Non-contact vital sign recognition and detection system

Abstract

The invention discloses a non-contact vital sign recognition and detection system, which relates to the technical field of vital sign monitoring, and comprises a time-sharing echo acquisition module for acquiring echo signals and constructing an original data matrix; the phase processing module is used for executing background subtraction and phase extraction processing on the original data matrix and resolving a phase signal; the vital sign separation module is used for processing the phase signal and generating a breathing waveform; the breathing event judgment module is used for converting the breathing waveform into a physical displacement waveform, and judging and generating an apnea early warning signal; the radar dynamic scheduling module is used for generating a radar switching instruction according to the calculated target position; and the power compensation module obtains the signal quality parameter and generates a power adjustment instruction. According to the invention, through cooperation of double radars and dynamic scheduling, global blind-area-free continuous monitoring is realized, obstacles can be penetrated and breath and heartbeat can be accurately separated by using an advanced signal processing method, and low power consumption and high environmental adaptability are realized while high precision is ensured.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of vital sign monitoring, in particular to a non-contact vital sign recognition and detection system. BACKGROUND

[0002] At present, the health monitoring technical solutions for the elderly mainly cover wearable devices, environmental sensors and radar technology applications. Wearable devices usually collect heart rate data through wristband sensors. Environmental sensor solutions mostly use a combination of cameras and infrared technology for behavior monitoring. Radar technology, especially millimeter wave radar, is also applied to health monitoring in nursing scenes. These technologies aim to remotely and real-time understand the health status of the elderly through different ways. The core principle is to capture physiological parameters or behavior patterns of the human body through various sensors, such as using ultra-wideband radar to emit nanosecond-level pulses and analyzing the return signal to perceive the weak displacement of the chest caused by human respiratory, heartbeat and other life activities. These displacements will cause periodic changes in the phase of the radar return signal, and by solving these changes, vital sign information can be extracted. However, existing technical solutions face a high abandonment rate in the context of monitoring solitary elderly people. The fundamental reasons are privacy invasion, the burden caused by device wearing, and insufficient adaptability to complex home environments.

[0003] Existing technologies have many defects in practical application. Non-ultra-wideband radar technology solutions have fundamental limitations. For example, millimeter wave radar has weak penetration ability, and the signal will be severely attenuated or even lost when encountering obstacles such as walls, making it impossible to achieve through-wall monitoring. Other sensors also have their own shortcomings. Infrared sensors rely on surface temperature and will fail when the target is covered by a quilt. Camera solutions are affected by light and cannot work in shaded areas. Contact sensors such as piezoelectric sensors require direct contact with the human body and can easily cause data interruption when the elderly turn over. Secondly, even if ultra-wideband radar with better penetration is used, traditional single-radar system architecture also has technical bottlenecks. First, there is a spatial coverage defect. The beam width of a single radar is limited, which can easily form a monitoring blind area in the room. Second, there is a dynamic response defect. When the monitoring target moves quickly, a single-radar system is difficult to quickly re-lock the target, resulting in long-term interruption of monitoring. In addition, when multiple devices are deployed to solve the blind area problem, signal interference will occur due to the same frequency operation, resulting in a significant decrease in measurement data accuracy. SUMMARY

[0004] The purpose of the present application is to provide a non-contact vital sign recognition and detection system that solves the problems in the background art.

[0005] To solve the above technical problems, the present application provides a kind of non-contact vital sign recognition detection system, comprising: time-sharing echo acquisition module, time-sharing multiplexing time sequence alternation control master ultra-wideband radar module and slave ultra-wideband radar module work, and the echo signal reflected by monitoring target is collected, and original data matrix is constructed; Phase processing module, background reduction and phase extraction processing are carried out to original data matrix, and phase signal is solved; Vital sign separation module, first band-pass filter processing is carried out to phase signal, and respiratory waveform is generated; Respiratory event determination module, respiratory waveform is converted into physical displacement waveform, and whether the amplitude and duration of physical displacement waveform meet the preset apnea condition is judged, and apnea warning signal is generated; Radar dynamic scheduling module, according to the target position solved from original data matrix, in combination with the preset monitoring area division, radar switching instruction for adjusting time-sharing multiplexing time sequence is generated; Power compensation module, signal quality parameter of echo signal is obtained, and power adjustment instruction for adjusting the transmission power of working radar module is generated according to the comparison result of signal quality parameter and preset quality threshold.

[0006] Preferably, the specific steps of background reduction processing carried out by phase processing module are as follows: The history average signal of original data matrix is calculated by using sliding average filter method, and background signal is obtained;The background signal is subtracted from the original data matrix to enhance the vital sign signal.

[0007] Preferably, it further includes data fusion processing, and the steps are as follows: The signal quality scores of echo signals collected by master ultra-wideband radar module and slave ultra-wideband radar module are calculated respectively;The original data matrices of master ultra-wideband radar module and slave ultra-wideband radar module are weighted and fused based on signal quality scores to generate fusion data matrix. Phase processing module solves phase signal based on fusion data matrix.

[0008] Preferably, the specific steps of respiratory event determination module are as follows: The phase change of respiratory waveform is converted into physical displacement signal of chest cavity based on the preset radar center frequency;The peak-to-peak value of physical displacement signal is calculated in sliding time window to obtain real-time displacement amplitude;When real-time displacement amplitude is continuously lower than the preset weak breathing threshold, and the duration exceeds the preset pause time threshold, apnea warning signal is generated;Otherwise, it is determined as normal breathing state.

[0009] Preferably, the specific steps of radar dynamic scheduling module are as follows: The monitoring space is divided into master radar responsible area, slave radar responsible area and transition area. predicting a time when the target will enter the transition zone based on the target position and the change; generating a radar switching instruction based on the predicted time.

[0010] Preferably, the specific steps of the power compensation module are: obtaining the received signal strength and the signal-to-noise ratio as the signal quality parameters; performing normalization processing and weighted summation on the signal quality parameters to generate a comprehensive quality index; setting a high quality threshold and a low quality threshold; when the comprehensive quality index is lower than the low quality threshold, generating a power adjustment instruction to increase the transmission power; when the comprehensive quality index is higher than the high quality threshold, generating a power adjustment instruction to reduce the transmission power; when the comprehensive quality index is between the low quality threshold and the high quality threshold, generating a power adjustment instruction to maintain the current transmission power.

[0011] Preferably, the vital sign separation module is also used to generate a heart rate waveform, and the steps are: using an adaptive filtering method to subtract the respiratory harmonic component from the phase signal with the respiratory waveform as a reference to obtain a heart rate phase signal; performing second band-pass filtering processing on the heart rate phase signal to generate a heart rate waveform.

[0012] Preferably, the radar switching instruction is used to control the relay timing, and the logic of the timing is: when the monitoring target is located in the main radar responsible area, opening the main radar relay channel and closing the slave radar relay channel; when the monitoring target is located in the slave radar responsible area, closing the main radar relay channel and opening the slave radar relay channel; when the monitoring target is located in the transition zone, alternately controlling the main radar relay channel and the slave radar relay channel in a pulse width modulation manner.

[0013] Compared with the prior art, the present application has the following beneficial effects: 1. The present application realizes all-around and uninterrupted vital sign monitoring of dynamic targets in a complex home environment by constructing an intelligent collaborative perception system. With the aid of radar technology, it exhibits excellent physical penetration ability and can effectively overcome the obstruction and interference of daily home obstacles such as walls and quilts. At the same time, the system establishes a closed-loop adaptive energy management mechanism that can automatically adjust the transmission power according to real-time signal quality, significantly reducing the overall energy consumption of the system on the premise of ensuring stable and reliable signal acquisition, making it particularly suitable for home monitoring scenarios that require long-term uninterrupted operation.

[0014] 2、The application ensures complete coverage of the whole space and continuous data tracking of the moving target by the prospective manner of pre-judging and seamless switching of the user's position through the cooperative deployment and dynamic scheduling of the dual radars, and discards the mechanical scanning structure, which not only improves the stability and reliability of the system, but also ensures that the vital sign data can be continuously and stably captured when the monitored object moves in different rooms, greatly improving the reliability of dynamic monitoring.

[0015] 3、The application adopts the dynamic background elimination algorithm, can efficiently separate the weak signal generated by human life activities from the complex static environment clutter, greatly improves the signal-to-noise ratio of the effective signal, and converts the abstract phase signal into the chest displacement with clear physical meaning, so that the determination standard of the respiratory abnormal event can be directly related to the clinical definition, greatly improving the accuracy and medical reference value of the early warning, realizing the deep mapping from abstract data to accurate physiological indicators at the signal processing algorithm level, and ensuring the high precision and high reliability of the monitoring result. BRIEF DESCRIPTION OF DRAWINGS

[0016] In order to more clearly illustrate the technical solutions in the embodiments of the present application or the prior art, the following will briefly introduce the drawings needed to be used in the embodiment or prior art description. Obviously, the drawings in the following description are only some embodiments of the present application, and other drawings can be obtained by those skilled in the art without creative labor. Figure 1 The logic block diagram of the non-contact vital sign recognition and detection system of the present application. DETAILED DESCRIPTION

[0017] The technical solutions in the embodiments of the present application will be described clearly and completely in combination with the drawings in the embodiments of the present application. Obviously, the described embodiments are only some embodiments of the present application, not all the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative labor are within the scope of protection of the present application.

[0018] Embodiment 1: Please refer to Figure 1 The present application provides a non-contact vital sign recognition and detection system, which comprises: a time-sharing echo acquisition module, which adopts time-sharing multiplexing time sequence alternation control to work the master ultra-wideband radar module and the slave ultra-wideband radar module, acquires the echo signal reflected by the monitoring target, and constructs an original data matrix; A phase processing module performs background subtraction and phase extraction processing on the original data matrix to solve the phase signal; A vital sign separation module performs first band-pass filtering processing on the phase signal to generate a respiratory waveform; a respiratory event determination module, which converts the respiratory waveform into a physical displacement waveform, and determines and generates a respiratory pause warning signal based on whether the amplitude and duration of the physical displacement waveform meet preset respiratory pause conditions; a radar dynamic scheduling module, which generates radar switching instructions for adjusting the time division multiplexing sequence according to the target position calculated from the original data matrix and in combination with a preset monitoring area division; a power compensation module, which obtains a signal quality parameter of the echo signal, and generates a power adjustment instruction for adjusting the transmission power of the working radar module according to the comparison result of the signal quality parameter and a preset quality threshold; The present application realizes non-invasive, continuous and high-precision detection of the vital signs of the monitoring target through the cooperative work of multiple modules. The time division echo acquisition module aims to solve the problem of same frequency interference when deploying two radars through a time division multiplexing mechanism, and to build a basic data set for subsequent signal processing; in this embodiment, the module generates a time division multiplexing sequence through a central processing unit, and the sequence alternately controls the transmission and reception states of the master ultra-wideband radar module and the slave ultra-wideband radar module at a preset period (for example, 200 milliseconds); at any moment, only one radar module is in a working state, transmits a nanosecond-level ultra-short pulse, and receives the echo signal reflected by the monitoring target (such as a human chest cavity); the collected echo signal is digitized and constructed into an original data matrix ; , wherein represents a slow time axis, which is the time sequence index of the radar pulse and is derived from a radar pulse counter; represents a fast time axis, which is the distance gate corresponding to the sampling point in each pulse and is derived from the sampling clock of an analog-to-digital converter; The phase processing module aims to accurately extract the weak phase change caused by life activities from the original data containing a large amount of static clutter; in this embodiment, the processing flow of the module is as follows: determining the core distance region where the target is located; the target position information can be calculated from the original data matrix by the radar dynamic scheduling module of the present application, and the result is the distance gate index where the target is located; performing background subtraction processing on the entire original data matrix to obtain a difference signal matrix ; extracting the difference signal on the target distance gate , which is a complex vector: ; ​ is the differential signal on the target range gate, a complex vector; is the in-phase signal component on the target range gate; is the quadrature signal component on the target range gate; is the imaginary unit; is the slow time axis, representing the time index of radar transmitted pulses; is the range gate index where the target is located; By taking the inverse tangent of the in-phase component of the vector and the quadrature component , the phase signal directly reflecting the micro-motion change of the target is calculated , whose calculation formula is: ; is the phase signal directly reflecting the micro-motion change of the target; is the four-quadrant inverse tangent function; is the in-phase signal component on the target range gate; is the quadrature signal component on the target range gate; The value of the phase in the interval can be determined, avoiding the angle ambiguity problem of the traditional inverse tangent function; Vital sign separation module, whose purpose is to separate the waveform corresponding to a specific physiological rhythm (such as respiration) from the mixed phase signal; in this embodiment, the module receives the phase signal output by the phase processing module and performs first band-pass filtering on it; the passband range of the filter is set to cover the interval of typical human respiratory frequency, for example, 0.1 Hz to 0.6 Hz; the result of the filtering process is to generate a respiratory waveform , the time sequence pattern of which directly corresponds to the periodic fluctuation of the chest cavity of the monitored target; Respiratory event judgment module, whose purpose is to monitor and warn possible respiratory abnormal events (such as apnea) in real time based on the extracted respiratory waveform; in this embodiment, the module first converts the respiratory waveform into a physical displacement waveform , which represents the real physical fluctuation of the chest cavity; then, the module continuously analyzes the amplitude and duration of the physical displacement waveform; when the amplitude of the waveform is below a preset weak breathing threshold for a continuous period of time, and the duration exceeds a preset pause time threshold, the system determines that an apnea event has occurred, and generates an apnea warning signal; the preset apnea condition here is a set of logical judgment rules established in combination with clinical medical standards and experimental data statistics, which comprehensively considers the amplitude and duration of respiratory movement to identify potential health risks.

[0019] A radar dynamic scheduling module, which aims to intelligently select the optimal radar for data collection according to the position change of the monitoring target, to realize uninterrupted continuous monitoring without blind area; in this embodiment, the module first calculates the real-time position of the target from the original data matrix ; specifically, the calculation process includes the following steps: first, variance calculation is performed on the original data matrix along the slow time axis to obtain the variance vector of each distance gate ; since human life activities (respiration, heartbeat, etc.) will cause regular fluctuations of signals in specific distance gates, and the signal fluctuations of static environment are smaller, therefore, the signal variance caused by the target reflection is usually the largest; by finding the peak value in the variance vector , the distance gate where the target is located can be determined; the real-time position (distance) of the target can be obtained by multiplying the distance gate index by the distance resolution of the radar, that is ; is the real-time position (distance) of the target; is the distance gate index where the target is located, determined by finding the peak value of the variance vector; is the distance resolution of the radar; When the moving speed of the target needs to be tracked, the difference operation can be performed on the target positions calculated at continuous multiple time points to estimate the target moving speed, that is ; is the target distance varying with time; is the time interval when the speed is calculated; is the estimated target moving speed; The location is compared with a preset monitoring area division, which is a spatial coordinate partitioning pre-set in the system based on radar beam characteristics and typical home environment layout. Based on the comparison results, the module generates radar switching instructions to adjust the time-division multiplexing sequence. For example, when a target enters the area of ​​responsibility of a secondary radar, the instruction will increase the proportion of working time of the secondary radar.

[0020] The power compensation module aims to dynamically adjust the transmit power of the operating radar based on real-time signal quality, thereby achieving a balance between ensuring data quality and reducing system power consumption. In this embodiment, the module acquires and analyzes the signal quality parameters of the echo signal in real time, such as the received signal strength and signal-to-noise ratio. The preset quality threshold is a boundary value used to distinguish between good and bad signal quality, calibrated based on a large amount of experimental data. The module compares the real-time signal quality parameters with the preset quality threshold and generates a power adjustment command based on the comparison result to adjust the transmit power of the current operating radar module. For example, when the signal quality is lower than the threshold, the command will increase the transmit power.

[0021] The technical effect is that, through the organic combination of the above modules, this system achieves comprehensive and uninterrupted vital sign monitoring of dynamic targets; the time-sharing acquisition and dynamic scheduling mechanism fundamentally solves the interference and blind zone problems of multi-radar deployment, while phase processing and vital sign separation ensure the accuracy of signal extraction; combined with respiratory event judgment and power compensation, the system achieves high-precision and high-reliability monitoring while possessing the ability of intelligent early warning and low-power operation, thus overcoming the shortcomings of existing technologies such as limited monitoring range, dynamic tracking failure, and insufficient environmental adaptability.

[0022] Example 2: The specific steps of background subtraction performed by the phase processing module are as follows: The historical average signal of the original data matrix is ​​calculated using the moving average filtering method to obtain the background signal; the background signal is then subtracted from the original data matrix to enhance the vital signs signal. The specific method by which the phase processing module performs background subtraction is further defined; its underlying logic lies in using a moving average filtering method to calculate the historical average signal of the original data matrix, thereby obtaining the background signal; specifically, in order to estimate the background signal at the 1st... pulses, distance is Background signal at the location The system will calculate the previous time window. The average signal value within; ; For background signals; This is the original data; is the window length; is the time; is the distance; The historical average signal is the background signal calculated by this method The selection of the window length W is technically based on the need to completely cover 2 to 3 breathing cycles, which can ensure that while effectively smoothing random noise and filtering out static reflections, the breathing itself as the target signal is not attenuated; after obtaining the background signal , the system subtracts it from the current raw data matrix , to obtain the difference signal: ; is the difference signal matrix; is the current raw data matrix; is the calculated background signal; , thereby significantly enhancing the micro-motion signal component caused by human life activities; This specific embodiment provides a dynamic, adaptive and accurate estimation model for the background signal by introducing a moving average filter; compared with simple static background subtraction, this method can better adapt to small and slow changes in the environment, thereby achieving more than 90% effective elimination of static clutter and improving the signal-to-noise ratio of the chest micro-motion signal by 10 to 30 times, laying a solid data foundation for subsequent high-precision phase extraction.

[0023] Embodiment 3: Further comprising a data fusion processing step, which is: Respectively calculate the signal quality scores of the echo signals collected by the master and slave ultra-wideband radar modules; based on the signal quality scores, weight the raw data matrices of the master and slave ultra-wideband radar modules to generate a fused data matrix; In another embodiment of the present application, the data fusion processing step is further included, which aims to integrate the collection information of the dual radar to cope with the situation of instantaneous decline of single radar signal quality, and generate a more stable and reliable data stream than a single source; The processing step is executed after the target distance gate is determined, and the specific steps are as follows: Respectively extract the complex signal vectors of the master and slave radar after background subtraction at the target distance gate , denoted as and ; Calculate the signal quality score of two complex signal vectors respectively With ; the score is a quantitative index, which must be calculated based on the information available at the current processing stage; one specific score calculation method is: first, the complex signal vector (where is a or b) is Fourier transformed in a preset breathing frequency band (for example, 0.1-0.6 Hz), and the ratio of the signal energy in the frequency band to the out-of-band noise energy is taken as the estimation of signal-to-noise ratio ; then the estimation is normalized, and the signal quality score is obtained ; Based on the signal quality score, the two complex signal vectors are weighted and fused to generate a fused complex signal vector , whose calculation formula is: ; is the fused complex signal vector is the complex signal vector of the main radar after background subtraction in the target range gate is the complex signal vector of the slave radar after background subtraction in the target range gate is the quality score of the main radar signal is the quality score of the slave radar signal The subsequent phase processing module will be based on this fused complex signal vector , rather than the signal of a single radar, to solve the phase signal ; for example: ; is the phase signal solved based on the fused data is the fused complex signal vector is the imaginary part of the complex number is the real part of the complex number After introducing data fusion processing, the system can dynamically and intelligently prioritize radar data sources with better current signal quality. When the signal quality of any radar deteriorates due to target attitude, angle, or temporary obstruction, the fusion algorithm will automatically reduce its weight, thereby effectively suppressing the interference of noise and abnormal data on the overall result. This ensures that the output phase signal has higher time continuity and fidelity, and ultimately strictly controls the heart rate error caused by multi-device interference to within ±2 beats / minute.

[0024] Example 4: The specific steps of the respiratory event determination module are as follows: Based on the preset radar center frequency, the phase change of the breathing waveform is converted into a physical displacement signal of the chest cavity; the peak-to-peak value of the physical displacement signal is calculated within the sliding time window to obtain the real-time displacement amplitude; when the real-time displacement amplitude is continuously lower than the preset weak breathing threshold and the duration exceeds the preset pause time threshold, a breathing pause warning signal is generated; otherwise, it is determined to be a normal breathing state. The specific steps of the respiratory event determination module have been further refined; this module first bases its operation on the radar center frequency used by the system. (In this embodiment, the frequency is 6.5 GHz), the breathing waveform The phase change is converted into a physical displacement signal of the thoracic cavity. ; ; For thoracic cavity displacement; Wavelength; This is the respiratory phase; For time; The wavelength of the radar signal; wavelength by the speed of light With center frequency Calculation yields ( (Approximately 4.6 cm); after obtaining the physical displacement signal, the system calculates the peak-to-peak value of the signal within a sliding time window (e.g., 10 seconds) to obtain the real-time displacement amplitude. When this real-time displacement amplitude is continuously lower than a preset weak breathing threshold At this time, the system starts timing; the preset weak breathing threshold is based on the clinical medical definition of respiratory attenuation, for example, set to 0.2 cm according to clinical standards, to distinguish between normal breathing fluctuations and weak or stopped breathing; if the duration of this state exceeds a preset pause time threshold... (the setting basis is to refer to the published medical standards, such as the American Sleep Medicine Association AASM standard, and is set to 10 seconds), the system finally determines that the apnea event occurs, and generates an apnea warning signal; otherwise, the system will determine that it is a normal breathing state; This specific embodiment converts abstract phase signals into physical displacements that can be directly understood, so that the determination criteria (amplitude and time) of respiratory events can be directly aligned with clinical standards; This physical quantity-based, quantitative determination logic greatly improves the accuracy and reliability of the warning, effectively avoiding false positives caused by signal fluctuations, and makes the warning result have clear clinical significance.

[0025] Embodiment 5: The specific steps of the radar dynamic scheduling module are: The monitoring space is divided into a main radar responsible area, a slave radar responsible area, and a transition area; Based on the target position and change, the time when the target will enter the transition area is predicted; Based on the predicted time, a radar switching instruction is generated; The radar switching instruction is used to control the timing of the relay, and the logic of the timing is: When the monitoring target is located in the main radar responsible area, open the main radar relay channel and close the slave radar relay channel; When the monitoring target is located in the slave radar responsible area, close the main radar relay channel and open the slave radar relay channel; When the monitoring target is located in the transition area, the main radar relay channel and the slave radar relay channel are alternately controlled in a pulse width modulation manner; The specific implementation of the radar dynamic scheduling module and how the radar switching instruction generated by the module is executed are described in detail; To achieve this function, the radar dynamic scheduling module divides the entire monitoring space into three logical areas: the main radar responsible area where the main radar has the best monitoring effect due to its installation location and orientation; the slave radar responsible area where the slave radar has a better effect; and the transition area at the junction of the two responsible areas; This pre-set monitoring area division is based on physical space coordinates (for example, in an 8m x 6m room, the area with an x coordinate less than 4.5m is the main radar responsible area), and the division scheme is established when the system is initialized; When the monitoring target moves in each area, the system will predict the time when the target will enter the transition area based on the target position and its change trend calculated from the echo signal; The calculation of the prediction time will consider the target's current moving speed and the distance from the transition area boundary ; Based on this predicted time, the module will generate a radar switching instruction in advance; To execute the instruction, the system realizes the physical switching of the radars by controlling the relay timing, and the control logic thereof is accurately defined according to the area where the target is located: 1. When the monitored target is located in the area responsible by the main radar, the instruction will turn on the relay channel of the main radar and turn off the relay channel of the slave radar, ensuring the independent work of the main radar; 2. When the monitored target is located in the area responsible by the slave radar, the instruction will turn off the relay channel of the main radar and turn on the relay channel of the slave radar, switching to the independent work of the slave radar; 3. When the monitored target is located in the transition area, the instruction will alternately control the relay channels of the main and slave radars in the form of pulse width modulation (PWM); the advantage of this form is that, in one working cycle, the time lengths of the work of the two radars can be finely distributed and rapidly rotated by adjusting the duty cycles of the “on” signals given to the main and slave radars, forming a kind of “soft switching” effect; By combining the area division, trajectory prediction and relay physical control logic, the system realizes a forward-looking and seamless radar switching mechanism; it not only solves the problem of static blind area, but more importantly, through the special processing (PWM alternate scanning) of the transition area, effectively avoids the possible data interruption (the interruption time is less than 0.4 seconds) of the target at the moment of area switching; this soft switching strategy ensures the data continuity in the dynamic tracking process, greatly improving the monitoring reliability of the moving target compared with the simple hard switching in the prior art.

[0026] The specific steps of the power compensation module are as follows: The received signal strength and signal-to-noise ratio are obtained as signal quality parameters; The signal quality parameters are normalized and weighted summed to generate a comprehensive quality index; High and low quality thresholds are set; When the comprehensive quality index is lower than the low quality threshold, a power adjustment instruction to increase the transmission power is generated; When the comprehensive quality index is higher than the high quality threshold, a power adjustment instruction to reduce the transmission power is generated; When the comprehensive quality index is between the low and high quality thresholds, a power adjustment instruction to maintain the current transmission power is generated; To make the purpose, technical scheme and advantages of the present application clearer, the specific steps of the power compensation module of the present application will be described below; The module first obtains multiple signal quality parameters reflecting the current detection performance; in this embodiment, the core parameters include the received signal strength and signal-to-noise ratio calculated by the radar front-end signal processing unit; to further improve the accuracy of the evaluation, the system can also calculate the packet loss rate in the signal transmission process and waveform confidence derived from rhythmic analysis of isolated vital sign waveforms (e.g. respiratory waveform) ; To comprehensively evaluate these parameters of different dimensions, the system first normalizes them and then performs a weighted summation to generate a single comprehensive quality index , whose calculation formula is: ; is the signal-to-noise ratio; is the signal strength; is the packet loss rate; is the waveform confidence; is a normalization function, such as the min-max normalization function; are the preset weight coefficients corresponding to each parameter, which can be optimized through regression analysis on a large amount of measured data, for example to reflect the different importance of each parameter in evaluation; Subsequently, the system sets two key thresholds: a high-quality threshold and a low-quality threshold ; The setting of these two thresholds is based on statistical analysis of the values under normal working conditions, for example, taking the upper and lower boundaries of the 95% confidence interval to determine; The logic of power adjustment is as follows: When the comprehensive quality index is lower than the low-quality threshold , a power adjustment instruction to increase the transmission power is generated; When the comprehensive quality index is higher than the high-quality threshold , a power adjustment instruction to reduce the transmission power is generated; When the comprehensive quality index is between the two, a power adjustment instruction to maintain the current transmission power is generated; Normalization is to map the original parameter value to a unified [0, 1] interval for fair comparison and weighting; This normalization can be realized by the min-max normalization method; For a parameter , its normalized value is calculated as: ; is the normalized value of parameter ; is the original parameter value; and are the minimum and maximum values of the parameter observed in a preset time window (e.g., the last 60 seconds) in the past, respectively; This dynamic normalization method based on a sliding window can better adapt to slow changes in the environment and target state; This specific embodiment realizes the refinement and automation of power adjustment by constructing a quantitative, multi-parameter fusion signal quality evaluation model; it enables the system to adapt in real time to signal fluctuations caused by target distance, posture changes, or environmental obstructions (such as a cotton blanket or curtains), maximally reduces system power consumption (more than 50% compared to the full-power operation mode) while ensuring data acquisition reliability (85% reduction in packet loss rate), and is particularly suitable for home monitoring scenarios that require long-term uninterrupted operation.

[0027] The vital sign separation module is also used to generate a heart rate waveform, and the steps are: An adaptive filtering method is used to subtract the respiratory harmonic component from the phase signal with the respiratory waveform as a reference to obtain a heart rate phase signal; the heart rate phase signal is subjected to second band-pass filtering to generate a heart rate waveform; The function of the vital sign separation module is expanded so that it can also be used to generate a heart rate waveform; the extraction of the heart rate waveform needs to solve a core technical problem: the chest vibration caused by the heartbeat (0.1-0.5 mm) is much smaller than the fluctuation caused by respiration (0.1-1 cm), resulting in the heartbeat signal being easily overwhelmed by the respiration signal and its harmonics; to solve this problem, the present embodiment uses an adaptive filtering method when separating the heart rate signal; this method uses the respiratory waveform extracted in the previous step as a reference signal to adaptively filter out all components related to the respiratory waveform, including its harmonic components, from the original, mixed phase signal , thereby obtaining a heart rate phase signal from which the respiratory artifacts have been eliminated; specifically, the adaptive filtering can be implemented using the least mean square algorithm; the LMS algorithm adjusts the filter weights through iteration so that the mean square error between the filtered output signal and the reference signal (respiratory waveform) is minimized; the core formula for updating the weights is: ; is the filter weight for the next iteration; is the current filter weight; is the input signal (mixed phase signal ); is an error signal (i.e. the output of the subtraction filter); is a step factor that determines the convergence speed and stability; By this method, the component related to the respiratory waveform can be effectively subtracted from the mixed signal to obtain a relatively pure heart rate phase signal; based on this signal, a second band-pass filtering process is performed, and the passband range of the filter is set to cover the interval of the typical human heart rate, for example, 1.0 Hz to 2.5 Hz (corresponding to 60 to 150 times per minute), to adapt to the heart rate changes of the elderly in a resting or stress state; the result of the filtering is the heart rate waveform; By introducing the key step of adaptive filtering with the respiratory waveform as the reference, the present embodiment can actively eliminate the main interference source (respiratory harmonic) before filtering, rather than passively relying on band-pass filtering; this greatly improves the signal-to-noise ratio and accuracy of heart rate signal separation, enabling the system to simultaneously achieve reliable monitoring of both respiratory and heart rate, two key vital signs, under the same non-contact detection framework, significantly enhancing the overall functionality and clinical application value of the system.

[0028] The present application realizes a significant technical progress in non-contact, high-precision, and all-weather continuous monitoring of vital signs in a complex home environment through a series of closely coupled physical processes and signal processing mechanisms; its beneficial effects are rooted in its unique system architecture and algorithm design, which are embodied in the following aspects: Compared with the inherent monitoring blind area and dynamic response delay defects of the single radar or mechanical scanning structure in the prior art, the present application has made a fundamental breakthrough; Firstly, through the collaborative operation of the time-sharing echo acquisition module and the radar dynamic scheduling module, an intelligent dual-radar collaborative perception system is constructed; the time-sharing multiplexing mechanism ensures that there is no mutual interference of electromagnetic signals when the dual-radar modules work in the same frequency band, thereby eliminating the heart rate measurement error caused by channel conflict (reduced from ±8 times / minute to ±2 times / minute) under multi-device deployment; further, the radar dynamic scheduling module performs radar switching instructions in a forward-looking manner before the target enters the preset regional boundary based on real-time solving and trajectory prediction of the monitoring target position; this not only realizes seamless coverage of areas such as bathrooms and kitchens outside the field of view of the traditional single radar (the monitoring blind area is less than 0.3 square meters), but also ensures continuous data capture during target movement (the monitoring interruption window is less than 0.4 seconds), completely solving the technical bottleneck of long re-capture time (more than 2.8 seconds) of the prior art when the target moves; Secondly, the power compensation module establishes a closed-loop adaptive energy management system; the core of this module lies in a comprehensive quality index function. This function performs dimensionless processing on multiple physical quantities such as signal-to-noise ratio and signal strength, and then performs a weighted summation. Its weighting coefficients... It was established based on statistical optimization of over 200 sets of measured data; this indicator This provides the system with a real-time, quantitative assessment of the current radar detection effectiveness. Based on this assessment, the system can dynamically adjust the transmission power, automatically increasing the power to penetrate obstacles when the target enters a weak signal area (such as being blocked by a thick blanket), and reducing power consumption when the signal quality is good. This mechanism enables the system to reduce overall power consumption by more than 50% compared to the full-power mode of a single radar while ensuring the stability of the data link, greatly improving the practicality and economy of long-term uninterrupted monitoring.

[0029] This invention achieves precise conversion from abstract data to specific physiological indicators at the signal processing level, and its core lies in the deep understanding and application of physical models. The background subtraction step in the phase processing module employs a moving average filtering method, which essentially establishes a dynamic, time-varying background reference frame; the calculated background signal... It is not a static value, but rather the mean of environmental reflection that evolves smoothly over time; from the original signal Subtracting this dynamic background allows us to capture the subtle dynamic components generated by vital activities such as chest rise and fall. Effectively isolated from the macroscopic static universe, its signal-to-noise ratio can be improved by 10 to 30 times; The physical significance of the respiratory event determination module is particularly prominent; its core conversion formula has a solid physical foundation; this formula originates from the basic principle of wave optics: when an electromagnetic wave is reflected by a moving target, the change in its round-trip distance. With phase change The relationship between them is By combining these two equations, the displacement conversion formula can be derived. This derivation process ensures that the system output is no longer a dimensionless phase signal, but a real chest cavity displacement in centimeters with clear physical meaning. The threshold for apnea judgment based on this physical displacement (such as an amplitude of less than 0.2 cm and a duration of more than 10 seconds) is directly aligned with the clinical diagnostic criteria, thus giving the warning signal unprecedented medical reference value and reliability. In addition, the adaptive filtering method used in the vital sign separation module for heart rate extraction is a deep insight into the relationship between signal and noise. In this scenario, the respiratory signal and its harmonics, which have much more energy than the heartbeat signal, constitute the main "noise" source when extracting the heart rate signal. The adaptive filtering uses the extracted respiratory waveform as a template to accurately "remove" all energy components related to it from the original mixed signal, thereby greatly purifying the signal background without damaging the weak heart rate signal. This makes it possible to reliably extract the heart rate within the same technical framework.

[0030] Compared with non-UWB radar technology (such as millimeter wave radar CN115295161A): The ultra-wideband radar used in the present application exhibits excellent penetration ability due to its lower center frequency (6.5 GHz) and nanosecond-level pulse characteristics. Actual measurements show that the present application can penetrate a brick wall with a thickness of more than 20 centimeters with a decay of no more than 3 decibels, while the millimeter wave radar has already decayed more than 15 decibels when penetrating a 10 centimeter thick brick wall, resulting in signal loss. At the same time, the present application has a sub-millimeter level of micro-motion resolution, which can capture the tiny vibrations caused by heartbeats, while the comparative millimeter wave radar can only detect displacements greater than 2 millimeters, and cannot effectively monitor respiration and heartbeat. Compared with traditional sensor solutions (such as infrared, camera, piezoelectric sensor): The present application completely eliminates the physical limitations of traditional solutions; it is not affected by light and temperature changes, has no privacy leakage risk, and does not require direct contact with the human body. In clinical tests, the traditional solution has a respiration detection error of more than 12% under the condition of cotton cover, while the present application still has an accuracy of more than 98% under the same conditions, showing unparalleled environmental adaptability in real elderly care monitoring scenarios. Compared with single UWB radar solution: The present application fundamentally solves the inherent defects of single radar solution through a dual radar cooperative architecture; the single radar solution has a high monitoring dead angle detection rate of up to 34%, while the dynamic scheduling mechanism of the present application achieves almost complete spatial coverage; in terms of dynamic tracking capability, the present application achieves a non-sensing switching time of less than 0.4 seconds, which is much better than the recapture time of more than 2.8 seconds of the single radar solution; in terms of anti-interference, the time division multiplexing mechanism of the present application effectively avoids channel conflict, reducing the heart rate error under multi-device interference from ±8 times / minute to within ±2 times / minute.

[0031] In summary, the present application is not a simple combination or improvement of existing technology, but provides a non-contact life detection solution that comprehensively surpasses existing technology in terms of detection range, dynamic performance, measurement accuracy and environmental adaptability through innovation of system architecture and deepening of core algorithm.

[0032] The above merely describes the preferred embodiments of the present application, and is not intended to limit the present application in other forms. Any skilled person in the art can modify or change the disclosed technical content into equivalent embodiments with equivalent changes, and apply them to other fields. However, any simple modification, equivalent change and modification made to the above embodiments according to the technical essence of the present application, without departing from the technical solution content of the present application, still falls within the protection scope of the present application.

Claims

1. A non-contact vital sign recognition and detection system, characterized in that, include: The time-division echo acquisition module uses time-division multiplexing to alternately control the operation of the main ultra-wideband radar module and the slave ultra-wideband radar module, to acquire the echo signals reflected by the monitored target and construct the original data matrix; The phase processing module performs background subtraction and phase extraction on the original data matrix to calculate the phase signal; The vital signs separation module performs a first bandpass filter on the phase signal to generate a respiratory waveform; The respiratory event determination module converts the respiratory waveform into a physical displacement waveform, and determines and generates a breathing apnea warning signal based on whether the amplitude and duration of the physical displacement waveform meet the preset breathing apnea conditions. The radar dynamic scheduling module generates radar switching commands for adjusting the time-division multiplexing sequence based on the target position calculated from the original data matrix and the preset monitoring area division. The power compensation module acquires the signal quality parameters of the echo signal and generates a power adjustment command for adjusting the transmit power of the working radar module based on the comparison results between the signal quality parameters and the preset quality threshold.

2. The non-contact vital sign recognition and detection system according to claim 1, characterized in that, The specific steps of background subtraction performed by the phase processing module are as follows: The historical average signal of the original data matrix is ​​calculated using the moving average filtering method to obtain the background signal; the background signal is then subtracted from the original data matrix to enhance the vital signs signal.

3. The non-contact vital sign recognition and detection system according to claim 1, characterized in that, It also includes data fusion processing, the steps of which are: Calculate the signal quality scores of the echo signals collected by the main UWB radar module and the slave UWB radar module respectively; and perform weighted fusion of the original data matrices of the main UWB radar module and the slave UWB radar module based on the signal quality scores to generate a fused data matrix. The phase processing module calculates the phase signal based on the fused data matrix.

4. The non-contact vital sign recognition and detection system according to claim 1, characterized in that, The specific steps of the respiratory event determination module are as follows: Based on the preset radar center frequency, the phase change of the respiratory waveform is converted into a physical displacement signal of the chest cavity; the peak-to-peak value of the physical displacement signal is calculated within the sliding time window to obtain the real-time displacement amplitude. When the real-time displacement amplitude is continuously lower than the preset weak breathing threshold and the duration exceeds the preset pause time threshold, a breathing apnea warning signal is generated; otherwise, it is determined to be a normal breathing state.

5. The non-contact vital sign recognition and detection system according to claim 1, characterized in that, The specific steps of the radar dynamic scheduling module are as follows: The monitoring space is divided into the main radar area of ​​responsibility, the secondary radar area of ​​responsibility, and the transition area. Based on the target's location and changes, predict the time when the target will enter the transition zone; Based on the predicted time, a radar switching command is generated.

6. The non-contact vital sign recognition and detection system according to claim 1, characterized in that, The specific steps of the power compensation module are as follows: The received signal strength and signal-to-noise ratio are obtained as signal quality parameters; The signal quality parameters are normalized and weighted summed to generate a comprehensive quality index. Set high-quality and low-quality thresholds; When the overall quality index is below the low quality threshold, a power adjustment command to increase the transmission power is generated. When the overall quality index is higher than the high quality threshold, a power adjustment command to reduce the transmission power is generated. When the overall quality index is between the low quality threshold and the high quality threshold, a power adjustment command is generated to maintain the current transmit power.

7. The non-contact vital sign recognition and detection system according to claim 1, characterized in that, The vital signs separation module is also used to generate heart rate waveforms, the steps of which are as follows: An adaptive filtering method is used to obtain the heart rate phase signal by subtracting the respiratory harmonic components from the phase signal with the respiratory waveform as a reference. The heart rate phase signal is then subjected to a second bandpass filter to generate the heart rate waveform.

8. A non-contact vital sign recognition and detection system according to claim 5, characterized in that, The radar switching command is used to control the relay timing. The timing logic is as follows: When the monitored target is located in the area of ​​responsibility of the main radar, the main radar relay channel is turned on and the slave radar relay channel is turned off. When the monitored target is located in the area of ​​responsibility of the slave radar, the main radar relay channel is closed and the slave radar relay channel is opened. When the monitored target is in the transition zone, the main radar relay channel and the slave radar relay channel are alternately controlled by pulse width modulation.

Images