms to cope with multi-user scenarios; each mobile navigation terminal's UWB communication module is assigned a specific working frequency band or time slot; the system monitors the signal quality of each UWB anchor point in real time, and dynamically selects the three to four anchor points with the strongest signals for each terminal to form the optimal positioning set, which is used to perform high-precision fusion positioning calculations, thereby reducing signal interference between multiple terminals and optimizing network load.
[0028] By allocating independent operating frequency bands or time slots to each mobile terminal, co-channel interference between multiple UWB signals is fundamentally avoided, ensuring the independence and accuracy of ranging data from each terminal. Real-time signal quality monitoring and dynamic anchor point selection mechanisms enable each terminal to consistently perform positioning calculations based on the three to four anchor points with the strongest signals in the current environment. This effectively overcomes the problem of decreased positioning accuracy caused by people moving around, objects obstructing the view, or local signal attenuation, ensuring the continuous reliability of centimeter-level positioning services. Simultaneously, this mechanism, through intelligent network load distribution, avoids the imbalance of some anchor points being overloaded while others are idle, optimizing the overall network resource utilization rate. This ensures that even during peak museum visitor times, the system can still provide stable, accurate, and non-interfering collaborative positioning services to a large number of blind users.
[0029] The system includes a group guided tour collaborative mode. When multiple mobile guided tour terminals are set up for the same group, the cloud management platform plans a common tour route for the entire group and calculates the relative positions of each member in real time. The system provides members with prompts to maintain formation through the human-computer interaction devices of each terminal. When a member deviates from the predetermined formation or the safe distance is too large, a warning message is sent to both the member and the leader's terminal.
[0030] By using a cloud platform to plan the optimal public tour route for the entire group, problems such as route conflicts or inconsistent goals among members are avoided, ensuring the integrity and coordination of the group's activities. The system calculates and monitors the relative positions of each member in real time, and can provide each member with directions to follow through wristband vibration in the multi-modal human-computer interaction device to maintain formation. This allows members of the blind group, who lack visual communication, to still perceive their positional relationships and maintain a tight formation. When a member is detected to have deviated from the predetermined formation or exceeded the safe distance, the system will send an alert to both the member and the leader's terminal. This not only promptly reminds the person who has deviated to correct the route, but also allows the leader to grasp the overall situation and take intervention measures, thereby effectively preventing members from getting separated, ensuring group safety, and significantly improving the manageability, safety, and experience quality of blind groups visiting museums together.
[0031] The mobile navigation terminal is equipped with a one-click help module. Once triggered, the terminal will immediately send a help data packet containing its precise real-time location, current sensor status, and several seconds of ambient audio recordings to the cloud management platform via wireless network. The cloud management platform will push the help information and the terminal's real-time movement trajectory to the staff monitoring terminal and establish a voice call connection between the staff terminal and the help terminal.
[0032] When users feel lost, encounter obstacles, or experience accidents, they can automatically initiate a distress call by pressing a single button. The system immediately packages and uploads a complete contextual data package containing centimeter-level accurate real-time location, current sensor status snapshots, and environmental audio recordings. This allows remote staff to accurately grasp the user's precise location, surrounding physical environment, and possible event background in a timely manner, greatly shortening the information verification and positioning time in the rescue response. The cloud platform simultaneously pushes the distress call information and real-time movement trajectory to the staff's terminals, facilitating staff to quickly arrive and continuously track the user's movements. The automatically established voice call connection enables direct two-way communication between the user and staff, allowing staff to provide real-time reassurance, remote guidance, or situation confirmation, while the user can also receive timely voice assistance. This creates a low-latency, end-to-end safety guarantee pathway from user-initiated triggering to precise staff intervention in emergency situations, significantly enhancing the confidence and sense of security of blind users exploring the museum independently.
[0033] The system supports self-learning and map calibration in the museum environment. After initial deployment or exhibition changes, staff can carry mobile guide terminals in learning mode along preset routes. The system automatically collects UWB base station signal characteristics, inertial data, and key landmarks identified by other sensors, and combines them with architectural drawings to semi-automatically generate or update high-precision 3D maps. During daily operation, the system continuously monitors the stability of fixed anchor point signals and automatically calibrates for minor offsets caused by environmental changes.
[0034] Through simple operation by staff carrying terminals and walking in learning mode, the system can automatically integrate collected UWB base station signal characteristics, inertial trajectory data, and visually recognized landmark information, and combine them with existing architectural drawings to semi-automatically generate or update centimeter-level precision 3D digital maps. This greatly reduces the cumbersome manual surveying and map reconstruction costs after initial deployment and exhibition changes, and shortens the system update cycle. In daily operation, the system continuously monitors the stability of anchor point signals and automatically calibrates for minor offsets caused by environmental changes, ensuring the long-term reliability and consistency of the positioning benchmark. This effectively overcomes the problem of positioning accuracy attenuation that may be caused by factors such as temperature and humidity changes, equipment aging, or minor adjustments to the exhibition hall layout. It ensures that the system can continuously provide stable and accurate positioning and navigation services in the complex and potentially changing physical environment of the museum, greatly improving the system's robustness, maintainability, and service quality throughout its lifecycle.
[0035] During use, after the system starts, the mobile navigation terminal connects to the UWB positioning base station network. Users select a target or join a group. The terminal uses frequency division multiplexing and dynamic anchor point selection mechanisms to measure distances with multiple base stations with optimal signals. It then combines data from the inertial measurement unit, millimeter-wave radar, and lightweight vision module, employing an extended Kalman filter algorithm for multi-sensor fusion to achieve centimeter-level real-time positioning and environmental perception. The cloud management platform, based on a high-precision 3D map containing obstacle information, uses an improved A* algorithm that integrates real-time dynamic obstacle prediction, user personalized preferences, and path turning costs to plan an optimal path with a safe width for individuals or groups. Planning instructions and real-time perception information are communicated through multiple... The human-computer interaction device transforms into coded vibrations from the wristband, distance ratio warnings from the guide cane, and directional voice from the bone conduction headphones, collaboratively guiding users along the path and providing gradual corrections when users deviate. The user behavior learning module continuously records and updates the personalized model to optimize subsequent navigation strategies. In case of emergency, the one-click help module can instantly upload precise location and environmental information to establish a voice connection with staff. After initial deployment or exhibition changes, the system can easily update the map through self-learning mode and automatically calibrate base station signals during daily operation, thus providing blind users with a complete guide solution from accurate positioning, intelligent obstacle avoidance, personalized guidance to safety assurance and sustainable maintenance.
[0036] Although embodiments of the invention have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit, the scope of which is defined by the appended claims and their equivalents.
Claims
1. A UWB-based precise positioning and obstacle avoidance navigation system for blind museums, characterized in that, include: Cloud management platform, UWB positioning base station network and mobile navigation terminal; The cloud management platform stores a high-precision 3D map containing obstacle information and a path planning engine. The UWB positioning base station network consists of multiple UWB anchor points deployed in spatial density, used to interact with the terminal to perform positioning and ranging. The mobile terminal integrates a UWB module, an inertial measurement unit, a millimeter-wave radar, and a multi-mode human-computer interaction device. The UWB module is used to achieve centimeter-level positioning; the inertial measurement unit is used to maintain short-term positioning when the signal is blocked; the millimeter-wave radar is used to detect moving obstacles; the multi-mode human-computer interaction device includes a wristband-type haptic feedback unit that provides directional encoded vibration, bone conduction headphones that play navigation voice, and a smart guide cane with integrated proximity sensors.
2. The UWB-based precise positioning and obstacle avoidance navigation system for blind museums according to claim 1, characterized in that: The mobile navigation terminal also integrates a lightweight visual processing module. This module extracts edge contour features from the images captured by the terminal camera and matches the extracted contours with local environmental feature templates issued by the cloud management platform. This helps to identify large glass or mirror-like fixed obstacles that are difficult to reflect effectively by UWB signals. The identification results are used as supplementary information to be input into the extended Kalman filter algorithm and obstacle avoidance decision module.
3. The UWB-based precise positioning and obstacle avoidance navigation system for blind museums according to claim 1, characterized in that: The wristband-type haptic feedback unit in the multi-mode human-computer interaction device uses a ring array vibration motor, which can encode and generate various haptic commands for indicating forward movement, left turn, right turn, stop, and fine-tuning direction by controlling the vibration timing and mode of motors in different positions. The handle of the smart guide cane has a built-in haptic warning device. When an obstacle is detected in the near-field emergency layer, the warning device will generate a vibration with an intensity inversely proportional to the distance from the obstacle. The navigation voice and exhibit explanation information broadcast by the bone conduction headphones are processed by directional sound field to create a spatial sense of virtual sound source to assist users in judging direction.
4. The UWB-based precise positioning and obstacle avoidance navigation system for blind museums according to claim 1, characterized in that: The cloud management platform's path planning algorithm engine adopts an improved A* algorithm framework. When calculating path costs, it not only considers geometric distance but also introduces real-time dynamic obstacle prediction costs, user historical walking preference costs, and path turning frequency costs. The algorithm outputs a path with width generated based on a safe area and provides guidance and correction through increasingly stronger tactile or voice prompts when the user deviates from the center line of the output path.
5. The UWB-based precise positioning and obstacle avoidance navigation system for blind museums according to claim 1, characterized in that: The system has context-adaptive capabilities. The mobile navigation terminal has a built-in user behavior learning module that can continuously record and analyze the user's average walking speed, turning range, and response sensitivity parameters to various prompts, and form a personalized user model. The personalized user model is uploaded to the cloud management platform for personalized adjustment of path planning strategies, advance warning prompts, and the intensity and frequency of various feedbacks in the multi-modal human-computer interaction device.
6. The UWB-based precise positioning and obstacle avoidance guidance system for blind museums according to claim 1, characterized in that: The UWB positioning base station network adopts frequency division multiplexing and dynamic anchor point selection mechanism to cope with multi-user scenarios; each mobile navigation terminal's UWB communication module is assigned a specific working frequency band or time slot; the system monitors the signal quality of each UWB anchor point in real time, and dynamically selects the three to four anchor points with the strongest signals for each terminal to form the optimal positioning set, which is used to perform high-precision fusion positioning calculation, thereby reducing signal interference between multiple terminals and optimizing network load.
7. The UWB-based precise positioning and obstacle avoidance navigation system for blind museums according to claim 1, characterized in that: The system includes a group guided tour collaborative mode. When multiple mobile guided tour terminals are set up for the same group, the cloud management platform plans a common tour route for the entire group and calculates the relative positions of each member in real time. The system provides members with prompts to maintain formation through the human-computer interaction devices of each terminal. When a member deviates from the predetermined formation or the safe distance is too large, a warning message is sent to both the member and the leader's terminal.
8. The UWB-based precise positioning and obstacle avoidance guidance system for blind museums according to claim 1, characterized in that: The mobile guide terminal is equipped with a one-click help module. When triggered, the terminal will immediately send a help data packet containing its precise real-time location, current sensor status and environmental recordings from a few seconds before and after to the cloud management platform via wireless network. The cloud management platform pushes the request for help information and the real-time movement trajectory of the terminal to the staff monitoring terminal, and establishes a voice call connection between the staff terminal and the terminal requesting help.
9. A UWB-based precise positioning and obstacle avoidance navigation system for blind museums according to claim 1, characterized in that: The system supports self-learning and map calibration in the museum environment. After initial deployment or exhibition changes, staff can carry mobile guide terminals in learning mode along preset routes. The system automatically collects UWB base station signal characteristics, inertial data, and key landmarks identified by other sensors, and combines them with architectural drawings to semi-automatically generate or update the high-precision 3D map. During daily operation, the system continuously monitors the stability of fixed anchor point signals and automatically calibrates for minor offsets caused by environmental changes.