UWB life signal micro-motion identification and positioning method for mine rescue robot

Abstract

This invention provides a UWB life signal micro-motion identification and positioning method for mine rescue robots, belonging to the field of micro-motion identification and positioning technology, and is used for mine rescue robots. The method includes: acquiring echo signals transmitted and received by UWB radar in a mine environment, and generating a reflection path template based on a three-dimensional structural model of the mine roadway characterizing the mine environment; performing parallel time-domain and frequency-domain processing on the echo signals to generate a preprocessed time-domain signal and a time-frequency energy distribution map; calculating a spectral concentration based on the time-frequency energy distribution map, and setting a time-domain energy weight and a frequency-domain energy weight based on the spectral concentration to generate a dynamic weight function; and performing a ternary collaborative optimization process, using the dynamic weight function as the evaluation criterion.

Description

Technical Field

[0001] This invention relates to the field of micro-motion recognition and positioning technology, and in particular to a UWB life signal micro-motion recognition and positioning method for a mine rescue robot. Background Technology

[0002] In modern emergency rescue, especially in mine disaster rescue scenarios, using robots equipped with non-contact life detection devices is a crucial technique for searching for survivors. Ultra-wideband (UWB) radar, due to its strong penetration and high resolution, is widely used to detect vital signs under rubble. The UWB radar echo signals collected by rescue robots are complex, non-stationary time-series data, containing weak periodic modulations caused by the rise and fall of the human chest cavity and strong noise interference from the environment. Advanced digital signal processing techniques are needed to analyze these echo signals to extract valid life characteristics. Current technologies typically employ signal decomposition algorithms, such as Variational Mode Decomposition (SVMD), to process such UWB radar echo signals. This algorithm can decompose a complex signal into several intrinsic mode components with specific frequency characteristics. In application, a set of fixed core parameters, such as a penalty factor and the number of mode components, are usually pre-set for the SVMD algorithm. Then, by analyzing the spectral characteristics of these decomposed mode components, it is determined whether there are components that match the human breathing frequency range. For location, after extracting suspected life signals, the approximate distance to the target is estimated by calculating the signal's time of flight.

[0003] First, existing signal decomposition methods have poor parameter adaptability. The decomposition effect of algorithms such as VMD is highly dependent on the setting of their internal parameters, and the optimal values ​​of these parameters are highly correlated with the signal-to-noise ratio and interference characteristics of the echo signal. In the typical time-varying and highly interference-prone environment of a mine, using a fixed set of pre-set parameters is simply unable to adapt to environmental changes. When the signal quality degrades or the noise characteristics change, using fixed parameters for VMD decomposition is very likely to lead to the phenomenon of "modal aliasing," that is, the real human breathing micro-motion signal and the harmonic components of environmental noise and multipath reflections are incorrectly decomposed into the same or overlapping modal components, making it impossible for the judgment logic to effectively distinguish between the real target and the interference, thus causing serious false alarms or missed detections. Secondly, existing positioning methods do not effectively utilize tunnel structure information and lack resistance to multipath interference. Mine tunnels are typical confined spaces where sidewalls, roofs, and internal facilities generate strong, structured multipath reflections of UWB electromagnetic waves. These reflected signals are received along with the direct wave, forming complex multipath superposition waveforms. Traditional time-of-flight positioning methods only use the first or strongest peak of the signal for ranging. However, in multipath environments, the positions of these peaks are very likely caused by a strong reflected wave rather than the actual direct wave. The distance calculated in this way will produce "false targets" or serious ranging errors. Existing technologies lack a closed-loop signal processing mechanism that can actively utilize the tunnel geometry to suppress such structured multipath interference, failing to meet the needs of high-precision rescue positioning. Therefore, how to adaptively and stably separate life-threatening micro-motion signals from highly time-varying and multipath-rich mine echo signals, and on this basis achieve high-precision three-dimensional positioning with multipath resistance, is a technical problem that urgently needs to be solved in this field, but a satisfactory solution has yet to be found.

[0004] Reference document CN115356714B, entitled "An Experimental Platform for UWB Radar Wave Penetration," discloses an experimental platform for UWB radar wave penetration, comprising a UWB radar wave transmitting system, an adjustable coal and rock mass filling system, a temporary personnel monitoring system, and a UWB radar wave receiving system. This platform enables the study of radar wave transmission attenuation parameters under different coal and rock types, detection thicknesses, and detection distances, with and without loaded life information. By adjusting the coal and rock mass filling system in conjunction with the UWB radar wave transmitting and receiving systems, the platform can study the radar wave transmission attenuation signals of different types of coal and rock masses. Adjusting the coal and rock mass volume control device allows for the study of radar wave attenuation signals under different coal and rock mass thicknesses. Furthermore, it enables the study of radar wave attenuation signals loaded with life information when personnel enter the detection area. Adjusting the servo motor status of the temporary personnel monitoring system allows for the study of radar wave attenuation signals under different personnel postures. Reference document CN116058818A, entitled "A Method for Heart Rate Detection in Ultra-Wideband Radar Based on Multi-Sequence WOA-VMD Algorithm," belongs to the field of ultra-wideband radar technology. This invention aims to utilize the vital sign information carried by ultra-wideband radar reflected echoes to obtain multi-sequence vital sign vectors, suppress respiratory interference, and accurately separate the heartbeat signal to obtain the heart rate value. This invention is based on multi-sequence WOA... The ultra-wideband radar heart rate detection method based on the VMD algorithm has the following steps: The radar transmits pulses at a certain pulse repetition period, and the receiver samples and records data at a certain time interval; the radar echo signal is preprocessed with clutter suppression, linear trend suppression, and signal-to-noise ratio improvement; the variance of the slow-time direction data is calculated to obtain the target position and obtain a multi-sequence vital sign vector; the Whale Algorithm (WOA) and Variational Mode Decomposition (VMD) algorithm are fused to obtain the heart rate value. This invention is mainly applied to situations where ultra-wideband radar is used to obtain vital sign information. Its optimization dimension is limited to the penalty factor and the number of modes of VMD, forming a binary open-loop optimization. No parameters related to the multipath propagation characteristics of the channel are introduced. The preprocessing of this type of method only uses conventional clutter suppression and other means, without using the environmental geometry to model and actively eliminate multipath reflections. The signal quality drops sharply in confined spaces such as mines and tunnels. At the same time, the method outputs at most a one-dimensional radial distance or heart rate value, lacks a fusion mechanism with the three-dimensional environmental model, and cannot provide the three-dimensional coordinates of the trapped person in the real space. The entire processing flow is an open-loop structure, lacking adaptive quality verification and parameter re-optimization feedback for the decomposition results. Reference document CN114279477A, ​​entitled "A Sensor Assembly and Display Device," discloses a sensor assembly and display device. The sensor assembly, for direct fixation to the lower surface of a cover plate, includes: a chip comprising a first surface and a second surface disposed opposite to each other, the first surface being fixedly connected to the lower surface of the cover plate; a transmitter and a receiver disposed on the chip, the upper surfaces of the transmitter and the receiver not exceeding the first surface of the chip. The sensor assembly provided in this application is directly disposed on the lower surface of the cover plate, eliminating the need to consider installation accuracy, which helps reduce background noise and thus improves the accuracy of the sensor sensing signal. It also eliminates the testing and assembly risks associated with the entire device, and the design of the sensor assembly can eliminate other structures found in existing technologies, significantly reducing costs and assembly costs. Summary of the Invention

[0005] This invention provides a UWB life signal micro-motion identification and positioning method for mine rescue robots.

[0006] To achieve the above objectives, the present invention adopts the following technical solution: Firstly, a UWB-based micro-motion identification and positioning method for mine rescue robots is provided. This method is applied to a query terminal and includes the following steps: The echo signals transmitted and received by UWB radar in the mine environment are acquired, and a reflection path template is generated based on the three-dimensional structural model of the mine roadway that characterizes the mine environment. Parallel time-domain and frequency-domain processing is performed on the echo signal to generate a preprocessed time-domain signal and a time-frequency energy distribution map; The spectral concentration is calculated based on the time-frequency energy distribution map, and the time-domain energy weight and frequency-domain energy weight are set according to the spectral concentration to generate a dynamic weight function. A ternary collaborative optimization process is performed, using the dynamic weight function as the evaluation standard, to simultaneously optimize the penalty factor, the number of modes, and the multipath attenuation coefficient, so as to obtain the optimal penalty factor, the optimal number of modes, and the optimal attenuation coefficient. Based on the optimal penalty factor and the optimal number of modes, variational mode decomposition is performed on the preprocessed time-domain signal to obtain multiple intrinsic mode components, and target mode components whose spectral main peak is located in the frequency band corresponding to physiological respiration are extracted from the multiple intrinsic mode components to generate candidate respiration signals; Based on the optimal attenuation coefficient and the reflection path template, a temporal deconvolution operation is performed on the candidate respiratory signal to generate a compensated respiratory signal; The radial distance is calculated based on the compensated respiratory signal, and the three-dimensional coordinates of the trapped personnel are determined by combining the three-dimensional structural model of the mine roadway. First, the raw signal containing multipath effects is acquired using UWB radar, and a reflection path template is synchronously generated using a 3D model of the mine, laying the physical foundation for subsequent interference suppression. A strategy of parallel processing in the time and frequency domains is adopted to separate the preprocessed time-domain signal and time-frequency energy distribution map from the raw signal, achieving synchronous characterization of the signal waveform and dynamic energy distribution. The most critical innovation lies in the ternary collaborative optimization process, which utilizes the whale optimization algorithm. Instead of optimizing the penalty factor and mode number of variational mode decomposition in isolation, it couples them with a path attenuation coefficient and introduces a dynamic weight function composed of time-frequency energy entropy. This allows the optimization process to adjust the optimization direction in real time according to the complexity of the signal itself, transforming from an open-loop search to a closed-loop feedback with self-correction capabilities. Finally, signal decomposition, phase compensation, and time calculation are performed using optimal parameters, achieving the entire process of accurately separating micro-motion signals from a strong interference background and performing high-precision 3D positioning.

[0007] Optionally, the step of performing parallel time-domain and frequency-domain processing on the echo signal includes: Low-pass filtering is performed on the echo signal in the time domain path to filter out high-frequency noise and retain the waveform containing human body micro-movements and low-frequency multipath interference, thus obtaining the preprocessed time domain signal; A short-time Fourier transform is performed on the echo signal in the frequency domain path to reveal the dynamic characteristics of the signal frequency components changing over time, thereby obtaining the time-frequency energy distribution map. The time-frequency energy distribution map provides a basis for subsequent coarse estimation of respiratory rate and interference localization. For a discrete echo signal x[n], its time-frequency transformation result X[m, k] can be expressed as: ; Where n is the index of the time sample point, m is the time index of the window function shift, k is the discrete frequency index, w is the window function, and N is the number of points in the Fourier transform. The time-frequency energy distribution map E[m, k] is obtained by calculating the square of the modulus of the transform result. ; This matrix is ​​a real number matrix, and its element values ​​reflect the energy intensity of the signal at a specific time m and a specific frequency k. This parallel processing mechanism ensures that the waveform fidelity in the time domain and the dynamic energy distribution in the frequency domain can be captured simultaneously without interference. Ultimately, the preprocessed time-domain signal provides a clean input source for the subsequent variational mode decomposition, while the time-frequency energy distribution map serves as a dynamic basis for calculating the spectral concentration and evaluating the decomposition effect, providing crucial prior information for subsequent adaptive optimization.

[0008] Optionally, the step of generating the dynamic weight function includes: The time-frequency energy distribution map is used to calculate the variance of its energy distribution on the time axis in order to obtain the time-domain energy uniformity. The energy uniformity U in this time domain can be expressed as: ; Where E represents a vector whose elements Let be the sum of the energies of all frequency components within the m-th time window of the time-frequency energy matrix. and U represents the mean and standard deviation of vector E, respectively. A U value close to 1 indicates that the energy is uniformly distributed over time, while a value close to 0 indicates that the distribution is non-uniform. The spectral concentration and the temporal energy uniformity are normalized and then weighted and fused to generate a signal complexity index; Signal complexity index The calculation formula is as follows: ; Where a and b are preset fusion weights, their sum is 1, and they can usually be set to an equal value of 0.5. The value range is also between 0 and 1, a high The value indicates that the signal energy exhibits a high degree of non-uniformity in both the time and frequency domains, meaning the signal is more complex. The spectral concentration F is obtained by calculating the ratio of the geometric mean to the arithmetic mean of the time-frequency energy matrix on the frequency axis. It reflects the degree of concentration of signal energy in the frequency domain, and its value is between 0 and 1. The lower the value, the more concentrated the energy is in a few frequency bands. The time-domain energy weight and the frequency-domain energy weight are generated by inverse linear mapping based on the signal complexity index, such that the frequency-domain energy weight increases accordingly when the signal complexity index increases, thereby enhancing the optimization of the purity of the decomposition result.

[0009] Optionally, the step of performing the ternary collaborative optimization process includes: Initialize a whale population containing multiple individuals, where each individual's position vector includes a candidate penalty factor, the number of candidate modes, and a candidate decay coefficient; For each individual in the population, a temporary variational mode decomposition and deconvolution reconstruction are performed on the preprocessed time-domain signal using its corresponding candidate parameters to generate a reconstructed signal. The reconstructed signal is scored using the dynamic weighting function to obtain the fitness value of the individual; Based on the fitness values ​​of all individuals, the position of the whale population is updated through the encirclement predation, bubble net attack and random search mechanisms of the whale optimization algorithm, and this process is repeated until the convergence condition is met, so as to output the optimal penalty factor, the optimal number of modes and the optimal decay coefficient.

[0010] Optionally, the step of performing the temporal deconvolution operation includes: Construct a deconvolution kernel, wherein the reflection path template defines the time offset sequence of the deconvolution kernel, and the optimal attenuation coefficient defines the pulse attenuation factor representing the multipath replica in the deconvolution kernel; The candidate respiratory signal is convolved with the deconvolution kernel to remove the multipath copies represented by the reflection path template point by point, thereby obtaining the compensated respiratory signal, wherein the energy ratio of the direct wave component to the first reflection wave component of the compensated respiratory signal is improved. To accurately remove multipath interference caused by the mine roadway environment from the initially extracted candidate respiratory signals, this method performs a time-domain deconvolution operation based on an adaptive kernel, aiming to recover the original micro-motion waveform of life uncontaminated by multipath effects. The core of this operation is to construct a deconvolution kernel that can accurately characterize the multipath channel response. The deconvolution kernel is designed as a time offset sequence, and its time offset is directly derived from the reflection path template data generated in the previous step. The template records the arrival time difference of the main reflected wave relative to the direct wave. At the same time, the pulse amplitude representing each multipath replica in the deconvolution kernel is set by the optimized multipath attenuation coefficient γ as its key attenuation factor. When considering the main multipath reflections, the deconvolution kernel h[n] can be expressed in the discrete-time domain as: ; L represents the number of primary reflection paths, where δ[n] is the unit impulse function, representing the direct wave path. This represents the number of delayed sampling points for the main reflection path extracted from the reflection path template. This is the optimal attenuation coefficient for the path, with a value range between 0 and 1. It reflects the degree of energy attenuation of the reflected wave relative to the direct wave. This time-domain deconvolution process is implemented through a filtering algorithm that removes multipath copies point by point. The candidate respiratory signal is processed with the constructed deconvolution kernel to eliminate the delay component recorded in the reflection path template in reverse, and finally generate the compensated respiratory signal.

[0011] Optionally, before performing the temporal deconvolution operation, the method further includes: The candidate respiratory signal and the preprocessed time-domain signal are respectively subjected to bandpass filtering, with the filtering passband being the frequency band corresponding to physiological respiration. Then, the maximum value of the normalized cross-correlation function of the two filtered waveforms is calculated, and the maximum value is used as the consistency coefficient. Determine whether the consistency coefficient is lower than the consistency threshold; If the consistency coefficient is lower than the consistency threshold, the optimal parameter combination obtained in the current optimization process is used as the initial value, and the ternary collaborative optimization process is restarted until the consistency coefficient corresponding to the newly generated candidate respiratory signal is not lower than the consistency threshold.

[0012] Optionally, the step of calculating the radial distance based on the compensated respiratory signal includes: Envelope detection is performed on the compensated respiratory signal to obtain the signal envelope; The peak position of the signal envelope is extracted as the arrival time of the direct wave; The radial distance is calculated based on the time difference between the arrival time of the direct wave and the pulse emission time, and in conjunction with the refractive index that characterizes the properties of the tunnel medium.

[0013] Optionally, the step of determining the three-dimensional coordinates of the trapped person includes: Based on the pose information of a rescue robot and the radial distance, a spherical constraint equation is generated; Based on the wall reflection surface equation in the three-dimensional structural model of the mine roadway and the multipath length in the reflection path template, a set of multipath geometric constraint equations is constructed. Solve the spherical constraint equation and the multipath geometric constraint equation simultaneously to obtain multiple candidate coordinate points, and identify the three-dimensional coordinates from the multiple candidate coordinate points based on the energy attenuation characteristics of the candidate respiratory signals.

[0014] Optionally, the frequency band corresponding to physiological respiration covers the human respiratory frequency in a resting state. Specifically, to ensure that this method can accurately capture the weak signals generated by human physiological activities, multiple processing steps in the system are designed and configured around a key preset respiratory frequency band. The respiratory rate of adults at rest or under mild stress is usually stable within a specific range. This method precisely sets the preset respiratory frequency band to 0.2 Hz to 0.4 Hz, which corresponds to a respiratory rate of 12 to 24 breaths per minute, covering the possible physiological states of most trapped individuals. The setting of this frequency band is the benchmark for all subsequent signal processing steps. In the dual-domain preprocessing stage, the frequency response range of the designed low-frequency time-domain signal must completely cover this frequency band to ensure that while filtering out high-frequency noise, no attenuation of the target respiratory signal is caused. The resolution of the generated time-frequency energy matrix on the frequency axis must also be high enough to clearly distinguish the energy peaks located within this frequency band.

[0015] In a second aspect, a UWB life signal micro-motion identification and positioning system for a mine rescue robot is provided, comprising: a processor and a memory; the memory is used to store a computer program, and when the processor executes the computer program, the electronic device performs the UWB life signal micro-motion identification and positioning method for a mine rescue robot as described in the first aspect.

[0016] The signal acquisition and template generation module is used to acquire echo signals and generate reflection path templates based on the three-dimensional structural model of the mine roadway. A dual-domain preprocessing module is used to perform parallel time-domain and frequency-domain processing on the echo signal to generate a preprocessed time-domain signal and a time-frequency energy distribution map; An adaptive function construction module is used to generate a dynamic weighting function based on the time-frequency energy distribution map; The collaborative optimization module is used to perform a ternary collaborative optimization process using the dynamic weight function as the evaluation criterion to obtain the optimal penalty factor, the optimal number of modes, and the optimal decay coefficient. The signal separation module is used to perform variational mode decomposition on the preprocessed time-domain signal based on the optimal penalty factor and the optimal number of modes to generate candidate respiratory signals; An interference suppression module is used to perform a temporal deconvolution operation on the candidate respiratory signal based on the optimal attenuation coefficient and the reflection path template to generate a compensated respiratory signal. The positioning and calculation module is used to calculate the radial distance based on the compensated breathing signal and, in conjunction with the three-dimensional structural model of the mine roadway, determine the three-dimensional coordinates of the trapped personnel.

[0017] Thirdly, a computer-readable storage medium is provided, comprising: a computer program or instructions; when the computer program or instructions are run on a computer, the computer causes the computer to execute the investment research report processing system based on deep semantic understanding described in the first aspect.

[0018] In summary, the above methods and systems have the following technical effects: 1. This invention significantly improves the accuracy and robustness of identifying micro-signals of life in a strongly interfering mining environment by constructing a multi-dimensional collaborative optimization signal processing flow. The method utilizes a three-element collaborative optimization process guided by a dynamic weight function, which enables the signal decomposition algorithm to adapt to the inherent complexity of the echo signal in real time. This effectively avoids the problems of mode aliasing or feature loss caused by environmental changes in traditional fixed parameter methods, thereby accurately extracting weak respiratory signal features in a complex background intertwined with dust, vibration, and multipath effects.

[0019] 2. This invention introduces a closed-loop verification and adaptive parameter reconstruction mechanism, and automatically optimizes SVMD parameters using WOA, achieving a high degree of intelligence and environmental adaptability in the signal processing flow. This method is not a one-time open-loop processing, but evaluates the quality of the decomposition result by calculating the consistency coefficient between the candidate signal and the original signal, and automatically restarts the optimization process when the result is not ideal. This feedback mechanism eliminates the dependence on manual intervention and experience settings, ensuring that the algorithm can achieve robust performance in rescue scenarios with different roadway structures and different noise levels, greatly enhancing the system's on-site practicality and deployment efficiency.

[0020] 3. This invention achieves high-precision three-dimensional positioning of trapped personnel by deeply integrating signal processing with a physical environment model and using pure UWB signal processing throughout the process. Traditional time positioning has large errors in the multipath environment of mines. This invention constructs a reflection path template and actively suppresses multipath interference using an optimized attenuation coefficient, restoring the pure form of the direct wave. On this basis, it combines spherical constraints and multipath geometric constraints for joint calculation, effectively overcoming the ambiguity of single ranging information, locking the spatial position of the trapped personnel within a more precise range, and providing key decision support for subsequent rescue operations. Attached Figure Description

[0021] Figure 1 This is a schematic diagram of the entire processing flow provided in the embodiments of the present invention; Figure 2 This is a schematic diagram of the signal acquisition and preprocessing process provided in an embodiment of the present invention; Figure 3 This is a schematic diagram of the ternary collaborative optimization and closed-loop verification process provided in an embodiment of the present invention; Figure 4This is a schematic diagram of the signal separation and enhancement process provided in an embodiment of the present invention; Figure 5 This is a flowchart illustrating the interference suppression, localization calculation, and final localization result provided in an embodiment of the present invention. Detailed Implementation

[0022] The technical solution of the present invention will now be described with reference to the accompanying drawings.

[0023] In this embodiment of the invention, "instruction" can include direct and indirect instructions, as well as explicit and implicit instructions. The information indicated by a certain piece of information is called the information to be instructed. In specific implementation, there are many ways to instruct the information to be instructed, such as, but not limited to, directly instructing the information to be instructed, such as the information to be instructed itself or its index. It can also indirectly instruct the information to be instructed by instructing other information, where there is a correlation between the other information and the information to be instructed. It can also instruct only a part of the information to be instructed, while the other parts are known or pre-agreed upon. For example, the instruction of specific information can be achieved by using a pre-agreed (e.g., protocol-defined) arrangement of various pieces of information, thereby reducing instruction overhead to some extent. Simultaneously, common parts of various pieces of information can be identified and uniformly indicated to reduce the instruction overhead caused by individually indicating the same information.

[0024] Furthermore, the specific indication method can also be any existing indication method, such as, but not limited to, the above-mentioned indication methods and their various combinations. Specific details of various indication methods can be found in existing technologies, and will not be elaborated upon here. As described above, for example, when multiple pieces of information of the same type need to be indicated, the indication methods for different pieces of information may differ. In specific implementation, the required indication method can be selected according to specific needs. This embodiment of the invention does not limit the selected indication method; therefore, the indication methods involved in this embodiment of the invention should be understood to cover various methods that enable the party to be indicated to obtain the information to be indicated.

[0025] It should be understood that the information to be indicated can be sent as a whole or divided into multiple sub-information messages sent separately, and the sending period and / or timing of these sub-information messages can be the same or different. The specific sending method is not limited in this embodiment of the invention. The sending period and / or timing of these sub-information messages can be predefined, for example, according to a protocol, or configured by the sending device by sending configuration information to the receiving device.

[0026] "Predefined" or "pre-configured" can be achieved by pre-saving corresponding codes, tables, or other means that can be used to indicate relevant information in the device. This embodiment of the invention does not limit the specific implementation method. "Saving" can refer to saving in one or more memories. These memories can be separate installations or integrated into the encoder, decoder, processor, or electronic device. Alternatively, some memories can be separately installed, while others are integrated into the decoder, processor, or electronic device. The type of memory can be any form of storage medium, and this embodiment of the invention does not limit this.

[0027] In the embodiments of this invention, "protocol" may refer to a protocol family in the field of communication, a standard protocol with a similar protocol family frame structure, or a related protocol to be applied in future systems. The embodiments of this invention do not specifically limit this.

[0028] In this embodiment of the invention, descriptions such as "when," "under the circumstances," "if," and "if" all refer to the device making corresponding processing under certain objective circumstances, and are not limited to a specific time. They do not require the device to make a judgment action during implementation, nor do they imply any other limitations.

[0029] In the description of the embodiments of the present invention, unless otherwise stated, " / " indicates that the objects before and after are in an "or" relationship. For example, A / B can represent A or B. "And / or" in the embodiments of the present invention is merely a description of the relationship between the related objects, indicating that three relationships can exist. For example, A and / or B can represent: A alone, A and B simultaneously, and B alone, where A and B can be singular or plural. Furthermore, in the description of the embodiments of the present invention, unless otherwise stated, "multiple" refers to two or more. "At least one of the following" or similar expressions refer to any combination of these items, including any combination of single or plural items. For example, at least one of a, b, or c can represent: a, b, c, ab, ac, bc, or abc, where a, b, and c can be single or multiple. Additionally, to facilitate a clear description of the technical solutions of the embodiments of the present invention, the terms "first" and "second" are used in the embodiments of the present invention to distinguish identical or similar items with essentially the same function and effect. Those skilled in the art will understand that the terms "first," "second," etc., do not limit the quantity or order of execution, and that "first," "second," etc., are not necessarily different. Furthermore, in the embodiments of this invention, words such as "exemplary" or "for example" are used to indicate that something is being described as an example, illustration, or description. Any embodiment or design scheme described as "exemplary" or "for example" in the embodiments of this invention should not be construed as being more preferred or advantageous than other embodiments or design schemes. Specifically, the use of words such as "exemplary" or "for example" is intended to present the relevant concepts in a concrete manner for ease of understanding.

[0030] The network architecture and business scenarios described in the embodiments of this invention are for the purpose of more clearly illustrating the technical solutions of the embodiments of this invention, and do not constitute a limitation on the technical solutions provided by the embodiments of this invention. As those skilled in the art will know, with the evolution of network architecture and the emergence of new business scenarios, the technical solutions provided by the embodiments of this invention are also applicable to similar technical problems.

[0031] To facilitate understanding of the embodiments of this application, the communication system applicable to the embodiments of this application will be described in detail first, taking the control system as an example.

[0032] Please see Figures 1-5 This application provides a UWB life signal micro-motion identification and positioning method for a mine rescue robot, the processing method including: The echo signals transmitted and received by UWB radar in the mine environment are acquired, and a reflection path template is generated based on the three-dimensional structural model of the mine roadway that characterizes the mine environment. Parallel time-domain and frequency-domain processing is performed on the echo signal to generate a preprocessed time-domain signal and a time-frequency energy distribution map; A spectral concentration is calculated based on the time-frequency energy distribution map, and time-domain energy weights and frequency-domain energy weights are set according to the spectral concentration to generate a dynamic weighting function; A ternary collaborative optimization process is performed, using the dynamic weight function as the evaluation standard, to simultaneously optimize the penalty factor, the number of modes, and the multipath attenuation coefficient, so as to obtain the optimal penalty factor, the optimal number of modes, and the optimal attenuation coefficient. Based on the optimal penalty factor and the optimal number of modes, variational mode decomposition is performed on the preprocessed time-domain signal to obtain multiple intrinsic mode components, and target mode components whose spectral main peak is located in the frequency band corresponding to physiological respiration are extracted from the multiple intrinsic mode components to generate candidate respiration signals; Based on the optimal attenuation coefficient and the reflection path template, a temporal deconvolution operation is performed on the candidate respiratory signal to generate a compensated respiratory signal; The radial distance is calculated based on the compensated respiratory signal, and the three-dimensional coordinates of the trapped personnel are determined by combining the three-dimensional structural model of the mine roadway. First, the raw signal containing multipath effects is acquired using UWB radar, and a reflection path template is synchronously generated using a 3D model of the mine, laying the physical foundation for subsequent interference suppression. A strategy of parallel processing in the time and frequency domains is adopted to separate the preprocessed time-domain signal and time-frequency energy distribution map from the raw signal, achieving synchronous characterization of the signal waveform and dynamic energy distribution. The most critical innovation lies in the ternary collaborative optimization process, which utilizes the whale optimization algorithm. Instead of optimizing the penalty factor and mode number of variational mode decomposition in isolation, it couples them with the path attenuation coefficient and introduces a dynamic weight function composed of time-frequency energy entropy. This allows the optimization process to adjust the optimization direction in real time according to the complexity of the signal itself, transforming from an open-loop search to a closed-loop feedback with self-correction capabilities. Finally, signal decomposition, phase compensation, and time calculation are performed using optimal parameters, achieving the entire process of accurately separating micro-motion signals from a strong interference background and performing high-precision 3D positioning.

[0033] Optionally, the step of performing parallel time-domain and frequency-domain processing on the echo signal includes: Low-pass filtering is performed on the echo signal in the time domain path to filter out high-frequency noise and retain the waveform containing human body micro-movements and low-frequency multipath interference, thus obtaining the preprocessed time domain signal; A short-time Fourier transform is performed on the echo signal in the frequency domain path to reveal the dynamic characteristics of the signal frequency components changing over time, thereby obtaining the time-frequency energy distribution map. The time-frequency energy distribution map provides a basis for subsequent coarse estimation of respiratory rate and interference localization. For a discrete echo signal x[n], its time-frequency transformation result X[m, k] can be expressed as: ; Where n is the index of the time sample point, m is the time index of the window function shift, k is the discrete frequency index, w is the window function, and N is the number of points in the Fourier transform. The time-frequency energy distribution map E[m, k] is obtained by calculating the square of the modulus of the transform result. ; This matrix is ​​a real number matrix, and its element values ​​reflect the energy intensity of the signal at a specific time m and a specific frequency k. This parallel processing mechanism ensures that the waveform fidelity in the time domain and the dynamic energy distribution in the frequency domain can be captured simultaneously without interference. Ultimately, the preprocessed time-domain signal provides a clean input source for the subsequent variational mode decomposition, while the time-frequency energy distribution map serves as a dynamic basis for calculating the spectral concentration and evaluating the decomposition effect, providing crucial prior information for subsequent adaptive optimization.

[0034] Optionally, the step of generating the dynamic weight function includes: The time-frequency energy distribution map is used to calculate the variance of its energy distribution on the time axis in order to obtain the time-domain energy uniformity. The energy uniformity U in this time domain can be expressed as: ; Where E represents a vector whose elements Let be the sum of the energies of all frequency components within the m-th time window of the time-frequency energy matrix. and U represents the mean and standard deviation of vector E, respectively. A U value close to 1 indicates that the energy is uniformly distributed over time, while a value close to 0 indicates that the distribution is non-uniform. The spectral concentration and the temporal energy uniformity are normalized and then weighted and fused to generate a signal complexity index; Signal complexity index The calculation formula is as follows: ; Where a and b are preset fusion weights, their sum is 1, and they can usually be set to an equal value of 0.5. The value range is also between 0 and 1, a high The value indicates that the signal energy exhibits a high degree of non-uniformity in both the time and frequency domains, meaning the signal is more complex. The spectral concentration F is obtained by calculating the ratio of the geometric mean to the arithmetic mean of the time-frequency energy matrix on the frequency axis. It reflects the degree of concentration of signal energy in the frequency domain, and its value is between 0 and 1. The lower the value, the more concentrated the energy is in a few frequency bands. The time-domain energy weight and the frequency-domain energy weight are generated by inverse linear mapping based on the signal complexity index, such that the frequency-domain energy weight increases accordingly when the signal complexity index increases, thereby enhancing the optimization of the purity of the decomposition result.

[0035] Optionally, the step of performing the ternary collaborative optimization process includes: Initialize a whale population containing multiple individuals, where each individual's position vector includes a candidate penalty factor, the number of candidate modes, and a candidate decay coefficient; For each individual in the population, a temporary variational mode decomposition and deconvolution reconstruction are performed on the preprocessed time-domain signal using its corresponding candidate parameters to generate a reconstructed signal. The reconstructed signal is scored using the dynamic weighting function to obtain the fitness value of the individual; Based on the fitness values ​​of all individuals, the position of the whale population is updated through the encirclement predation, bubble net attack and random search mechanisms of the whale optimization algorithm, and this process is repeated until the convergence condition is met, so as to output the optimal penalty factor, the optimal number of modes and the optimal decay coefficient.

[0036] Optionally, the step of performing the temporal deconvolution operation includes: Construct a deconvolution kernel, wherein the reflection path template defines the time offset sequence of the deconvolution kernel, and the optimal attenuation coefficient defines the pulse attenuation factor representing the multipath replica in the deconvolution kernel; The candidate respiratory signal is convolved with the deconvolution kernel to remove the multipath copies represented by the reflection path template point by point, thereby obtaining the compensated respiratory signal, wherein the energy ratio of the direct wave component to the first reflection wave component of the compensated respiratory signal is improved. To accurately remove multipath interference caused by the mine roadway environment from the initially extracted candidate respiratory signals, this method performs a time-domain deconvolution operation based on an adaptive kernel, aiming to recover the original micro-motion waveform of life uncontaminated by multipath effects. The core of this operation is to construct a deconvolution kernel that can accurately characterize the multipath channel response. The deconvolution kernel is designed as a time offset sequence, and its time offset is directly derived from the reflection path template data generated in the previous step. The template records the arrival time difference of the main reflected wave relative to the direct wave. At the same time, the pulse amplitude representing each multipath replica in the deconvolution kernel is set by the optimized multipath attenuation coefficient γ as its key attenuation factor. When considering the main multipath reflections, the deconvolution kernel h[n] can be expressed in the discrete-time domain as: ; L represents the number of primary reflection paths, where δ[n] is the unit impulse function, representing the direct wave path. This represents the number of delayed sampling points for the main reflection path extracted from the reflection path template. This is the optimal attenuation coefficient for the path, with a value range between 0 and 1. It reflects the degree of energy attenuation of the reflected wave relative to the direct wave. This time-domain deconvolution process is implemented through a filtering algorithm that removes multipath copies point by point. The candidate respiratory signal is processed with the constructed deconvolution kernel to eliminate the delay component recorded in the reflection path template in reverse, and finally generate the compensated respiratory signal.

[0037] Optionally, before performing the temporal deconvolution operation, the method further includes: The candidate respiratory signal and the preprocessed time-domain signal are respectively subjected to bandpass filtering, with the filtering passband being the frequency band corresponding to physiological respiration. Then, the maximum value of the normalized cross-correlation function of the two filtered waveforms is calculated, and the maximum value is used as the consistency coefficient. Determine whether the consistency coefficient is lower than the consistency threshold; If the consistency coefficient is lower than the consistency threshold, the optimal parameter combination obtained in the current optimization process is used as the initial value, and the ternary collaborative optimization process is restarted until the consistency coefficient corresponding to the newly generated candidate respiratory signal is not lower than the consistency threshold.

[0038] Optionally, the step of calculating the radial distance based on the compensated respiratory signal includes: Envelope detection is performed on the compensated respiratory signal to obtain the signal envelope; The peak position of the signal envelope is extracted as the arrival time of the direct wave; The radial distance is calculated based on the time difference between the arrival time of the direct wave and the pulse emission time, and in conjunction with the refractive index that characterizes the properties of the tunnel medium.

[0039] Optionally, the step of determining the three-dimensional coordinates of the trapped person includes: Based on the rescue robot's own pose information and the radial distance, a spherical constraint equation is generated; Based on the wall reflection surface equation in the three-dimensional structural model of the mine roadway and the multipath length in the reflection path template, a set of multipath geometric constraint equations is constructed. Solve the spherical constraint equation and the multipath geometric constraint equation simultaneously to obtain multiple candidate coordinate points, and identify the three-dimensional coordinates from the multiple candidate coordinate points based on the energy attenuation characteristics of the candidate respiratory signals.

[0040] Optionally, the frequency band corresponding to physiological respiration covers the human respiratory frequency in a resting state. Specifically, to ensure that this method can accurately capture the weak signals generated by human physiological activities, multiple processing steps in the system are designed and configured around a key preset respiratory frequency band. The respiratory rate of adults at rest or under mild stress is usually stable within a specific range. This method precisely sets the preset respiratory frequency band to 0.2 Hz to 0.4 Hz, which corresponds to a respiratory rate of 12 to 24 breaths per minute, covering the possible physiological states of most trapped individuals. The setting of this frequency band is the benchmark for all subsequent signal processing steps. In the dual-domain preprocessing stage, the frequency response range of the designed low-frequency time-domain signal must completely cover this frequency band to ensure that while filtering out high-frequency noise, no attenuation of the target respiratory signal is caused. The resolution of the generated time-frequency energy matrix on the frequency axis must also be high enough to clearly distinguish the energy peaks located within this frequency band.

[0041] Optionally, the signal acquisition and template generation module is used to acquire an echo signal and generate a reflection path template based on the three-dimensional structural model of the mine roadway; A dual-domain preprocessing module is used to perform parallel time-domain and frequency-domain processing on the echo signal to generate a preprocessed time-domain signal and a time-frequency energy distribution map; An adaptive function construction module is used to generate a dynamic weighting function based on the time-frequency energy distribution map; The collaborative optimization module is used to perform a ternary collaborative optimization process using the dynamic weight function as the evaluation criterion to obtain the optimal penalty factor, the optimal number of modes, and the optimal decay coefficient. The signal separation module is used to perform variational mode decomposition on the preprocessed time-domain signal based on the optimal penalty factor and the optimal number of modes to generate candidate respiratory signals; An interference suppression module is used to perform a temporal deconvolution operation on the candidate respiratory signal based on the optimal attenuation coefficient and the reflection path template to generate a compensated respiratory signal. The positioning and calculation module is used to calculate the radial distance based on the compensated breathing signal and, in conjunction with the three-dimensional structural model of the mine roadway, determine the three-dimensional coordinates of the trapped personnel.

[0042] In this test, the core operating parameters of the UWB radar system were set as follows: pulse repetition frequency of 10 MHz and sampling frequency of 100 Hz. The rescue robot started from the alleyway entrance (origin) and moved to the position (50, 10, 2) (unit: meters), at which point the system activated its detection.

[0043] 1. Signal Acquisition and Dual-Domain Preprocessing: The robot acquired a composite echo signal lasting 20 seconds. This signal included micro-motion signals generated by the breathing simulator, mechanical vibration noise from the robot itself, and strong multipath reflections generated by the tunnel wall. Simultaneously, based on the pre-loaded three-dimensional structural model of the tunnel, the system generated a reflection path template and identified a main reflection path with a delay of 15 nanoseconds relative to the direct wave.

[0044] Subsequently, the composite echo signal is fed into a dual-domain parallel preprocessing module. In the time domain path, the signal passes through a fourth-order Butterworth low-pass filter with a cutoff frequency set to 0.8 Hz, generating a low-frequency time-domain signal that retains key micro-motion information. In the frequency domain path, the system performs a Short-Time Fourier Transform (STFT) on the signal, with parameters set to a 256-point Hamming window and a 75% overlap. This process generates a time-frequency energy matrix, providing a data foundation for subsequent complexity analysis.

[0045] 2. Construction of dynamic weight function number: Based on the generated time-frequency energy matrix, the system begins to calculate the intrinsic complexity of the signal. First, by analyzing the energy distribution of the matrix along the time axis, its mean energy is calculated. =2.8, standard deviation is =1.54. Based on this, the time-domain energy uniformity is obtained. The lower The value indicates that the signal energy is unevenly distributed over time, and there is sudden interference.

[0046] Next, the system calculated the spectral concentration. ,get This indicates that the signal energy is concentrated in a few frequency bands, and the frequency domain is highly discrete.

[0047] Subsequently, the system merges these two metrics to generate a signal complexity index. =0.615.

[0048] because = 0.615, which is higher than the preset threshold of 0.6. The system determines that the current signal environment is complex and requires a greater focus on the purity of the decomposition. Therefore, the weights of the fitness function are dynamically generated: frequency domain energy weights. = 0.615, time-domain energy weight .

[0049] Table 1. Data on UWB signal preprocessing and complexity analysis 3. Triadic collaborative optimization and closed-loop verification: The system initiates the Whale Optimization Algorithm to perform ternary collaborative optimization, with the WOA population size set to 30 and the maximum number of iterations set to 80. The parameter search range is set to: penalty factor. [500, 5000], number of modes [2, 10], multipath attenuation coefficient [0, 1].

[0050] On the 45th iteration, the algorithm converges to obtain a set of candidate optimal parameters: The system uses these parameters to generate a candidate respiratory signal and then proceeds to the closed-loop verification stage. The consistency coefficient is obtained by calculating the energy correlation between this candidate signal and a low-frequency time-domain signal within a preset respiratory frequency band (0.2 Hz to 0.4 Hz). .

[0051] This value is below the preset consistency threshold of 0.85, so the system determines this to be a pseudo-optimal solution and triggers a re-optimization. The system will ( This serves as the initial reference point for a new round of searching, and the process continues iterating.

[0052] In the 72nd iteration, the algorithm converges to a new optimal combination of parameters: optimal penalty factor α_opt = 2350, and optimal number of modes. Optimal multipath attenuation coefficient The consistency coefficient of the new candidate respiratory signals generated based on this combination is... The result is higher than the threshold of 0.85. The system confirms the result is valid, and the optimization process ends.

[0053] Table 2. Parameter optimization and closed-loop verification process data 4. Signal separation, interference suppression, and localization calculation: The system uses the validated optimal parameters to perform variational mode decomposition on the low-frequency time-domain signal, obtaining four intrinsic mode components. One of these components has a spectral peak at 0.3 Hz, which is identified as the final candidate respiratory signal.

[0054] Subsequently, based on the optimal multipath attenuation coefficient Construct a deconvolution kernel with a 15 nanosecond delay in the reflection path template. Where d corresponds to the number of sampling points in 15 nanoseconds. After performing temporal deconvolution on the candidate respiratory signals, a compensated respiratory signal is generated. Analysis shows that the energy ratio of its direct wave component to the first reflected wave component increases from 5 dB before processing to 18 dB.

[0055] Next, Time-of-Flight (TOF) calculations are performed. The system performs envelope detection on the compensated respiratory signal to pinpoint the arrival time of the direct wave. The time difference between it and the launch time t_t is calculated as follows: Nanoseconds. Combined with the refractive index of the tunnel medium calibrated on-site. Calculate the radial distance rice.

[0056] Finally, the three-dimensional coordinates were calculated by simultaneously solving the spherical constraint equations and the multipath geometric constraint equations constructed based on the multipath path length L_m = 23.75 meters and the tunnel wall equation y=0. Two candidate coordinate points were calculated: P1=(65.8, 4.1, 2.5) and P2=(43.1, -3.5, 3.1). By comparing the theoretical attenuation value with the actual received signal energy using an attenuation model, the theoretical attenuation value of P1 best matched the actual value. Therefore, the three-dimensional coordinates of the trapped personnel were finally determined to be (65.8, 4.1, 2.5).

[0057] Table 3 Key data in the location solution process As can be seen from the detailed process and data derivation of the above embodiments, this invention, by constructing a dynamic weighted fitness function, can adaptively adjust the optimization strategy according to the complexity of the signal environment (as shown in Table 1, high complexity leads to a higher emphasis on purity). Its closed-loop verification mechanism (as shown in Table 2) effectively avoids the algorithm getting trapped in local optima, ensuring the authenticity of the extracted breathing signals. Finally, by combining multipath compensation and multiple geometric constraints, this invention realizes the entire process from weak signal detection under strong interference background to high-precision three-dimensional positioning (as shown in Table 3), verifying its reliability and advanced nature in complex mining environments.

[0058] Those skilled in the art will recognize that the units and algorithm steps of the various examples described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementations should not be considered beyond the scope of this invention.

[0059] Those skilled in the art will understand that, for the sake of convenience and brevity, the specific working processes of the systems, devices, and units described above can be referred to the corresponding processes in the foregoing method embodiments, and will not be repeated here.

[0060] In the embodiments provided by this invention, it should be understood that the disclosed systems, apparatuses, and methods can be implemented in other ways. For example, the apparatus embodiments described above are merely illustrative; for instance, the division of units is only a logical functional division, and in actual implementation, there may be other division methods. For example, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the coupling or direct coupling or communication connection shown or discussed may be through some interfaces; the indirect coupling or communication connection between devices or units may be electrical, mechanical, or other forms.

[0061] The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the units can be selected to achieve the purpose of this embodiment according to actual needs.

[0062] In addition, the functional units in the various embodiments of the present invention can be integrated into the processing unit, or each unit can exist physically separately, or two or more units can be integrated into the unit.

[0063] If the aforementioned functions are implemented as software functional units and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this invention, or the part that contributes to the prior art, or a portion of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of this invention. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0064] The above description is merely a specific embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the technical scope disclosed in the present invention should be included within the scope of protection of the present invention. Therefore, the scope of protection of the present invention should be determined by the scope of the claims.

Claims

1. A method for UWB life signal micro-motion identification and positioning of a mine rescue robot, characterized in that, Includes the following steps: The echo signals transmitted and received by UWB radar in the mine environment are acquired, and a reflection path template is generated based on the three-dimensional structural model of the mine roadway that characterizes the mine environment. Parallel time-domain and frequency-domain processing is performed on the echo signal to generate a preprocessed time-domain signal and a time-frequency energy distribution map; A spectral concentration is calculated based on the time-frequency energy distribution map, and a time-domain energy weight and a frequency-domain energy weight are set according to the spectral concentration to generate a dynamic weight function. A ternary collaborative optimization process is performed, using the dynamic weight function as the evaluation standard, to simultaneously optimize the penalty factor, the number of modes, and the multipath attenuation coefficient, so as to obtain the optimal penalty factor, the optimal number of modes, and the optimal attenuation coefficient. Based on the optimal penalty factor and the optimal number of modes, variational mode decomposition is performed on the preprocessed time-domain signal to obtain multiple intrinsic mode components, and target mode components whose spectral main peak is located in the frequency band corresponding to physiological respiration are extracted from the multiple intrinsic mode components to generate candidate respiration signals; Based on the optimal attenuation coefficient and the reflection path template, a temporal deconvolution operation is performed on the candidate respiratory signal to generate a compensated respiratory signal; The radial distance is calculated based on the compensated respiratory signal, and the three-dimensional coordinates of the trapped personnel are determined by combining the three-dimensional structural model of the mine roadway.

2. The UWB life signal micro-motion identification and positioning method for a mine rescue robot according to claim 1, characterized in that, The step of performing parallel time-domain and frequency-domain processing on the echo signal includes: Low-pass filtering is performed on the echo signal in the time domain path to filter out high-frequency noise and retain the waveform containing human body micro-movements and low-frequency multipath interference, thus obtaining the preprocessed time domain signal; A short-time Fourier transform is performed on the echo signal in the frequency domain path to reveal the dynamic characteristics of the signal frequency components changing over time, thereby obtaining the time-frequency energy distribution map. The time-frequency energy distribution map provides a basis for subsequent coarse estimation of respiratory rate and interference localization.

3. The UWB life signal micro-motion identification and positioning method for a mine rescue robot according to claim 1, characterized in that, The step of generating the dynamic weight function includes: The time-frequency energy distribution map is used to calculate the variance of its energy distribution on the time axis in order to obtain the time-domain energy uniformity. The spectral concentration and the temporal energy uniformity are normalized and then weighted and fused to generate a signal complexity index; The time-domain energy weight and the frequency-domain energy weight are generated by inverse linear mapping based on the signal complexity index, such that the frequency-domain energy weight increases accordingly when the signal complexity index increases, thereby enhancing the optimization of the purity of the decomposition result.

4. The UWB life signal micro-motion identification and positioning method for a mine rescue robot according to claim 1, characterized in that, The steps for performing the ternary collaborative optimization process include: Initialize a whale population containing multiple individuals, where the position vector of each individual includes a candidate penalty factor, the number of candidate modes, and a candidate decay coefficient; For each individual in the population, a temporary variational mode decomposition and deconvolution reconstruction are performed on the preprocessed time-domain signal using its corresponding candidate parameters to generate a reconstructed signal. The reconstructed signal is scored using the dynamic weighting function to obtain the fitness value of the individual; Based on the fitness values ​​of all individuals, the position of the whale population is updated through the encirclement predation, bubble net attack and random search mechanisms of the whale optimization algorithm, and this process is repeated until the convergence condition is met, so as to output the optimal penalty factor, the optimal number of modes and the optimal decay coefficient.

5. The UWB life signal micro-motion identification and positioning method for a mine rescue robot according to claim 4, characterized in that, The step of performing the temporal deconvolution operation includes: Construct a deconvolution kernel, wherein the reflection path template defines the time offset sequence of the deconvolution kernel, and the optimal attenuation coefficient defines the pulse attenuation factor representing the multipath replica in the deconvolution kernel; The candidate respiratory signal is convolved with the deconvolution kernel to remove the multipath copies represented by the reflection path template point by point, thereby obtaining the compensated respiratory signal, wherein the energy ratio of the direct wave component to the first reflected wave component of the compensated respiratory signal is improved.

6. The UWB life signal micro-motion identification and positioning method for a mine rescue robot according to claim 5, characterized in that, Before performing the temporal deconvolution operation, the method further includes: The candidate respiratory signal and the preprocessed time-domain signal are respectively subjected to bandpass filtering, with the filtering passband being the frequency band corresponding to physiological respiration. Then, the maximum value of the normalized cross-correlation function of the two filtered waveforms is calculated, and the maximum value is used as the consistency coefficient. Determine whether the consistency coefficient is lower than the consistency threshold; If the consistency coefficient is lower than the consistency threshold, the optimal parameter combination obtained in the current optimization process is used as the initial value, and the ternary collaborative optimization process is restarted until the consistency coefficient corresponding to the newly generated candidate respiratory signal is not lower than the consistency threshold.

7. The UWB life signal micro-motion identification and positioning method for a mine rescue robot according to claim 1, characterized in that, The step of calculating the radial distance based on the compensated respiratory signal includes: Envelope detection is performed on the compensated respiratory signal to obtain a signal envelope; The peak position of the signal envelope is extracted as the arrival time of the direct wave; The radial distance is calculated based on the time difference between the arrival time of the direct wave and the emission time of a pulse, and in conjunction with the refractive index that characterizes the properties of the tunnel medium.

8. The UWB life signal micro-motion identification and positioning method for a mine rescue robot according to claim 7, characterized in that, The step of determining the three-dimensional coordinates of the trapped personnel includes: Based on the pose information of a rescue robot and the radial distance, a spherical constraint equation is generated; Based on the wall reflection surface equation in the three-dimensional structural model of the mine roadway and the multipath length in the reflection path template, a set of multipath geometric constraint equations is constructed. Solve the spherical constraint equation and the multipath geometric constraint equation simultaneously to obtain multiple candidate coordinate points, and identify the three-dimensional coordinates from the multiple candidate coordinate points based on the energy attenuation characteristics of the candidate respiratory signals.

9. A method for UWB life signal micro-motion identification and positioning of a mine rescue robot according to claim 8, characterized in that, The frequency band corresponding to physiological respiration covers the human respiratory frequency in a resting state.

10. A UWB life signal micro-motion identification and positioning system for a mine rescue robot, applied to the UWB life signal micro-motion identification and positioning method for a mine rescue robot as described in any one of claims 1-9, characterized in that, Specifically, it includes the following modules: The signal acquisition and template generation module is used to acquire echo signals and generate reflection path templates based on the three-dimensional structural model of the mine roadway. A dual-domain preprocessing module is used to perform parallel time-domain and frequency-domain processing on the echo signal to generate a preprocessed time-domain signal and a time-frequency energy distribution map; An adaptive function construction module is used to generate a dynamic weighting function based on the time-frequency energy distribution map; The collaborative optimization module is used to perform a ternary collaborative optimization process using the dynamic weight function as the evaluation criterion to obtain the optimal penalty factor, the optimal number of modes, and the optimal decay coefficient. The signal separation module is used to perform variational mode decomposition on the preprocessed time-domain signal based on the optimal penalty factor and the optimal number of modes to generate candidate respiratory signals; An interference suppression module is used to perform a temporal deconvolution operation on the candidate respiratory signal based on the optimal attenuation coefficient and the reflection path template to generate a compensated respiratory signal. The positioning and calculation module is used to calculate the radial distance based on the compensated breathing signal and, in conjunction with the three-dimensional structural model of the mine roadway, determine the three-dimensional coordinates of the trapped personnel.

Images