A method and system for analyzing the position state of a trapped person

Abstract

The application provides a trapped personnel position state analysis method and system, and relates to the field of data processing. In the method, a background entropy flow base state is formed by constructing the building structure distribution, conductive fluid flow channel and metal component distribution in the disaster target area, and a background entropy flow field is generated by statistically analyzing the noise strength probability distribution, channel joint fluctuation relationship and time stability index. In the search and rescue stage, the background entropy flow field is converted to a controlled mode by active disturbance, and the multi-physical channel noise response before and after the disturbance is collected by relying on a multi-field passive sensing network to construct a multi-dimensional local entropy response vector. The initial suspected area is identified and the entropy anomaly core area is extracted by combining the consistency, directionality and duration of multiple noise channels. Finally, the metabolic activity and posture restriction condition are inferred according to the local entropy response coordination relationship under different disturbance modes, and the trapped personnel position state result is output. The technical solution provided by the application is convenient for accurately identifying the position state of trapped personnel.

Description

Technical Field

[0001] This application relates to the technical field of data processing, specifically to a method and system for analyzing the location and status of trapped personnel. Background Technology

[0002] In typical disaster relief scenarios such as earthquake collapses, underground utility tunnel fires, urban tunnel explosions, and offshore oil and gas platform accidents, trapped personnel are often in dangerous environments that are highly fragmented, partially enclosed, and whose structures are constantly evolving. The real-time monitoring of the location and status of trapped personnel is crucial for rescue operations.

[0003] However, such disaster environments commonly involve leaks of large amounts of conductive fluids, such as fire sprinkler water, process liquids, groundwater, and accumulations of steel bars or metal fragments from equipment. Residual electromagnetic energy from incompletely de-energized electrical systems also contributes to the high uncertainty of electromagnetic propagation behavior in space. Signals undergo multiple scattering, strong coupling, and random resonance along their propagation path, causing the signal-to-noise ratio and target distance to no longer satisfy the conventional monotonic attenuation relationship. This can even lead to areas of abnormally enhanced noise and deep target echo fading. Against this backdrop, traditional passive radar and wireless sensing technologies based on electromagnetic attenuation laws struggle to accurately identify the location and status of trapped personnel.

[0004] Therefore, there is an urgent need for a method and system for analyzing the location and status of trapped personnel. Summary of the Invention

[0005] This application provides a method and system for analyzing the location status of trapped personnel, which facilitates accurate identification of the location status of trapped personnel.

[0006] The first aspect of this application provides a method for analyzing the location status of trapped personnel. The method includes: acquiring the distribution of building structures, conductive fluid flow channels, and metal components in a disaster target area to construct a background entropy flow ground state; based on the statistical noise intensity probability distribution, channel joint fluctuation relationship, and time stability index of the background entropy flow ground state, forming a background entropy flow field covering the disaster target area; during the search and rescue phase, using an active disturbance source to transfer the background entropy flow field from a natural equilibrium noise mode to a controlled non-equilibrium noise mode; simultaneously with the active disturbance triggered by the active disturbance source, collecting noise responses from multiple physical channels before and after the disturbance through a multi-field passive sensor network deployed in the disaster target area, and constructing local entropy response features to form a multidimensional local entropy response vector; identifying an initial suspected trapped personnel area based on the consistency, directionality, and duration of the multidimensional local entropy response vector in multiple noise channels, and determining a continuously stable offset area as the entropy anomaly core area from the initial suspected trapped personnel area; and determining the metabolic activity and posture restriction status of the trapped personnel based on the local entropy response coordination relationship of the entropy anomaly core area under different disturbance modes to obtain the location status result of the trapped personnel.

[0007] A second aspect of this application provides a system for analyzing the location and status of trapped personnel. The system includes an acquisition module and a processing module. The acquisition module is used to acquire the distribution of building structures, conductive fluid flow channels, and metal components in a disaster target area to construct a background entropy flow ground state. The processing module is used to statistically analyze the probability distribution of noise intensity, the joint fluctuation relationship of channels, and time stability indices based on the background entropy flow ground state to form a background entropy flow field covering the disaster target area. The processing module is further used to transfer the background entropy flow field from a natural equilibrium noise mode to a controlled non-equilibrium noise mode during the search and rescue phase using an active disturbance source. The processing module is also used to... Simultaneously with the active disturbance triggering, a multi-field passive sensor network deployed in the disaster target area collects noise responses from multiple physical channels before and after the disturbance, and constructs local entropy response features to form a multidimensional local entropy response vector. The processing module is further used to identify the initial suspected trapped personnel area based on the consistency, directionality, and duration of the multidimensional local entropy response vector across multiple noise channels, and to determine the continuously stable offset area as the core region of entropy anomaly from the initial suspected trapped personnel area. The processing module is also used to determine the metabolic activity and posture restriction status of the trapped personnel based on the coordinated relationship of the local entropy response of the core region of entropy anomaly under different disturbance modes, so as to obtain the position status result of the trapped personnel.

[0008] A third aspect of this application provides an electronic device including a processor, a memory, a user interface, and a network interface. The memory is used to store instructions, and both the user interface and the network interface are used to communicate with other devices. The processor is used to execute the instructions stored in the memory to cause the electronic device to perform the method described above.

[0009] A fourth aspect of this application provides a non-transitory computer-readable storage medium storing instructions that, when executed, perform the method described above.

[0010] In summary, one or more technical solutions provided in this application have at least the following technical effects or advantages:

[0011] By constructing a background entropy flow ground state and background entropy flow field coupled with the disaster environment, subsequent noise shifts can be clearly attributed to additional disturbances introduced by trapped personnel, avoiding analytical distortions that are prone to occur in traditional signal attenuation positioning under conditions such as random resonance and electromagnetic anomaly propagation. By introducing multi-channel active perturbations during the search and rescue phase to trigger controlled non-equilibrium noise modes, and simultaneously acquiring the noise responses of multiple physical channels before and after the perturbation, the local entropy shifts caused by the presence of trapped personnel are detectable and traceable in both the time and channel dimensions. By analyzing the consistency, directionality, and duration characteristics of the multi-dimensional local entropy response vector, the trapped personnel area can be stably identified even under complex obstruction, signal scattering, and metallic coupling conditions. High-precision positioning is achieved by filtering out core entropy anomaly areas through stable shifts. Furthermore, by inferring metabolic activity and attitude restriction status based on the synergistic relationship of local entropy responses under different perturbation modes, the acquired information not only includes location but also quantifies the risk to vital states, helping to improve the prioritization, path planning, and effectiveness of rescue strategies. Therefore, it facilitates accurate identification of the location and status of trapped personnel. Attached Figure Description

[0012] Figure 1 A flowchart illustrating a method for analyzing the location status of trapped personnel, provided in an embodiment of this application;

[0013] Figure 2 A schematic diagram of a trapped person location status analysis system provided in this application embodiment;

[0014] Figure 3 This is a schematic diagram of the structure of an electronic device provided in an embodiment of this application.

[0015] Explanation of reference numerals in the attached figures: 21. Acquisition module; 22. Processing module; 31. Processor; 32. Communication bus; 33. User interface; 34. Network interface; 35. Memory. Detailed Implementation

[0016] To enable those skilled in the art to better understand the technical solutions in this specification, the technical solutions in the embodiments of this specification will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, and not all embodiments.

[0017] In the description of the embodiments of this application, the words "for example" or "for instance" are used to indicate examples, illustrations, or explanations. Any embodiment or design that is described as "for example" or "for instance" in the embodiments of this application should not be construed as being more preferred or advantageous than other embodiments or design options. Rather, the use of the words "for example" or "for instance" is intended to present the relevant concepts in a specific manner.

[0018] In the description of the embodiments of this application, the term "multiple" means two or more. For example, multiple systems means two or more systems, and multiple screen terminals means two or more screen terminals. Furthermore, the terms "first" and "second" are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the indicated technical features. Thus, a feature defined with "first" or "second" may explicitly or implicitly include one or more of that feature. The terms "comprising," "including," "having," and variations thereof all mean "including but not limited to," unless otherwise specifically emphasized.

[0019] To address the aforementioned technical problems, this application provides a method for analyzing the location and status of trapped personnel, referring to... Figure 1 , Figure 1 This is a flowchart illustrating a method for analyzing the location status of trapped personnel, provided in an embodiment of this application. The method is applied to a server and includes steps S110 to S160, as follows:

[0020] S110. Obtain the distribution of building structures, conductive fluid flow channels, and metal components in the disaster target area to construct the background entropy flow ground state.

[0021] Specifically, a server refers to a computing device or cluster that performs centralized computing, data fusion, and model inference during disaster relief. This can be a high-performance physical server deployed in a command center or a virtual server in an edge data center. The server connects to multiple passive sensor networks, active disturbance source control terminals, and on-site information acquisition terminals to receive building information, environmental information, and sensor data. It then performs modeling and analysis under a unified time reference and spatial coordinate system. Therefore, the server is the core computing node in this method for constructing the background entropy flow ground state, statistically analyzing the background entropy flow field, and analyzing the location and status of trapped personnel. The disaster target area refers to the spatial range where focused searches for trapped personnel and risk assessments are needed during an accident or disaster. This could be a few floors of a partially collapsed high-rise building, an area of ​​a subway station where an explosion occurred, a section of an underground utility tunnel where a fire occurred, or a partially damaged section of a chemical plant.

[0022] Building structure distribution refers to the spatial layout and relative relationships of all load-bearing and enclosing components within the disaster target area. This includes the location, size, shape, and interconnection of columns, beams, slabs, shear walls, stairwells, machine room partitions, underground foundations, and any possible reinforcement components in three-dimensional space. Building structure distribution is derived from a combination of building information models, structural construction drawings, and on-site scanning data. For example, the column grid layout, beam-slab system, and escalator opening locations of a subway station concourse level collectively constitute the building structure distribution of that area. Conductive fluids refer to liquid media with significant electrical conductivity within the disaster target area. Examples include water containing impurities, salt water, corrosive solutions, and conductive coolants generated by fire sprinklers, pipe ruptures, or equipment leaks. The presence of dissolved ions or conductive components in these conductive fluids significantly affects electromagnetic fields and current transmission. Conductive fluid flow channels guide the actual or potential flow and accumulation paths of electrical fluids within the disaster target area. These paths are strongly correlated with factors such as building structure distribution, pipeline layout, and ground elevation differences. For example, cable trenches in underground utility tunnels, gaps between equipment foundations and the ground, drainage ditches on basement floors, and water accumulation areas at the bottom of stairwells can all become part of conductive fluid flow channels.

[0023] Metal components refer to components made of metallic materials within the disaster target area that significantly affect electromagnetic fields, structural mechanics, and electrical conductivity. These include reinforcing bars, steel beams, steel columns, and metal supports in building structures, as well as equipment casings, cable trays, pipes, valves, and metal brackets in electromechanical equipment. It also includes areas of accumulated metal debris formed after a disaster. The distribution of metal components refers to the spatial location, density, and connectivity of these components within the disaster target area. For example, the arrangement of high-voltage switchgear, the routing of busbars, and the internal steel mesh of reinforced concrete walls in an underground substation collectively constitute the distribution of metal components in that area. Background entropy flow ground state refers to the statistical equilibrium state of multi-physics noise formed solely by the combined effects of building structure distribution, conductive fluid flow channels, metal component distribution, and natural environmental disturbances, before the introduction of trapped personnel as new energy, material, and structural disturbance sources. It describes the spatial generation, transmission, and dissipation patterns of energy and disorder in the disaster target area under natural conditions.

[0024] Furthermore, when constructing the initial distribution framework of the building structure based on historical engineering data, building information model (BIM) data, and pipeline layout archives of the disaster target area, the BIM data, structural construction drawings, and pipeline layout archives are first imported into the server. Building components such as columns, beams, slabs, shear walls, stairwells, machine room partitions, and underground foundations are represented as three-dimensional entities with geometric boundaries and material properties in the model coordinate system. Simultaneously, the pipeline layout archives are imported to label spatial units interwoven with the building structure, such as pipe corridors, cable trenches, and reserved openings for equipment foundations. Then, the three-dimensional laser scanning point cloud, structural radar detection reflection points, and dense point sets reconstructed from multi-view imaging are projected onto the same reference coordinate system. The registration relationship between the scanning coordinate system and the model coordinate system is established through feature point matching or manual selection of control points. The rigid body transformation matrix is ​​solved in the server to spatially align the scanning data with the BIM model. For example, the rotation matrix and displacement vector are obtained by solving the following least-squares registration problem, which are used to map the scanning point cloud into the BIM coordinate system, as follows:

[0025]

[0026] in, This represents the first result obtained from three-dimensional laser scanning or structured radar inversion. The three-dimensional coordinates of a spatial feature point in the scanning coordinate system This represents the three-dimensional coordinates in the building information model corresponding to the feature point. Let represent the three-dimensional rotation matrix to be determined, where each element takes real values ​​and satisfies orthogonality constraints. The vector to be determined is a three-dimensional translation vector, and the values ​​of its components are real numbers. By minimizing the registration residual between the scanned feature points and the model feature points, the scanned data is spatially aligned with the initial distribution frame of the building structure. Then, the scanned data and the model data are compared, and the areas where local collapse, component breakage, component displacement and space blockage occur are marked in the server. Building units that form closed cavities, narrow gaps and local isolation spaces are marked as potential trapped space candidate areas, thus obtaining the building structure distribution that has been corrected for post-disaster status and includes potential trapped space candidate areas.

[0027] When identifying potential flow paths of conductive fluids based on water supply and drainage pipeline data, fire sprinkler system data, and underground pipe gallery layout data of the disaster target area, the server first extracts the geometric paths of water supply pipes, drainage pipes, sprinkler risers, sprinkler branch pipes, and channels and culverts in the underground pipe gallery from the water supply and drainage design drawings and fire sprinkler system layout drawings. These paths are then mapped onto a three-dimensional coordinate system identical to the building structure distribution, forming the initial framework of potential conductive fluid flow paths. Subsequently, low-frequency conductivity measuring devices and contact electrodes are deployed on-site to collect conductivity scanning data and contact electrode probe signals at the pipe gallery floor, floor slab recesses, stair bottoms, and around equipment foundations. The server then interpolates and spatially reconstructs these measuring points to form conductivity data. The system is designed to identify areas where conductive fluids accumulate or seep, using conductive fluid trace sensing data, such as the outputs of water trace sensors, humidity sensors, and electrolyte residue detectors. Based on the conductivity distribution field and conductive fluid trace distribution, the server identifies line segments that are continuously connected to high conductivity areas and obvious trace areas in the potential flow path of conductive fluids as main conductive fluid channels, line segments that are connected to the main channels but have lower conductivity as conductive fluid branch channels, and areas with only local high conductivity that are not connected to the main channels as conductive fluid seepage areas. This process yields the actual connectivity of the conductive fluid flow channels and expresses these channels in the server as graph structures or mesh attributes.

[0028] When obtaining the theoretical distribution of metal components based on the reinforcement layout diagram, steel structure construction drawings, and electromechanical equipment layout data of the disaster target area, the server first parses the reinforcement mesh layout method in each floor, wall, and slab from the reinforcement layout diagram, including the reinforcement spacing, diameter, protective layer thickness, and anchoring method, recording the reinforcement as a linear metal unit embedded inside the concrete component; simultaneously, it parses the spatial position and connection method of steel components such as steel beams, steel columns, support beams, and node plates from the steel structure construction drawings, and parses the three-dimensional position and occupied volume of electromechanical equipment such as switch cabinets, transformers, pumps, fans, pipes, valves, and metal cable trays from the electromechanical equipment layout data, thus forming the theoretical distribution of metal components; based on this, it uses ground magnetic field anomaly measurement... The server obtains on-site magnetic field anomaly data and reflection intensity data through quantitative analysis, electromagnetic induction scanning, and multi-band ground-penetrating radar imaging. This data is then compared with the theoretical locations of metal components. If a strong magnetic field anomaly or high-intensity electromagnetic reflection exists in a spatial area but no metal component is recorded in the theoretical model, it is determined that the area may contain accumulated metal debris or displaced equipment after a collapse. If a metal component exists in the theoretical model but the on-site electromagnetic response is significantly weakened, it is determined that the component may have been damaged or its location may have shifted. The server corrects the spatial location of each metal component by comparing the theoretical metal component coordinates with the metal reflection structure derived from on-site observations, and uses a weighted integration method, such as the linear fusion method described below, to combine the theoretical and measured distributions.

[0029]

[0030] in, Indicates a point in three-dimensional coordinates The final metal component distribution indicator at the location can be either metal density or the probability of metal presence. This indicates the theoretical distribution of metal components obtained from reinforcement layout drawings, steel structure construction drawings, and mechanical and electrical equipment layout data. This indicates the distribution indication of observed metallic components obtained from ground magnetic field anomaly measurements, electromagnetic induction scanning, and multi-band imaging inversion by ground-penetrating radar. The weighting coefficient between the theoretical distribution and the observed distribution is between zero and one. It is used to adjust the influence ratio of the two in the fusion result according to the credibility of the field observation. By performing the above fusion on all spatial points, the server obtains a distribution of metal components that is consistent with the actual field situation.

[0031] When mapping the distribution of building structures, conductive fluid flow channels, and metal components to a unified three-dimensional coordinate system to form a multi-material spatial decomposition result, the server first determines the boundary range of the disaster target area in three-dimensional space. This range is then divided into regular three-dimensional spatial grids according to a preset resolution, with each grid cell corresponding to a volume element. For each grid cell, the server detects its overlap with the building structure geometry, the conductive fluid flow channels, and the distribution of metal components. Whether it is occupied by concrete, masonry, or cavities is used as a building structure distribution attribute; whether it is located in the main, branch, or seepage zone of the conductive fluid is used as a conductive fluid flow channel attribute; and metal density or the probability of metal presence is used as a metal component distribution attribute. During the coordinate mapping process, the three-dimensional continuous space can be mapped to a discrete grid through the relationship between voxel coordinates and physical coordinates. For example, the following index relationship can be used to map continuous coordinates... Mapped to grid index :

[0032]

[0033] in, These represent the three-dimensional coordinate components of a point in physical space. These represent the minimum coordinate values ​​of the disaster target area along the three coordinate axes, respectively. These represent the step size for dividing the spatial grid cells along the three coordinate axes, and their values ​​are preset according to resolution requirements and computing power. This represents the floor function, used to map continuous coordinates to integer grid indices. The corresponding spatial grid cell index is used. Through the above mapping, the server can assign building structure distribution attributes, conductive fluid flow channel attributes, and metal component distribution attributes to each spatial grid cell, and regard these attributes as state labels describing the local multi-material spatial environment. Subsequently, based on the long-term noise statistics collected by the multi-field passive sensor network, the noise statistical characteristics of each spatial grid cell are associated with its multi-material attributes to form a set of local entropy generation, entropy transmission, and entropy dissipation characteristics under the condition of no trapped personnel. The combination of the local entropy characteristics of multiple spatial grid cells in space constitutes the background entropy flow ground state.

[0034] S120. Based on the probability distribution of background entropy flow ground state statistical noise intensity, channel joint fluctuation relationship and time stability index, a background entropy flow field covering the disaster target area is formed.

[0035] Specifically, the noise intensity probability distribution refers to the probability distribution obtained by statistically analyzing the noise amplitude in a certain spatial grid cell or physical channel under the background entropy flow ground state. It is used to describe the relative probability and typical level of different noise intensities. The channel joint fluctuation relationship refers to the cooperative relationship and coupling degree of noise changes over time between different physical channels under the background entropy flow ground state. It is used to characterize whether electromagnetic noise, mechanical vibration noise, thermal noise, and gas noise exhibit synchronous enhancement, synchronous weakening, or mutual compensation in the same spatial grid cell. The time stability index is a measure of whether the noise intensity probability distribution and channel joint fluctuation relationship under the background entropy flow ground state remain stable over long time scales. It is used to distinguish which statistical characteristics of background noise belong to long-term stable features and which belong to slow evolution or short-term shifts caused by occasional events.

[0036] Background entropy flow field refers to a three-dimensional field obtained by spatially splicing and interpolating the noise intensity probability distribution, channel joint fluctuation relationship, and time stability index of each spatial grid cell within the disaster target area. It is used to describe the overall spatial distribution and statistical characteristics of multiphysics noise in the region under the background entropy flow ground state. Each spatial location in the background entropy flow field not only contains the noise intensity probability distribution on the electromagnetic noise, mechanical vibration noise, thermal noise, and gas noise channels at that location, but also the channel joint fluctuation relationship between these channels and the time stability index evolving over time. Therefore, the background entropy flow field is a spatially continuous, statistically significant high-dimensional field that reflects the natural flow pattern of energy and disorder.

[0037] Furthermore, the server first spatially maps each spatial grid cell to several physical sensors or interpolated observation areas based on the aforementioned multi-material spatial decomposition results. On the time axis, it acquires electromagnetic noise amplitude sequences, mechanical vibration noise amplitude sequences, thermal noise amplitude sequences, and gas concentration noise amplitude sequences at uniform sampling intervals, and performs time synchronization and missing value imputation. Subsequently, a sliding time window method is used to divide each noise time series into multiple statistical windows. Under the condition that the window length and window step size can cover daily environmental fluctuations without excessively smoothing short-term changes, the continuous time axis is divided into multiple statistical windows. Channels in each statistical window The noise samples are represented as follows:

[0038]

[0039] in, Indicates in the channel Upper Within the first statistical window Noise amplitude at each sampling point, channel Used to distinguish between electromagnetic noise channels, mechanical vibration noise channels, thermal noise channels, and gas concentration noise channels. Indicates the index of the statistics window. This indicates the sample index within the window. This represents the number of samples contained in each statistical window, and its value is determined by both the sampling frequency and the window length. Through this sliding time window division method, the server obtains multiple sets of noise segments with temporal locality on each spatial grid cell, providing a basic sample set for subsequently constructing the probability distribution of noise intensity and the joint fluctuation relationship of channels at the statistical window scale.

[0040] The server operates on each spatial grid cell, targeting channels within each statistical window. sample set First, select several noise amplitude intervals. Then, accumulate the occurrence counts of all samples within the window according to whether their amplitude falls into different intervals to obtain the frequency of occurrence for each noise amplitude interval. Next, normalize the frequencies to construct the probability distribution of noise intensity for the corresponding channel within the window. Using histogram statistics, the frequency of occurrence for the first... The probability estimate for each noise amplitude interval is defined as follows:

[0041]

[0042] in, Indicates in the channel and statistics window The lower noise amplitude falls into the first Probability estimates for each interval, Display window Central Channel The noise sample falls into the first The number of samples in each interval Display window Central Channel The total number of noise samples, the number of noise amplitude intervals, and their boundaries are determined by a preset strategy or an adaptive rule based on the sample distribution. For cases where the noise distribution is relatively smooth or the sample size is large, the server can also use kernel density estimation. By selecting the kernel function and bandwidth parameters, a smooth probability density curve is constructed for continuous values ​​of the noise amplitude, thereby more precisely characterizing the tail features and multi-peak structure of the noise strength probability distribution.

[0043] The server is on the same spatial grid cell, in each statistical window In this process, samples from the electromagnetic noise channel, mechanical vibration noise channel, thermal noise channel, and gas concentration noise channel are aligned one-to-one according to sampling time, forming multiple time-corresponding sample pairs for each channel. For each channel... With channel The server first calculates the sample mean for each channel within the statistical window. Then, it constructs a covariance based on the deviation between the sample and the mean to characterize the linear linkage between the noise amplitudes of the two channels on the same time scale. After obtaining the variance, the server calculates the standardized cross-correlation coefficient, ensuring that the joint fluctuation relationship between channels is independent of the absolute noise amplitude scale and only reflects the synchronicity of enhancement or weakening between channels. Furthermore, for scenarios where there may be significant nonlinear linkage or threshold effects, the server can also accumulate statistics on the cases where the amplitudes of two channels simultaneously fall into a specified interval based on the joint occurrence frequency of noise amplitude interval combinations within the statistical window, forming a joint occurrence frequency distribution to characterize the cooperative occurrence pattern of multiple channels under specific intensity combinations. For example, the server can statistically analyze the frequency of the combination of "high amplitude electromagnetic noise and medium amplitude mechanical vibration noise," using this joint occurrence frequency to reflect the common triggering characteristics of multi-channel noise under specific operating conditions.

[0044] On each spatial grid cell, the server compares pairwise the estimated noise intensity probability distributions obtained from multiple statistical windows of the same physical channel. The degree of difference between these distributions measures the stability of the noise statistical characteristics over time. This difference can be represented by distance or similarity measures between the probability distributions. The time stability index is obtained by averaging or statistically aggregating these distances or similarities across multiple statistical window combinations, thus characterizing the stability of the noise distribution over a long period. Similarly, for channel joint fluctuation relationships, the server compares and analyzes the covariance matrix, cross-correlation coefficient set, or joint occurrence frequency distribution obtained from different statistical windows over time. When the changes in these joint statistical characteristics between different statistical windows are small, the corresponding channel joint fluctuation relationship has high time stability; when the changes are large or significant structural changes occur within a specific time period, the time stability index is low. In this way, the server generates a time stability index for each spatial grid cell, each physical channel, and channel combinations to distinguish between background stable noise structures and noise structures that drift slowly over time or are affected by occasional events.

[0045] The server constructs a set of statistical feature vectors for each spatial grid cell. The noise intensity probability distribution parameters, channel joint fluctuation relationship parameters, and time stability indices for each spatial grid cell across electromagnetic noise, mechanical vibration noise, thermal noise, and gas concentration noise channels are concatenated in a fixed order. Simultaneously, the server incorporates the building structure distribution attributes, conductive fluid flow channel attributes, and metal component distribution attributes of the spatial grid cell into the same feature vector, ensuring that each spatial grid cell possesses a complete fusion description of structural and noise statistical information. Subsequently, the server concatenates the feature vectors of all spatial grid cells according to spatial indices within a unified three-dimensional coordinate system. When spatial continuity is required, spatial interpolation methods are used to smooth the statistical features between grid cells, making the transition of background entropy flow characteristics near grid boundaries more continuous. Through this process, the server constructs a multidimensional background entropy flow field across the entire disaster target area. This background entropy flow field can provide a noise intensity probability distribution, channel joint fluctuation relationship, and time stability index that are consistent with the local building structure distribution attributes, conductive fluid flow channel attributes, and metal component distribution attributes at any spatial location. This provides a unified statistical background reference for subsequent identification of entropy anomalies, locating trapped personnel areas, and analyzing the location status of trapped personnel.

[0046] S130. During the search and rescue phase, the background entropy flow field is transferred from the natural equilibrium noise mode to the controlled non-equilibrium noise mode by actively disturbing sources.

[0047] Specifically, an active disturbance source refers to a device or combination of devices controlled by the rescue system or server during the search and rescue phase, intentionally applying minute external disturbances within or at the boundary of the disaster target area. These external disturbances target specific physical channels, such as mechanical, thermal, electromagnetic, or gas channels, by injecting stimulus signals with limited amplitude, controllable frequency, and programmable timing, causing a repeatable and identifiable response in a multi-physics field that was originally in a state of natural noise. Active disturbance sources include miniature vibratory actuators installed on structural components, miniature resistance heaters installed on walls or equipment surfaces, electromagnetic disturbance coils for injecting weak electromagnetic signals, and gas release units that release trace amounts of inert trace gases. For example, in an underground utility tunnel, small vibratory actuators can be attached to local beams and slabs to excite a structural response using low-amplitude frequency sweep vibration, while simultaneously releasing a small amount of inert gas at one end of the tunnel to alter the local gas concentration field, such as using helium as a trace inert trace gas, thereby generating controlled responses in both mechanical and gas noise channels.

[0048] Furthermore, based on the previously constructed multi-material spatial decomposition results, the server first filters all spatial mesh elements in a unified three-dimensional coordinate system. Spatial mesh elements located on the surface of key building structural components, at the intersection of conductive fluid flow channels, or near densely packed metal components, and whose structural safety margin meets the requirements, are marked as candidate spatial nodes. A comprehensive weight is calculated for each candidate spatial node to balance structural sensitivity, environmental representativeness, and construction accessibility. The importance of the candidate spatial nodes can be described by the following weight function:

[0049]

[0050] in, Indicates the first The comprehensive weight of each candidate spatial node. This is a sensitivity index representing the building structure where the spatial node is located. It reflects the structural components' ability to transmit vibrations, temperature changes, and electromagnetic disturbances. The value is a dimensionless value between zero and one. This indicates the degree of coupling between the spatial node and the conductive fluid flow channel, reflecting the connectivity between the main and branch channels of the conductive fluid near the node. The value is a dimensionless value between zero and one. This index represents the density of metal components near a spatial node, reflecting the concentration and connectivity of metal components around that node. It is also a dimensionless value between zero and one. The adjustment coefficients representing structural sensitivity, conductive fluid coupling degree, and metal component density in the weight calculation are real numbers between zero and one, satisfying the constraint that their sum is one. The server sorts the nodes from highest to lowest weight. Under the premise of meeting on-site construction feasibility and equipment deployment constraints, a set of spatial nodes is selected from the candidate spatial nodes. Miniature vibration actuators are fixed to the structural surface capable of transmitting vibration, miniature resistance heaters are attached to components with good thermal conductivity, electromagnetic disturbance units are arranged at positions with effective coupling to the metal frame or metal cable tray, and gas release units are arranged near conductive fluid flow channels or in relatively enclosed spaces, thus forming a set of active disturbance nodes targeting different physical channels in space.

[0051] The server designs perturbation sequences simultaneously in both temporal and spatial dimensions, ensuring that the perturbations are sufficient to elicit observable responses across multiple physical channels without causing structural risks or exceeding environmental limits. To this end, the server first normalizes the expected impact of each active perturbation node on different physical channels, forming a node-channel influence matrix. Then, a pseudo-random sequence generator produces a long-term perturbation control sequence, mapping the pseudo-random sequence to node and channel weights to determine whether each active perturbation node is activated at each discrete scheduling moment, the type of physical channel used upon activation, and the duration of its effect. Based on a preset channel priority list and a pseudo-random perturbation channel selection factor, the server selects the type of physical channel activated by the node at the current moment. For example, it prioritizes mechanical perturbations at locations with high structural stiffness and electromagnetic or gaseous perturbations at the junctions of conductive fluids, thus forming an active perturbation scheduling strategy that is approximately random in time but closely correlated with environmental properties in space and channels.

[0052] The server integrates the active disturbance scheduling strategy with the sampling plan of the multi-field passive sensor network. At each scheduling moment, it records the start / stop status, disturbance channel type, and expected duration of all active disturbance nodes, and injects a limited amplitude of disturbance energy into the active nodes. For the mechanical channel, the micro-vibration actuator generates narrow-band or broadband micro-vibrations on the connected structural components, causing mechanical vibration noise to propagate in the components and the connected conductive fluid flow channels and metal components, changing the statistical characteristics of local mechanical noise. For the thermal channel, the micro-resistance heater slowly changes the local temperature field in a gradual increase and decrease manner, causing a slight shift in thermal noise in a short period of time. For the electromagnetic channel, the electromagnetic disturbance unit deploys low-amplitude electromagnetic signals within a limited frequency range, forming a coupling path with the conductive fluid through the metal components, causing controllable enhancement or attenuation of electromagnetic noise. For the gas channel, the gas release unit locally releases a small amount of inert gas, causing a identifiable disturbance in gas concentration noise within a certain time window. A multi-field passive sensor network acquires the background response under a natural equilibrium noise mode before the disturbance, continuously acquires the disturbed noise response during the disturbance, and continues to acquire noise changes during the recovery process after the disturbance ends. The server records these data at various times. By comparing the noise statistical characteristics of the corresponding spatial grid cells within and before the time window with the pre-constructed background entropy flow ground state, it was found that the local entropy shift generated by multi-physical channel noise under controlled disturbance was observed. As a series of active disturbance events continued to execute on the time axis, the background entropy flow field gradually expanded from a natural equilibrium noise pattern that only reflected the equilibrium state of natural noise to a controlled non-equilibrium noise pattern that also included the response characteristics to active disturbances. This provides rich and identifiable information for subsequent identification of entropy anomaly regions related to trapped personnel based on multi-dimensional local entropy response vectors.

[0053] S140. At the same time as the active disturbance source triggers the active disturbance, the noise response of multiple physical channels before and after the disturbance is collected by a multi-field passive sensor network deployed in the disaster target area, and the local entropy response characteristics are constructed to form a multi-dimensional local entropy response vector.

[0054] Specifically, the multidimensional local entropy response vector refers to a vector form with multidimensional components that, for a given spatial grid cell, combines the local entropy response features of that grid cell across different physical channels in a predetermined order under an active perturbation event. Each dimension of this vector corresponds to the local entropy shift features of the electromagnetic noise channel, mechanical vibration noise channel, thermal noise channel, and gas concentration noise channel, respectively. If necessary, it can be further subdivided into sub-dimensions such as amplitude variation features, structural variation features, and linkage variation features within the same channel. "Multidimensional" emphasizes that the vector simultaneously carries response information from multiple physical channels, while "local entropy response vector" emphasizes that it is a vectorized expression of local entropy response features, facilitating subsequent comparison, clustering, and identification along the time and spatial axes.

[0055] Furthermore, the server first performs a pre-programmed perturbation scheduling in the active perturbation scheduling module for the first... Each active disturbance establishes an event record, encapsulating the start and stop times of the disturbance, the corresponding spatial index of the active disturbance node, and the activated disturbance physical channel type into an event identifier, and associating the event identifier with a unified time reference. The unified time reference is broadcast by the server to various sensors and active disturbance units in the multi-field passive sensing network via a time synchronization signal or a local high-precision clock, synchronously correcting the internal time references of the micro-vibration actuators, micro-resistance heaters, electromagnetic disturbance units, and gas release units, ensuring that these active disturbance units execute disturbance outputs according to the unified clock at preset start and stop times. Based on this, the electromagnetic sensing units, structural vibration sensing units, micro-heat flow and temperature / humidity sensing units, and chemical gas sensing units in the multi-field passive sensing network continuously sample each spatial grid unit at a unified sampling interval. The server, based on the start and stop times recorded in the event identifier, samples the same spatial grid unit at the [missing information - likely a specific time reference]. The disturbance event is divided into three time periods: before the disturbance, during the disturbance, and after the disturbance. A corresponding noise response time series is generated for each physical channel, for example, the first... Under the secondary disturbance event, a certain spatial grid cell in the channel The noise response time series can be denoted as:

[0056]

[0057] in, Indicates the first The channel within the time window corresponding to the secondary active disturbance event In the Noise amplitude sampling points at each sampling time, channel Used to distinguish between electromagnetic noise channels, mechanical vibration noise channels, thermal noise channels, and gas concentration noise channels. This represents the index of active disturbance events. This represents the discrete sample number within the event's time window. Represented as the first The total number of sampling points allocated for each active disturbance event is determined by both the disturbance duration and the sensor sampling frequency. In this way, the server obtains a noise response time series covering multiple physical channels across all spatial grid cells at the same time each active disturbance is triggered, encompassing the three stages before, during, and after the disturbance. This provides a unified temporal and spatial index for subsequently constructing local entropy response features.

[0058] For each spatial grid cell and each physical channel, the server reads the long-term statistical mean, statistical variance, and typical joint fluctuation pattern of that spatial grid cell under that channel from the background entropy flow ground state, and uses these statistics as a benchmark reference for that spatial location. For the th A spatial grid cell channel under a secondary active disturbance event noise response time series The server first utilizes the mean value provided by the background entropy flow ground state. and standard deviation After amplitude normalization, a standardized response sequence is obtained that eliminates the influence of absolute noise levels and highlights the relative offset, as follows:

[0059]

[0060] in, Indicates the first Channel under secondary active disturbance event The standardized noise response value, This indicates that the spatial grid cell is in the channel. The long-term average amplitude of the background noise is obtained from the ground-state statistics of the background entropy flow, and its value range is a real number. This indicates that the spatial grid cell is in the channel. The standard deviation of the background noise amplitude is also obtained from the ground state statistics of the background entropy flow. Its value is a real number greater than zero, which is used to characterize the natural fluctuation scale of the background noise.

[0061] The server then aggregates the standardized sequences along the time dimension, for example, by calculating the average offset within the complete event time window, and assigns the channel to the [missing information - likely a specific data point or feature]. The overall entropy shift intensity under the perturbation event is extracted as a scalar feature, as follows:

[0062]

[0063] in, Indicates the first Channel under secondary active disturbance event The local entropy response intensity characteristic is used to reflect the average deviation of the channel noise statistical state from the background entropy flow ground state within the event time window. The absolute value operation is used to unify the measurement of positive and negative offsets on the same scale. As mentioned above, this represents the total number of sampling points for the event. Regarding changes in the joint fluctuation relationship and temporal stability of the channels, the server can re-estimate the covariance structure and instantaneous fluctuation patterns among the multiple channels within the event time window, and compare them with the corresponding statistical structure in the background entropy flow ground state to obtain additional response features describing the degree of joint structure rearrangement and the degree of change in temporal fluctuation morphology. These additional response features are then incorporated into the channel data. In the set of local entropy response features.

[0064] Finally, the server, according to the pre-agreed dimensional order, places a certain spatial grid cell in the [number]th [dimension]. The local entropy response features of the four physical channels under the secondary active perturbation event are stacked to form a multidimensional local entropy response vector, as follows:

[0065]

[0066] in, This indicates that the spatial grid cell is in the first... The multidimensional local entropy response vector under a secondary active perturbation event is in column vector form, with components... This represents the local entropy response intensity characteristics of an electromagnetic noise channel. This represents the local entropy response intensity characteristics of the mechanical vibration noise channel. This represents the local entropy response intensity characteristics of the thermal noise channel. The local entropy response intensity characteristics of the gas concentration noise channel are represented by non-negative real numbers, which are used to characterize the cooperative entropy shift behavior of each physical channel relative to the background entropy flow ground state under this disturbance event. By repeating the above process for all spatial grid cells and all active disturbance events, the server obtains a set of multidimensional local entropy response vectors covering the disaster target area and the disturbance time axis, providing quantitative input for subsequent identification of the initial suspected trapped personnel area and the core area of ​​entropy anomaly based on the consistency, directionality, and duration of multiple noise channels.

[0067] S150. Based on the consistency, directionality and duration of the multidimensional local entropy response vector in multiple noise channels, identify the initial suspected trapped personnel area, and determine the continuously stable offset area from the initial suspected trapped personnel area as the core area of ​​entropy anomaly.

[0068] Specifically, the entropy anomaly core region refers to the core spatial subset obtained by further shrinking the space after separating the continuously stable offset region from the initial suspected trapped personnel region. This subset represents the center of gravity of the trapped personnel's location or their most important area of ​​influence. The entropy anomaly core region is regarded as a high-density clustering center of multidimensional local entropy response offset in the disaster target area. This region has the most significant synergistic relationship of local entropy response under various disturbance modes and maintains the strongest continuous offset feature in time. Therefore, the entropy anomaly core region can serve as a key spatial anchor point in the analysis of the trapped personnel's location status, used to infer the specific location, metabolic activity level, and degree of posture restriction of the trapped personnel, and to provide spatial positioning basis for subsequent formulation of rescue routes and structural load reduction plans.

[0069] Furthermore, the server first targets spatial grid cells. Physical channels and active disturbance events Obtain the local entropy response intensity Based on the event time sequence, a normalized response sequence with time decay weights is constructed, and an adaptive gating factor is introduced to form a channel consistency index, which is described by the following formula:

[0070]

[0071] in, Represents spatial grid cells In the passage The channel consistency index, with a value between zero and one, is used to measure whether the channel's response to multiple active disturbance events is stable and significant over a long period of time. This represents the total number of active disturbance events included in the statistics, and is a positive integer. This represents the time decay factor, which ranges from zero to one. The closer the value is to one, the more memory is retained for earlier events, and the smaller the value, the more emphasis is placed on recent events. Represents an S-shaped gated function, for example A monotonic mapping is used to map real numbers to zero and one, which is used to map response intensity to the activation probability of soft events. Represents spatial grid cells In the passage The above is aimed at the first The local entropy response intensity of the second active perturbation event is a non-negative real number. Indicates channel The response baseline threshold is a non-negative real number used to limit the minimum offset that is considered a "valid response"; Indicates channel The scaling factor, a positive real number, is used to control the smoothness of the response intensity mapping to the gating function; Represents spatial grid cells In the passage The binary pattern entropy of the response event sequence is obtained through statistical analysis. The output is obtained by thresholding and binarizing the probability distributions of "response events" and "non-response events," with values ​​ranging from zero to... between, The most random response pattern is the one that corresponds to the most random response pattern. This indicates the degree of regularity in the response pattern; a higher value indicates a more stable response time distribution pattern for that channel. The above structure is equivalent to performing an exponentially weighted average of all events over the time axis, while using pattern entropy to suppress channels with highly random responses, making the channel consistency index focus more on long-term repeatable responses.

[0072] To characterize whether the local entropy response is biased towards a single direction or frequently switches between positive and negative directions within the same spatial grid cell and physical channel, a directional index based on weighted vector synthesis is introduced. This index considers both the sign and magnitude of the local entropy response in each event, constructs a directional vector, and performs normalization. The directional index is described by the following formula:

[0073]

[0074] in, Represents spatial grid cells In the passage The directional indicator on the surface has a value range between zero and one. The closer the value is to one, the more consistent the offset direction is across multiple events. The closer the value is to zero, the more the offset direction alternates between positive and negative and lacks stability. Represents spatial grid cells In the passage Upper The signed offset after removing the background mean for this event is a real number. The sign function outputs one when the input is greater than zero, negative one when the input is less than zero, and zero when the input is equal to zero. It is used to extract the response offset direction. It still represents the intensity of the local entropy response. By multiplying the sign by the intensity, both the offset direction and the offset magnitude are included in the directionality calculation. The unit vector can be viewed as a simplified representation of the net offset direction in one-dimensional space, characterized by its length. This vector is introduced to reflect the form of vector sum in the formula structure. This indicates the time decay weight for earlier events, consistent with the meaning of the previous formula; This represents a very small positive number, used to avoid a denominator of zero. The physical meaning of this index is to weight and accumulate the offset direction and offset intensity of all events on the time axis, and then divide by the weighted sum. If the offset direction is always consistent, the numerator and denominator are close, and the index is close to one. If the offset direction is frequently reversed, the accumulated vector almost cancels out, and the index approaches zero.

[0075] Regarding duration metrics, to reflect the longest continuous hold time of significant entropy shifts in the event sequence, a normalized metric for the length of continuous segments based on an event-level response gating function is introduced, which normalizes the channel length. Upper-level response probability Considered The output defines an event-level activation gating sequence, and searches for the longest consecutive activation interval with a probability greater than a certain threshold on this gating sequence. The duration metric is described by the following formula:

[0076]

[0077] in, Represents spatial grid cells In the passage The duration index, ranging from zero to one, indicates that the significant entropy shift of the channel is maintained continuously for a longer period of time in the event sequence. Represents the set of all possible consecutive event intervals in the sequence, each interval It consists of several adjacent event indexes; This indicates the use of gating functions from The obtained event-level response probability ranges from zero to one. Indicates channel The probability threshold for determining "continuous activation" is between zero and one. This is an indicator function, which is one when the gating probability is not lower than the threshold, and zero otherwise; Indicates a continuous interval The number of events that meet the "continuous activation" condition. This means selecting the longest interval from all consecutive intervals and then dividing by the total number of events. Normalization is achieved. The above construction ensures that the duration index not only considers whether a response occurs, but also requires continuity over the event sequence, thus characterizing the long-term stable entropy shift caused by trapped personnel.

[0078] After obtaining the channel consistency index of each spatial grid cell on each channel. directional indicators and duration indicators Subsequently, to obtain multi-channel comprehensive indicators at the spatial grid cell level, the server introduces a weighted generalized mean structure. Indices from different channels are aggregated through power functions to construct comprehensive channel consistency, comprehensive directionality, and comprehensive duration indicators, respectively. For example, the comprehensive channel consistency indicator can be expressed as:

[0079]

[0080] in, Represents spatial grid cells The comprehensive channel consistency index has a value range between zero and one. Indicates channel The weighting coefficient is used to reflect the difference in importance of different physical channels in the environment. It takes the value of a non-negative real number and the sum of the weights of each channel is one. The parameter represents the power, and is a real number greater than or equal to one. The larger the channel set, the more pronounced its effect on high consistency. These represent the electromagnetic noise channel, the mechanical vibration noise channel, the thermal noise channel, and the gas concentration noise channel, respectively. Similarly, the comprehensive directional index... Compared with the overall duration index It can be constructed in the same way, simply by... Replace with the corresponding or .

[0081] When selecting spatial grid cells with stable cooperative migration characteristics on multiple noise channels to form the initial suspected trapped personnel area based on channel consistency, directionality, and duration indicators, the server constructs a joint stability scoring function for each spatial grid cell. This function maps the three comprehensive indicators to a range of zero to one through nonlinear fusion, describing the overall level of entropy migration stability of the spatial location under multi-channel, multi-event conditions. The joint stability score can be constructed using the following formula:

[0082]

[0083] in, Represents spatial grid cells The joint stability score ranges from zero to one. The closer the value is to one, the more significant the entropy shift characteristics of the spatial grid cell in terms of channel consistency, directionality and duration are. These represent the scaling factors of the three types of indicators in the index mapping, which are non-negative real numbers used to control the relative influence of each indicator in the overall score; These represent the nonlinear amplification powers of the corresponding indicators, and are real numbers greater than or equal to one, used to enhance the distinction between high-indicator regions and medium-to-low-indicator regions; the exponential function is used to compress unlimited linear combinations to a scale between zero and one, while strengthening high-value regions. The server will process all spatial grid cells... With joint stability threshold Compare and satisfy The spatial grid cells are extracted from a unified three-dimensional coordinate system. Connectivity analysis is performed on grid cells that are spatially adjacent to each other or strongly coupled through the distribution of building structures, conductive fluid flow channels, and metal components. Multiple spatially connected regions are formed through graph clustering, region growing, or connected component extraction. These spatially connected regions are defined as the initial suspected trapped personnel areas, making these areas represent spatial blocky regions that exhibit significant, stable, and multi-channel cooperative entropy shifts over a long period under multiple active perturbation conditions.

[0084] The server targets the spatial grid cells within each initial area where people are suspected to be trapped. The multidimensional local entropy response vector under all active perturbation events The time-weighted average is used to construct the time-aggregated multidimensional local entropy response vector, as follows:

[0085]

[0086] in, Represents spatial grid cells The time-aggregated multidimensional local entropy response vectors over all events are represented by components, each corresponding to the weighted average entropy response intensity of the electromagnetic noise channel, mechanical vibration noise channel, thermal noise channel, and gas concentration noise channel, respectively. Indicates the first Spatial grid unit under the sub-event The multidimensional local entropy response vector; This is a time decay weight, consistent with the meaning described above. Based on this, the server will combine it with the stability score. Channel consistency index directional indicators and duration indicators The composite score is constructed using a logarithmic weighted geometric mean, which suppresses the overall score when any dimension index is too low. The composite score can be expressed by the following formula:

[0087]

[0088] in, Represents spatial grid cells The overall score ranges from zero to around one (depending on the individual indicators and...). (specific numerical value); These represent the weighting coefficients of the joint stability score, channel consistency index, directionality index, and duration index in the overall score, respectively. The values ​​are non-negative real numbers and can be normalized as needed. It is a very small positive number, used to avoid taking the logarithm of zero; the structure of logarithm and re-exponential corresponds to the weighted form of geometric mean, so that when any index approaches zero, the comprehensive score is significantly reduced, thus ensuring that the core region of entropy anomaly must show high values ​​in multiple dimensions at the same time.

[0089] To characterize the similarity of multidimensional local entropy response vectors among neighboring spatial grid cells, the server performs a test on any pair of neighboring spatial grid cells within the initial suspected trapped personnel area. and The weighted Mahalanobis distance between their time-aggregated multidimensional local entropy response vectors is calculated and mapped to similarity weights, which can be specifically described by the following formula:

[0090]

[0091] in, Represents spatial grid cells With spatial grid units The similarity index in multidimensional local entropy response features ranges from zero to one. The closer the value is to one, the more similar the entropy response patterns of the two are. and Representing spatial grid cells and spatial grid units Time-aggregated multidimensional local entropy response vector; This represents the covariance matrix of the multidimensional local entropy response vector statistically obtained within the initial suspected trapped personnel area. It is a symmetric positive definite matrix used to normalize the distance when the scale and correlation of each component are large. This indicates the transpose operation. Essentially, this similarity index measures the statistical distance between two vectors using Mahalanobis distance, and then uses an exponential mapping to form similarity weights between zero and one.

[0092] After obtaining the overall score Similarity Subsequently, the server first selects all spatial grid cells in the initial suspected trapped personnel area that satisfy the following criteria as candidate seed grid cells for entropy anomalies:

[0093]

[0094] in, This represents the overall score threshold, a real number between zero and one, used to ensure that only grid cells with a high overall multidimensional index are retained. The L2 norm of the time-aggregated multidimensional local entropy response vector is used to summarize the overall offset magnitude. This represents the lower limit of the offset amplitude. It is a non-negative real number and is used to filter out weak offset units that are close to background noise. This represents the upper limit of the offset amplitude and is a positive real number. It is used to filter out extremely high offset cells that may be caused by strong equipment interference or local abnormal noise sources. For seed mesh cells that meet the above conditions, the server further expands the region along the spatial neighborhood, only in adjacent mesh cells. and Between In the case of merging them into the same connected region, where The similarity threshold, ranging from zero to one, is used to ensure a high degree of consistency in the multidimensional local entropy response patterns within the region. Ultimately, each spatially connected region selected through simultaneous filtering by comprehensive score constraints, offset amplitude constraints, and similarity constraints is defined as an entropy anomaly core region. Spatially, this core region corresponds to the location where the trapped personnel's impact on the environment is most concentrated; statistically, it corresponds to a multi-channel, long-term, and strongly collaborative entropy offset aggregate, providing a high-confidence spatial anchor point for subsequent output of the trapped personnel's location and status results.

[0095] S160. Based on the coordinated relationship of local entropy response in the core region of entropy anomaly under different perturbation modes, determine the metabolic activity and posture restriction status of the trapped personnel to obtain the position status results of the trapped personnel.

[0096] Specifically, the location status results of trapped personnel refer to the comprehensive analysis results outputting the spatial location of the core area of ​​entropy anomaly, the level of metabolic activity of the trapped personnel, and the degree of posture restriction. This includes the location coordinates or range of the trapped personnel in a unified three-dimensional coordinate system, qualitative or graded judgments of metabolic activity status and posture restriction degree, as well as optional risk warning information. The location status results not only provide spatial information about where the person is, but also a comprehensive description of their vital state and posture constraints. This allows the rescue command system to optimize rescue routes, select demolition methods, and prioritize rescue sequences. For example, it can prioritize rescuing trapped personnel whose metabolic activity is still relatively good but whose posture is highly restricted and who are at risk of suffocation, or avoid applying strong vibration impacts to the core area of ​​entropy anomaly during demolition to reduce the risk of secondary injury. By establishing a correspondence between the synergistic relationship of the local entropy response of the core area of ​​entropy anomaly under different disturbance modes and the metabolic activity and posture restriction status of the trapped personnel, the location status results of trapped personnel achieve a complete mapping from multi-physics noise statistical characteristics to specific rescue decision information.

[0097] Furthermore, the server targets each spatial grid cell within the core region of entropy anomaly. They will belong to different disturbance modes. and different active disturbance events Multidimensional local entropy response vector Mapping to a unified time axis, by interpolating or resampling the timestamps recorded under different perturbation modes, the thermal noise channel and the gas concentration noise channel are made comparable on the same time index; based on this, from Extracting the local entropy response component of the thermal noise channel Local entropy response component of gas concentration noise channel A collaborative response index for the metabolic activities of trapped personnel was constructed. The temporal co-permanent shifts of the thermal noise channel and the gas concentration noise channel under different disturbance modes were weighted, integrated, and normalized to form a metabolic activity intensity index, as follows:

[0098]

[0099] in, Represents spatial grid cells The metabolic activity intensity index is used to quantify the intensity of the coordinated continuous offset of the thermal noise channel and the gas concentration noise channel of the grid cell under all perturbation modes. The larger the value, the more active the metabolic activity. This represents the normalization factor, which takes the value of a positive real number and is used to scale the metabolic activity intensity index to a preset range. This indicates all perturbation modes Summation is performed, and the perturbation modes include those dominated by mechanical perturbations, those dominated by thermal perturbations, those dominated by electromagnetic perturbations, and those dominated by gas perturbations, etc. Indicates the perturbation mode The weighting coefficients, which are non-negative real numbers and can be normalized according to the importance of the pattern, are used to express that certain patterns are more sensitive to assessing metabolic activities. Indicating in perturbation mode The length of the time window used for evaluation is a positive real number. Indicating in perturbation mode Next, Time Spatial grid unit The local entropy response intensity of the thermal noise channel is a statistical measure of the difference between the local temperature field and the background entropy flow ground state. Indicating in perturbation mode Next, Time Spatial grid unit The local entropy response intensity of the gas concentration noise channel is a statistical measure of the local gas composition and its fluctuations relative to the background state. and These represent the response reference thresholds for the thermal noise channel and the gas concentration noise channel, respectively. They are non-negative real numbers used to eliminate weak noise fluctuations. and These are positive real numbers representing the scaling factors for the thermal noise channel and the gas concentration noise channel, respectively, used to control the smoothness of the response intensity mapping to the gating function. and These represent monotonically nonlinear gating functions designed for the thermal noise channel and the gas concentration noise channel, respectively. They map the continuous response intensity to an effective activation metric between zero and one, ensuring that the thermal noise channel and the gas concentration noise channel only activate when they consistently exceed a threshold and have significant temporal overlap. It makes a significant contribution; through the above integration and weighting combination, the server performs time alignment and perturbation pattern aggregation on the multidimensional local entropy response vector in the core region of entropy anomaly, and forms a metabolic activity intensity index that can reflect the persistence and coordination of the metabolic activities of trapped personnel, which is used to determine whether there is continuous life activity in the region.

[0100] The server abstracts the effects of active disturbance sources under different disturbance directions into a set of disturbance directions. For example, micro-vibrations and electromagnetic disturbances applied along different structural principal axes, different beam and slab normals, or different wall normals; targeting spatial grid units within the core region of entropy anomalies. The server in each disturbance direction Local entropy response intensity of vibration noise channel in statistical mechanics With the local entropy response intensity of the electromagnetic noise channel With disturbance amplitude The changing response curves are used to extract directional sensitivity and nonlinear constrained behavior. Directional sensitivity can be measured by normalizing the average response difference under each perturbation direction, and an attitude anisotropy index is constructed as follows:

[0101]

[0102] in, Represents spatial grid cells The attitude anisotropy index, the larger the value, the more significant the response pattern of the corresponding region of the grid cell to different disturbance directions, suggesting that there is orientation-selective contact or constraint between the trapped person's body and the surrounding structure. Indicates the direction of the disturbance The following is a two-dimensional vector composed of the local entropy responses of the electromagnetic noise channel and the mechanical vibration noise channel, for example... ,in and These represent the directions of the disturbance, respectively. The average entropy response intensity of the electromagnetic noise channel and the mechanical vibration noise channel; Indicates the perturbation direction. The directional average response vector obtained after averaging; Represents the Euclidean norm, used to measure the size of a vector; This is a very small positive number, used to avoid the denominator being zero. Nonlinear constrained behavior is characterized by analyzing the curvature and saturation characteristics of the entropy response curve under varying perturbation amplitudes. The server can... and Second-order curvature analysis was performed to extract the nonlinear constrained index reflecting the "gradual saturation of response increase", as follows:

[0103]

[0104] in, Represents spatial grid cells The nonlinear constraint index, the larger the value, the more obvious the bending and saturation of the response curves of the mechanical vibration noise channel and the electromagnetic noise channel appear as the disturbance amplitude increases, inferring that the limbs or torso of the trapped person are strongly constrained in the corresponding area. This represents the number of perturbation directions in the set of perturbation directions, and is a positive integer. and These represent the local entropy response of the mechanical vibration noise channel relative to the disturbance amplitude, respectively. The first and second derivatives are used to characterize the local slope and curvature of the response as the amplitude of the disturbance changes; and The meaning is similar, used to characterize the nonlinear changes in the local entropy response of an electromagnetic noise channel; and These represent the minimum and maximum values ​​of the disturbance amplitude, respectively. This represents a weighting function that weights different perturbation amplitude ranges, used to emphasize the importance of certain amplitude ranges, such as the low-amplitude safe perturbation range, in attitude-constrained inference. The server uses the attitude anisotropy index... With nonlinear restricted index By combining thresholds and mapping rules, a pose-constrained level is constructed, for example, when... Significantly large and When the disturbances are all too high in multiple directions, it is determined that the trapped person's posture and height are restricted; when medium and When there is a significant difference in direction, the attitude is determined to be moderately restricted; when both are low, the attitude is determined to be slightly restricted.

[0105] The server's metabolic activity intensity index across all spatial grid cells within the entropy anomaly core region. Anisotropy index With nonlinear restricted index A joint assessment was conducted, incorporating a comprehensive vital signs score, as follows:

[0106]

[0107] in, Represents spatial grid cells The comprehensive vital status score is used to integrate metabolic activity levels and postural limitation status on the same scale. and These represent the weighting coefficients of metabolic activity intensity and postural limitation in the comprehensive vital status assessment, respectively. They are non-negative real numbers and can be set according to the rescue strategy. This represents a mapping function that combines the attitude anisotropy index and the nonlinear constraint index. For example, it normalizes the two and then performs a weighted sum or a nonlinear combination to output a scalar representing the degree of attitude constraint. The server, in a unified three-dimensional coordinate system, identifies the spatial grid cell with the highest overall life state score located within the entropy anomaly core region, or its spatial centroid. Output as the trapped position, and simultaneously according to Within a preset grading threshold range, metabolic activity levels are divided into "active," "weak," and "very weak" levels, based on... Within a preset grading threshold range, the posture restriction level is divided into "highly restricted", "moderately restricted" and "slightly restricted". This results in a position status result of the trapped person, which includes three parts: the location coordinates or coordinate range, the metabolic activity level, and the posture restriction level. This information can be directly used by the rescue command system in path planning, demolition strategies, and rescue priority ranking.

[0108] This application also provides a system for analyzing the location and status of trapped personnel, referring to... Figure 2 , Figure 2This is a schematic diagram of a trapped person location status analysis system provided in an embodiment of this application. The system is a server, which includes an acquisition module 21 and a processing module 22. The acquisition module 21 is used to acquire the distribution of building structures, conductive fluid flow channels, and metal components in the disaster target area to construct a background entropy flow ground state. The processing module 22 is used to statistically analyze the noise intensity probability distribution, channel joint fluctuation relationship, and time stability index based on the background entropy flow ground state to form a background entropy flow field covering the disaster target area. The processing module 22 is also used to transfer the background entropy flow field from a natural equilibrium noise mode to a controlled non-equilibrium noise mode during the search and rescue phase using an active disturbance source. The processing module 22 is also used to actively disturb the background entropy flow field simultaneously with the active disturbance source. At the same time, the multi-field passive sensor network deployed in the disaster target area collects the noise response of multiple physical channels before and after the disturbance, and constructs local entropy response features to form a multidimensional local entropy response vector; the processing module 22 is also used to identify the initial suspected trapped personnel area based on the consistency, directionality and duration of the multidimensional local entropy response vector in multiple noise channels, and to determine the continuously stable offset area as the core area of ​​entropy anomaly from the initial suspected trapped personnel area; the processing module 22 is also used to determine the metabolic activity and posture restriction status of the trapped personnel based on the local entropy response coordination relationship of the core area of ​​entropy anomaly under different disturbance modes, so as to obtain the position status result of the trapped personnel.

[0109] It should be noted that the above embodiments of the apparatus are only illustrated by the division of the above functional modules. In practical applications, the above functions can be assigned to different functional modules as needed, that is, the internal structure of the device can be divided into different functional modules to complete all or part of the functions described above. In addition, the apparatus and method embodiments provided in the above embodiments belong to the same concept, and the specific implementation process can be found in the method embodiments, which will not be repeated here.

[0110] This application also provides an electronic device, with reference to... Figure 3 , Figure 3 This is a schematic diagram of the structure of an electronic device provided in an embodiment of this application. The electronic device may include: at least one processor 31, at least one network interface 34, a user interface 33, a memory 35, and at least one communication bus 32.

[0111] The communication bus 32 is used to enable communication between these components.

[0112] The user interface 33 may include a display screen and a camera. Optionally, the user interface 33 may also include a standard wired interface and a wireless interface.

[0113] The network interface 34 may optionally include a standard wired interface or a wireless interface (such as a Wi-Fi interface).

[0114] The processor 31 may include one or more processing cores. The processor 31 connects to various parts of the server via various interfaces and lines, executing instructions, programs, code sets, or instruction sets stored in the memory 35, and calling data stored in the memory 35 to perform various server functions and process data. Optionally, the processor 31 may be implemented using at least one hardware form of Digital Signal Processing (DSP), Field-Programmable Gate Array (FPGA), or Programmable Logic Array (PLA). The processor 31 may integrate one or a combination of several of the following: Central Processing Unit (CPU), Graphics Processing Unit (GPU), and modem. The CPU primarily handles the operating system, user interface, and applications; the GPU is responsible for rendering and drawing the content to be displayed on the screen; and the modem handles wireless communication. It is understood that the modem may also not be integrated into the processor 31 and may be implemented as a separate chip.

[0115] The memory 35 may include random access memory (RAM) or read-only memory. Optionally, the memory 35 may include a non-transitory computer-readable storage medium. The memory 35 can be used to store instructions, programs, code, code sets, or instruction sets. The memory 35 may include a program storage area and a data storage area, wherein the program storage area may store instructions for implementing an operating system, instructions for at least one function (such as touch function, sound playback function, image playback function, etc.), instructions for implementing the above-described method embodiments, etc.; the data storage area may store data involved in the above-described method embodiments, etc. Optionally, the memory 35 may also be at least one storage device located remotely from the aforementioned processor 31. Figure 3 As shown, the memory 35, which serves as a computer storage medium, may include an operating system, a network communication module, a user interface module, and an application program for analyzing the location and status of trapped personnel.

[0116] exist Figure 3In the electronic device shown, the user interface 33 is mainly used to provide an input interface for the user and to obtain the user input data; while the processor 31 can be used to call an application program stored in the memory 35 for analyzing the location status of trapped personnel. When executed by one or more processors, the electronic device executes one or more methods as described in the above embodiments.

[0117] It should be noted that, for the sake of simplicity, the foregoing method embodiments are all described as a series of actions. However, those skilled in the art should understand that this application is not limited to the described order of actions, as some steps may be performed in other orders or simultaneously according to this application. Furthermore, those skilled in the art should also understand that the embodiments described in the specification are preferred embodiments, and the actions and modules involved are not necessarily essential to this application.

[0118] This application also provides a non-transitory computer-readable storage medium storing instructions. When executed by one or more processors, these instructions cause an electronic device to perform one or more of the methods described in the above embodiments.

[0119] The foregoing description is merely an exemplary embodiment of this disclosure and should not be construed as limiting the scope of this disclosure. Any equivalent changes and modifications made in accordance with the teachings of this disclosure shall still fall within the scope of this disclosure. Those skilled in the art will readily conceive of other embodiments of this disclosure upon considering the specification and the disclosure of practical truth. This application is intended to cover any variations, uses, or adaptations of this disclosure that follow the general principles of this disclosure and include common knowledge or customary techniques in the art not described in this disclosure. The specification and embodiments are considered exemplary only, and the scope and spirit of this disclosure are defined by the claims.

Claims

1. A method for analyzing the location and status of trapped personnel, characterized in that, The method includes: The distribution of building structures, conductive fluid flow channels, and metal components in the disaster target area is obtained to construct the background entropy flow ground state; Based on the probability distribution of the strength of the background entropy flow ground state statistical noise, the joint fluctuation relationship of the channels, and the time stability index, a background entropy flow field covering the disaster target area is formed; During the search and rescue phase, the background entropy flow field is transferred from a natural equilibrium noise mode to a controlled non-equilibrium noise mode by actively disturbing sources. At the same time as the active disturbance source is triggered, the noise response of multiple physical channels before and after the disturbance is collected by a multi-field passive sensor network deployed in the disaster target area, and local entropy response features are constructed to form a multi-dimensional local entropy response vector. Based on the consistency, directionality, and duration of the multidimensional local entropy response vector across multiple noise channels, the initial suspected trapped personnel region is identified, and the continuously stable offset region within the initial suspected trapped personnel region is determined as the core region of entropy anomaly. Based on the coordinated relationship of local entropy response in the core region of entropy anomaly under different perturbation modes, the metabolic activity and posture restriction status of the trapped personnel are determined to obtain the position status of the trapped personnel. The step of identifying the initial suspected trapped personnel region based on the consistency, directionality, and duration of the multidimensional local entropy response vector across multiple noise channels, and determining the persistently stable offset region as the core region of entropy anomaly from the initial suspected trapped personnel region, specifically includes: Based on the multidimensional local entropy response vector, the channel consistency index, directionality index, and duration index of each spatial grid cell in the electromagnetic noise channel, mechanical vibration noise channel, thermal noise channel, and gas concentration noise channel are calculated respectively. Based on the channel consistency index, the directionality index, and the duration index, spatial grid cells with stable cooperative offset characteristics on the multi-noise channels are selected to form the initial suspected trapped personnel area; By ranking the comprehensive scores of each multidimensional local entropy response vector within the initial suspected trapped personnel area, and considering the similarity of multidimensional local entropy response vectors between adjacent spatial grid cells, the local region that meets the preset threshold for comprehensive score and whose multidimensional local entropy response vector offset amplitude meets the preset range during multiple active disturbances is determined as the core region of entropy anomaly.

2. The method for analyzing the location and status of trapped personnel according to claim 1, characterized in that, The acquisition of building structure distribution, conductive fluid flow channels, and metal component distribution in the disaster target area to construct the background entropy flow ground state specifically includes: Based on historical engineering data, building information model data and pipeline layout archives of the disaster target area, an initial building structure distribution framework is constructed. The initial building structure distribution framework is then corrected for post-disaster status by combining three-dimensional laser scanning, structural radar detection and multi-view imaging to obtain a building structure distribution that includes candidate areas of potential trapped spaces. Based on the water supply and drainage pipeline data, fire sprinkler system data and underground pipe gallery layout data of the disaster target area, potential flow paths of conductive fluid are identified, and the actual connectivity of the conductive fluid flow channels is obtained by combining low-frequency conductivity scanning data, contact electrode test signals and conductive fluid trace sensing data. The theoretical distribution of metal components is obtained based on the steel reinforcement layout diagram, steel structure construction diagram and electromechanical equipment layout data of the disaster target area, and the distribution of the metal components is verified by ground magnetic field anomaly measurement, electromagnetic induction scanning and ground-penetrating radar multi-band imaging. The distribution of the building structure, the conductive fluid flow channel, and the distribution of the metal components are mapped to a unified three-dimensional coordinate system to form a multi-material spatial decomposition result. Based on the multi-material spatial decomposition result, the system is divided into spatial grid cells with building structure distribution attributes, conductive fluid flow channel attributes, and metal component distribution attributes to generate the background entropy flow ground state.

3. The method for analyzing the location and status of trapped personnel according to claim 2, characterized in that, The process of forming a background entropy flow field covering the disaster target area based on the probability distribution of background entropy flow ground-state statistical noise intensity, channel joint fluctuation relationship, and time stability index specifically includes: For each spatial grid cell within the disaster target area, corresponding electromagnetic noise time series, mechanical vibration noise time series, thermal noise time series, and gas concentration noise time series are collected, and the noise time series are divided into multiple statistical windows by using a sliding time window. Histogram statistics or kernel density estimation are performed on the noise time series based on each of the statistical windows to obtain the probability distribution of noise intensity. Based on the synchronous alignment of the noise amplitude of each physical channel within each statistical window, the covariance, cross-correlation coefficient, or joint occurrence frequency are calculated to obtain the joint fluctuation relationship of the channels. A time stability analysis is performed on the relationship between the probability distribution of noise intensity and the joint fluctuation of the channel across different statistical windows to obtain the time stability index. The noise intensity probability distribution, the channel joint fluctuation relationship, and the time stability index are associated and encoded with the corresponding building structure distribution attributes, conductive fluid flow channel attributes, and metal component distribution attributes. The spatial grid cells are then spatially spliced ​​and interpolated in a unified three-dimensional coordinate system to form a background entropy flow field covering the disaster target area.

4. The method for analyzing the location and status of trapped personnel according to claim 1, characterized in that, The process of transferring the background entropy flow field from a natural equilibrium noise mode to a controlled non-equilibrium noise mode through active disturbance sources during the search and rescue phase specifically includes: Within the disaster target area, spatial nodes are selected and miniature vibration actuators, miniature resistance heaters, electromagnetic disturbance units, and gas release units are deployed to form active disturbance nodes; An active disturbance scheduling strategy is constructed based on the building structure distribution attributes, conductive fluid flow channel attributes, and metal component distribution attributes of the background entropy flow field. The active disturbance scheduling strategy is used to control the disturbance start and stop time, disturbance effect location, and disturbance physical channel type of the active disturbance node through a pseudo-random sequence. Based on the start and stop times of the disturbance, the location of the disturbance effect, and the type of the disturbance physical channel, the background entropy flow field is transferred from the natural equilibrium noise mode to the controlled non-equilibrium noise mode.

5. The method for analyzing the location and status of trapped personnel according to claim 4, characterized in that, Simultaneously with the active disturbance source triggering the disturbance, a multi-field passive sensor network deployed in the disaster target area is used to collect noise responses from multiple physical channels before and after the disturbance, and to construct local entropy response features to form a multi-dimensional local entropy response vector. Specifically, this includes: Each active disturbance is assigned an event identifier that includes the disturbance start and stop time, the disturbance location, and the disturbance physical channel type. Based on a unified time reference, the micro vibration actuator, the micro resistance heater, the electromagnetic disturbance unit, and the gas release unit are time-synchronized and acquired, so that the same spatial grid cell obtains the noise response time series of electromagnetic noise, mechanical vibration noise, thermal noise, and gas concentration noise in the three stages before, during, and after the disturbance. By aligning and normalizing the noise response time series with the noise strength probability distribution, channel joint fluctuation relationship and time stability index under the background entropy flow ground state, the differences in noise statistical features are extracted to construct local entropy response features. The local entropy response features of different physical channels are stacked in a fixed dimension order to form a multidimensional local entropy response vector to characterize the multi-channel cooperative entropy shift behavior.

6. The method for analyzing the location and status of trapped personnel according to claim 1, characterized in that, The determination of the trapped personnel's metabolic activity and posture restriction status based on the coordinated relationship of local entropy response in the core region under different perturbation modes, in order to obtain the trapped personnel's position status, specifically includes: Within the core region of the entropy anomaly, the multidimensional local entropy response vector is time-aligned and perturbation mode aggregated. The metabolic activity of the trapped personnel is determined based on the cooperative and continuous shift of the local entropy response components corresponding to the thermal noise channel and the gas concentration noise channel under different perturbation modes. The attitude constraint status is determined based on the directional sensitivity and nonlinear constraint behavior of the local entropy response components of the electromagnetic noise channel and the mechanical vibration noise channel under different disturbance directions; Combining the metabolic activity of the trapped personnel with their posture restriction status, the system outputs the trapped personnel's position status in a unified three-dimensional coordinate system, including their location, metabolic activity level, and posture restriction level.

7. A system for analyzing the location and status of trapped personnel, characterized in that, The system is used to execute the method for analyzing the location status of trapped personnel as described in any one of claims 1 to 6, and the system includes an acquisition module and a processing module, wherein... The acquisition module is used to acquire the distribution of building structures, conductive fluid flow channels, and metal components in the disaster target area in order to construct the background entropy flow ground state. The processing module is used to form a background entropy flow field covering the disaster target area based on the probability distribution of the strength of the background entropy flow ground state statistical noise, the joint fluctuation relationship of the channels, and the time stability index. The processing module is also used to transfer the background entropy flow field from a natural equilibrium noise mode to a controlled non-equilibrium noise mode during the search and rescue phase by using an active disturbance source. The processing module is also used to collect the noise response of multiple physical channels before and after the disturbance by a multi-field passive sensor network deployed in the disaster target area at the same time as the active disturbance source is triggered, and to construct local entropy response features to form a multi-dimensional local entropy response vector. The processing module is also used to identify the initial suspected trapped personnel area based on the consistency, directionality and duration of the multidimensional local entropy response vector in multiple noise channels, and to determine the continuously stable offset area as the core region of entropy anomaly from the initial suspected trapped personnel area. The processing module is also used to determine the metabolic activity and posture restriction status of the trapped personnel based on the coordinated relationship of local entropy response under different perturbation modes in the core region of entropy anomaly, so as to obtain the location status result of the trapped personnel.

8. An electronic device, characterized in that, The electronic device includes a processor, a memory, a user interface, and a network interface. The memory is used to store instructions. The user interface and the network interface are both used to communicate with other devices. The processor is used to execute the instructions stored in the memory to cause the electronic device to perform the method as described in any one of claims 1 to 6.

9. A non-transitory computer-readable storage medium, characterized in that, The non-transitory computer-readable storage medium stores instructions that, when executed, perform the method as described in any one of claims 1 to 6.

Images