A flexible wearable fire evacuation path planning and guiding device and method based on UWB sensing

Abstract

The application discloses a flexible wearable fire evacuation path planning and guiding device and method based on UWB sensing, and relates to the technical field of fire safety evacuation path planning.The system comprises a flexible wearable main body, a positioning communication module, an environment sensing module, a central control unit, a flexible driving module, a multi-modal interaction module and a building fire control system; wherein the positioning communication module fuses UWB, IMU, BIM constraints and SLAM positioning to realize failure completion, the central control unit performs multi-source fusion risk coupling positioning, regional risk assessment and old age adaptive dynamic obstacle avoidance path planning, the flexible driving module and the multi-modal interaction module cooperatively output power assistance, tactile, visual and auditory guidance, and link lighting, smoke exhaust, fire doors and other facilities, and are suitable for safe, efficient and continuous evacuation of people with difficulty in movement in a complex building fire environment, and improve path decision accuracy and linkage control reliability.

Description

Technical Field

[0001] This invention relates to the field of fire safety evacuation technology, specifically to a distributed fire evacuation path planning and guidance device that integrates ultra-wideband (UWB) positioning, multi-dimensional environmental perception, flexible protection drive, and multi-modal guidance functions. It also relates to a fire evacuation method adapted to the system, which is particularly suitable for the safe evacuation of the elderly and other people with mobility impairments in complex fire scenarios. Background Technology

[0002] Fire is a sudden, rapidly evolving safety event in the built environment, characterized by highly coupled causative factors. During a fire, high temperatures, dense smoke, obstructed visibility, and complex building structures severely hinder evacuation. The elderly, disabled individuals, and other groups with mobility impairments face even greater difficulties in evacuation, resulting in low efficiency and extremely high safety risks. Therefore, developing efficient, intelligent, and reliable auxiliary evacuation technologies and equipment is crucial for ensuring the safety of these vulnerable groups.

[0003] While existing fire evacuation technologies and equipment have made some progress, they still suffer from the following core shortcomings in dealing with complex and dynamic real-world fire scenarios, especially in meeting the needs of special groups: First, the guidance methods are passive and simplistic, relying mainly on passive equipment such as fixed evacuation signs and emergency lighting. These become completely ineffective when dense smoke obscures visibility or signs are damaged, lacking proactive and precise guidance capabilities. Second, positioning accuracy is insufficient. Traditional positioning technologies (such as GPS and Bluetooth) are easily interfered with in indoor fire scenarios, resulting in large positioning errors (usually exceeding 1 meter), failing to provide accurate location data for evacuation route planning. Third, protective performance is inadequate. Existing auxiliary evacuation equipment does not address the high temperatures and smoke of fires. The existing technology suffers from several shortcomings. First, it lacks specialized protective structures for fog environments, making it difficult to guarantee the personal safety of users in hazardous environments. Second, it has poor adaptability, failing to fully consider the mobility characteristics of the elderly (such as slow gait, weak balance, and slow reaction), and its path planning does not take into account both the difficulty of movement and the need for obstacle avoidance response, resulting in insufficient flexible assistive driving force. Third, it lacks environmental perception and risk quantification, failing to integrate multi-dimensional environmental data such as temperature, smoke, and obstacles in real time, making it difficult to accurately assess the risk level of the area, leading to a lack of scientific basis for evacuation decisions. In addition, the existing technology also suffers from weak system linkage, with most devices operating independently and unable to communicate with the building fire protection system in real time to obtain fire data, causing evacuation decisions to lag behind the dynamic development of the fire.

[0004] In summary, existing fire evacuation technologies generally suffer from insufficient active guidance capabilities, low indoor positioning stability, inadequate risk perception and quantitative assessment, weak adaptability to people with mobility impairments, and limited integration with building fire protection systems. Therefore, developing a distributed fire evacuation path planning and guidance device and method that is suitable for complex fire environments and can simultaneously achieve accurate perception, dynamic decision-making, individual adaptation, and collaborative evacuation, thereby addressing the pain points of existing technologies and enabling the safe and rapid evacuation of people with mobility impairments in fire environments, is of significant practical importance for improving the safety and efficiency of fire evacuation. Summary of the Invention

[0005] (a) Purpose of the invention The core objective of this invention is to overcome the aforementioned deficiencies of existing technologies and provide a flexible wearable fire evacuation path planning and guidance device and method based on UWB sensing. This device achieves precise positioning in fire scenarios by combining UWB positioning with adaptive switching between Simultaneous Localization and Mapping (SLAM) modes to control positioning errors and provide accurate location support for evacuation decisions. It offers comprehensive fire and heat insulation protection through a multi-layered composite flexible wearable body design that balances comfort and mobility. It enables quantitative environmental risk assessment by integrating multi-dimensional environmental data such as temperature, smoke concentration, and obstacle distance to establish a risk assessment model and accurately classify regional risk levels. Adapting to the mobility characteristics of the elderly, it develops an elderly-adaptive path planning algorithm that considers path distance, environmental risk, movement difficulty, and obstacle avoidance response time, providing flexible auxiliary driving force. It constructs a multimodal active guidance system that integrates visual, auditory, and tactile guidance methods to ensure effective transmission of guidance information in complex scenarios such as dense smoke and noise. Finally, it achieves linkage with building fire control systems to obtain real-time fire dynamic data, dynamically adjust evacuation strategies, and improve the timeliness and accuracy of evacuation decisions.

[0006] (II) Technical Solution To achieve the objective of this invention and solve its technical problems, the present invention adopts the following technical solution: The first objective of this invention is to provide a flexible wearable fire evacuation route planning and guidance device based on UWB sensing. This device mainly includes a flexible wearable body and a central control unit, a positioning and communication module, an environmental sensing module, a flexible drive module, and a multimodal interaction module distributed on the flexible wearable body. The central control unit is communicatively connected to each module and the building fire control system. The flexible wearable body is an open vest structure made of multi-layer composite flexible fabric and extending from the shoulders to the thighs. The wearable body is also provided with at least a shape memory alloy skeleton for providing biomimetic support and a wear fit sensor for detecting the wear fit status. The environmental perception module includes distributed lidar, temperature sensors and smoke concentration sensors, which are used to collect real-time information on obstacles, ambient temperature and smoke concentration around the wearer. The positioning communication module includes at least a UWB positioning tag and an inertial measurement unit (IMU) for collecting the wearer's location data and IMU attitude data, and for executing a UWB positioning threshold dynamic calibration mechanism based on environmental parameters, a UWB multi-user interference adaptive avoidance mechanism based on time slot allocation, and a three-level emergency completion mechanism that combines IMU inertial estimation, Building Information Modeling (BIM) constraints, and SLAM positioning when UWB signal fails. The central control unit is connected to each module and the building fire control system. It is used to generate fused positioning coordinates, comprehensive environmental risk values ​​and risk direction vectors with risk weights by fusion positioning and environmental data, and to generate evacuation decision instructions based on evacuation targets and impassable areas. The flexible drive module includes artificial muscles distributed along the lower limbs and an array of vibration motors distributed in the shoulder and waist areas, used to output auxiliary driving force and directional tactile feedback. The multimodal interaction module includes a projector and headphones, which are used to output visual and auditory guidance information according to evacuation decision instructions, and work with the flexible drive module to complete multimodal evacuation guidance.

[0007] The second objective of this invention is to provide a fire evacuation route planning and guidance method based on flexible wearable devices and UWB sensing. Based on the aforementioned flexible wearable fire evacuation route planning and guidance device, building coordination facilities, and interaction mechanism based on UWB sensing, this invention achieves accurate, safe, and efficient evacuation in fire scenarios through a five-step closed-loop process of wake-up, positioning, assessment, planning, and guidance, combined with building facility linkage. The method includes at least the following steps: SS1. Wearable verification wake-up: The central control unit receives the alarm signal sent by the building fire control system and wakes up the functional modules of the path planning and guidance equipment. It reads the wear fit status data, determines that the wear meets the standards, and then starts the environmental perception module and the positioning communication module. SS2. Positioning Data Fusion: The positioning communication module collects UWB positioning data and IMU attitude data, while the environmental perception module simultaneously collects ambient temperature, smoke concentration, and obstacle distance data. It performs UWB positioning threshold dynamic calibration, multi-user interference adaptive avoidance, and positioning failure emergency completion. Through multi-source fusion-risk coupling positioning processing, it generates fused positioning coordinates, comprehensive environmental risk values, and risk direction vectors. When the UWB signal fails, the emergency positioning coordinates are used to replace the fused positioning coordinates. SS3. Environmental Risk Assessment: Based on fused positioning coordinates or emergency positioning coordinates, and combined with lidar point cloud data collected by the environmental perception module, obstacles are segmented and their coordinates, dimensions, and distances are marked; regional risk levels are divided according to preset intervals of comprehensive environmental risk values; grid cells are divided with the user's current positioning coordinates as the center and a preset range as the radius, and the risk level of each grid cell is matched; the center coordinates of high-risk areas are derived based on the risk direction unit vector and the user's current positioning coordinates. SS4. Path planning decision: The central control unit calls the pre-stored BIM model data and combines it with the real-time fire data, candidate list of safe target points and data of inaccessible areas issued by the building fire control system to select the safe target point that is closest and meets the risk level; it executes the aging-adaptive-dynamic obstacle avoidance cost algorithm to determine the evacuation route and dynamically replans the local obstacle avoidance route according to the environmental changes. SS5. Multimodal guidance and building linkage: The central control unit generates evacuation decision instructions based on the determined evacuation route, outputs visual, auditory, and tactile guidance information, and controls the building fire protection facilities in conjunction with the building fire control system until the guidance stops after detecting that the wearer has reached the safe area.

[0008] (III) Technical Effects Compared with the prior art, the flexible wearable fire evacuation path planning and guidance device and method based on UWB sensing of the present invention has the following beneficial and significant technical effects: (1) Based on the adaptive switching of UWB positioning and SLAM mode, this invention adds a dynamic calibration mechanism for UWB positioning threshold to realize real-time adaptation of mode switching to fire environment and avoid unreasonable switching caused by fixed threshold; the multi-user interference adaptive avoidance mechanism eliminates signal conflict when multiple people are evacuated and ensures reliable transmission of positioning data; the positioning failure emergency completion mechanism quickly generates emergency positioning coordinates through a three-level strategy to solve the problem of positioning interruption in extreme scenarios.

[0009] (2) The environmental perception module comprehensively collects multi-dimensional data such as temperature, smoke concentration, and obstacles through distributed sensors. Combined with precise positioning coordinates, it constructs a comprehensive environmental risk value model through relevant algorithms to achieve regional risk classification and precise positioning of high-risk areas. This provides quantitative and real-time risk data support for path planning and effectively prevents users from accidentally entering dangerous areas. The age-adaptive algorithms fully consider the characteristics of the elderly, such as slow movement and weak balance. The artificial muscles of the flexible drive module provide lower limb auxiliary driving force, and the vibration motor array provides intuitive tactile feedback, reducing the physical exertion and operational difficulty of the elderly during evacuation and improving the evacuation safety of the elderly population.

[0010] (3) Integrating visual, auditory, and tactile multimodal guidance to adapt to complex environments such as obstructed vision and noise interference at fire scenes; during emergency positioning, guidance instructions and positioning coordinates are dynamically adapted to maintain guidance continuity and ensure that users do not lose their way even when UWB signals fail. The central control unit communicates with the building fire control system in real time to synchronously acquire fire data and dynamically adjust evacuation strategies; the modular integrated design facilitates maintenance and function upgrades and is suitable for scenarios where multiple people evacuate simultaneously; the system is lightweight and easy to wear, and can be widely used in fire evacuation of various buildings such as office buildings, nursing homes, and residential buildings. Attached Figure Description

[0011] The above and / or additional aspects and advantages of the present invention will become apparent and readily understood in conjunction with the following description of the embodiments, in which: Figure 1 The diagram shown is a connection diagram of the functional modules in the flexible wearable fire evacuation path planning and guidance device based on UWB sensing of the present invention. Figure 2 The diagram shown is a front view of the flexible wearable body in this invention. Figure 3 The diagram shown is a schematic diagram of the back structure of the flexible wearable body in this invention; Figure 4 The diagram shows the implementation flowchart of the fire evacuation route planning and guidance method based on flexible wearable devices and UWB sensing according to the present invention.

[0012] Figure labeling: 1-Shoulder strap, 2-Shape memory alloy skeleton, 3-Artificial muscle, 4-Circuit channel, M-Magnetic self-locking buckle, V-Vibration motor, P-Wear fit sensor, LP-Laser projector, R-LiDAR, T-Temperature sensor, S-Smoke concentration sensor, CCU-Central control unit, BCH-Earphones. Detailed Implementation

[0013] This invention aims to provide a flexible wearable fire evacuation route planning and guidance device and method based on UWB sensing. To make the objectives, technical solutions, and advantages of this invention clearer, the technical solutions in the embodiments of this invention will be described in more detail below with reference to the accompanying drawings. The described embodiments are some, but not all, embodiments of this invention, and are exemplary, intended to explain the invention, and should not be construed as limiting the invention. All other embodiments obtained by those skilled in the art based on the embodiments of this invention without creative effort are within the scope of protection of this invention.

[0014] Example 1: Fire evacuation route planning and guidance equipment As a specific example, the flexible wearable fire evacuation route planning and guidance device based on UWB sensing provided in this embodiment of the invention is used in complex fire scenarios (such as a sudden fire in a nursing home inpatient building, where high temperatures and dense smoke rapidly spread to evacuation routes, and evacuation signs inside the building are obscured by smoke and become ineffective) to provide support for the elderly and other people with mobility difficulties in terms of positioning, risk assessment, route planning, and evacuation guidance. Figure 1 As shown, the path planning and guidance device adopts a modular integrated structure, mainly including a flexible wearable body and a central control unit, positioning and communication module, environmental perception module, flexible drive module and multimodal interaction module distributed on the flexible wearable body. Each module establishes a wired connection through embedded circuits. The central control unit serves as the core hub to communicate with each module and the building fire control system. The overall structure is lightweight and easy to wear.

[0015] In embodiments of the present invention, such as Figure 2 , Figure 3 As shown, the flexible wearable body is made of multi-layered composite flexible fabric and designed as an open vest structure extending from the shoulders to the thighs. It is formed through segmented cutting and elastic sewing processes. The body also has at least a shape memory alloy skeleton 2 for providing biomimetic support, magnetic self-locking buckles M for quick putting on and taking off and locking, and a wear fit sensor P for detecting the wearing fit status. The shoulder straps 1, waist and thigh straps on the flexible wearable body are designed with a multi-segment telescopic structure to adapt to different body types by adjusting the strap length. The magnetic self-locking buckles M are distributed at the shoulder, chest, waist and thigh strap connection points (preferably 8 buckles, preferably including male and female buckles, the male buckle has a built-in permanent magnet, and the female buckle has a built-in magnetic sheet and spring-driven claw to automatically lock when fitted, with a single magnetic force ≥8kg and an unlocking force ≤5N), ensuring that the body fits closely to the human body. Wearable fit sensors P (preferably 6 FSR402 flexible pressure sensors) are distributed on the fit surfaces of the main body and the human body's shoulders, back, and waist to collect fit data in real time. The measurement range is 0-10N, the sampling frequency is 10Hz, and after confirming that the wear meets the standards, the data is fed back to the central control unit CCU and the entire path planning and guidance device module is activated.

[0016] In the flexible wearable body, the shape memory alloy skeleton 2 is made of Ni-Ti alloy (preferably with a diameter of 1.2 mm and a phase transition temperature of 60°C). It is distributed and set in the front and back shoulder straps and side waist positions of the open vest structure (preferably 4 sets). It is also arranged along the sides of the spine, the transition area from the shoulder to the hip, and the force path on the outside of the thigh. It is used to automatically tighten when the phase transition is triggered by temperature rise to maintain the stable fit of the wearable body to the human body. It provides biomimetic support and posture stability constraint during standing, starting, crossing, and turning. While maintaining the stability of the main body structure, it does not restrict the wearer's limb movement and ensures the flexibility of movement during evacuation. At room temperature, it returns to a soft shape to reduce the feeling of wearing restraint.

[0017] The multi-layered composite structure of the flexible wearable body simultaneously provides protection and comfort. The multi-layered flexible fabric, from the outside in, includes a fire-resistant outer layer, a heat-insulating middle layer, a flexible circuit and actuator embedding layer, and a skin-friendly and breathable inner lining layer. These layers are bonded and hot-pressed together to form a single structure. The fire-resistant outer layer is preferably made of modified polybenzimidazole (PBI) material with a flame-retardant coating on its surface, and has a surface density of 280 g / m³. 2 With an extreme temperature resistance of 600℃ and a flame-retardant coating with an oxygen index ≥38%, it effectively resists external high temperatures and flame attacks. The heat insulation intermediate layer is preferably a 5mm thick aerogel heat insulation felt layer with a thermal conductivity ≤0.02W / (m•K), which can effectively block heat transfer and prevent wearers from being burned by high temperatures. The flexible circuit and driver embedding layer preferably uses a polyimide flexible substrate with a thickness of 0.8mm. The skin-friendly and breathable inner lining layer is preferably made of Modal spandex blend material with a weight of 120g / m². 2 With a breathability of ≥800mm / s, it can effectively reduce the stuffiness caused by wearing it for a long time.

[0018] More preferably, the fabrication of the flexible wearable body mainly includes the following steps: (1) Preparation of multi-layer composite structure: The open vest structure extending from the shoulder to the thigh is formed by using piece cutting and elastic sewing process, and a multi-segment elastic structure of shoulder straps, waist straps and thigh straps is formed; each layer is stacked in sequence from the outside to the inside, and hot-pressed for 30 minutes at 120℃ and 0.8MPa. The layers are bonded with high temperature resistant silicone adhesive to form a multi-layer composite body that is laminated in one piece; (2) Wearable fit sensor integration: The FSR402 flexible pressure sensor is formed by atomization spraying process, embedded in the pre-installed bag of the flexible wearable body, and fixed to the fit surface of the shoulder, back and waist with elastic stitching to form a distributed detection unit for detecting the fit state between the flexible wearable body and the human body. (3) Embedded circuit channel and hardware installation interface: An embedded circuit channel 4 is arranged inside the multi-layer composite body. The embedded circuit channel adopts a flame-retardant polyimide corrugated tube structure. A hardware installation interface for installing sensors, positioning communication components and driving components is formed at the preset installation position of the main body. The hardware installation interface is a snap-on waterproof structure with a protection level of IP65. (4) Preparation and fixing of shape memory alloy skeleton: The Ni-Ti alloy shape memory skeleton is prepared by laser powder bed fusion process, forming a spiral sheet topology structure. The surface is covered with polyimide film for insulation and protection. It is embedded in the preset cavity of the flexible wearable body and locked by high temperature resistant buckle to form a support component for providing biomimetic support force and maintaining the shape of the main body structure. (5) Structure and installation of magnetic self-locking buckle: The buckle consists of a male buckle, a female buckle and a self-locking mechanism. The male buckle has a built-in permanent magnet, and the female buckle has a built-in magnetic conductive sheet and a spring-driven claw. The male buckle is fixed to the strap by high-frequency heat sealing, and the female buckle is bolted to the flexible wearable body, so that the connection of the straps at the shoulder, chest, waist and thigh forms a quick-wearing and locking structure that automatically locks when the straps fit.

[0019] In this embodiment of the invention, the environmental perception module comprehensively monitors the surrounding environment, providing multi-dimensional data support for evacuation decisions. It includes at least distributed LiDARs R, temperature sensors T, and smoke concentration sensors S. Preferably, two miniature solid-state LiDARs R are mounted on the shoulders and one on the waist, fixed to preset mounting positions on the flexible wearable body via a bracket. The bracket is made of lightweight aluminum alloy in one piece, secured to the main body mounting interface with bolts, and allows for fine-tuning of the angle. Each LiDAR measures 30mm × 25mm × 15mm, weighs ≤20g, and has an IP64 protection rating. The shoulder-mounted LiDAR is used to detect obstacles facing diagonally upwards and forwards, preferably with a horizontal field of view of 120°, a vertical field of view of 30°, and a detection direction at a 30° angle between diagonally upwards and forwards and the horizontal direction. The waist-mounted LiDAR is used for… The system detects obstacles facing forward, ideally with a horizontal field of view of 120° and a vertical field of view of 20°, with the detection direction horizontally forward. It accurately captures the coordinates, size, and distance of obstacles within a range of 0.5-10m, facilitating obstacle avoidance path planning. Six temperature sensors (T) cover at least the shoulder, chest, and outer waist areas, measuring a temperature range of -20℃ to 800℃ with an accuracy of ±1℃ and a response time ≤50ms, collecting ambient temperature data in real time. Four smoke concentration sensors (S) cover at least the near-field breathing area in front of the chest, measuring a range of 0-1000ppm with an accuracy of ±5ppm and a response time ≤100ms, simultaneously monitoring smoke concentration. All environmental data is transmitted to the central control unit (CCU) via an embedded flame-retardant polyimide corrugated tube circuit channel, ensuring secure and stable data transmission. The CCU establishes a directional mapping relationship for the environmental data based on the spatial installation positions of the lidar (R), temperature sensors (T), and smoke concentration sensors (S), converting obstacle distance, temperature gradient, and smoke concentration gradient into risk direction vectors corresponding to the human body's orientation.

[0020] In this embodiment of the invention, the positioning and communication module integrates a built-in UWB positioning tag and an IMU (Inertial Measurement Unit) in a dedicated mounting box on the same side of the central control unit. The mounting box is preferably detachably connected to the main body via Velcro, and its interior is filled with flame-retardant cushioning foam to prevent damage to the module from vibrations during a fire. The UWB positioning tag preferably operates at a frequency of 3.5-6.5 GHz, with a positioning accuracy of ±10 cm and a communication distance of ≤100 m. It integrates two printed circuit board (PCB) antennas and uses a Time Difference of Arrival (TDOA) positioning algorithm to obtain positioning information between itself and a pre-set UWB base station array within the building. This allows for accurate collection of the wearer's real-time location data, with a data update frequency of 50 Hz, providing accurate location data for path planning. The IMU includes a three-axis accelerometer, a three-axis gyroscope, and a three-axis magnetometer, with a sampling frequency of 100 Hz. It synchronously collects the wearer's attitude data to assist in correcting positioning deviations.

[0021] The positioning and communication module is configured to implement a UWB positioning threshold dynamic calibration mechanism based on environmental parameters, a UWB multi-user interference adaptive avoidance mechanism based on time slot allocation, and a three-level emergency completion mechanism combining IMU inertial estimation, BIM constraints, and SLAM positioning when UWB signals fail. Due to increased ambient temperature and smoke concentration, the path planning and guidance equipment activates the UWB positioning threshold dynamic calibration mechanism, dynamically adjusting the Received Signal Strength Indicator (RSSI) switching threshold based on environmental data. adj To ensure the positioning mode adapts to changes in the fire environment, and given that multiple wearable users trigger the system simultaneously, the UWB multi-user interference adaptive avoidance mechanism dynamically allocates positioning time slots to ensure a positioning data reception success rate of ≥98%, preventing signal conflicts that could lead to positioning failure when multiple users are using the device. If the UWB signal is completely lost due to severe interference from dense smoke, a three-level emergency completion mechanism is activated, consisting of IMU inertial estimation, BIM model constraints, and SLAM fusion, to ensure emergency positioning accuracy of ≥±0.2m and uninterrupted positioning. Specifically: The UWB positioning threshold dynamic calibration mechanism is as follows: A mapping relationship is established between environmental parameters (temperature, smoke concentration) and UWB signal attenuation, and the RSSI switching threshold is dynamically adjusted. adj Its calculation formula is When RSSI < RSSI adj And when the duration exceeds the first preset duration (e.g., 500ms), the system automatically switches to SLAM positioning mode, when RSSI ≥ RSSI. adj And if the duration exceeds the second preset duration (e.g., 300ms), the system automatically switches back to UWB positioning mode. Here, RSSI0 is the initial fixed switching threshold (preferably RSSI0 = -85dBm), κ1 and κ2 are the temperature influence coefficient and smoke concentration influence coefficient, respectively, calibrated through numerous fire simulation environment experiments; T(t) and S(t) are the real-time ambient temperature and smoke concentration collected by the environmental sensing module, respectively; T0 and T... max These are the safe threshold and extreme danger threshold for ambient temperature, S0 and S1, respectively. max These are the safe threshold and the extreme danger threshold for smoke concentration, respectively. The higher the ambient temperature and the greater the smoke concentration, the lower the RSSI switching threshold should be. adj When environmental conditions improve, the switching threshold RSSI is increased. adj Ensure that UWB mode is used preferentially to guarantee the sensitivity of positioning mode switching in response to changes in the fire environment.

[0022] The UWB multi-user interference adaptive avoidance mechanism is as follows: For scenarios where multiple positioning communication modules work concurrently in the same building space, a time slot dynamic allocation mechanism based on the TDOA positioning algorithm is used to count the number of active UWB positioning tags in real time and dynamically adjust the UWB positioning time slot length, time slot allocation order, and tag access scheduling strategy to reduce mutual interference of UWB signals during group evacuation and ensure that the positioning data reception success rate is ≥98%. The three-level emergency recovery mechanism for UWB positioning failure is as follows: When a complete UWB signal failure is detected, the most recent valid UWB positioning coordinates are used as the initial reference coordinates. A three-level emergency recovery strategy is initiated, involving IMU inertial estimation, BIM model constraints, and SLAM fusion. First, inertial estimation is performed based on the real-time acceleration, angular velocity, and attitude data output by the IMU to obtain the predicted displacement of the wearer's current position. This is then combined with the initial reference coordinates to generate initial emergency prediction coordinates. Next, spatial constraint correction is applied to the initial emergency prediction coordinates based on pre-stored BIM model data, eliminating abnormal coordinates falling into impassable areas to obtain BIM-constrained corrected coordinates. Finally, SLAM positioning is performed based on the point cloud data collected in real-time by the lidar in the environmental perception module. The SLAM-corrected coordinates are then fused with the BIM-constrained corrected coordinates to generate the emergency positioning coordinates P. emg (t), and ensure emergency positioning accuracy ≥ ±0.2m.

[0023] In this embodiment of the invention, the Central Control Unit (CCU) is integrated into the flexible wearable body and communicates with various modules and the building fire protection system. It is used to fuse positioning and environmental data to generate fused positioning coordinates with risk weights, a comprehensive environmental risk value, and a risk direction vector. Based on evacuation targets and impassable areas, it generates evacuation decision commands. The CCU, as the core hub, is integrated into a dedicated mounting position on the back of the flexible wearable body. It is secured to the body via four M4 threaded holes. A buffer pad is installed at the mounting position to reduce vibration. Its outer shell is preferably made of flame-retardant ABS material through one-piece injection molding, with closed sealing grooves at the seams and an internal silicone sealing ring. The dimensions are 120mm × 80mm × 30mm, and the weight is ≤300g. It communicates with the building control system via a wired interface, achieving an IP65 protection rating. Inside the shell, a data processing chip, multi-protocol communication interfaces, and a power management module are installed. All components are integrated via a PCB board with a 4-layer wiring design, separating the power layer and ground layer. The data processing chip uses a high-performance microcontroller with a matching FPGA chip. The multi-protocol communication interfaces include a CAN bus interface, a UART interface, an SPI interface, and an I2C interface. The power management module uses a 6000mAh lithium battery pack with a voltage of 14.8V. It integrates a power management chip, supports multiple voltage outputs, has an output ripple of ≤50mV, and features overcharge, over-discharge, and overcurrent protection. Furthermore, its internal high-performance microcontroller and FPGA chip rapidly process location and environmental data, and communicate in real-time with the building's fire control system via a multi-protocol communication interface to obtain dynamic fire data and provide a basis for adjusting evacuation strategies.

[0024] The central control unit is configured to perform multi-source fusion-risk coupling localization processing, specifically including: The environmental sensing module collects real-time data on ambient temperature T(t), smoke concentration S(t), and distance d from the nearest obstacle. obs (t) The data is preprocessed and the comprehensive environmental risk value R is calculated. total Risk impact factors are determined based on IMU attitude data. λ And derive the risk direction unit vector e r Simultaneously, when the UWB signal is normal, the original coordinates P of the UWB positioning are determined based on the UWB signal strength. UWB (t), SLAM positioning coordinates P SLAM (t) and IMU attitude correction ΔP IMU The multi-source positioning weighting coefficients corresponding to (t) are dynamically adjusted to generate fused positioning coordinates P(t) with risk weights. In the event of UWB signal failure, the emergency positioning coordinates P generated by the positioning communication module are used. emg (t) replaces the fusion positioning coordinates P(t); and based on this, according to the comprehensive environmental risk value R total(t) is used to divide the area into risk levels based on the preset interval. Risk grid cells are constructed with the current fused or emergency positioning coordinates as the center, and the risk level of each grid cell is matched. This is combined with the risk direction unit vector e. r The center coordinates of the high-risk area are derived from the current location coordinates.

[0025] Furthermore, the central control unit is also configured to call up the floor topology, passage geometry and stair connectivity in the pre-stored BIM, and combine it with the real-time fire data, candidate list of safety target points and data of impassable areas issued by the building fire control system to perform elderly-adaptive dynamic obstacle avoidance cost path planning: comprehensively consider the spatial distance between the current fused or emergency positioning coordinates and the evacuation target point, the spatial relationship between the comprehensive environmental risk value and the center coordinates of the high-risk area, the path turning angle, the difficulty coefficient of elderly gait and the obstacle avoidance response time margin, to select the path with the lowest cost as the initial optimal path, and perform local dynamic replanning when new obstacles or local risk increases are detected.

[0026] In this embodiment of the invention, the flexible drive module includes artificial muscles 3 distributed along the lower limbs and a vibration motor V array distributed in the shoulder and waist areas, which are used to output auxiliary driving force and directional tactile feedback, respectively. Specifically, the flexible drive module is activated according to the instructions of the central control unit. The artificial muscle 3 is a twisted and coiled nylon 66 fiber artificial muscle distributed along the lower limb muscle group of the wearer. It is preferably 0.8mm in diameter and arranged with 2 on the front of the thigh, 2 on the back, and 1 on the outside. The maximum contraction of a single muscle is 30%, the maximum output force is 50N, and the response time is ≤100ms. By adjusting the driving sequence, driving amplitude, and duration of the artificial muscle, it provides flexible auxiliary driving force to the wearer, reducing the physical exertion of walking and helping people with mobility impairments to complete evacuation more easily. The vibration motor array adopts a symmetrical distribution on the shoulders and waist, and a layered distribution to form a directional tactile output matrix. Preferably, there are 2 flat micro vibration motors on each side of the shoulders and 2 on each side of the waist. They are fixed in the preset groove of the main body. The surface of the groove is covered with a breathable protective film. The rated voltage of the motor is 5V, the vibration frequency is 200-500Hz, and the vibration force is 0.5-1.5g.

[0027] Furthermore, the flexible drive module preferably integrates four H-bridge drive chips, supporting pulse width modulation (PWM) signal control at a frequency of 1kHz and a duty cycle of 0-100%. The drive chips are fixed to the polyimide flexible substrate by soldering and are linked with the central control unit through embedded circuitry. Based on the guidance direction and evacuation decision commands, the corresponding side vibration motors are driven independently to provide directional tactile feedback, helping the wearer identify the direction of travel when vision is obstructed. This works in conjunction with the flexible auxiliary driving force output by artificial muscles and the visual and auditory guidance information output by the multimodal interaction module to improve directional recognition ability and the continuity of evacuation guidance.

[0028] Preferably, the flexible drive module mainly includes the following components during fabrication: (1) Artificial muscle preparation and fixation: Nylon 66 fiber is dried, drawn, twisted and heat-set, then inserted into a silicone tube for encapsulation, embedded into the main channel according to the preset layout, and fixed with a high-temperature buckle to form a flexible assistive execution structure distributed along the lower limb muscle group of the wearer. (2) Vibration motor array assembly: After cleaning the surface of the pre-set groove of the flexible wearable body, apply double-sided tape and fix the flat micro vibration motors to the corresponding positions on both sides of the shoulder and waist, so that they form a symmetrical left and right and layered upper and lower tactile output dot matrix. Then cover with a protective film, and after the wires are soldered, arrange them along the reserved channel so that the vibration motor array and the subsequent drive circuit form an electrical connection. (3) Driver chip integration: The polyimide flexible substrate of the flexible wearable body is laser cut and line etched to form a wiring structure that is compatible with artificial muscle drive and vibration motor drive. Then, the H-bridge driver chip and interface are welded on the flexible substrate and packaged and fixed. (4) Overall assembly and debugging: Connect the components, test the artificial muscle contraction amount, output force and motor vibration parameters, and confirm that the flexible drive module meets the requirements of auxiliary drive and directional tactile feedback.

[0029] In this embodiment of the invention, the multimodal interaction module includes a laser projector (LP) and a bone conduction headset (BCH), used to output visual and auditory guidance information based on evacuation decisions and to synchronously execute guidance tasks with the flexible drive module. Specifically, the chest-mounted laser projector (LP) is fixed to the chest of the main body via an adjustable bracket that can rotate 360°; the projector casing is made of flame-retardant PC material and has a built-in micro cooling fan with a speed of 3000 rpm, allowing air circulation through ventilation holes in the casing; the laser projector uses a green semiconductor laser, preferably with a power of 50mW and a brightness of 1000-5000 cd / m². 2Adjustable projection angle (30°), projection distance (0.5-5m), projecting dynamic arrows in the safe direction (arrow length adjustable from 10-30cm), and movement speed (0.5-2m / s adjustable as needed), providing clear visual guidance even in situations where dense smoke obscures vision. The bone conduction headphones (BCH) on the outer shoulder are secured with elastic straps, the ends of which connect to magnetic fasteners on the main body, allowing for adjustable tightness. Vibration frequency is 20-20000Hz, output power is 10mW, and volume is adjustable from 20-80dB. An integrated noise reduction chip provides ≥25dB ​​noise reduction, and 32MB of storage space is provided for pre-stored navigation commands. Simultaneously, it filters noise from the fire scene, ensuring users clearly receive navigation information. Combined with tactile feedback from the flexible drive module, a three-in-one guidance system integrating vision, hearing, and touch is formed, ensuring users accurately receive evacuation information and smoothly move to a safe area in complex fire environments. Throughout the evacuation process, the system generates fused positioning coordinates with risk weights through a multi-source fusion-risk coupling positioning algorithm, providing precise positional support for guiding actions; the protective structure of the flexible wearable body always resists high temperature and smoke invasion, ensuring the safety of wearers in all aspects.

[0030] Example 2: Building Fire Control System Based on Embodiment 1 above, Embodiment 2 further provides and describes a building fire control system that works in conjunction with the flexible wearable fire evacuation path planning and guidance device based on UWB positioning perception described above. Its core components include a fire monitoring and alarm module, a data storage and management module, a communication interaction module, an evacuation decision support module, and a linkage control module. The building fire control system and the building's fire protection facilities, together with the flexible wearable fire evacuation path planning and guidance device worn within the building, collaboratively initiate the evacuation support process, creating a safe environment for evacuees and improving evacuation efficiency. Specifically: The fire monitoring and alarm module includes distributed smoke sensors, temperature sensors, flame sensors, high-definition infrared video monitoring units, and audible and visual alarm units installed within the building. The distributed smoke and temperature sensors are used to collect real-time fire parameters such as smoke concentration and temperature in various areas of the building; the flame sensor captures the trajectory of flame spread; the high-definition infrared video monitoring unit is used to capture fire images and confirm the fire range and spread trend; when the fire monitoring data exceeds a preset threshold, an alarm signal is triggered and sent to each path planning and guidance device, while simultaneously activating the audible and visual alarm devices within the building.

[0031] The data storage and management module includes an industrial-grade solid-state drive, a data management server, and dual backup storage units. It is used to store high-precision BIM model data of buildings (including detailed information such as building structure, evacuation routes, safety exits, equipment layout, etc.), historical fire data, sensor calibration parameters, wearable user adaptation information, and to update fire dynamic data in real time. It also classifies and manages the data and backs it up regularly, and supports real-time access and data writing for various path planning and guidance devices.

[0032] The communication interaction module, including a multi-protocol communication gateway, a snap-on waterproof wired interface, and a signal amplification and anti-interference unit, serves as a data interaction bridge, realizing protocol conversion, signal enhancement, and anti-interference processing. It receives data uploaded by the path planning and guidance device and sends relevant information to achieve bidirectional data interaction with the path planning and guidance device. The data transmission frequency is adjustable, and the Cyclic Redundancy Check 32 (CRC32) algorithm is used for data verification, with a transmission delay of ≤100ms.

[0033] The evacuation decision support module, including a high-performance computing unit and an evacuation strategy database, is used to filter a candidate list of safe target points based on fire monitoring data, BIM model data, and user status data uploaded by various path planning and guidance devices. It prioritizes marking safe exits and refuge floors that are far from fire areas and have unobstructed passageways, allowing wearers to quickly pinpoint safe evacuation directions. It also marks impassable areas based on the fire's spread trend, such as passageways blocked by flames and areas with excessive smoke concentrations, preventing wearers from accidentally entering dangerous areas. Furthermore, it provides basic path planning suggestions to support the wearable system's precise path planning. It dynamically assesses changes in the fire situation, generating and issuing path adjustment instructions when new hazards appear on the original planned paths, ensuring that evacuation paths remain safe and feasible, and coordinating with building facilities to adapt to evacuation needs.

[0034] The linkage control module, including relay control units, elevator linkage interfaces, emergency lighting control units, smoke extraction equipment control units, and fire door control units, receives instructions from the evacuation decision support module and coordinates the control of building facilities to provide a safe environment for evacuation. For example, the relay control unit dispatches elevators through the elevator linkage interface, moving them to safe floors and preventing them from going to fire areas or affected floors, thus avoiding elevator malfunctions that could trap people. The emergency lighting control unit activates emergency lights along evacuation routes, ensuring illumination and allowing wearers to clearly identify passageways even in smoky conditions. The smoke extraction equipment control unit activates smoke extraction fans in corresponding areas, accelerating smoke removal, reducing smoke concentration in passageways, and improving the evacuation environment. The fire door control unit closes fire doors in fire areas and along evacuation routes, preventing the spread of fire and smoke and creating a safety barrier for evacuation routes. During evacuation, the linkage control module dynamically adjusts the operational status of building facilities based on real-time instructions from the evacuation decision support module, continuously creating a safe evacuation environment for wearers and helping them quickly reach safe areas.

[0035] The interaction mechanism between the path planning and guidance equipment and the building fire control system is as follows: the path planning and guidance equipment acts as the terminal execution unit, while the building fire control system acts as the data support and collaborative decision-making unit, following the core logic of two-way communication, data interoperability, collaborative decision-making, and dynamic adjustment. Triggered by a fire alarm, the building fire control system sends core data to each path planning and guidance equipment, and the path planning and guidance equipment uploads real-time data to the building fire control system. Based on a unified protocol and algorithm logic, both parties collaboratively complete risk assessment, path planning, and evacuation guidance. The building fire control system simultaneously links internal facilities to provide environmental protection, forming a closed-loop interactive system.

[0036] In terms of interactive connection, the central control unit of each path planning and guidance device can be connected to the snap-on waterproof interface of the communication interaction module of the building fire control system through a dedicated flame-retardant PVC communication cable. The interface has a built-in O-ring seal and the protection level reaches IP65. It preferably adopts the fire evacuation dedicated communication protocol based on CAN2.0A / B extension. The data frame format includes frame header, data type identifier, data length, data content, check code, and frame tail. The protocol supports data priority division, with alarm signals and fire emergency data having the highest priority.

[0037] Furthermore, the interaction process between the path planning and guidance equipment and the building fire control system mainly includes the interaction during the fire triggering stage, the location assessment stage, the path planning stage, and the guidance execution stage, specifically: During the fire triggering phase, after the fire monitoring and alarm module detects that the fire parameters exceed the standard, it sends an alarm signal through the communication interaction module to wake up the path planning and guidance equipment. The central control unit reads the wear fit sensor data and determines that the wear meets the standard before uploading it to the building fire control system. During the location assessment phase, the path planning and guidance equipment uploads location, environmental perception and user status data. The building fire control system calls the BIM model data to correct and issue the fire spread trend and regional risk level. Based on this, the path planning and guidance equipment performs dynamic calibration of the location threshold, adaptive avoidance of multi-user interference, and emergency completion of location failure to generate fused or emergency location coordinates. During the route planning phase, the route planning guidance device uploads the fusion or emergency positioning coordinates and comprehensive environmental risk values. The building fire control system issues a candidate list of safety target points and information on impassable areas. Based on the issued information, the route planning guidance device executes evacuation route planning and uploads it to the building fire control system for verification. If there are conflicts, route adjustment suggestions are issued. During the guidance and execution phase, the path planning and guidance equipment outputs visual, auditory, and tactile guidance information according to the verified and adjusted evacuation route and uploads user movement status data in real time. Based on the movement status data and real-time fire situation changes, the building fire control system issues path replanning instructions and updates data, and controls elevators, emergency lighting, smoke extraction equipment, fire doors, and emergency broadcast facilities in conjunction with these controls. Upon reaching the safe area, the path planning and guidance equipment stops outputting guidance signals and uploads an arrival signal to the building fire control system. After receiving the signal, the building fire control system updates the user's evacuation status, shuts down redundant emergency equipment, and records the evacuation data.

[0038] During the interaction, when a communication interruption is detected, the path planning and guidance device uses the pre-stored complete BIM model and emergency evacuation strategy to independently complete path planning and guidance, and continuously attempts to reconnect; the building fire control system marks the user status and continuously sends connection requests; after the communication connection is restored, both parties synchronously re-transmit data and correct the strategy; when there is a conflict between the data of the two parties, data verification is initiated; if the conflict still exists, the data of the path planning and guidance device shall prevail, and the building fire control system marks the corresponding sensor as a suspected anomaly and focuses on monitoring it.

[0039] Example 3: Fire Evacuation Route Planning and Guidance Method Based on embodiments 1 and 2 above, embodiment 3 further provides a fire evacuation path planning and guidance method based on flexible wearable devices and UWB sensing. This method is implemented using the flexible wearable fire evacuation path planning and guidance device, building coordination facilities, and interaction mechanism described above. Through steps such as wake-up, positioning, evaluation, planning, and guidance, combined with building facility linkage, it achieves accurate, safe, and efficient evacuation in fire scenarios. The specific process is as follows: Figure 4 As shown: SS1. Wearable verification wake-up: The central control unit receives alarm signals from the building fire control system and activates the functional modules of the path planning and guidance device. It reads the wearability data and, after determining that the wearability meets the standards, activates the environmental sensing module and the positioning communication module. Specifically, after the fire monitoring and alarm module of the building fire control system detects that fire parameters exceed the standards, it sends an alarm command at a frequency of 10Hz to the flexible wearable fire evacuation path planning and guidance device worn by the user via the communication interaction module. The central control unit of the path planning and guidance device listens for alarm signals at a frequency of 10Hz. After identifying signals conforming to the preset protocol format, it immediately activates the modules of the path planning and guidance device to ensure that the device quickly enters the working state. Simultaneously, it reads data from each wearability sensor, sampling each sensor multiple times (preferably 10 times or more) and averaging the results. After determining that the average values ​​of all sensors meet the preset standards... When the fit condition is set, the system confirms that the wearer meets the standards. Once the standards are met, a preheating command is sent to the shape memory alloy skeleton to provide biomimetic support, and a standard signal is transmitted to the building fire control system to avoid affecting the system's function due to improper wear. After receiving the standard signal, the building fire control system sends a preheating command and environmental baseline data to the path planning and guidance device. If the standards are not met, prompts are played repeatedly through bone conduction headphones until the wearer is determined to meet the standards. Then, the environmental perception module, positioning and communication module, flexible drive module, and multimodal interaction module in the path planning and guidance device are activated. At the same time, preheating calibration is initiated to ensure that each module works accurately and stably in the future. After completion, the device enters the working state.

[0040] SS2. Location Data Fusion: The flexible wearable fire evacuation route planning and guidance device uses a positioning and communication module to collect UWB positioning data and IMU attitude data, while an environmental perception module simultaneously collects ambient temperature, smoke concentration, and obstacle distance data. It performs dynamic calibration of UWB positioning thresholds, adaptive avoidance of multi-user interference, and emergency compensation for positioning failures. Through multi-source fusion-risk coupling positioning processing, it generates fused positioning coordinates, a comprehensive environmental risk value, and a risk direction vector. In the event of UWB signal failure, emergency positioning coordinates are used to replace the fused positioning coordinates. Specifically, the environmental perception module collects environmental data such as temperature, smoke concentration, and obstacle distance at a set frequency; the UWB positioning tag of the positioning and communication module collects the original positioning coordinates; the IMU inertial measurement unit collects the wearer's attitude data; the central control unit integrates this data and uploads it to the building system at 50Hz to ensure the building system has real-time access to the situation; the building system calls the latest BIM model data to correct the fire spread trend and regional risk level and feeds it back to the flexible wearable system, providing a more accurate reference for system positioning and route planning.

[0041] The flexible wearable system initiates a dynamic calibration mechanism for UWB positioning thresholds, dynamically adjusting the RSSI switching threshold based on real-time temperature and smoke concentration to better adapt the positioning mode to changes in the fire environment. It uses a UWB multi-user interference adaptive avoidance mechanism to count the number of active UWB tags and dynamically allocate positioning time slots, ensuring a positioning data reception success rate of ≥98% and avoiding signal conflicts during multi-person evacuation. If a complete UWB signal failure is detected, a three-level emergency completion strategy involving IMU inertial estimation, BIM model constraints, and SLAM fusion is activated to generate emergency positioning coordinates, ensuring uninterrupted positioning. Simultaneously, the system executes a multi-source fusion-risk coupled positioning algorithm: preprocessing environmental perception data to calculate a comprehensive environmental risk value; determining risk impact factors through IMU attitude data and deriving the risk direction unit vector; dynamically adjusting the multi-source positioning weighting coefficients based on UWB signal strength, with the weights summed to a fixed value; and finally generating fused positioning coordinates with risk weights (emergency positioning coordinates are used when UWB signals fail), which are uploaded to the building system for record-keeping, providing accurate location and risk data support for subsequent risk assessment and path planning.

[0042] Furthermore, the determination of the risk-weighted fused positioning coordinates P(t) is based on the multi-source fusion-risk coupling positioning formula. The comprehensive environmental risk value R is calculated. total The formula for calculating (t) is: Where ω1, ω2, and ω3 are weighting coefficients for multi-source positioning data, and P UWB (t) represents the original coordinates for UWB positioning, P SLAM (t) represents the SLAM positioning coordinates, ΔP IMU (t) represents the IMU attitude correction. λ As a risk impact factor determined using IMU attitude data, e r Let e ​​be the unit vector in the direction of risk. r Based on the spatial distribution of ambient temperature, smoke concentration, and obstacle distances around the current positioning coordinates, and combined with the current orientation corresponding to the IMU attitude data, a unit vector pointing towards the local high-risk direction is derived. α , β , γ Here, T(t) represents the risk weighting coefficient, S(t) represents the real-time ambient temperature, and T0 and S0 represent the safety thresholds for ambient temperature and smoke concentration, respectively. max S max d represents the extreme hazard thresholds for ambient temperature and smoke concentration. obs (t) represents the distance to the nearest obstacle. ε To prevent zero correction, the fused positioning coordinates P(t) are used when the UWB signal is normal, and the emergency positioning coordinates P(t) are used when the UWB signal fails. emg (t).

[0043] SS3. Environmental Risk Assessment: Based on fused positioning coordinates or emergency positioning coordinates, and combined with lidar point cloud data collected by the environmental perception module, obstacles are segmented and their coordinates, dimensions, and distances are marked; regional risk levels are divided according to preset intervals of comprehensive environmental risk values; grid units are divided with the user's current positioning coordinates as the center and a preset range as the radius, and the risk level of each grid unit is matched; based on the risk direction unit vector and the user's current positioning coordinates, the center coordinates of high-risk areas are derived.

[0044] Specifically, the flexible wearable fire evacuation route planning and guidance device receives refined LiDAR point cloud data collected by the environmental perception module based on risk-weighted fused positioning coordinates or emergency positioning coordinates. Through data processing, it segments obstacles and accurately marks their coordinates, dimensions, and distances, allowing the system to clearly understand the obstacle situation along the route. According to a preset range of comprehensive environmental risk values, the current area is divided into low, medium, and high risk levels. Grid units are divided with the wearer's current positioning coordinates as the center and a preset range as the radius, and each grid unit is matched with a corresponding risk level, making the risk distribution more intuitive. Based on the risk direction unit vector and the current positioning coordinates, the center coordinates of high-risk areas are derived, clarifying the location of the area with the most serious fire hazards and preventing wearers from moving to high-risk areas. Finally, the device integrates fused positioning coordinates, obstacle information, regional risk levels, and center coordinates of high-risk areas to generate a standardized data package, which is uploaded to the building fire control system to provide a comprehensive understanding of the on-site risk distribution and support for collaborative decision-making.

[0045] SS4. Path Planning Decision: The central control unit calls up the pre-stored BIM model data and combines it with the real-time fire data, candidate list of safe target points and data of inaccessible areas issued by the building fire control system to select the safe target point that is closest and meets the risk level standard; it executes the aging-adaptive-dynamic obstacle avoidance cost algorithm to determine the evacuation route and dynamically replans the local obstacle avoidance route according to the environmental changes.

[0046] Specifically, the flexible wearable fire evacuation route planning and guidance device calls up the pre-stored simplified BIM model of the residential building, and combines it with the real-time fire data, candidate list of safe target points and data on inaccessible areas issued by the building fire control system to select the safe target point that is closest and meets the risk level standard, namely the outdoor safe area on the first floor, thereby shortening the evacuation distance and reducing the evacuation risk. The path planning and guidance equipment executes an elderly-adaptive dynamic obstacle avoidance cost algorithm: starting from the current fused positioning coordinates and ending at a selected safe target point, it constructs a path planning space based on the BIM model, marking impassable areas (such as staircases blocked by flames or passages with excessive smoke concentration) to avoid planning dangerous paths; it substitutes all feasible paths into the algorithm to calculate path cost values ​​(the smaller the value, the better the path). The cost calculation comprehensively considers factors such as the Euclidean distance of the path space, the spatial Euclidean distance between the comprehensive environmental risk value and the center coordinates of the high-risk area, the path turning angle, the difficulty coefficient of elderly gait, and the obstacle avoidance response time margin, making the planned path more suitable for the mobility characteristics of the elderly and reducing walking difficulty and physical exertion; it selects the path with the lowest cost as the initial optimal path, outputs the key point coordinates and turning and walking information, and uploads it to the building fire control system for verification; after the building fire control system verifies that there are no conflicts, it provides feedback confirmation information, and the path planning and guidance equipment determines the final path; if there are conflicts, it receives adjustment suggestions from the building fire control system and recalculates to determine the final path. During the evacuation, environmental data is continuously monitored. If any emergencies are detected (such as new obstacles or increased risk levels in local areas), the data is promptly uploaded to the building's fire control system. After receiving adjustment suggestions, local obstacle avoidance routes are dynamically replanned to ensure that the routes are always safe and feasible.

[0047] Furthermore, the mathematical expression for the aging-adaptive-dynamic obstacle avoidance cost algorithm is: Where Cost(t) is the path cost, k d k r k m k o These are the cost weighting coefficients, D(·) represents the spatial Euclidean distance function, and P(t) represents the fused positioning coordinates. goal R represents the coordinates of the evacuation target point. total (t) represents the comprehensive environmental risk value, Z high Let θ be the center coordinate of the high-risk area. turn (t) represents the path turning angle, μ represents the difficulty coefficient of elderly gait, and Δt obs The obstacle avoidance response time margin is δ, which is the zero-prevention correction term.

[0048] SS5. Multimodal Guidance and Building Interconnection: The central control unit generates evacuation decision instructions based on the determined evacuation route, outputs visual, auditory, and tactile guidance information, and controls the building's fire protection facilities in conjunction with the building fire control system until it detects that the user wearing the device has reached a safe area and then stops guiding.

[0049] Specifically, the central control unit of the flexible wearable fire evacuation route planning and guidance device breaks down the final route decision command into visual, auditory, and tactile execution signals, controlling the multimodal interaction module and flexible drive module to execute guidance actions. For visual guidance, a laser projector on the chest projects dynamic arrows along the route, with brightness automatically adjusted according to the ambient light intensity to ensure clear visibility for the wearer. For auditory guidance, bone conduction headphones on the shoulders play noise-reduced navigation voice prompts with a preset priority order of emergency avoidance alerts > route turning alerts > movement status alerts, allowing the wearer to receive key information first. For tactile guidance, twisted artificial muscles in the lower limbs contract to apply guiding torque according to the route requirements, while the corresponding vibration motor activates tactile feedback to assist the wearer in adjusting their direction and pace. These three guidance methods work together to ensure the wearer accurately receives guidance information even in complex environments.

[0050] Meanwhile, based on the real-time location data of wearable users uploaded by the path planning and guidance equipment, the building fire control system coordinates and controls related facilities: activating emergency lighting along evacuation routes to ensure well-lit passageways; activating smoke extraction equipment along the routes to accelerate smoke removal; and closing fire doors around fire areas and evacuation routes to prevent the spread of fire and smoke, creating a safe evacuation environment for wearable users. The building system receives the movement status data of wearable users at a frequency of 20Hz, continuously monitoring changes in the fire situation. If new risks emerge along the original route, a replanning instruction and updated data are immediately issued. Upon receiving the instruction, the path planning and guidance equipment quickly adjusts the local route, and the building fire control system synchronously adjusts the working status of the linked facilities. If the system detects that the wearable user's movement speed is too slow, the flexible wearable system uploads a signal, and the building fire control system triggers an emergency broadcast to alert nearby personnel to provide assistance. Once the wearable user's location coordinates fall within the first-floor outdoor safe zone, the path planning and guidance equipment stops all guidance signal output and uploads a signal indicating arrival at the safe zone to the building's fire control system. Upon receiving the signal, the building's fire control system updates the wearable user's evacuation status (i.e., safe evacuation), shuts down redundant emergency facilities along the evacuation route (such as excess emergency lighting and smoke extraction equipment) to avoid resource waste, and records the evacuation time, route, and environmental data to provide a reference for subsequent optimization of evacuation strategies. The entire evacuation process is then complete.

[0051] The objectives of this invention have been fully and effectively achieved through the above embodiments. Those skilled in the art will understand that this invention includes, but is not limited to, the contents described in the accompanying drawings and the specific embodiments described above. Although the invention has been described with reference to what is currently considered the most practical and preferred embodiments, it should be understood that the invention is not limited to the disclosed embodiments, and any modifications that do not depart from the functional and structural principles of the invention will be included within the scope of the claims.

Claims

1. A flexible wearable fire evacuation path planning and guidance device based on UWB sensing, characterized in that, It includes a flexible wearable body and a central control unit, a positioning and communication module, an environmental sensing module, a flexible driving module, and a multimodal interaction module distributed on the flexible wearable body, wherein: The flexible wearable body is an open vest structure made of multi-layer composite flexible fabric and extending from the shoulders to the thighs. It is also provided with at least a shape memory alloy skeleton for providing biomimetic support and a wear fit sensor for detecting the wear fit status. The environmental perception module includes distributed lidar, temperature sensors and smoke concentration sensors, which are used to collect real-time information on obstacles, ambient temperature and smoke concentration around the wearer. The positioning and communication module includes a UWB positioning tag and an IMU inertial measurement unit, which are used to collect the location data and IMU attitude data of the wearer, and to execute a UWB positioning threshold dynamic calibration mechanism based on environmental parameters, a UWB multi-user interference adaptive avoidance mechanism, and a three-level emergency completion mechanism that combines IMU inertial estimation, BIM constraints and SLAM positioning when the UWB signal fails. The central control unit is connected to each module and the building fire control system. It is used to generate fused positioning coordinates, comprehensive environmental risk values ​​and risk direction vectors with risk weights by fusion positioning and environmental data, and to generate evacuation decision instructions based on evacuation targets and impassable areas. The flexible drive module includes artificial muscles distributed along the lower limbs and an array of vibration motors distributed in the shoulder and waist areas, which are used to output auxiliary driving force and directional tactile feedback, respectively. The multimodal interaction module includes a laser projector and bone conduction headphones, which are used to output visual and auditory guidance information based on evacuation decisions, and work in conjunction with the flexible drive module to complete multimodal evacuation guidance.

2. The flexible wearable fire evacuation path planning and guidance device based on UWB sensing according to claim 1, characterized in that, The flexible wearable body features multi-segment telescopic structures for its shoulder straps, waist straps, and thigh straps. The multi-layered composite flexible fabric, from the outside in, includes a fire-resistant outer layer, a heat-insulating middle layer, a flexible circuit and driver embedding layer, and a skin-friendly and breathable inner lining. The shape memory alloy skeleton provides biomimetic support and is distributed along the front and back shoulder straps and side waist of the flexible wearable body, as well as along the sides of the spine, the transition area from the shoulder to the hip, and the outer thigh force path. The flexible wearable body also features multiple magnetic self-locking buckles for quick on / off and locking, distributed at the shoulder, chest, waist, and thigh strap connections. Wear fit sensors are installed on the contact surfaces between the flexible wearable body and the human shoulder, back, and waist to collect real-time wear fit data.

3. The flexible wearable fire evacuation path planning and guidance device based on UWB sensing according to claim 1, characterized in that, In the environmental perception module, a lidar placed on the shoulder is used to detect obstacles in the area facing diagonally upward, and a lidar placed on the waist is used to detect obstacles in the area facing forward; a temperature sensor covers at least the temperature field of the shoulder, chest and outer waist area, and a smoke concentration sensor covers at least the near-field breathing area in front of the chest; the central control unit establishes the directional mapping relationship of environmental data according to the spatial installation position of the lidar, temperature sensor and smoke concentration sensor, and converts the obstacle distance, temperature gradient and smoke concentration gradient into risk direction vectors corresponding to the human body's orientation.

4. The flexible wearable fire evacuation path planning and guidance device based on UWB sensing according to claim 1, characterized in that, The positioning and communication module is integrated in the back area on the same side as the central control unit. The UWB positioning tag adopts a dual-antenna differential receiving structure and obtains positioning information between itself and the preset UWB base station array in the building based on the TDOA method. The IMU inertial measurement unit is used to collect the attitude data of the wearer in real time. The UWB positioning threshold dynamic calibration mechanism establishes a mapping relationship between ambient temperature, smoke concentration, and UWB signal attenuation, and dynamically adjusts the RSSI switching threshold. adj When RSSI < RSSI adj And when the duration exceeds the first preset duration, switch to SLAM positioning mode, when RSSI ≥ RSSI adj And if the duration exceeds the second preset duration, it will switch back to UWB positioning mode; and among them, RSSI adj satisfy T(t) and S(t) are the ambient temperature and smoke concentration collected in real time, respectively, and T0 and T... max These are the safe threshold and extreme danger threshold for ambient temperature, S0 and S1, respectively. max κ1 and κ2 are the safe threshold and extreme danger threshold of smoke concentration, respectively; κ1 and κ2 are the temperature influence coefficient and smoke concentration influence coefficient calibrated in the experiment, respectively; RSSI0 is the initial fixed threshold. The UWB multi-user interference adaptive avoidance mechanism is designed for scenarios where multiple positioning and communication modules work concurrently within the same building space. Based on the TDOA positioning algorithm, it uses a dynamic time slot allocation method to count the number of active UWB positioning tags in real time and dynamically adjust the UWB positioning time slot length, time slot allocation order, and tag access scheduling strategy to reduce mutual interference of UWB signals during group evacuation. The three-level emergency replenishment mechanism, upon detecting a UWB signal failure, uses the most recent valid UWB positioning coordinates as the initial reference coordinates and initiates a three-level emergency replenishment positioning strategy based on IMU inertial estimation, BIM constraints, and SLAM fusion. First, based on the real-time acceleration, angular velocity, and attitude data output by the IMU, inertial estimation is performed to obtain the predicted displacement of the wearer's current position, and this is combined with the initial reference coordinates to generate initial emergency prediction coordinates. Then, based on pre-stored BIM model data, spatial constraints are corrected on the initial emergency prediction coordinates, eliminating abnormal coordinates falling into impassable areas to obtain BIM constraint-corrected coordinates. Finally, SLAM positioning is performed based on the point cloud data collected in real-time by the lidar in the environmental perception module, obtaining SLAM-corrected coordinates, which are then fused with the BIM constraint-corrected coordinates to generate the emergency positioning coordinates P. emg (t).

5. The flexible wearable fire evacuation path planning and guidance device based on UWB sensing according to claim 4, characterized in that, The central control unit is integrated into the back of the flexible wearable body and is configured to perform multi-source fusion-risk coupling positioning processing: The environmental sensing module collects real-time data on ambient temperature T(t), smoke concentration S(t), and distance d from the nearest obstacle. obs (t) The data is preprocessed and the comprehensive environmental risk value R is calculated. total Risk impact factors are determined based on IMU attitude data. λ And derive the risk direction unit vector e r Simultaneously, when the UWB signal is normal, the original coordinates P of the UWB positioning are determined based on the UWB signal strength. UWB (t), SLAM positioning coordinates P SLAM (t) and IMU attitude correction ΔP IMU The multi-source positioning weighting coefficients corresponding to (t) are dynamically adjusted to generate fused positioning coordinates P(t) with risk weights. In the event of UWB signal failure, the emergency positioning coordinates P generated by the positioning communication module are used. emg (t) replaces the fusion positioning coordinates P(t); and based on this, according to the comprehensive environmental risk value R total (t) defines the risk level of the area within the preset interval, constructs a risk grid cell centered on the current fused or emergency positioning coordinates, and matches the risk level of each grid cell, combined with the risk direction unit vector e. r The center coordinates of the high-risk area are derived from the current location coordinates.

6. The flexible wearable fire evacuation path planning and guidance device based on UWB sensing according to claim 5, characterized in that, The central control unit is also configured to call the floor topology, passage geometry and stair connectivity in the pre-stored BIM, and combine the real-time fire data, candidate list of safe target points and data of impassable areas issued by the building fire control system to perform elderly-adaptive dynamic obstacle avoidance cost path planning: the path planning comprehensively considers the spatial distance between the current fused or emergency positioning coordinates and the evacuation target point, the spatial relationship between the comprehensive environmental risk value and the center coordinates of the high-risk area, the path turning angle, the difficulty coefficient of elderly gait and the obstacle avoidance response time margin, to select the path with the lowest cost as the initial optimal path, and to perform local dynamic replanning when new obstacles or local risk increases are detected.

7. The flexible wearable fire evacuation path planning and guidance device based on UWB sensing according to claim 1, characterized in that, In the flexible drive module, the artificial muscles are twisted and coiled artificial muscles distributed along the muscle groups of the lower limbs. By adjusting the driving timing, driving amplitude, and duration of the artificial muscles, a flexible auxiliary driving force is output. The vibration motor array is arranged in a symmetrical and layered manner on the shoulders and waist to form a directional tactile output matrix. According to the evacuation decision command, the vibration motors on each corresponding side are driven independently to output directional tactile feedback. In the multimodal interaction module, the laser projector is fixed at the chest position and supports dynamic arrow projection. The bone conduction headphones are fixed on the outside of the shoulder and integrate a noise reduction chip and a pre-stored navigation voice storage unit.

8. The flexible wearable fire evacuation path planning and guidance device based on UWB sensing according to claim 6, characterized in that, The building fire control system includes at least a fire monitoring and alarm module, a data storage and management module, a communication and interaction module, an evacuation decision support module, and a linkage control module. Specifically: the fire monitoring and alarm module collects fire parameters from various areas of the building in real time and sends alarm signals to the path planning and guidance equipment, while simultaneously activating audible and visual alarm devices within the building; the data storage and management module stores BIM model data, historical fire data, sensor calibration parameters, wearable user adaptation information, and updates dynamic fire data in real time; the communication and interaction module enables bidirectional data interaction with the path planning and guidance equipment; the evacuation decision support module generates a candidate list of safe target points and information on impassable areas based on fire monitoring data, BIM model data, and user status data uploaded by the path planning and guidance equipment, and sends path adjustment commands to the path planning and guidance equipment; and the linkage control module controls building facilities in conjunction with the commands output by the evacuation decision support module.

9. The flexible wearable fire evacuation route planning and guidance device based on UWB positioning perception according to claim 8, characterized in that, The interaction between the path planning and guidance equipment and the building fire control system includes: During the fire triggering phase, after the fire monitoring and alarm module detects that the fire parameters exceed the standard, it sends an alarm signal through the communication interaction module to wake up the path planning and guidance equipment. The central control unit reads the wear fit sensor data and determines that the wear meets the standard before uploading it to the building fire control system. During the location assessment phase, the path planning and guidance equipment uploads location, environmental perception and user status data. The building fire control system calls the BIM model data to correct and issue the fire spread trend and regional risk level. Based on this, the path planning and guidance equipment performs dynamic calibration of the location threshold, adaptive avoidance of multi-user interference, and emergency completion of location failure to generate fused or emergency location coordinates. During the route planning phase, the route planning guidance device uploads the fusion or emergency positioning coordinates and comprehensive environmental risk values. The building fire control system issues a candidate list of safety target points and information on impassable areas. Based on the issued information, the route planning guidance device executes evacuation route planning and uploads it to the building fire control system for verification. If there are conflicts, route adjustment suggestions are issued. During the guidance and execution phase, the path planning and guidance equipment outputs visual, auditory, and tactile guidance information according to the verified and adjusted evacuation path and uploads user movement status data in real time. Based on the movement status data and real-time fire situation changes, the building fire control system issues path replanning instructions and updates data, and controls elevators, emergency lighting, smoke extraction equipment, fire doors, and emergency broadcast facilities in conjunction with these systems.

10. A fire evacuation route planning and guidance method based on flexible wearable devices and UWB sensing, based on the flexible wearable fire evacuation route planning and guidance device based on UWB positioning sensing as described in any one of claims 1 to 9, characterized in that, It should include at least the following steps: SS1. The path planning and guidance device receives alarm signals from the building fire control system, reads the wear fit status data, and activates the environmental perception module and positioning communication module after determining that the wear meets the standards. SS2. Based on the collected UWB positioning data, IMU attitude data, and ambient temperature, smoke concentration and obstacle distance data, perform UWB positioning threshold dynamic calibration, multi-user interference adaptive avoidance and positioning failure emergency completion, generate fused positioning coordinates, comprehensive environmental risk value and risk direction vector, and use emergency positioning coordinates to replace fused positioning coordinates when UWB signal fails. SS3. Based on fused or emergency positioning coordinates, and combined with lidar point cloud data collected by the environmental perception module, obstacles are segmented and their coordinates, dimensions, and distances are marked; the regional risk level is divided according to the preset range of comprehensive environmental risk value. Divide the system into grid cells centered on the user's current location coordinates and with a preset range as the radius, and match the risk level of each grid cell. Based on the risk direction unit vector and the user's current location coordinates, the center coordinates of the high-risk area are derived. SS4. Call the pre-stored BIM model data and combine it with the real-time fire data, candidate list of safe target points and data of inaccessible areas issued by the building fire control system to select the safe target points that are closest and meet the risk level; execute the aging-adaptive-dynamic obstacle avoidance cost algorithm to determine the evacuation route, and dynamically replan the local obstacle avoidance route according to the environmental changes; SS5. Generate evacuation decision instructions based on the evacuation route, output visual, auditory, and tactile guidance information, and link with the control of building fire protection facilities until the guidance stops after detecting that the wearer has reached a safe area.

11. The method according to claim 10, characterized in that, In step SS1, the central control unit listens to the alarm signals sent by the building fire control system at a fixed frequency and only responds to alarm signals that conform to the preset protocol format to wake up the path planning and guidance equipment; it reads the data of each wearable fit sensor, takes the average value after sampling each sensor multiple times, and determines that the wear meets the preset fit conditions when the average value of all sensors meets the preset fit conditions. After meeting the conditions, it sends a preheating command to the shape memory alloy skeleton to provide bionic support force, and at the same time sends feedback to the building fire control system to indicate that the wear meets the conditions and the preheating is complete. If the conditions are not met, it plays prompt information in a loop through the bone conduction headphones until the wear meets the conditions, and then starts the environmental perception module, positioning and communication module, flexible drive module and multimodal interaction module to enter the linkage working state.

12. The method according to claim 10, characterized in that, In step SS2, the multi-source fusion-risk coupling localization formula is used. Determine the risk-weighted fused positioning coordinates P(t) and the comprehensive environmental risk value R. total The formula for calculating (t) is: Where ω1, ω2, and ω3 are weighting coefficients for multi-source positioning data, and P UWB (t) represents the original coordinates for UWB positioning, P SLAM (t) represents the SLAM positioning coordinates, ΔP IMU (t) represents the IMU attitude correction. λ As a risk impact factor determined using IMU attitude data, e r Let e ​​be the unit vector in the direction of risk. r Based on the spatial distribution of ambient temperature, smoke concentration, and obstacle distances around the current positioning coordinates, and combined with the current orientation corresponding to the IMU attitude data, a unit vector pointing towards the local high-risk direction is derived. α , β , γ Here, T(t) represents the risk weighting coefficient, S(t) represents the real-time ambient temperature, and T0 and S0 represent the safety thresholds for ambient temperature and smoke concentration, respectively. max S max d represents the extreme hazard thresholds for ambient temperature and smoke concentration. obs (t) represents the distance to the nearest obstacle. ε To prevent zero correction, the fused positioning coordinates P(t) are used when the UWB signal is normal, and the emergency positioning coordinates P(t) are used when the UWB signal fails. emg (t).

13. The method according to claim 12, characterized in that, In step SS4, the mathematical expression of the aging adaptation-dynamic obstacle avoidance cost algorithm is: Where Cost(t) is the path cost, k d k r k m k o These are the cost weighting coefficients, D(·) represents the spatial Euclidean distance function, and P(t) represents the fused positioning coordinates. goal R represents the coordinates of the evacuation target point. total (t) represents the comprehensive environmental risk value, Z high Let θ be the center coordinate of the high-risk area. turn (t) represents the path turning angle, μ represents the difficulty coefficient of elderly gait, and Δt obs The obstacle avoidance response time margin is δ, which is the zero-prevention correction term. Furthermore, during path planning, the current location coordinates are used as the starting point and the selected safe target point is used as the ending point. Based on the BIM model data, a path planning space is constructed and inaccessible areas are marked. All feasible paths are substituted into the elderly-adaptive-dynamic obstacle avoidance cost algorithm to calculate the path cost value, and the path with the lowest cost value is selected as the initial optimal path. When new obstacles are detected, the local risk level increases, or the building fire control system issues path adjustment information, the local planning space is reconstructed with the updated current location coordinates, and local dynamic replanning is performed while retaining the existing accessible path segments.

14. The method according to claim 10, characterized in that, In step SS5, the central control unit breaks down the evacuation decision command into visual execution signals, auditory execution signals, and tactile execution signals, and controls the multimodal interaction module and flexible drive module to perform guidance actions respectively. Among them, the visual execution signal is used to control the laser projector to project dynamic arrows in the direction of the evacuation path, the auditory execution signal is used to control the bone conduction headphones to play navigation voice, and the tactile execution signal is used to control the artificial muscle contraction to apply guiding torque and control the corresponding side vibration motor to start tactile feedback. At the same time, the building fire control system controls the emergency lighting, smoke exhaust equipment and fire doors on the evacuation path according to the user's current location, and stops guidance and records the evacuation status after detecting that the user's positioning coordinates have fallen into the safe area.

Images