An intelligent alarm system and method for abnormal behavior based on ultra-wideband spatial positioning and multimodal perception.

Abstract

This invention discloses an intelligent alarm system and method for abnormal behavior based on ultra-wideband spatial positioning and multimodal sensing, relating to an alarm system. The system acquires the three-dimensional spatial coordinates of a target object through an ultra-wideband spatial positioning subsystem and locks the absolute spatial coordinates when a suspected fall event is detected. Height coordinate data is extracted to calculate the absolute height difference between the target object's chest cavity and the ground. The physical phase difference is calculated based on the absolute height difference and the initial transmission frequency of the ultrasonic sensing subsystem to determine if the target object is in a standing wave cancellation blind zone. If in a blind zone, the absolute height difference is used as a feedforward control parameter to calculate a frequency offset command to fine-tune the ultrasonic transmission frequency, disrupting the physical interference conditions of standing wave cancellation. Micro-motion vital signs are extracted at the fine-tuned transmission frequency, triggering an alarm. This invention overcomes the standing wave blind zone in ground-level conditions, avoids the loss of vital sign signals caused by blind frequency scanning, and achieves high-precision intelligent alarms in extreme scenarios.

Description

Technical Field

[0001] This invention relates to an alarm system, and more particularly to an intelligent alarm system and method for abnormal behavior based on ultra-wideband spatial positioning and multimodal perception. Background Technology

[0002] With the aging population, the demand for alarm systems for abnormal behavior in private settings such as nursing homes and hospital wards is increasing. Due to strict privacy restrictions, traditional visual monitoring equipment cannot be deployed in high-risk areas such as bathrooms or bedrooms. Therefore, the industry is gradually adopting non-contact radio frequency and acoustic multimodal sensing technologies to replace visual monitoring. Among these technologies, the combination of ultra-wideband spatial positioning technology and ultrasonic vital sign monitoring technology has become a research hotspot in the field of intelligent alarms.

[0003] In existing multimodal alarm systems, a data-level fusion control architecture is commonly used. The ultra-wideband positioning subsystem is responsible for tracking the spatial position of the target object and determining whether a fall or other violent movement has occurred. The ultrasonic sensing subsystem independently emits sound waves into the target area and monitors weak physiological signs such as respiration and heart rate by extracting the Doppler frequency shift of the echo. However, in extremely high-risk scenarios such as when the target object falls and lies flat on the ground, existing ultrasonic vital sign monitoring often fails, easily leading to the inability to extract respiratory signals, resulting in false alarms or missed alarms about life-threatening situations.

[0004] Regarding the issue of weak and easily lost vital signs signals when the target object is close to the ground, those skilled in the art generally believe that the cause lies in the multipath effect and standing wave interference caused by ground reflection. To overcome this physical phenomenon, existing conventional techniques generally employ broadband frequency sweeping technology or blindly increase transmission power, attempting to avoid the blind zone of standing wave cancellation by drastically changing the transmission frequency. However, respiratory fluctuations are extremely low-frequency and minimally sized micro-motion signals. When ultrasonic sensing devices perform large-scale broadband frequency sweeps to avoid standing waves, their residence time at a single frequency becomes extremely short, thus disrupting the temporal phase continuity of the micro-Doppler signal. This leads to a significant decrease in the effective sampling rate of chest cavity micro-movements, and the already weak respiratory signal is submerged in the broadband noise introduced by the frequency sweep. Furthermore, existing ultrasonic sensing devices lack spatial geometric sensing capabilities at the physical level, making it impossible to determine the true height of the target object from the reflecting ground. Therefore, they can only rely on probabilistic full-band frequency sweeping trial and error, failing to achieve precise anti-interference control.

[0005] In summary, existing multimodal sensing systems suffer from irreconcilable contradictions in their underlying physical architecture. The core technical challenge that urgently needs to be addressed in this field is how to overcome the standing wave phase cancellation blind zone when the target object is close to the ground, while simultaneously avoiding the loss of phase continuity of weak vital signs during blind broadband frequency sweeping, thereby achieving high-precision intelligent alarm for abnormal behavior in extreme scenarios. Summary of the Invention

[0006] This invention overcomes the shortcomings of the prior art and provides an intelligent alarm system and method for abnormal behavior based on ultra-wideband spatial positioning and multimodal perception.

[0007] To achieve the above objectives, the technical solution adopted by this invention is: an intelligent alarm method for abnormal behavior based on ultra-wideband spatial positioning and multimodal perception, comprising the following steps: S1. The three-dimensional spatial coordinates of the target object are obtained in real time through the ultra-wideband spatial positioning subsystem, and the current absolute spatial coordinates are locked when it is determined that the target object has experienced a suspected fall event. S2. Extract the height coordinate data from the absolute spatial coordinates and calculate the absolute height difference between the chest cavity of the target object and the reflecting ground. S3. Based on the absolute height difference and the initial transmission frequency of the ultrasonic sensing subsystem, calculate the physical phase difference corresponding to the optical path difference caused by ground reflection, and determine whether the target object is in the standing wave phase cancellation blind zone based on the physical phase difference. S4. If it is determined that the system is in the standing wave cancellation blind zone, the absolute height difference is used as the feedforward control parameter to calculate and output the frequency offset command in order to finely adjust the transmission frequency of the ultrasonic sensing subsystem and destroy the physical interference conditions of standing wave cancellation. S5. At the finely adjusted transmission frequency, the ultrasonic sensing subsystem extracts the micro-motion characteristics of the target object and triggers a corresponding alarm action based on the micro-motion characteristics.

[0008] In a preferred embodiment of the present invention, the step of calculating the physical phase difference corresponding to the optical path difference caused by ground reflection based on the absolute height difference and the initial transmission frequency of the ultrasonic sensing subsystem, and determining whether the target object is in the standing wave cancellation blind zone based on the physical phase difference, specifically includes: Obtain the product of the absolute height difference and the initial transmission frequency; The product is divided by the speed of sound in the current environment and multiplied by a preset phase constant to calculate the physical phase difference. When the physical phase difference is determined to be close to an odd multiple of half the wavelength, the target object is determined to be currently in the standing wave phase cancellation blind zone.

[0009] In a preferred embodiment of the present invention, the step of calculating and outputting a frequency offset command to fine-tune the transmission frequency of the ultrasonic sensing subsystem specifically includes: The target frequency offset is calculated based on the absolute height difference, so that the updated physical phase difference corresponding to the fine-tuned transmission frequency approaches an even multiple of half the wavelength to form in-phase superposition, or approaches a non-integer multiple of half the wavelength to form orthogonality. The target frequency offset is limited to a preset percentage threshold of the initial transmission frequency to output the frequency offset command while maintaining the resonant efficiency of the ultrasonic transducer. The preset percentage threshold is ≤3%.

[0010] In a preferred embodiment of the present invention, the step of locking the current absolute spatial coordinates when determining that the target object has experienced a suspected fall specifically includes: Real-time monitoring of the rate of change of the three-dimensional spatial coordinates within a preset time window; When the decrease in the height coordinate data exceeds a preset height threshold and the displacement speed of the horizontal plane is lower than a preset speed threshold, a suspected fall signal is generated, and the current absolute spatial coordinates of the target object are locked in response to the suspected fall signal.

[0011] In a preferred embodiment of the present invention, before the step of dividing the product by the speed of sound in the current environment, the method further includes: Obtain real-time ambient temperature data within the monitored area; The sound velocity is calculated by adding the product of the base sound velocity, the temperature compensation coefficient, and the real-time ambient temperature data, in order to eliminate the error in the calculation of physical phase difference caused by ambient temperature drift.

[0012] An intelligent alarm system for abnormal behavior based on ultra-wideband spatial positioning and multimodal perception, comprising: An ultra-wideband spatial positioning subsystem is used to acquire the three-dimensional spatial coordinates of a target object in real time. The ultrasonic sensing subsystem is used to transmit ultrasonic waves to the target area and receive the echoes to extract micro-motion characteristics. The edge computing and alarm control host is communicatively connected to the ultra-wideband spatial positioning subsystem and the ultrasonic sensing subsystem, respectively. The edge computing and alarm control host is configured to execute the intelligent alarm method for abnormal behavior based on ultra-wideband spatial positioning and multimodal perception as described in any of the preceding items.

[0013] In a preferred embodiment of the present invention, the ultrasonic sensing subsystem includes: A high-frequency piezoelectric ceramic transducer array is installed at the top of the monitoring area to perform ultrasonic wave transmission and reception. The Doppler echo receiving circuit is electrically connected to the high-frequency piezoelectric ceramic transducer array and is used to extract the micro-Doppler frequency shift signal; The frequency synthesis drive module, electrically connected to the high-frequency piezoelectric ceramic transducer array, includes a voltage-controlled oscillator or a direct digital frequency synthesizer, and is used to receive frequency offset commands issued by the edge computing and alarm control host and dynamically adjust the transmission frequency.

[0014] In a preferred embodiment of the present invention, the ultra-wideband spatial positioning subsystem includes: Multiple ultra-broadband base stations are distributed and deployed within the monitoring area; An ultra-wideband tag is configured on the target object; The ultra-wideband base station and the ultra-wideband tag communicate using a time-of-flight ranging algorithm or a time-difference-of-arrival ranging algorithm to calculate and output the three-dimensional spatial coordinates.

[0015] In a preferred embodiment of the present invention, an alarm linkage execution subsystem electrically connected to the edge computing and alarm control host is further included, the alarm linkage execution subsystem comprising: Audible and visual alarms are used to issue audible and visual warnings on-site when the highest level alarm command is received; Access control controllers are used to automatically unlock access to the monitored area when an alarm is triggered. A communication gateway is used to send a distress signal containing the absolute spatial coordinates and the micro-motion characteristics to a remote monitoring terminal.

[0016] In a preferred embodiment of the present invention, the edge computing and alarm control host includes: At least one processor; and a memory communicatively connected to the at least one processor; The memory stores a computer program that can be executed by the at least one processor, which, when executed by the at least one processor, causes the edge computing and alarm control host to implement the method described in any of the preceding methods.

[0017] This invention addresses the shortcomings of the prior art and has the following beneficial effects: This invention uses the absolute height difference acquired by the ultra-wideband spatial positioning subsystem as a feedforward control parameter to calculate and output a frequency offset command to fine-tune the transmission frequency of the ultrasonic sensing subsystem. It utilizes macroscopic spatial geometric boundary conditions to accurately calculate the optical path difference and physical phase difference caused by ground reflection, thereby actively disrupting the interference conditions of standing wave destructiveness at the physical level. This allows the ultrasonic waves to maintain a high sampling rate of single-frequency continuous wave even when the target object is close to the ground, enabling the extraction of micro-motion characteristics. Compared to existing technologies that blindly use wideband frequency sweeping for anti-interference, resulting in the disruption of the phase continuity of the micro-Doppler signal and a sharp drop in the sampling rate, this solution achieves precise obstacle avoidance without frequency sweeping. While avoiding phase blind zones, it achieves non-destructive extraction and high-precision alarm of weak breathing signals in extreme fall scenarios.

[0018] This invention constructs a low-level control architecture that connects the edge computing and alarm control host to the ultra-wideband and ultrasonic subsystems, respectively. This breaks down the data silos caused by the independent operation of each sensor, directly reducing the macroscopic three-dimensional spatial coordinates output by the ultra-wideband system to the physical input parameters of the ultrasonic low-level RF transmitter driver, thus achieving cross-modal signal-level feedforward physical control. Compared to existing technologies that only perform simple logical judgments on multimodal results at the central processing unit level, this solution obtains clean vital signs from the physical emission source, eliminating excessive reliance on complex and easily failing backend filtering algorithms, and significantly improving the reliability and response speed of the system's low-level data in complex environments.

[0019] This invention strictly limits the target frequency deviation to within 3% of the initial transmission frequency using a voltage-controlled oscillator or direct digital frequency synthesizer. Utilizing the extreme sensitivity of spatial geometric interference to wavelength, only a tiny wavelength change is needed to shift the phase difference from an odd multiple of destructive phase to an even multiple of in-phase superposition or a quadrature state. This successfully breaks the standing wave dead zone while maintaining the high quality factor and resonant efficiency of the high-frequency piezoelectric ceramic transducer. Compared to the prevailing technical bias that deviations from the center frequency of a high-frequency piezoelectric transducer lead to a precipitous drop in electroacoustic conversion efficiency, this solution not only overcomes the inherent bias of hardware frequency modulation but also obtains additional physical interference gain through in-phase superposition, resulting in an order-of-magnitude improvement in the overall signal-to-noise ratio.

[0020] This invention introduces ambient temperature data to dynamically compensate for the base sound velocity before calculating the physical phase difference. The product of the temperature compensation coefficient and the real-time ambient temperature is incorporated into the sound velocity calculation model, eliminating the interference of ambient temperature drift on the sound wave propagation speed and ensuring the accuracy of the optical path difference and physical phase difference calculations. Compared to existing technologies that use a fixed sound velocity constant, which easily leads to phase prediction deviations under different seasons or diurnal temperature variations, this solution enhances the system's robustness in complex and variable environments, ensuring the accurate issuance of frequency offset commands.

[0021] This invention generates suspected fall signals by real-time monitoring of the rate of change of three-dimensional spatial coordinates within a preset time window, combined with the sudden drop in height coordinate data and the near-zero characteristic of horizontal displacement velocity. It can accurately distinguish normal everyday actions such as squatting or bending over from actual falls, and locks the absolute spatial coordinates at the moment of confirmed fall, providing a reliable static geometric benchmark for subsequent ultrasonic directional scanning and frequency offset calculation. Compared to existing technologies that rely solely on a single acceleration threshold, which are prone to misjudgment or coordinate tailing, this solution effectively avoids the interference of spatiotemporal coordinate drift on multimodal fusion, constructing a solid and accurate triggering precondition for the entire intelligent alarm system. Attached Figure Description

[0022] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments recorded in the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort. Figure 1 This is a flowchart of an intelligent alarm method for abnormal behavior based on ultra-wideband spatial positioning and multimodal perception according to the present invention. Figure 2 This is a system architecture block diagram of an intelligent alarm system for abnormal behavior based on ultra-wideband spatial positioning and multimodal perception, according to the present invention. Detailed Implementation

[0023] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0024] Many specific details are set forth in the following description in order to provide a full understanding of the invention. However, the invention may also be practiced in other ways different from those described herein. Therefore, the scope of protection of the invention is not limited to the specific embodiments disclosed below.

[0025] Application Overview: Existing technologies, when addressing the problem of weak and easily lost vital signs signals when a target is close to the ground, generally employ broadband frequency sweeping techniques at the acoustic or radio frequency levels. These techniques attempt to circumvent multipath effects and standing wave interference blind zones caused by ground reflections by significantly altering the transmission frequency. While broadband frequency sweeping can avoid standing wave interference to a certain extent, it faces an irreconcilable physical contradiction between combating multipath fading and maintaining a high-frequency sampling rate for micro-motion characteristics when extracting extremely low-frequency, minimally displaced signals such as respiration. Furthermore, the broadband frequency sweeping mechanism shortens the residence time of ultrasound waves at a single frequency, disrupting the temporal phase continuity of the micro-Doppler signal and submerging weak vital signs signals in the broadband noise introduced by the frequency sweep. This results in multimodal sensing systems being unable to accurately extract respiratory features in extreme fall scenarios.

[0026] To address the physical contradiction between combating multipath fading and maintaining a high-frequency sampling rate for micro-motion characteristics, this invention proposes an intelligent alarm method and system for abnormal behavior based on ultra-wideband spatial positioning and multimodal sensing. The ultra-wideband spatial positioning subsystem acquires the three-dimensional spatial coordinates of the target object, extracts height coordinate data to calculate the absolute height difference between the target object's chest cavity and the reflected ground, and inputs this absolute height difference as a feedforward control parameter to the ultrasonic sensing subsystem. The subsystem calculates and outputs a frequency offset command to fine-tune the transmission frequency of the ultrasonic sensing subsystem. The feedforward intervention of ultra-wideband macroscopic spatial geometric data on the ultrasonic microscopic physical transmission frequency can precisely disrupt the interference conditions of standing wave cancellation based on the physical phase difference without broadband frequency sweeping. This allows the ultrasonic sensing subsystem to maintain a high sampling rate of single-frequency continuous waves while avoiding phase blind zones, enabling accurate extraction of weak respiratory signals and abnormal behavior alarms in fall scenarios.

[0027] Example 1:

[0028] like Figure 1 As shown, an intelligent alarm method for abnormal behavior based on ultra-wideband spatial positioning and multimodal perception includes the following steps: S1. The three-dimensional spatial coordinates of the target object are obtained in real time through the ultra-wideband spatial positioning subsystem, and the current absolute spatial coordinates are locked when it is determined that the target object has experienced a suspected fall event. S2. Extract the height coordinate data from the absolute spatial coordinates and calculate the absolute height difference between the chest cavity of the target object and the reflecting ground. S3. Based on the absolute height difference and the initial transmission frequency of the ultrasonic sensing subsystem, calculate the physical phase difference corresponding to the optical path difference caused by ground reflection, and determine whether the target object is in the standing wave phase cancellation blind zone based on the physical phase difference. S4. If it is determined that the system is in the standing wave cancellation blind zone, the absolute height difference is used as the feedforward control parameter to calculate and output the frequency offset command in order to finely adjust the transmission frequency of the ultrasonic sensing subsystem and destroy the physical interference conditions of standing wave cancellation. S5. At the finely adjusted transmission frequency, the ultrasonic sensing subsystem extracts the micro-motion characteristics of the target object and triggers a corresponding alarm action based on the micro-motion characteristics.

[0029] When integrating ultra-wideband spatial positioning technology with ultrasonic sensing technology at the underlying physical level, we face the dual technical challenges of cross-modal data mapping accuracy and acoustic hardware physical characteristics constraints. When converting macroscopic three-dimensional spatial coordinates into microscopic acoustic interference parameters, even small coordinate errors or fluctuations in environmental parameters can cause the physical phase difference calculation to deviate from the actual physical field state. Furthermore, ultrasonic sensing subsystems typically use high-quality piezoelectric ceramic transducers, and emissions deviating from the inherent center frequency will lead to a decrease in electroacoustic conversion efficiency. Achieving frequency fine-tuning to break standing wave interference without compromising the inherent resonant efficiency of piezoelectric ceramic transducers is a key technical hurdle that must be overcome to realize cross-modal feedforward control.

[0030] Preferably, this embodiment is executed based on a hardware architecture that includes an ultra-wideband spatial positioning subsystem, an ultrasonic sensing subsystem, and an edge computing and alarm control host.

[0031] Specifically, the ultra-wideband spatial positioning subsystem communicates with the tags worn by the target object through base stations deployed in the monitoring area, and outputs three-dimensional spatial coordinates containing X-axis, Y-axis and Z-axis information in real time; Specifically, the edge computing and alarm control host monitors the changes in three-dimensional spatial coordinates in real time. When it detects a sudden change in the spatial trajectory representing a fall, it generates a suspected fall event signal and records the three-dimensional spatial coordinates at the moment the suspected fall event signal is generated as the absolute spatial coordinates.

[0032] Specifically, the edge computing and alarm control host extracts the Z-axis data in absolute spatial coordinates as height coordinate data. Combined with the pre-calibrated ground zero point coordinates and the pre-entered tag wearing position height compensation value, the vertical distance from the target object's chest cavity surface to the reflecting ground is calculated. For example, when the ultra-wideband tag is worn on the target object's wrist, the system compensates and converts the wrist Z-axis coordinates into the chest cavity center coordinates based on the standard human skeletal proportion model, and defines the compensated vertical distance as the absolute height difference.

[0033] Furthermore, the ultrasonic sensing subsystem radiates continuous sound waves into the monitoring area at a fixed initial transmission frequency, and the sound waves propagate in a multipath pattern between the target object's chest cavity surface and the reflecting ground. The edge computing and alarm control host calculates the optical path difference between the sound wave reflection path in the chest cavity and the ground reflection path based on the absolute height difference. It also derives the physical phase difference formed by the optical path difference in the spatial physical field based on the initial transmission frequency. Based on the numerical characteristics of the physical phase difference, it determines whether the Doppler echo generated by the chest cavity micro-movement is canceled by the ground static echo. The spatial region where the echo is canceled is defined as the standing wave cancellation blind zone.

[0034] Preferably, when the target object is determined to be in the standing wave phase cancellation blind zone, the edge computing and alarm control host extracts the absolute height difference as the feedforward control parameter, and calculates the target frequency that can make the physical phase difference leave the cancellation state based on the law of physical interference, and generates a frequency deviation command containing the target frequency adjustment amount. The ultrasonic sensing subsystem receives frequency offset commands and changes the oscillation frequency of the underlying drive circuit to radiate sound waves at the finely tuned transmission frequency, thereby altering the wavelength distribution in the spatial physical field and disrupting the physical interference conditions that cause echo cancellation.

[0035] The ultrasonic sensing subsystem maintains a single-frequency continuous wave operating mode at the finely adjusted transmission frequency. The edge computing and alarm control host extracts the horizontal plane coordinate data of the absolute spatial coordinates, converts it into the beam pointing angle of the ultrasonic transducer array through a preset coordinate system mapping matrix, and calculates the phase delay of each transducer element in the array accordingly. It then controls the high-frequency piezoelectric ceramic transducer array to perform directional scanning of the target area and extracts the chest cavity undulation micro-Doppler frequency shift signal that is not destroyed by standing wave interference as a micro-motion characteristic. The edge computing and alarm control host performs a fast Fourier transform on the micro-movement vital signs to separate the respiratory frequency characteristics in the frequency band of 0.2Hz to 0.5Hz and the heart rate characteristics in the frequency band of 1.0Hz to 2.0Hz. The extracted respiratory frequency and heart rate are compared with the preset life safety threshold range. When abnormal breathing or abnormal heart rate is detected, an alarm control signal is generated to drive the peripheral linkage equipment to perform the corresponding alarm action.

[0036] Example 2:

[0037] Based on Example 1, this embodiment refines the underlying mathematical and physical models of the logic for determining the standing wave phase cancellation blind zone and the logic for generating the frequency offset command, and introduces an ambient temperature dynamic compensation algorithm to eliminate the error of ambient temperature drift in the calculation of physical phase difference.

[0038] The edge computing and alarm control host collects real-time ambient temperature data of the monitored area through temperature sensors. Internally, this real-time ambient temperature data is defined as a one-dimensional scalar time series. One-dimensional scalar time series This serves as the raw input data for the temperature compensation algorithm.

[0039] The edge computing and alarm control host uses the sound velocity and temperature compensation formula for core calculation logic processing. The sound velocity and temperature compensation formula is specifically expressed as follows: [Formula omitted for brevity]. The constant 331.4 represents the base sound speed under the current environment after temperature compensation, in m / s. The constant 0.6 represents the temperature compensation coefficient in the air medium. Both the base sound speed and the temperature compensation coefficient are set based on standard empirical values ​​in the field of air acoustics.

[0040] The edge computing and alarm control host will calculate and output the speed of sound in the current environment. As floating-point scalar data, it is input into the subsequent physical phase difference calculation module to replace the static sound velocity constant; floating-point scalar data The introduction of dynamic corrections to the propagation speed benchmark of sound waves in the space physical field eliminates the interference of winter and summer temperature differences on wavelength extrapolation.

[0041] The physical phase difference calculation module receives the absolute height difference. Initial transmission frequency of the ultrasonic sensing subsystem And the speed of sound in the current environment As input data; absolute height difference Initial transmission frequency Speed ​​of sound in the current environment Construct a floating-point input vector in memory containing three-dimensional features. .

[0042] The physical phase difference calculation module performs calculations based on the physical phase difference derivation formula. The physical phase difference derivation formula is specifically expressed as follows: , in the formula The physical phase difference is represented by the optical path difference caused by ground reflection, in radians. The constant 4π is a preset phase constant derived from the product of double optical path difference and wavenumber.

[0043] The physical phase difference calculation module synchronously executes the logic decision tree to calculate the physical phase difference. The absolute value of the phase deviation from the phase that is an odd multiple of half the wavelength. The absolute value of the phase deviation is specifically expressed as... , in the formula In order to pass through The integer obtained by rounding down.

[0044] The logic decision tree has a phase tolerance threshold set. Phase tolerance threshold The minimum detectable signal-to-noise ratio prior data based on the Doppler echo receiver circuit is set as follows: radian.

[0045] When the absolute value of the phase deviation is less than or equal to the phase tolerance threshold At that time, the physical phase difference calculation module outputs a Boolean state variable. .

[0046] Boolean state variables The reverse mapping is to the hardware interrupt trigger signal in the physical entity. The hardware interrupt trigger signal drives the edge computing and alarm control host to enter the frequency offset instruction calculation stage.

[0047] After receiving a hardware interrupt trigger signal, the edge computing and alarm control host extracts the absolute height difference. Speed ​​of sound in the current environment and initial transmission frequency As the original input scalar for frequency offset calculation.

[0048] Edge computing and alarm control host set target physical phase difference The aim is to create constructive interference by superimposing phases in the space physical field.

[0049] The edge computing and alarm control host calculates the target transmission frequency based on the target frequency back-calculation formula. The specific expression of the target frequency back-calculation formula is as follows: ; Edge computing and alarm control host based on formula Calculate the target frequency offset .

[0050] Edge computing and alarm control host synchronously execute frequency offset constraint logic to calculate target frequency offset. The absolute value of the initial transmission frequency The ratio is used as the frequency offset percentage; the frequency offset constraint logic has a preset percentage threshold. Preset percentage threshold Based on the negative 3 dB bandwidth characteristic of the impedance curve of the high-frequency piezoelectric ceramic transducer, it is set to 0.03.

[0051] If the frequency deviation percentage is greater than the preset percentage threshold The edge computing and alarm control host will determine the target physical phase difference. Switch to To achieve an orthogonal state; the edge computing and alarm control host recalculates the target frequency offset based on the target's physical phase difference under the orthogonal state. Ensure target frequency offset Strictly limited to a preset percentage threshold Internally, this avoids a sharp drop in the electroacoustic conversion efficiency of the ultrasonic transducer.

[0052] Edge computing and alarm control host based on formula Calculate the frequency control word, where The number of bits in the phase accumulator of a direct digital frequency synthesizer chip. The system reference clock frequency is used, and the calculated frequency control word is used. Convert to hexadecimal numeric control words; The edge computing and alarm control host sends hexadecimal digital control words to the direct digital frequency synthesizer chip in the ultrasonic sensing subsystem via a serial peripheral interface bus. The direct digital frequency synthesizer chip changes the frequency of the output drive waveform according to the hexadecimal digital control word, thus completing the underlying hardware fine-tuning of the physical transmission frequency of the ultrasonic transducer.

[0053] Example 3:

[0054] Based on Embodiments 1 and 2, this embodiment provides a detailed description of the underlying hardware architecture and linkage execution components of an intelligent alarm system for abnormal behavior based on ultra-wideband spatial positioning and multimodal perception.

[0055] The ultra-wideband spatial positioning subsystem includes multiple ultra-wideband base stations distributed across the monitoring area and ultra-wideband tags configured on the target objects.

[0056] Specifically, the ultra-wideband base station and the ultra-wideband tag communicate wirelessly via time-of-flight ranging algorithm or time difference of arrival ranging algorithm, and calculate and output three-dimensional spatial coordinates containing the absolute position information of the target object. The three-dimensional spatial coordinates are transmitted to the edge computing and alarm control host via Ethernet or wireless LAN.

[0057] The ultrasonic sensing subsystem includes a high-frequency piezoelectric ceramic transducer array installed at the top of the monitoring area, a Doppler echo receiving circuit, and a frequency synthesis drive module.

[0058] Specifically, the high-frequency piezoelectric ceramic transducer array is used to transmit ultrasonic waves to the target area and receive Doppler echoes generated by the micro-movement of the target object's chest cavity; The Doppler echo receiving circuit is electrically connected to the high-frequency piezoelectric ceramic transducer array, and the mixing local oscillator terminal of the Doppler echo receiving circuit is connected to the reference signal output terminal of the frequency synthesis driving module. During the fine-tuning of the transmission frequency, the frequency synthesis driving module synchronously adjusts the reference frequency output to the mixing local oscillator, so that the Doppler echo receiving circuit amplifies, filters and performs in-phase quadrature (I / Q) mixing on the Doppler echo, and extracts the micro-Doppler frequency shift signal after eliminating carrier frequency offset as a micro-motion characteristic.

[0059] The frequency synthesis drive module is electrically connected to the high-frequency piezoelectric ceramic transducer array, and the frequency synthesis drive module contains a direct digital frequency synthesizer chip.

[0060] Specifically, the direct digital frequency synthesizer chip integrates a phase accumulator, a waveform read-only memory, and a digital-to-analog converter. The edge computing and alarm control host transmits data containing the target frequency offset via a serial peripheral interface bus. The hexadecimal control word is written into the frequency control register of the direct digital frequency synthesizer chip.

[0061] The phase accumulator changes the phase accumulation step size according to the hexadecimal digital control word in the frequency control register, and the waveform read-only memory outputs discrete digital waveform amplitudes according to the phase accumulation step size.

[0062] A digital-to-analog converter converts discrete digital waveform amplitudes into analog sinusoidal drive voltages with finely tuned transmit frequencies.

[0063] A simulated sinusoidal driving voltage is applied to the two poles of a high-frequency piezoelectric ceramic transducer array, driving the piezoelectric ceramic material to generate mechanical vibrations at the corresponding frequency, thereby realizing the underlying hardware mapping of digital frequency offset commands to physical transmission frequencies.

[0064] The system also includes an alarm linkage execution subsystem that is electrically connected to the edge computing and alarm control host.

[0065] Specifically, the alarm linkage execution subsystem includes an audible and visual alarm, an access control controller, and a communication gateway. When the edge computing and alarm control host detects abnormal breathing or heart rate based on micro-movement vital signs, it generates the highest-level alarm command.

[0066] The audible and visual alarm receives the highest level alarm command and emits a high-decibel sound and flashing light signal at the monitoring site.

[0067] The access control controller receives the highest-level alarm command and controls the internal relay to close, automatically releasing the electromagnetic access control lock status of the monitored area.

[0068] The communication gateway receives distress signals containing absolute spatial coordinates and micro-motion characteristics, encapsulates the distress signals into transmission control protocol data packets, and sends them to the remote monitoring terminal via the wide area network.

[0069] Example 4:

[0070] This embodiment provides a specific simulation of an abnormal behavior intelligent alarm system based on ultra-wideband spatial positioning and multimodal perception in a winter nursing home bathroom fall scenario.

[0071] The target person wearing an ultra-wideband tag enters the bathroom area where the ambient temperature is 5°C; the ultrasonic sensing subsystem radiates continuous sound waves into the bathroom area at an initial transmission frequency of 40,000 Hz.

[0072] The target object falls and lies flat on the ground; the ultra-wideband spatial positioning subsystem detects a sudden drop in the Z-axis coordinate and a horizontal displacement velocity approaching 0 m / s, locks the target object's current absolute spatial coordinates, and calculates the absolute height difference between the target object's chest cavity and the reflected ground as 0.249 m.

[0073] The edge computing and alarm control host collects real-time ambient temperature data of 5℃, and uses the sound speed temperature compensation formula to calculate that the sound speed in the current environment is 334.4m / s.

[0074] The physical phase difference calculation module receives an absolute height difference of 0.249m, an initial transmission frequency of 40000Hz, and a sound speed of 334.4m / s. Based on the physical phase difference derivation formula, it calculates the physical phase difference to be 119πrad. 119πrad is an odd multiple of half the wavelength. The edge calculation and alarm control host determines that the target object is currently in the standing wave phase cancellation blind zone, and the Doppler echo generated by the chest cavity micro-movement is canceled by the ground static echo.

[0075] The edge computing and alarm control host sets 120πrad as the physical phase difference of the target to form in-phase superposition; the edge computing and alarm control host calculates the target transmission frequency as 40289.15Hz based on the target frequency back-calculation formula, and calculates the target frequency deviation as 289.15Hz.

[0076] The edge computing and alarm control host calculates the target frequency offset to the initial transmission frequency as a ratio of 0.72%. Since the 0.72% frequency offset ratio is strictly less than the preset percentage threshold of 3%, the edge computing and alarm control host converts the 289.15Hz target frequency offset into a hexadecimal digital control word and sends it to the direct digital frequency synthesizer chip.

[0077] A direct digital frequency synthesizer chip drives a high-frequency piezoelectric ceramic transducer array to radiate sound waves at a finely tuned transmission frequency of 40289.15Hz. The finely tuned transmission frequency forms a constructive interference with in-phase superposition at an absolute height difference of 0.249m. The ultrasonic sensing subsystem extracts the weak breathing characteristics of the target object under the in-phase superposition state. The edge computing and alarm control host determines that the target object is in a state of rapid breathing after falling based on the weak breathing characteristics, generates the highest level alarm command, and drives the access controller to unlock the bathroom door.

[0078] Example 5:

[0079] This embodiment provides an underlying hardware structure for an edge computing and alarm control host and a computer-readable storage medium.

[0080] The edge computing and alarm control host includes at least one processor and a memory communicatively connected to the at least one processor; the memory stores a computer program that can be executed by the at least one processor; when the computer program is executed by the at least one processor, the edge computing and alarm control host implements the intelligent alarm method for abnormal behavior based on ultra-wideband spatial positioning and multimodal perception described in Embodiments 1 and 2.

[0081] The computer-readable storage medium stores a computer program; when the computer program is executed by a processor, it implements the intelligent alarm method for abnormal behavior based on ultra-wideband spatial positioning and multimodal perception described in Embodiments 1 and 2; the computer-readable storage medium includes a non-volatile solid-state physical medium capable of storing program code, such as a read-only memory, random access memory, magnetic disk, or optical disk.

[0082] In summary, the intelligent alarm method and system for abnormal behavior based on ultra-wideband spatial positioning and multimodal perception provided by this invention obtains the absolute height difference of the target object through the ultra-wideband spatial positioning subsystem and inputs the absolute height difference as a feedforward control parameter to the ultrasonic sensing subsystem. Based on the law of physical interference, the frequency offset command is calculated and output to fine-tune the transmission frequency of the ultrasonic sensing subsystem. The feedforward intervention of ultra-wideband macroscopic spatial geometric data on the ultrasonic microscopic physical transmission frequency breaks the standing wave phase cancellation blind zone when the target object is close to the ground. Under the condition of maintaining the resonant efficiency of high-frequency piezoelectric ceramic transducer and high sampling rate of single-frequency continuous wave, the accurate extraction of weak breathing features in extreme fall scenarios is achieved. The cross-modal underlying control architecture established by edge computing and alarm control host eliminates the dependence on complex back-end filtering algorithms. Combined with the dynamic environmental temperature compensation algorithm and spatiotemporal multi-dimensional fall judgment logic, the reliability of the underlying data and the accuracy of abnormal behavior alarm of the multimodal perception system in complex environments are improved.

[0083] Based on the preferred embodiments of the present invention described above, those skilled in the art can make various changes and modifications without departing from the inventive concept. The technical scope of this invention is not limited to the contents of the specification, but must be determined according to the scope of the claims.

Claims

1. An intelligent alarm method for abnormal behavior based on ultra-wideband spatial positioning and multimodal perception, characterized in that, Includes the following steps: S1. The three-dimensional spatial coordinates of the target object are obtained in real time through the ultra-wideband spatial positioning subsystem, and the current absolute spatial coordinates are locked when it is determined that the target object has experienced a suspected fall event. S2. Extract the height coordinate data from the absolute spatial coordinates and calculate the absolute height difference between the chest cavity of the target object and the reflecting ground. S3. Based on the absolute height difference and the initial transmission frequency of the ultrasonic sensing subsystem, calculate the physical phase difference corresponding to the optical path difference caused by ground reflection, and determine whether the target object is in the standing wave phase cancellation blind zone based on the physical phase difference. S4. If it is determined that the system is in the standing wave cancellation blind zone, the absolute height difference is used as the feedforward control parameter to calculate and output the frequency offset command in order to finely adjust the transmission frequency of the ultrasonic sensing subsystem and destroy the physical interference conditions of standing wave cancellation. S5. At the finely adjusted transmission frequency, the ultrasonic sensing subsystem extracts the micro-motion characteristics of the target object and triggers a corresponding alarm action based on the micro-motion characteristics. The step of calculating the physical phase difference corresponding to the optical path difference caused by ground reflection based on the absolute height difference and the initial transmission frequency of the ultrasonic sensing subsystem, and determining whether the target object is in the standing wave deactivation blind zone based on the physical phase difference, specifically includes: Obtain the product of the absolute height difference and the initial transmission frequency; The product is divided by the speed of sound in the current environment and multiplied by a preset phase constant to calculate the physical phase difference. When the physical phase difference is determined to be close to an odd multiple of half the wavelength, the target object is determined to be currently in the standing wave phase cancellation blind zone. The step of calculating and outputting a frequency offset command to fine-tune the transmission frequency of the ultrasonic sensing subsystem specifically includes: The target frequency offset is calculated based on the absolute height difference, so that the updated physical phase difference corresponding to the fine-tuned transmission frequency approaches an even multiple of half the wavelength to form in-phase superposition, or approaches a non-integer multiple of half the wavelength to form orthogonality. The target frequency offset is limited to a preset percentage threshold of the initial transmission frequency to output the frequency offset command while maintaining the resonant efficiency of the ultrasonic transducer. The preset percentage threshold is ≤3%.

2. The intelligent alarm method for abnormal behavior based on ultra-wideband spatial positioning and multimodal perception according to claim 1, characterized in that, The step of locking the current absolute spatial coordinates when determining that the target object has experienced a suspected fall event specifically includes: Real-time monitoring of the rate of change of the three-dimensional spatial coordinates within a preset time window; When the decrease in the height coordinate data exceeds a preset height threshold and the displacement speed of the horizontal plane is lower than a preset speed threshold, a suspected fall signal is generated, and the current absolute spatial coordinates of the target object are locked in response to the suspected fall signal.

3. The intelligent alarm method for abnormal behavior based on ultra-wideband spatial positioning and multimodal perception according to claim 1, characterized in that, Before the step of dividing the product by the speed of sound in the current environment, the method further includes: Obtain real-time ambient temperature data within the monitored area; The sound velocity is calculated by adding the product of the base sound velocity, the temperature compensation coefficient, and the real-time ambient temperature data, in order to eliminate the error in the calculation of physical phase difference caused by ambient temperature drift.

4. An intelligent alarm system for abnormal behavior based on ultra-wideband spatial positioning and multimodal perception, characterized in that, include: An ultra-wideband spatial positioning subsystem is used to acquire the three-dimensional spatial coordinates of a target object in real time. The ultrasonic sensing subsystem is used to transmit ultrasonic waves to the target area and receive the echoes to extract micro-motion characteristics. The edge computing and alarm control host is communicatively connected to the ultra-wideband spatial positioning subsystem and the ultrasonic sensing subsystem, respectively. The edge computing and alarm control host is configured to execute the intelligent alarm method for abnormal behavior based on ultra-wideband spatial positioning and multimodal perception as described in any one of claims 1 to 3.

5. The intelligent alarm system for abnormal behavior based on ultra-wideband spatial positioning and multimodal perception according to claim 4, characterized in that, The ultrasonic sensing subsystem includes: A high-frequency piezoelectric ceramic transducer array is installed at the top of the monitoring area to perform ultrasonic wave transmission and reception. The Doppler echo receiving circuit is electrically connected to the high-frequency piezoelectric ceramic transducer array and is used to extract the micro-Doppler frequency shift signal; The frequency synthesis drive module, electrically connected to the high-frequency piezoelectric ceramic transducer array, includes a voltage-controlled oscillator or a direct digital frequency synthesizer, and is used to receive frequency offset commands issued by the edge computing and alarm control host and dynamically adjust the transmission frequency.

6. The intelligent alarm system for abnormal behavior based on ultra-wideband spatial positioning and multimodal perception according to claim 4, characterized in that, The ultra-wideband spatial positioning subsystem includes: Multiple ultra-broadband base stations are distributed and deployed within the monitoring area; An ultra-wideband tag is configured on the target object; The ultra-wideband base station and the ultra-wideband tag communicate using a time-of-flight ranging algorithm or a time-difference-of-arrival ranging algorithm to calculate and output the three-dimensional spatial coordinates.

7. The intelligent alarm system for abnormal behavior based on ultra-wideband spatial positioning and multimodal perception according to claim 4, characterized in that, It also includes an alarm linkage execution subsystem electrically connected to the edge computing and alarm control host, the alarm linkage execution subsystem comprising: Audible and visual alarms are used to issue audible and visual warnings on-site when the highest level alarm command is received; Access control controllers are used to automatically unlock access to the monitored area when an alarm is triggered. A communication gateway is used to send a distress signal containing the absolute spatial coordinates and the micro-motion characteristics to a remote monitoring terminal.

8. The intelligent alarm system for abnormal behavior based on ultra-wideband spatial positioning and multimodal perception according to claim 4, characterized in that, The edge computing and alarm control host includes: At least one processor; and a memory communicatively connected to the at least one processor; The memory stores a computer program that can be executed by the at least one processor, which, when executed by the at least one processor, causes the edge computing and alarm control host to implement the method as described in any one of claims 1 to 3.

Images