Multi-source positioning data fusion method for fire protection clothing

Abstract

A multi-source positioning data fusion method for a fire protection suit relates to the technical field of fire scene positioning, a miniature UWB base station is deployed in a target area, a multi-source sensing terminal is deployed on the fire protection suit, and the multi-source sensing terminal is used for collecting multi-source heterogeneous data, constructing a three-dimensional model and generating three-dimensional position information of the multi-source sensing terminal in the three-dimensional model; performing time difference analysis on the ultra-wideband signal in the multi-source heterogeneous data, and judging whether a distortion interference event is generated or not according to an analysis result; when the ultra-wideband signal generates a distortion interference event, carrying out collaborative feature analysis on the multi-source heterogeneous data to obtain a key feature, a first feature and a second feature, and carrying out critical positioning matching on the ultra-wideband signal based on the key feature, the first feature, the second feature and the three-dimensional position information, and according to a critical positioning matching result, positioning deviation compensation operation, critical characteristic internet surfing operation or rapid link selection operation is carried out, so that accurate and reliable personnel position information support is provided for fire rescue.

Description

Technical Field

[0001] This invention relates to the field of fire scene positioning technology, specifically a method for fusing multi-source positioning data of fire protective clothing. Background Technology

[0002] Chinese patent CN118764821A discloses a method for locating firefighters in a fire scene, comprising: deploying smoke detectors within a protected area, including a smoke sensing module and a positioning module; when the smoke sensing module detects smoke, in addition to issuing an alarm, it also activates the positioning module and battery-powered mode, and the positioning module periodically broadcasts positioning signals into the space; firefighters carry positioning terminals and enhanced positioning equipment into the fire area, and deploy enhanced positioning equipment along their route; the enhanced positioning equipment receives the positioning signals broadcast by the smoke detectors for self-positioning, and then broadcasts the positioning signals into the space; the positioning terminal receives the positioning signals broadcast by the smoke detectors and the enhanced positioning equipment, performs position fusion calculation and Kalman filtering on itself, uses the filtering result as the position estimate of the positioning terminal, and transmits it back to the command center through an emergency communication network.

[0003] The aforementioned solutions and existing technologies face several technical challenges in positioning fire protective suits: In complex building environments, ultra-wideband (UWB) signals are easily distorted by wall obstruction and multipath effects, making it difficult for existing technologies to accurately identify such distortion interference events, leading to significant positioning deviations; the collaborative utilization of multi-source heterogeneous data (inertial navigation data, UWB signals, and environmental data) is insufficient. Relying solely on UWB signals results in unreliable positioning when the signal is interrupted by interference, while relying solely on inertial navigation data leads to a sharp decline in positioning accuracy over long periods due to error accumulation; there is a lack of dynamic positioning deviation compensation mechanisms and scene feature self-learning capabilities, making it impossible to quickly correct current positioning deviations based on historical interference features, and difficult to incorporate new interference scene features into the database to optimize subsequent positioning; furthermore, when UWB signals are subject to prolonged interference, the optimal communication frequency band cannot be adaptively selected based on real-time scene features, making it difficult to guarantee positioning stability and continuity, and failing to meet the real-time and accurate positioning needs of firefighters in fire rescue operations. Summary of the Invention

[0004] To address the aforementioned technical problems, the present invention aims to provide a method for fusing multi-source positioning data of fire protective clothing, comprising the following steps: Step s1: Deploy a micro UWB base station in the target area and deploy a multi-source sensor terminal on the fire protective suit. The multi-source sensor terminal is used to collect multi-source heterogeneous data and mark the collection time, obtain the prior data of the building structure in the target area, build a three-dimensional model, and generate the three-dimensional position information of the multi-source sensor terminal in the three-dimensional model. Step s2: Perform time difference analysis on the ultra-wideband signal in the multi-source heterogeneous data, and determine whether distortion interference events are generated based on the analysis results; Step s3: When the ultra-wideband signal generates distortion interference events, perform collaborative feature analysis on multi-source heterogeneous data to obtain key features, first features, and second features. Based on the key features, first features, second features, and three-dimensional position information, perform critical positioning matching on the ultra-wideband signal, and perform positioning deviation compensation operation, critical feature Internet access operation, or fast link selection operation according to the critical positioning matching results.

[0005] Furthermore, the micro UWB base station communicates with multi-source sensor terminals to form a self-organizing network, and the multi-source heterogeneous data includes inertial navigation data, ultra-wideband signals and environmental data.

[0006] Furthermore, the process of acquiring prior data on the building structure of the target area, constructing a 3D model, and generating the stereoscopic position information of the multi-source sensing terminals within the 3D model includes: The system acquires prior data on the building structure of the target area, constructs a 3D model of the target area based on the prior data, and annotates the 3D model. It then generates the planar coordinates of the multi-source sensor terminal in the 3D model based on ultra-wideband (UWB) signals and inertial navigation data. (When no UWB signal distortion interference event occurs, the planar coordinates of the multi-source sensor terminal in the 3D model are generated based on the UWB signal; when a UWB signal distortion interference event occurs, the planar coordinates of the multi-source sensor terminal in the 3D model are generated based on the inertial navigation data during the duration of the distortion interference event.) Simultaneously, it acquires the elevation coordinates of the multi-source sensor terminal in the 3D model based on inertial navigation data and environmental data. Finally, it merges the planar coordinates and elevation coordinates to generate the stereoscopic position information of the multi-source sensor terminal in the 3D model.

[0007] Furthermore, the process of performing time-difference analysis on ultra-wideband signals in multi-source heterogeneous data and determining whether distortion interference events have occurred based on the analysis results includes: The system acquires the time difference data of the ultra-wideband signal, presets a normal threshold range for the time difference data of the ultra-wideband signal, compares the time difference data of the ultra-wideband signal at the current moment with the normal threshold range, and if the time difference data of the ultra-wideband signal is not within the normal threshold range, then the ultra-wideband signal at the current moment is determined to be a distortion interference event.

[0008] Furthermore, when a distortion interference event is generated, the process of performing collaborative feature analysis on multi-source heterogeneous data to obtain key features, first features, and second features includes: Set the analysis window for the time of the distortion interference event, perform waveform analysis on the ultra-wideband signal within the analysis window, and obtain the waveform feature sequence of the ultra-wideband signal within the analysis window; The system acquires the numerical time series sequences of various indicators in the inertial navigation data and environmental data within the analysis window. It performs time series feature analysis on the numerical time series sequences of each type of indicator, obtains the cooperative change features corresponding to the inertial navigation data and the cooperative change features corresponding to the environmental data, and marks the waveform feature sequence and time difference data corresponding to the ultra-wideband signal as key features. The cooperative change features corresponding to the inertial navigation data are marked as the first feature, and the cooperative change features corresponding to the environmental data are marked as the second feature.

[0009] Furthermore, the process of critical positioning matching for ultra-wideband signals includes: The key features, first feature, second feature, and stereo position information within the analysis window at the time of the distortion interference event are obtained. The key features, first feature, second feature, and stereo position information are then input into the critical database for matching, and the feature similarity of each critical feature package in the critical database is obtained. A preset feature similarity threshold is set. If the feature similarity of each critical feature package in the critical database is less than the feature similarity threshold, the critical location matching will fail. If there exists a critical feature packet with a feature similarity greater than or equal to the feature similarity threshold, then the positioning deviation compensation factor in the critical feature packet is used to compensate for the positioning deviation of the planar coordinates generated from the ultra-wideband signal.

[0010] Furthermore, the process of performing positioning deviation compensation, critical feature-based internet access, or fast link selection operations based on the critical positioning matching results includes: If the critical positioning match is successful, a positioning deviation compensation operation is performed on the planar coordinates generated by the ultra-wideband signal. If the critical positioning match fails, an inertial error duration window is set at the time of the distortion interference event, the duration of the distortion interference event generated by the ultra-wideband signal is obtained, and the planar coordinates of the multi-source sensing terminal in the three-dimensional model are generated based on the inertial navigation data within the duration of the distortion interference event. Based on the planar coordinates, the three-dimensional position information of the multi-source sensing terminal in the three-dimensional model is obtained. If the duration of the ultra-wideband signal distortion interference event is less than the inertial error duration window, then at the end timestamp of the distortion interference event, the planar coordinates of the multi-source sensor terminal in the 3D model generated by the ultra-wideband signal are selected. Based on the planar coordinates, the three-dimensional position information of the multi-source sensor terminal in the 3D model is obtained, and the critical feature Internet access operation is performed. If the duration of the ultra-wideband signal distortion interference event is greater than or equal to the inertial error duration window, then at the end timestamp of the inertial error duration window, the fast link selection operation is performed.

[0011] Furthermore, the process of performing borderline internet access operations includes: A critical database is constructed within a micro UWB base station to obtain key features, first features, second features, and stereo position information for each analysis window during the duration of distortion interference events. Positioning deviation compensation analysis is performed on the key features and stereo position information of each analysis window to obtain the positioning deviation compensation factor. The key features, first feature, second feature, stereo position information, and positioning deviation compensation factor of each analysis window during the duration of the distortion interference event are packaged into a critical feature package and uploaded to the critical database.

[0012] Furthermore, the process of obtaining the positioning deviation compensation factor includes: Based on the time difference data in the key features within the analysis window, obtain the planar coordinates of the multi-source sensing terminal in the 3D model based on the start and end timestamps of the analysis window. Obtain the planar coordinate vector based on the planar coordinates of the multi-source sensing terminal in the 3D model based on the start and end timestamps, and mark the planar coordinate vector as the first vector. Obtain the planar coordinates corresponding to the start and end timestamps of the 3D position information within the analysis window; obtain the planar coordinate vector based on the planar coordinates corresponding to the start and end timestamps of the 3D position information within the analysis window; and mark the planar coordinate vector as the second vector. The positioning deviation compensation factor is obtained based on the first vector and the second vector.

[0013] Furthermore, the process of performing fast link selection includes: A UWB frequency band feature library for different scenarios is pre-trained. The UWB frequency band feature library includes the optimal frequency bands corresponding to different key features, the first feature, and the second feature. The stereo position information of the multi-source sensing terminal in the three-dimensional model is obtained at the end timestamp of the inertial error duration window. The scene corresponding to the multi-source sensing terminal in the three-dimensional model is obtained based on the stereo position information. A UWB band feature library consistent with the scene corresponding to the multi-source sensing terminal in the 3D model is extracted. Key features, first features, and second features of each analysis window within the inertial error duration window are obtained. The key features, first features, and second features are matched with the UWB band feature library. The cosine similarity method is used to obtain the key features, first features, and second features with the highest cosine similarity to the key features, first features, and second features of each analysis window in the UWB band feature library. The optimal frequency band corresponding to the key features, first features, and second features with the highest cosine similarity is obtained, and the original frequency band of the ultra-wideband signal is replaced with the optimal frequency band.

[0014] Compared with the prior art, the beneficial effects of the present invention are: 1. By constructing a 3D model and generating stereo positioning information, and combining ultra-wideband signals, inertial navigation data, and environmental data, the limitations of traditional single-data source positioning are overcome. The joint calculation of planar and elevation coordinates enables precise positioning in 3D space, solving the problem of disconnect between planar positioning and elevation information in existing technologies. To address ultra-wideband signal distortion interference, critical positioning matching and deviation compensation operations can dynamically correct positioning errors caused by multipath effects and wall obstruction.

[0015] 2. An innovative time difference analysis and distortion event judgment mechanism is designed to identify signal anomalies in real time and trigger collaborative feature analysis, avoiding positioning failures caused by signal interference in existing technologies. When the ultra-wideband signal is continuously disturbed, positioning continuity is ensured through dynamic switching between temporary replacement of inertial navigation data and rapid link selection: for short-term interference (less than the inertial error window), the database is updated using critical features to achieve self-learning of the interference scenario; for long-term interference, the optimal frequency band is switched based on scene matching, solving the pain point of easy positioning interruption in complex building environments.

[0016] 3. A collaborative analysis framework is constructed, comprising key features (ultra-wideband signals), primary features (inertial navigation), and secondary features (environmental data), enabling complementary utilization of heterogeneous data: ultra-wideband signals provide an absolute position reference, inertial navigation compensates for signal interruptions, and environmental data (such as temperature and humidity) assists in scene recognition. Compared to the isolated data processing model in existing technologies, this approach provides multi-dimensional verification criteria for positioning results through the correlation analysis of three-dimensional position information and multiple features, reducing the misjudgment rate caused by single data anomalies.

[0017] 4. The dual-database design of the critical database and the UWB band feature database endows the system with self-evolution capabilities: the dynamic uploading of critical feature packets allows the database to accumulate compensation factors for various interference scenarios, and the positioning deviation compensation accuracy can be continuously optimized as the data volume increases; the pre-trained band feature database combined with real-time scene matching supports rapid adaptation to different building environments, avoiding the problem of repeated debugging for specific scenarios required by traditional methods. In addition, the self-organizing network architecture makes the deployment of micro base stations and sensing terminals more flexible and can adapt to diverse scenarios from small buildings to large complexes.

[0018] 5. Enhanced practicality in fire rescue scenarios: Addressing the real-world needs of firefighters, this method controls positioning response latency to the millisecond level. Automated handling of distortion event detection and switching operations reduces manual intervention costs. Three-dimensional location information accurately reflects the movement trajectory of firefighters in three-dimensional space such as floors and corridors. Combined with environmental data (e.g., high-temperature areas), it provides command centers with more comprehensive decision-making support. Compared to existing two-dimensional positioning methods, it better meets real-world needs, significantly improving the safety and efficiency of fire rescue. Attached Figure Description

[0019] Figure 1 This is a schematic diagram of a multi-source positioning data fusion method for fire protective clothing according to an embodiment of this application. Detailed Implementation

[0020] The technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of this application. Based on the embodiments of this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.

[0021] like Figure 1 As shown, a method for fusing multi-source positioning data of fire protective clothing includes the following steps: Step s1: Deploy a micro UWB base station in the target area and deploy a multi-source sensor terminal on the fire protective suit. The multi-source sensor terminal is used to collect multi-source heterogeneous data and mark the collection time, obtain the prior data of the building structure in the target area, build a three-dimensional model, and generate the three-dimensional position information of the multi-source sensor terminal in the three-dimensional model. Step s2: Perform time difference analysis on the ultra-wideband signal in the multi-source heterogeneous data, and determine whether distortion interference events are generated based on the analysis results; Step s3: When the ultra-wideband signal generates distortion interference events, perform collaborative feature analysis on multi-source heterogeneous data to obtain key features, first features, and second features. Based on the key features, first features, second features, and three-dimensional position information, perform critical positioning matching on the ultra-wideband signal, and perform positioning deviation compensation operation, critical feature Internet access operation, or fast link selection operation according to the critical positioning matching results.

[0022] It should be further explained that in the specific implementation process, the micro UWB base station communicates with the multi-source sensor terminal to form a self-organizing network. For example, micro UWB base stations (such as those carried by firefighting robots) are pre-deployed at the fire scene to form a "mobile base station network". The multi-source sensor terminal includes a UWB positioning module, an IMU inertial navigation module, a barometer, a temperature sensor, a gas sensor, and a thermocouple. The multi-source heterogeneous data includes inertial navigation data (the IMU inertial navigation module integrates a three-axis accelerometer and a three-axis gyroscope), which collects the firefighter's motion status in real time (such as walking acceleration and turning angular velocity, used to obtain relative position changes through dead reckoning (PDR)), ultra-wideband signals (including time difference data: by interacting with the UWB positioning module on the fire protective clothing through the UWB base station deployed in the target area, the signal arrival time difference (TDoA) is obtained, which is used to calculate the planar coordinates (X, Y) of the tag (firefighter)) and environmental data (including air pressure, temperature, smoke concentration, etc.).

[0023] It should be further explained that, in the specific implementation process, the process of acquiring prior data on the building structure of the target area, constructing a 3D model, and generating the stereoscopic position information of the multi-source sensing terminals in the 3D model includes: The system acquires prior data on the building structure of the target area, constructs a 3D model of the target area based on this prior data (generating physical entities of the building structure in physical space based on the current prior data, performing 3D modeling on these physical entities to generate a 3D model), and annotates the 3D model with scene data (annotating the corresponding building structures, such as rooms, corridors, metal escalators, etc.). It then generates the planar coordinates of the multi-source sensor terminal in the 3D model based on ultra-wideband signals and inertial navigation data (wherein, if no distortion interference event is generated by the ultra-wideband signal, the planar coordinates of the multi-source sensor terminal in the 3D model are generated based on the ultra-wideband signal; if a distortion interference event is generated by the ultra-wideband signal, the planar coordinates of the multi-source sensor terminal in the 3D model are generated based on the inertial navigation data during the duration of the distortion interference event). Simultaneously, it acquires the elevation coordinates of the multi-source sensor terminal in the 3D model based on inertial navigation data and environmental data, and merges the planar coordinates and elevation coordinates to generate the stereo position information of the multi-source sensor terminal in the 3D model.

[0024] The specific process of obtaining the elevation coordinates of the multi-source sensing terminal in the 3D model based on inertial navigation data and environmental data is as follows: ; in, The height after physical calibration. The original height measured by the barometer. Temperature compensation coefficient, The motion noise figure was calibrated experimentally. The change in temperature This is the vertical acceleration.

[0025] It should be further explained that, in the specific implementation process, the process of performing time difference analysis on ultra-wideband signals in multi-source heterogeneous data and determining whether distortion interference events have been generated based on the analysis results includes: The time difference data of the ultra-wideband signal is acquired. A normal threshold range for the time difference data of the ultra-wideband signal is preset (based on the fluctuation range of historical interference-free data). The time difference data of the ultra-wideband signal at the current moment is compared with the normal threshold range. If the time difference data of the ultra-wideband signal is not within the normal threshold range, the ultra-wideband signal at the current moment is determined to be a distortion interference event.

[0026] It should be further explained that, in the specific implementation process, when generating distortion interference events, the process of performing collaborative feature analysis on multi-source heterogeneous data to obtain key features, primary features, and secondary features includes: Set an analysis window for the time of the distortion interference event (e.g., within a time period of 5 times before and after the time of the distortion interference event), perform waveform analysis on the ultra-wideband signal within the analysis window, and obtain the waveform feature sequence of the ultra-wideband signal within the analysis window (waveform features include waveform peak offset and fluctuation amplitude). The waveform features reflect the intensity and pattern of multipath interference (e.g., metal reflection may cause sharp abrupt changes in the waveform, while concrete reflection may cause smooth tailing of the waveform). The system acquires the time-series numerical values ​​of various indicators (including acceleration and angular velocity in inertial navigation data, and temperature and smoke concentration in environmental data) from the inertial navigation data and environmental data within the analysis window. It then performs time-series feature analysis on these numerical time-series values ​​to obtain the coordinated change features corresponding to the inertial navigation data and the environmental data. This includes the coordinated change features of the inertial navigation data within the analysis window (e.g., whether the firefighter is stationary, turning, or accelerating—if the firefighter is stationary, a sudden change in UWB time difference is more likely to be multipath interference than a change in actual position) and the coordinated change features of the environmental data (e.g., current temperature and temperature conversion rate, current air pressure and air pressure conversion rate). The system then marks the waveform feature sequence and time difference data corresponding to the ultra-wideband signal as key features, marks the coordinated change features corresponding to the inertial navigation data as the first feature, and marks the coordinated change features corresponding to the environmental data as the second feature.

[0027] It should be further explained that, in the specific implementation process, the critical positioning and matching process for ultra-wideband signals includes: The key features, first feature, second feature, and stereo position information within the analysis window at the time of the distortion interference event are obtained. The key features, first feature, second feature, and stereo position information are then input into the critical database for matching (using cosine similarity for matching), and the feature similarity of each critical feature package in the critical database is obtained. A preset feature similarity threshold (set by those skilled in the art based on actual conditions or obtained through simulation with a large amount of data) is set. If the feature similarity of each critical feature package in the critical database is less than the feature similarity threshold, the critical location matching will fail. If there exists a critical feature packet with a feature similarity greater than or equal to the feature similarity threshold, then the positioning deviation compensation factor in the critical feature packet is used to compensate for the positioning deviation of the planar coordinates generated from the ultra-wideband signal.

[0028] It should be further explained that, in the specific implementation process, the process of performing positioning deviation compensation, critical feature-based internet access, or fast link selection operations based on the critical positioning matching results includes: If the critical positioning match is successful, a positioning deviation compensation operation is performed on the planar coordinates generated by the ultra-wideband signal. If the critical positioning match fails, an inertial error duration window is set for the moment of the distortion interference event (because the cumulative error of inertial navigation data increases rapidly over time, the error may reach several meters or even tens of meters within 3-5 minutes, and the inertial error duration window is usually within 5 to 30 seconds, that is, within 5 to 30 seconds after the moment of the distortion interference event). The duration of the distortion interference event generated by the ultra-wideband signal is obtained. Within the duration of the distortion interference event, the planar coordinates of the multi-source sensor terminal in the three-dimensional model are generated based on the inertial navigation data. The three-dimensional position information of the multi-source sensor terminal in the three-dimensional model is obtained based on the planar coordinates. If the duration of the ultra-wideband signal distortion interference event is less than the inertial error duration window, then at the end timestamp of the distortion interference event, the planar coordinates of the multi-source sensor terminal in the 3D model are selected based on the planar coordinates. The three-dimensional position information of the multi-source sensor terminal in the 3D model is obtained based on the planar coordinates, and the critical feature Internet access operation is performed. If the duration of the ultra-wideband signal distortion interference event is greater than or equal to the inertial error duration window, then a fast link selection operation is performed at the end timestamp of the inertial error duration window.

[0029] It should be further explained that, in the specific implementation process, the process of performing the critical characteristic internet access operation includes: A critical database is constructed within a micro UWB base station to obtain key features, first features, second features, and stereo position information (obtained from inertial navigation data) for each analysis window during the duration of distortion interference events. Positioning deviation compensation analysis is performed on the key features and stereo position information of each analysis window to obtain the positioning deviation compensation factor. The key features, first feature, second feature, stereo position information, and positioning deviation compensation factor of each analysis window during the duration of the distortion interference event are packaged into a critical feature package and uploaded to the critical database.

[0030] It should be further explained that, in the specific implementation process, the process of obtaining the positioning deviation compensation factor includes: Based on the time difference data in the key features within the analysis window, obtain the planar coordinates of the multi-source sensing terminal in the 3D model using the start and end timestamps of the analysis window. Based on the planar coordinates of the multi-source sensing terminal in the 3D model using the start and end timestamps, obtain the planar coordinate vector and mark the planar coordinate vector as the first vector. Obtain the planar coordinates corresponding to the start and end timestamps of the stereo position information within the analysis window. Based on the planar coordinates corresponding to the start and end timestamps of the stereo position information within the analysis window, obtain the planar coordinate vector and mark the planar coordinate vector as the second vector. The positioning deviation compensation factor is obtained based on the first vector and the second vector (using the second vector as the judgment standard, the vector direction deviation value and vector magnitude deviation value between the first vector and the judgment standard are obtained, and the vector direction deviation value and vector magnitude deviation value are marked as the positioning deviation compensation factor). This factor quantifies the magnitude and direction of the UWB positioning error caused by multipath interference (for example, if the compensation factor corresponding to a certain critical feature packet is +0.8 meters, it means that the UWB estimate needs to be corrected to the left by 0.8 meters).

[0031] It should be further explained that, in the specific implementation process, the process of performing the rapid link selection operation includes: Pre-trained UWB band feature libraries for different scenarios (such as rooms, corridors, and metal escalators). The UWB band feature library includes the best frequency bands corresponding to different key features, first features, and second features. The stereo position information of the multi-source sensing terminal in the 3D model is obtained at the end timestamp of the inertial error duration window. Based on the stereo position information, the scene corresponding to the multi-source sensing terminal in the 3D model is obtained. A UWB band feature library consistent with the scene corresponding to the multi-source sensing terminal in the 3D model is extracted. Key features, first features, and second features of each analysis window within the inertial error time window are obtained. The key features, first features, and second features are matched with the UWB band feature library. The cosine similarity method is used to obtain the key features, first features, and second features with the highest cosine similarity to the key features, first features, and second features of each analysis window in the UWB band feature library. The optimal frequency band corresponding to the key features, first features, and second features with the highest cosine similarity is obtained. The original frequency band of the ultra-wideband signal is replaced with the optimal frequency band. For example, in a high-temperature smoke environment, a 4GHz low-frequency band (such as CH3) is selected to maintain signal transmission by utilizing its smoke-penetrating properties. If this frequency band is interfered with by plasma generated by combustion, it is automatically switched to an 8GHz high-frequency band (such as CH9) to distinguish multipath signals through higher time resolution (<1ns). In areas with dense metal beams and columns, horizontally polarized and vertically polarized signals are used simultaneously, and reflection interference is suppressed by multi-band switching combination (such as CH5+CH7).

[0032] The above embodiments are only used to illustrate the technical methods of the present invention and are not intended to limit it. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can be made to the technical methods of the present invention without departing from the spirit and scope of the technical methods of the present invention.

Claims

1. A multi-source positioning data fusion method for fire protection clothing, characterized in that, Includes the following steps: Step s1: Deploy a micro UWB base station in the target area and deploy a multi-source sensor terminal on the fire protective suit. The multi-source sensor terminal is used to collect multi-source heterogeneous data and mark the collection time, obtain the prior data of the building structure in the target area, build a three-dimensional model, and generate the three-dimensional position information of the multi-source sensor terminal in the three-dimensional model. Step s2: Perform time difference analysis on the ultra-wideband signal in the multi-source heterogeneous data, and determine whether distortion interference events are generated based on the analysis results; Step s3: When the ultra-wideband signal generates distortion interference events, perform collaborative feature analysis on multi-source heterogeneous data to obtain key features, first features, and second features. Based on the key features, first features, second features, and three-dimensional position information, perform critical positioning matching on the ultra-wideband signal, and perform positioning deviation compensation operation, critical feature Internet access operation, or fast link selection operation according to the critical positioning matching results.

2. The multi-source positioning data fusion method for fire-fighting protective clothing according to claim 1, characterized in that, The micro UWB base station communicates with multi-source sensor terminals to form a self-organizing network. The multi-source heterogeneous data includes inertial navigation data, ultra-wideband signals, and environmental data.

3. The multi-source positioning data fusion method for fire-fighting protective clothing according to claim 2, characterized in that, The process of constructing a 3D model and generating the stereo position information of multi-source sensing terminals within the 3D model includes: A 3D model of the target area is constructed based on prior data of the building structure, and scene annotation is performed on the 3D model. The planar coordinates of the multi-source sensor terminal in the 3D model are generated based on ultra-wideband signal and inertial navigation data. At the same time, the elevation coordinates of the multi-source sensor terminal in the 3D model are obtained based on inertial navigation data and environmental data. The planar coordinates and elevation coordinates are merged to generate the stereo position information of the multi-source sensor terminal in the 3D model.

4. The multi-source positioning data fusion method for fire-fighting protective clothing according to claim 3, characterized in that, The process of performing time difference analysis on ultrawideband signals in multi-source heterogeneous data includes: The time difference data of the ultra-wideband signal is acquired. A normal threshold range for the time difference data of the ultra-wideband signal is preset. The time difference data of the ultra-wideband signal at the current moment is compared with the normal threshold range. If the time difference data of the ultra-wideband signal is not within the normal threshold range, the ultra-wideband signal at the current moment is determined to be a distortion interference event.

5. The multi-source positioning data fusion method for fire-fighting protective clothing according to claim 4, characterized in that, The process of collaborative feature analysis of multi-source heterogeneous data includes: Set the analysis window for the time of the distortion interference event, perform waveform analysis on the ultra-wideband signal within the analysis window, and obtain the waveform feature sequence of the ultra-wideband signal within the analysis window; The system acquires the numerical time series sequences of various indicators in the inertial navigation data and environmental data within the analysis window. It performs time series feature analysis on the numerical time series sequences of each type of indicator, obtains the cooperative change features corresponding to the inertial navigation data and the cooperative change features corresponding to the environmental data, and marks the waveform feature sequence and time difference data corresponding to the ultra-wideband signal as key features. The cooperative change features corresponding to the inertial navigation data are marked as the first feature, and the cooperative change features corresponding to the environmental data are marked as the second feature.

6. The multi-source positioning data fusion method for fire-fighting protective clothing according to claim 5, characterized in that, The process of critical location matching for ultra-wideband signals includes: The key features, first feature, second feature, and stereo position information within the analysis window at the time of the distortion interference event are obtained. The key features, first feature, second feature, and stereo position information are then input into the critical database for matching, and the feature similarity of each critical feature package in the critical database is obtained. A preset feature similarity threshold is set. If the feature similarity of each critical feature package in the critical database is less than the feature similarity threshold, the critical location matching will fail. If there exists a critical feature packet with a feature similarity greater than or equal to the feature similarity threshold, then the positioning deviation compensation factor in the critical feature packet is used to compensate for the positioning deviation of the planar coordinates generated from the ultra-wideband signal.

7. The multi-source positioning data fusion method for fire-fighting protective clothing according to claim 6, characterized in that, The process of performing positioning deviation compensation, critical feature-based internet access, or fast link selection based on critical positioning matching results includes: If the critical positioning match is successful, a positioning deviation compensation operation is performed on the planar coordinates generated by the ultra-wideband signal. If the critical positioning match fails, an inertial error duration window is set at the time of the distortion interference event, the duration of the distortion interference event generated by the ultra-wideband signal is obtained, and the planar coordinates of the multi-source sensing terminal in the three-dimensional model are generated based on the inertial navigation data within the duration of the distortion interference event. Based on the planar coordinates, the three-dimensional position information of the multi-source sensing terminal in the three-dimensional model is obtained. If the duration of the distortion interference event generated by the ultra-wideband signal is less than the inertial error duration window, then the planar coordinates of the multi-source sensor terminal in the three-dimensional model are selected based on the end timestamp of the distortion interference event, and the critical feature Internet access operation is performed. If the duration of the distortion interference event generated by the ultra-wideband signal is greater than or equal to the inertial error duration window, then the fast link selection operation is performed at the end timestamp of the inertial error duration window.

8. The multi-source positioning data fusion method for fire-fighting protective clothing according to claim 7, characterized in that, The process of performing borderline internet access operations includes: A critical database is constructed within a micro UWB base station to obtain key features, first features, second features, and stereo position information for each analysis window during the duration of distortion interference events. Positioning deviation compensation analysis is performed on the key features and stereo position information of each analysis window to obtain the positioning deviation compensation factor. The key features, first feature, second feature, stereo position information, and positioning deviation compensation factor of each analysis window during the duration of the distortion interference event are packaged into a critical feature package and uploaded to the critical database.

9. The multi-source positioning data fusion method for fire-fighting protective clothing according to claim 8, characterized in that, The process of obtaining the positioning deviation compensation factor includes: Based on the time difference data in the key features, the start and end timestamps of the analysis window are obtained, and the planar coordinates of the multi-source sensing terminal in the three-dimensional model are obtained. Based on the planar coordinates of the start and end timestamps, a planar coordinate vector is obtained, and the planar coordinate vector is marked as the first vector. Obtain the planar coordinates corresponding to the start and end timestamps of the stereoscopic position information in the analysis window; obtain the planar coordinate vector based on the planar coordinates corresponding to the start and end timestamps of the stereoscopic position information in the analysis window; and mark the planar coordinate vector as the second vector. The positioning deviation compensation factor is obtained based on the first vector and the second vector.

10. A method for fusing multi-source positioning data of fire protective clothing according to claim 9, characterized in that, The process of performing fast link selection includes: A UWB frequency band feature library for different scenarios is pre-trained. The UWB frequency band feature library includes the optimal frequency bands corresponding to different key features, the first feature, and the second feature. The stereo position information of the multi-source sensing terminal in the three-dimensional model is obtained at the end timestamp of the inertial error duration window. The scene corresponding to the multi-source sensing terminal in the three-dimensional model is obtained based on the stereo position information. Extract a UWB band feature library that matches the scene corresponding to the multi-source sensing terminal in the 3D model, obtain the key features, first features, and second features of each analysis window within the inertial error duration window, match the key features, first features, and second features with the UWB band feature library, obtain the optimal frequency band, and replace the original frequency band of the ultra-wideband signal with the optimal frequency band.

Images