Intelligent alarm method and system based on millimeter wave radar ranging

Abstract

The application relates to the field of radio ranging technology and discloses an intelligent alarm method and system based on millimeter wave radar ranging. The method comprises the following steps: scanning a safety area around a medical device by using a frequency-modulated continuous wave millimeter wave radar to obtain reflection signal data of a layered detection area; performing coherent superposition processing on the reflection signal data by using a medical safety ranging signal processing algorithm to obtain a device safety frequency domain signal; performing distance calculation on the frequency domain signal by using a medical device ranging frequency estimation algorithm to obtain distance measurement results and approaching speed data between a person and the device; performing threat prediction on the distance measurement results according to a medical safety trajectory tracking algorithm to obtain a person motion trajectory and an alarm triggering condition; and performing intelligent judgment on the person motion trajectory and a safety threshold to obtain a hierarchical alarm output signal. The application improves the accuracy of person approaching detection and the intelligent level of alarm response in a medical environment.

Description

Technical Field

[0001] This application relates to the field of radio ranging technology, and in particular to an intelligent alarm method and system based on millimeter-wave radar ranging. Background Technology

[0002] Existing medical device safety monitoring technologies primarily employ traditional infrared sensors, ultrasonic sensors, or video surveillance systems for personnel proximity detection. These technologies work by placing sensor arrays around the medical device, triggering an alarm signal when personnel enter a pre-defined safe area. Traditional infrared sensors, based on the principle of thermal radiation detection, can identify the heat source signal of the human body; ultrasonic sensors utilize the principle of sound wave reflection ranging to detect the presence and distance of objects; and video surveillance systems analyze the position and behavior of personnel through image recognition algorithms. These technologies are widely used in medical environments, providing fundamental monitoring capabilities for the safety protection of medical devices.

[0003] However, infrared sensors are susceptible to changes in ambient temperature and heat source interference, often resulting in false alarms and missed detections in complex medical environments. Ultrasonic sensors have limited detection accuracy, failing to provide high-precision distance measurements, and are easily affected by airflow and acoustic noise. While video surveillance systems can provide rich visual information, their performance degrades in low light or obstructed conditions, and they suffer from high computational complexity and slow response times. More importantly, these traditional technologies lack effective multi-target tracking capabilities and motion trajectory prediction functions, failing to provide early warnings of approaching personnel and only responding when a threat has already approached or occurred. Summary of the Invention

[0004] This application provides an intelligent alarm method and system based on millimeter-wave radar ranging, which solves the problems of low ranging accuracy, poor environmental adaptability, lack of trajectory prediction and intelligent threat assessment in existing medical equipment safety monitoring technologies, and improves the accuracy of personnel approach detection and the level of intelligence of alarm response in medical environments.

[0005] In a first aspect, this application provides an intelligent alarm method based on millimeter-wave radar ranging. The method includes: scanning a safe area surrounding a medical device using frequency-modulated continuous-wave millimeter-wave radar to obtain radio reflection signal data for a layered safety detection area; performing coherent superposition processing on the radio reflection signal data according to a medical safety ranging signal processing algorithm to obtain a device safety frequency domain signal; performing distance calculation processing on the device safety frequency domain signal using a medical device ranging frequency estimation algorithm to obtain distance measurement results and approach speed data between personnel and the device; performing safety threat prediction processing on the distance measurement results according to a medical safety trajectory tracking algorithm to obtain personnel movement trajectory and medical safety alarm triggering conditions; and performing intelligent judgment processing on the personnel movement trajectory and medical device safety thresholds to obtain a medical safety graded alarm output signal based on radio ranging.

[0006] Secondly, this application provides an intelligent alarm system based on millimeter-wave radar ranging, the intelligent alarm system based on millimeter-wave radar ranging comprising: The ranging module is used to perform radio wave ranging scans of the safety area around medical equipment using frequency-modulated continuous wave millimeter-wave radar to obtain radio reflection signal data of the layered safety detection area; The superposition module is used to coherently superimpose the radio reflection signal data according to the medical safety ranging signal processing algorithm to obtain the equipment safety frequency domain signal; The calculation module is used to perform distance calculation processing on the safe frequency domain signal of the equipment through the medical equipment ranging frequency estimation algorithm to obtain the distance measurement results and approach speed data between the personnel and the equipment; The prediction module is used to perform security threat prediction processing on the distance measurement results based on the medical safety trajectory tracking algorithm to obtain the personnel movement trajectory and medical safety alarm triggering conditions. The judgment module is used to intelligently judge the movement trajectory of the person and the safety threshold of the medical equipment to obtain a medical safety graded alarm output signal based on radio ranging.

[0007] Thirdly, a smart alarm device based on millimeter-wave radar ranging is provided, comprising: a memory and at least one processor, wherein the memory stores instructions; the at least one processor invokes the instructions in the memory to cause the smart alarm device based on millimeter-wave radar ranging to execute the aforementioned smart alarm method based on millimeter-wave radar ranging.

[0008] Fourthly, a computer-readable storage medium is provided, wherein instructions are stored in the computer and, when executed on a computer, cause the computer to perform the aforementioned intelligent alarm method based on millimeter-wave radar ranging.

[0009] The technical solution provided in this application utilizes frequency-modulated continuous wave millimeter-wave radar to perform radio wave ranging scans of the safety area surrounding medical equipment. This enables all-weather, high-precision distance measurement in complex medical environments, unaffected by lighting conditions, temperature changes, or electromagnetic interference. The obtained radio reflection signal data from the layered safety detection area provides a reliable foundation for subsequent accurate analysis. The medical safety ranging signal processing algorithm coherently superimposes the radio reflection signal data. Through the coordinated enhancement of multiple antenna signals, it significantly improves the signal-to-noise ratio and detection sensitivity. The resulting equipment safety frequency domain signal exhibits higher quality and stability, effectively suppressing the impact of environmental noise on detection accuracy. The medical equipment ranging frequency estimation algorithm performs distance calculation processing on the equipment safety frequency domain signal. Employing advanced frequency domain analysis technology, it achieves millimeter-level ranging accuracy. Simultaneously, the obtained distance measurement results between personnel and equipment, along with proximity speed data, provide dual-dimensional quantitative indicators for threat assessment, greatly improving the accuracy and reliability of safety monitoring.

[0010] The medical safety trajectory tracking algorithm performs security threat prediction processing on distance measurement results. By establishing a motion model and trajectory analysis mechanism, it can accurately predict the movement trend and future location of personnel. The resulting personnel movement trajectory and medical safety alarm triggering conditions achieve a technological breakthrough from passive detection to proactive early warning, providing sufficient response time for medical safety protection. The threat cascading decision mechanism, which intelligently processes personnel movement trajectories and medical equipment safety thresholds, effectively reduces false alarm and missed alarm rates through multi-level security assessments and dynamic threshold adjustments. The resulting medical safety graded alarm output signal based on radio ranging can provide differentiated alarm responses according to the threat level, ensuring timely security protection while avoiding excessive alarms that interfere with medical work, achieving intelligent and humanized security monitoring results. Attached Figure Description

[0011] To more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings used in the description of the embodiments will be briefly introduced below. Obviously, the drawings described below are some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0012] Figure 1 This is a schematic diagram of an embodiment of the intelligent alarm method based on millimeter-wave radar ranging in this application. Figure 2 This is a schematic diagram of an embodiment of an intelligent alarm system based on millimeter-wave radar ranging in this application. Figure 3This is a schematic block diagram of the intelligent alarm device based on millimeter-wave radar ranging in an embodiment of the present invention. Detailed Implementation

[0013] This application provides an intelligent alarm method and system based on millimeter-wave radar ranging. The terms "first," "second," "third," "fourth," etc. (if present) in the specification, claims, and accompanying drawings are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such data can be interchanged where appropriate so that the embodiments described herein can be implemented in a sequence other than that illustrated or described herein. Furthermore, the terms "comprising" or "having" and any variations thereof are intended to cover a non-exclusive inclusion; for example, a process, method, system, product, or device that includes a series of steps or units is not necessarily limited to those steps or units explicitly listed, but may include other steps or units not explicitly listed or inherent to such processes, methods, products, or devices.

[0014] For ease of understanding, the specific process of the embodiments of this application is described below. Please refer to [link / reference]. Figure 1 One embodiment of the intelligent alarm method based on millimeter-wave radar ranging in this application includes: Step S101: Use frequency modulated continuous wave millimeter wave radar to perform radio wave ranging scans on the safety area around the medical equipment to obtain radio reflection signal data of the layered safety detection area; Step S102: Perform coherent superposition processing on the radio reflection signal data according to the medical safety ranging signal processing algorithm to obtain the equipment safety frequency domain signal; Step S103: The distance calculation of the equipment safety frequency domain signal is performed by the medical equipment ranging frequency estimation algorithm to obtain the distance measurement results and approach speed data between the personnel and the equipment. Step S104: Based on the medical safety trajectory tracking algorithm, perform security threat prediction processing on the distance measurement results to obtain the personnel movement trajectory and medical safety alarm triggering conditions; Step S105: Perform intelligent judgment processing on the personnel movement trajectory and the safety threshold of medical equipment to obtain a medical safety graded alarm output signal based on radio ranging.

[0015] It is understood that the executing entity of this application can be an intelligent alarm system based on millimeter-wave radar ranging, or it can be a terminal or a server; the specific implementation is not limited here. This application's embodiments use a server as an example for illustration.

[0016] Specifically, when using frequency-modulated continuous wave (FM-CVT) millimeter-wave radar to perform radio wave ranging scans of the safe area surrounding medical equipment, the FM-CVT millimeter-wave radar transmitter first generates a linear frequency-modulated (LFM) signal. The sweep bandwidth of this signal determines the ranging resolution, and the sweep period parameter controls the measurement update frequency. A multi-antenna array directionally transmits the millimeter-wave signal around the medical equipment, forming a three-layer detection range covering the near-field danger zone, the mid-field warning zone, and the far-field monitoring zone. Each layer corresponds to different safety level requirements: the near-field danger zone covers a distance of 0 to 2 meters, the mid-field warning zone covers a distance of 2 to 5 meters, and the far-field monitoring zone covers a distance of 5 to 10 meters. When a person or object enters the detection area, the radio waves are reflected, and the reflected radio wave signal carries distance and speed information. The receiving antenna array captures these reflected signals, and after signal processing, obtains an intermediate frequency (IF) signal containing beat frequency information. The frequency of this beat frequency signal is directly proportional to the target distance; that is, the farther the distance, the higher the beat frequency, and the closer the distance, the lower the beat frequency. Therefore, the near-field danger zone corresponds to a low-frequency beat frequency, the mid-field warning zone corresponds to a mid-frequency beat frequency, and the far-field monitoring zone corresponds to a high-frequency beat frequency. Through analog-to-digital conversion and digitization, the original analog signal is converted into digitized radio reflection signal data, where each data point contains a timestamp, amplitude value, and phase information.

[0017] When the medical safety ranging signal processing algorithm performs coherent superposition processing on radio reflection signal data, it first inputs the signal into an I / Q demodulator for mixing, down-converting the high-frequency reflected signal into a beat frequency signal containing target distance information. The medical safety weight allocation rule assigns threat assessment priority coefficients based on the safety importance of different spatial areas. This priority coefficient is independent of the beat frequency value and is determined solely by the threat level of the target's spatial location. Near-field danger zones, being closest to medical equipment and possessing the highest threat level, are assigned the highest safety priority coefficient of 3.0 in subsequent threat assessments. Mid-field warning zones are assigned a medium safety priority coefficient of 2.0, and far-field monitoring zones are assigned the lowest safety priority coefficient of 1.0. Although near-field areas correspond to lower beat frequency values ​​(below the frequency band), this does not affect their priority in safety assessments because the safety priority coefficient is applied to threat scoring calculations only after the target distance and its location are determined, rather than being directly multiplied by the beat frequency value.

[0018] The multi-antenna coherent superposition formula performs phase compensation on the beat frequency signals received by each antenna, eliminating the phase difference caused by antenna spacing, and performs amplitude superposition processing to superimpose the signals received by multiple antennas in phase, enhancing the useful signal and suppressing random noise. During this process, signals in each region are processed according to the same signal processing flow, without involving differentiated applications of regional weights. A Fast Fourier Transform (FFT) converts the composite beat frequency signal in the time domain to the frequency domain, obtaining the frequency domain amplitude spectrum and frequency domain phase spectrum. The frequency domain amplitude spectrum shows the intensity distribution of different frequency components, where the low-frequency band corresponds to near-field targets, the mid-frequency band to mid-field targets, and the high-frequency band to far-field targets. The frequency domain phase spectrum contains the target's phase information. The medical security frequency domain optimization algorithm filters the frequency domain data, removing high-frequency noise and low-frequency drift, and enhances the signal to highlight the target's frequency peaks. The optimized frequency domain signal is used for subsequent distance calculation and target identification. Only after the target's region is determined through distance calculation is the threat score weighted according to the region's security priority coefficient.

[0019] When the medical equipment ranging frequency estimation algorithm processes the distance of the equipment's safe frequency domain signal, the peak detection algorithm searches for the frequency component with the largest amplitude in the frequency domain. These peaks correspond to the target positions with the strongest reflected signals, and the corresponding frequency amplitude values ​​are recorded for subsequent signal-to-noise ratio analysis. The Chirp-Z converter refines the frequency of the detected target frequency peaks, performing high-resolution frequency analysis within a small range near the peaks, improving the original frequency resolution from several kilohertz to tens of kilohertz. The millimeter-wave ranging formula converts the precise target frequency value into the actual physical distance. This conversion is based on the fundamental ranging principle of frequency-modulated continuous wave radar, i.e., the beat frequency is proportional to the target distance. The Doppler frequency shift extraction algorithm separates the frequency shift caused by target motion from the frequency domain signal. By analyzing the frequency difference between two consecutive scans, the radial and tangential velocity components of the target are calculated. Data fusion processing comprehensively analyzes the instantaneous distance value and the radial velocity component to form a description of the target's motion state.

[0020] When the medical safety trajectory tracking algorithm performs security threat prediction processing on distance measurement results, the Kalman filter, a recursive digital filter, is used to estimate the state of a dynamic system from a series of noisy measurement data. In this invention, it is used to process continuous distance and velocity measurement data to obtain a smooth target state vector containing the target's position, velocity, and acceleration information. A cubic spline interpolation algorithm mathematically fits the target state vector across multiple consecutive measurement cycles to construct a continuous trajectory function. This function describes the target's motion pattern along the time axis and calculates key indicators such as average velocity and rate of change of acceleration as motion trend parameters. Based on the current motion trajectory function, the trajectory prediction model predicts the target's possible position in future time periods using mathematical extrapolation. The trajectory prediction algorithm divides the personnel motion trajectory function into time steps, setting multiple prediction time nodes within a preset time interval. The position extrapolation module uses motion trend parameters for recursive calculation, progressively calculating the predicted distance and angle values ​​corresponding to each time node. The medical equipment coordinate positioning algorithm converts the predicted distance and angle values ​​in polar coordinates into three-dimensional position coordinates in a Cartesian coordinate system, facilitating subsequent spatial position analysis. The trajectory continuity check algorithm examines the continuity between adjacent predicted location points. If abnormal jumps or unreasonable movement trajectories are detected, smoothing corrections are performed. The collision detection process performs a spatial geometric comparison between the predicted personnel position sequence and the safety zone boundary of the medical equipment, calculating potential collision time points and corresponding collision probability values. The threat assessment matrix comprehensively considers multiple factors such as collision probability, target approach speed, and direction of movement, providing a comprehensive score for each target to form a quantified threat level index.

[0021] When intelligent decision processing compares and analyzes personnel movement trajectories with medical device safety thresholds, the threat cascade decision unit employs a three-level decision structure for multi-layered security assessment. The initial decision responds quickly based on real-time distance and speed data, outputting a threat warning when the target distance is less than 2 meters and the approach speed exceeds 1 meter per second. The intermediate decision verifies stability by considering data consistency over 5 to 10 consecutive measurement cycles; only when continuous measurements show a threat is the threat confirmed. The advanced decision performs in-depth analysis based on long-term movement trends over 30 seconds, comprehensively considering the target's movement intentions and behavioral patterns. The medical device safety grading standard classifies the decision results into extremely high threat, high threat, and medium threat levels. Each level corresponds to different alarm intensities and response measures. The extremely high threat level requires simultaneous confirmation at all three decision levels and the target is expected to contact the device within 3 seconds. The high threat level requires confirmation at least at two decision levels and the target continuing to approach. The medium threat level requires only one decision level to indicate a minor threat.

[0022] The dynamic threshold adjustment algorithm continuously adjusts the preset safety thresholds for medical devices based on current environmental conditions, resulting in real-time safety threshold parameters adapted to the current environment. The algorithm's input parameters include: background noise level quantified by signal-to-noise ratio (SNR), personnel density quantified by the number of targets per unit area, temperature and humidity values ​​directly measured by sensors, and the device's current operating status. The algorithm's output parameters include: adjusted real-time distance thresholds and real-time speed thresholds. The specific adjustment rules are as follows: when the SNR corresponding to the background noise level is higher than 20 dB, it is considered a low-noise environment; the real-time distance threshold is adjusted to 0.9 times the baseline distance threshold to improve sensitivity, and the real-time speed threshold is also adjusted accordingly to 0.9 times the baseline speed threshold. When the SNR is between 10 and 20 dB, it is considered a normal environment; the real-time threshold remains the same as the baseline threshold. When the SNR is lower than 10 dB, it is considered a high-noise environment; the real-time distance threshold is adjusted to 1.2 times the baseline distance threshold to reduce false alarms, and the real-time speed threshold is also adjusted accordingly to 1.2 times the baseline speed threshold. When the population density exceeds 5 people per 10 square meters, regardless of the noise level, the real-time distance and speed thresholds are multiplied by a factor of 0.85 to increase the alert level. Temperature changes affect the propagation speed of electromagnetic waves. When the temperature deviates from the standard temperature by 20 degrees Celsius, the distance measurement results need to be multiplied by a temperature correction factor, which is adjusted at a rate of 0.01% per degree Celsius based on the temperature difference.

[0023] The comparison and matching process logically compares the threat level classification identifier with real-time security threshold parameters. Based on the threat level and the threshold setting, it generates corresponding alarm output control commands and alarm intensity parameters. When the target distance for an extremely high threat level is less than 50% of the real-time distance threshold, the highest-level alarm control command is generated, and the alarm intensity parameter is set to 100% of the full scale. When the target distance for a high threat level is less than 70% of the real-time distance threshold, a medium-level alarm control command is generated, and the alarm intensity parameter is set to 60% of the full scale. When the target distance for a medium threat level is less than the real-time distance threshold, a low-level alarm control command is generated, and the alarm intensity parameter is set to 30% of the full scale. The graded alarm output control module generates different forms of alarm signals based on the control commands and alarm intensity parameters, including multiple output methods such as audible alarms, optical alarms, and data communication alarms. The volume, frequency, and repetition count of the audible alarm, and the brightness, color, and flashing mode of the optical signal are all adjusted accordingly based on the alarm intensity parameters.

[0024] In one specific embodiment, the process of performing step S101 may specifically include the following steps: A linear frequency modulated signal is generated by a frequency modulated continuous wave millimeter-wave radar transmitter and transmitted, resulting in a millimeter-wave signal with sweep bandwidth and sweep period parameters. Millimeter-wave signals are transmitted directionally around medical devices via a multi-antenna array to obtain radio waves covering near-field danger zones, mid-field warning zones, and far-field monitoring zones. Based on the reflection of radio waves by the target object, a reflected radio wave signal carrying distance and speed information is obtained; The reflected radio wave signal is input into the receiving antenna array for signal acquisition processing to obtain an intermediate frequency signal containing beat frequency information; The intermediate frequency signal is converted from analog to digital and then digitized to obtain radio reflection signal data of the layered security detection area.

[0025] Specifically, when a frequency-modulated continuous wave (FM-CRW) millimeter-wave radar transmitter generates a linearly frequency-modulated (LFM) signal for transmission, the voltage-controlled oscillator (VCO) inside the transmitter generates a millimeter-wave signal whose frequency varies linearly with time according to preset frequency modulation parameters. This signal's frequency is periodically scanned within a set frequency range in a sawtooth wave pattern. The sweep bandwidth parameter determines the radar's range resolution, while the sweep period parameter controls the measurement update rate; together, they constitute the basic operating parameters of the FM-CRW radar. The LFM signal is amplified to a suitable transmission power level by a power amplifier, and simultaneously, the fundamental frequency signal is multiplied to the millimeter-wave band by a frequency multiplier, forming a millimeter-wave signal with a specific sweep bandwidth and sweep period. The phase-locked loop (PLL) circuit inside the signal generator ensures the linearity and stability of the frequency scan, avoiding the impact of frequency drift on ranging accuracy.

[0026] When millimeter-wave signals are directionally transmitted around medical devices using a multi-antenna array, the array consists of multiple microstrip antennas or waveguide horn antennas, each with a specific radiation pattern and gain characteristics. Directional transmission processes control the phase and amplitude of each antenna element to form a beam pointing in a specific direction, covering a predetermined area around the medical device. The near-field danger zone corresponds to the area closest to the medical device, typically defined as within a few meters of the device surface; detection in this area requires the highest sensitivity and rapid response. The mid-field warning zone, located outside the danger zone, serves as a buffer and early warning area, providing operators with sufficient reaction time. The far-field monitoring zone covers a larger area for early detection and tracking of potentially approaching personnel or objects. Radio waves propagate in space according to the laws of electromagnetic wave propagation, with their energy density decreasing inversely with the square of the distance.

[0027] When radio waves are reflected by a target object, reflection occurs when the radio waves encounter objects with different dielectric constants. The intensity of the reflection depends on the target object's radar cross-section, material properties, and surface roughness. As a primary detection target, the human body's radar cross-section changes according to its posture, orientation, and distance, resulting in corresponding variations in the intensity of the reflected signal. The reflected radio wave signal carries not only distance information but also velocity information due to the Doppler effect. When the target object moves radially relative to the radar, the frequency of the reflected signal shifts, and the magnitude of this shift is proportional to the target's radial velocity. Reflection processing is a passive physical process; the surface characteristics, shape, and material composition of the target object all influence the characteristics of the reflected signal.

[0028] When the reflected radio wave signal is input into the receiving antenna array for signal acquisition processing, the receiving antenna array adopts a similar structure to the transmitting antenna array, but is specifically optimized for receiving weak reflected signals. Signal acquisition processing includes multiple stages such as antenna reception, low-noise amplification, and down-conversion, each affecting the final signal quality. The received reflected signal is first amplified by a low-noise amplifier, which has an extremely low noise figure, minimizing noise introduction while amplifying the useful signal. Down-conversion mixing mixes the high-frequency millimeter-wave reflected signal with the local oscillator signal to obtain a lower-frequency intermediate frequency (IF) signal, which retains the range and velocity information from the original reflected signal. Beat frequency information is the core information of frequency-modulated continuous wave radar; its frequency directly corresponds to the target's range. Beat frequency is generated due to the frequency difference caused by the time delay between the transmitted and received signals.

[0029] When performing analog-to-digital conversion (ADC) and digitization on intermediate frequency (IF) signals, the ADC converts the continuous analog IF signal into a discrete digital signal. The sampling frequency of the conversion must satisfy the Nyquist sampling theorem to ensure that all frequency components in the signal are correctly sampled. Digitization includes two steps: quantization and encoding. Quantization converts continuous amplitude values ​​into a finite number of discrete levels, and encoding represents these discrete levels as digital codes. The concept of layered security detection zones is reflected in the digitization stage. By processing and labeling frequency bands corresponding to different distance ranges separately, data groups corresponding to three security zones are formed. The data structure of radio reflection signal data includes timestamps, frequency values, amplitude values, and phase information. This data provides the information foundation for subsequent signal processing and target identification. Digitization also includes preliminary filtering and preprocessing operations to remove obvious noise and interference signals, ensuring that the data quality meets the requirements of subsequent processing.

[0030] In one specific embodiment, the process of performing step S102 may specifically include the following steps: The radio reflection signal data is input into the I / Q demodulator for mixing to obtain a beat frequency signal containing target distance information; Based on the medical safety zone division rules, the beat frequency signal is processed by frequency band region identification. According to the frequency band of the beat frequency, its corresponding near-field danger zone, mid-field warning zone or far-field monitoring zone is identified. The beat frequency is proportional to the distance. The near-field danger zone corresponds to the low-frequency beat frequency, the mid-field warning zone corresponds to the mid-frequency beat frequency, and the far-field monitoring zone corresponds to the high-frequency beat frequency. The phase compensation and amplitude superposition processing of the beat frequency signals of each region are performed according to the multi-antenna coherent superposition formula to obtain the composite beat frequency signal after coherent superposition. The composite beat frequency signal is input into a fast Fourier transform for frequency domain conversion processing to obtain the frequency domain amplitude spectrum and frequency domain phase spectrum; The frequency domain amplitude spectrum and frequency domain phase spectrum are filtered and enhanced based on the medical safety frequency domain optimization algorithm to obtain the equipment safety frequency domain signal.

[0031] Specifically, when radio reflection signal data is input into the I / Q demodulator for mixing, the I / Q demodulator decomposes the RF signal into two independent channels: in-phase and quadrature components. Frequency down-conversion is achieved by mixing these components with the local oscillator signal. The frequency of the beat frequency signal directly reflects the distance to the target; the farther the distance, the higher the beat frequency, and the closer the distance, the lower the beat frequency. Therefore, near-field danger zones correspond to low-frequency beat frequencies, mid-field warning zones correspond to mid-frequency beat frequencies, and far-field monitoring zones correspond to high-frequency beat frequencies. When processing beat frequency signals for frequency band region identification based on medical safety zone division rules, the system determines the corresponding spatial region based on the frequency range of the beat frequency, identifying the low-frequency band as near-field, the mid-frequency band as mid-field, and the high-frequency band as far-field. At this stage, only region identification is performed without applying safety weights, and all region signals are processed using the same procedure.

[0032] When performing phase compensation and amplitude superposition processing on the beat frequency signals of each region using the multi-antenna coherent superposition formula, the propagation distance of the reflected signal from the same target to different antennas varies slightly due to the different spatial positions of the antenna elements in the antenna array, resulting in phase differences between the received signals of each antenna. Multi-antenna coherent superposition processing first calculates the phase difference between the received signals of each antenna, and then performs phase rotation correction on the signal of each antenna to align the phases of multiple signals from the same target to the same reference phase. After phase alignment, the signals from multiple antennas are vector-superimposed and summed. Since the phases of the target signals are now consistent, they accumulate and strengthen in the same direction, while random noise signals on each antenna cancel each other out due to their random phases. This results in a composite beat frequency signal with higher amplitude and better signal-to-noise ratio compared to a single-antenna signal. During this superposition process, regardless of whether the target originates from the near-field, mid-field, or far-field region, all region signals use the same phase compensation and amplitude superposition algorithm, without any differentiated processing based on security levels.

[0033] The composite beat frequency signal is input into a Fast Fourier Transform (FFT) for frequency domain transformation, yielding a frequency domain amplitude spectrum and a frequency domain phase spectrum. Low-frequency peaks correspond to near-field targets, mid-frequency peaks to mid-field targets, and high-frequency peaks to far-field targets. A medical safety frequency domain optimization algorithm is used to filter and enhance the frequency domain amplitude and phase spectra, removing noise interference and highlighting target peaks to obtain the equipment safety frequency domain signal. The peaks in different frequency bands of this signal are identified to correspond to specific spatial regions, providing a data foundation for subsequent distance calculation. Only after the target's specific location is determined through distance calculation will the system apply corresponding safety priority coefficients to weight the threat score during the threat assessment phase, based on the target's region.

[0034] In one specific embodiment, the process of executing step S103 may specifically include the following steps: Based on the peak detection algorithm, the spectrum peak search processing of the equipment's safe frequency domain signal is performed to obtain the target frequency peak position and the corresponding frequency amplitude value; The target frequency peak position is input into the Chirp-Z converter for frequency refinement to obtain a high-resolution, accurate target frequency value. The instantaneous distance between personnel and medical equipment is obtained by performing distance conversion on the precise target frequency value according to the millimeter wave ranging formula. Based on the Doppler frequency shift extraction algorithm, velocity component separation processing is performed on the equipment safety frequency domain signal to obtain the target radial velocity component and tangential velocity component; The instantaneous distance value and radial velocity component are fused to obtain the distance measurement results and approach speed data between personnel and equipment.

[0035] Specifically, the entire frequency spectrum is scanned, the amplitude value of each frequency point is calculated, and compared with the amplitude values ​​of its neighboring frequency points. When the amplitude value of a frequency point is greater than the amplitude values ​​of a preset number of neighboring points on both its left and right sides, that point is identified as a peak candidate point. To avoid false peaks caused by noise, the algorithm sets minimum peak height thresholds and minimum peak spacing thresholds; only candidate points that meet these conditions are confirmed as true target peaks. The target frequency peak position records the specific frequency coordinates of each detected target in the frequency domain, which directly correspond to the distance relationship between the target and the radar. The corresponding frequency amplitude value reflects the target's reflection intensity; large targets or targets with strong reflection characteristics will produce higher amplitude values, while small targets or targets with weak reflection will produce relatively lower amplitude values. The peak search processing also includes a peak clustering function, which merges multiple small peaks that are very close together into a single main peak, avoiding repeated detection of the same target.

[0036] When the target frequency peak position is input into a Chirp-Z transformer for frequency refinement, the Chirp-Z transformer, an extension of the Fast Fourier Transform, is specifically designed for high-resolution analysis of a specific frequency range in the frequency domain. The transformer receives a coarse frequency position provided by a peak detection algorithm, sets a small frequency range window around that position, and then performs dense frequency sampling within this window. The core principle of frequency refinement is to improve the accuracy of frequency measurement by increasing the density of frequency domain sampling points. The frequency resolution of the original Fast Fourier Transform is limited by the signal length and sampling frequency, while the Chirp-Z transformer can achieve arbitrarily high resolution within a specified frequency range. The transformation process includes multiple steps such as window function weighting, complex multiplication, and inverse transform, each step aimed at obtaining more accurate frequency information within the specified frequency range. High-resolution, accurate target frequency values ​​improve the frequency resolution from the kilohertz level to tens of hertz or even higher precision, directly translating into a significant improvement in distance measurement accuracy. Refinement processing can also separate multiple targets that are very close together, solving the problem of traditional methods being unable to distinguish dense targets.

[0037] When converting the precise target frequency value into distance using the millimeter-wave ranging formula, the formula establishes a mathematical relationship between the beat frequency and the target distance. In a frequency-modulated continuous wave radar system, the target distance is directly proportional to the beat frequency, with the proportionality coefficient determined by the radar's system parameters. The mathematical expression for distance conversion can be described as: target distance equals beat frequency multiplied by the speed of light, then multiplied by the sweep period, and finally divided by twice the sweep bandwidth. Here, the beat frequency is the precise target frequency value obtained from frequency domain analysis, the speed of light is the constant of electromagnetic wave propagation in a vacuum, the sweep period is the frequency modulation period of the radar signal, and the sweep bandwidth is the frequency variation range of the radar signal. The reason for dividing by two is that the electromagnetic wave needs to travel twice the propagation distance from the radar to the target and back to the radar. The distance conversion process also needs to consider the influence of environmental factors on the propagation speed of electromagnetic waves; in practical applications, the speed of light constant is fine-tuned based on environmental parameters such as temperature and humidity. The instantaneous distance value between personnel and medical equipment is a real-time distance measurement result obtained through precise calculation; the accuracy of this value directly affects the performance of the entire security monitoring system. When performing velocity component separation processing on the device's security frequency domain signal based on the Doppler frequency shift extraction algorithm, the Doppler frequency shift is a frequency change phenomenon caused by the target's motion relative to the radar. The extraction algorithm calculates the change in the target's peak frequency by comparing two or more consecutive frequency domain measurements; this change directly corresponds to the target's velocity. Velocity component separation processing decomposes the target's total velocity into two independent vector components: radial velocity and tangential velocity. The radial velocity component is the target's velocity along the radar's line of sight; a positive value indicates the target is moving away from the radar, while a negative value indicates the target is approaching the radar. The tangential velocity component is the target's velocity perpendicular to the radar's line of sight; this component does not produce a Doppler frequency shift and needs to be obtained through angle measurement and geometric calculation. The relationship between the Doppler frequency shift and the radial velocity can be expressed as: the frequency shift magnitude equals twice the radial velocity multiplied by the radar's operating frequency, then divided by the speed of light. The algorithm also includes a velocity direction discrimination function, determining whether the target is approaching or moving away from the radar by analyzing the positive or negative direction of the frequency shift. The accuracy of velocity measurement depends on the accuracy of frequency measurement and the sensitivity of the Doppler effect. Millimeter-wave radar is very sensitive to velocity changes due to its high operating frequency.

[0038] When fusing instantaneous distance values ​​and radial velocity components, data fusion is a process of comprehensively analyzing information from different measurement dimensions. The fusion process first aligns the instantaneous distance values ​​and radial velocity components in time, ensuring that the two sets of data correspond to the target state at the same moment. Distance measurement results include the target's current position information and historical position change trends, which are used to build a trajectory model of the target. Approach velocity data includes not only the instantaneous value of the radial velocity component but also the rate of change of velocity, i.e., acceleration information; these parameters are crucial for predicting the target's future motion state. The data fusion algorithm uses a weighted average method, assigning appropriate weight coefficients based on the reliability of different measurement parameters. Distance measurements typically have higher accuracy and reliability, so they are assigned a larger weight; velocity measurements are relatively more affected by noise, so they are assigned a smaller weight. The fused data undergoes consistency checks and outlier removal to ensure the accuracy and stability of the output results. The distance measurement results between personnel and equipment, along with the approach velocity data, constitute a description of the target's motion state.

[0039] In one specific embodiment, the process of executing step S104 may specifically include the following steps: The distance measurement results and approach velocity data are input into a Kalman filter for state estimation processing to obtain a target state vector containing position, velocity and acceleration. Based on the cubic spline interpolation algorithm, the target state vector of multiple consecutive measurement cycles is subjected to trajectory fitting to obtain the personnel motion trajectory function and motion trend parameters. Based on the trajectory prediction model, the future position of the personnel movement trajectory function is extrapolated to obtain the personnel position sequence within the prediction time window; Collision detection is performed between the personnel location sequence and the boundary of the medical equipment safety zone to obtain potential collision time points and collision probability values. Based on the threat assessment matrix, the collision probability value and motion trend parameters are comprehensively scored to obtain the personnel movement trajectory and medical safety alarm triggering conditions.

[0040] Specifically, when range measurement results and approach velocity data are input into a Kalman filter for state estimation, the Kalman filter is a recursive digital filtering algorithm specifically designed to estimate the true state of a dynamic system from noisy measurement data. The filter establishes a mathematical model to describe the target's motion state, which includes three state variables: position, velocity, and acceleration. These variables constitute the system's state vector. The Kalman filtering process consists of two phases: prediction and update. The prediction phase predicts the current state based on the system's dynamic model and the state estimate from the previous moment. The update phase uses the current observation data to correct the prediction. Range measurement results are used as the position observation input filter, and approach velocity data are used as the velocity observation input filter. The filter sets the corresponding measurement noise covariance parameters based on the noise characteristics of these observation data. The state estimation process obtains the optimal state estimate by minimizing the mean square value of the estimation error, effectively suppressing the influence of measurement noise on the state estimation. The target state vector, containing position, velocity, and acceleration, provides a complete description of the target's motion state at the current moment. The position component represents the target's spatial coordinates relative to the radar, the velocity component represents the target's velocity vector, and the acceleration component represents the target's changing motion trend.

[0041] When performing trajectory fitting on target state vectors across multiple measurement periods using cubic spline interpolation, cubic spline interpolation, a numerical analysis method, constructs a smooth interpolation function through a series of cubic polynomial segments. The algorithm uses the state vectors from multiple measurement periods as interpolation nodes, constructing piecewise cubic polynomial functions between these nodes, with each polynomial segment maintaining second-order continuity and differentiability within its domain. Trajectory fitting first preprocesses the state vector data in the time series, removing obvious outliers and unreasonable data points, then arranges the remaining valid data points in chronological order. The construction of the cubic spline function must satisfy boundary and continuity conditions. Boundary conditions typically use natural or complete splines, while the continuity condition requires that the function value, first derivative, and second derivative be continuous at each node. The personnel movement trajectory function is the fitted mathematical expression, which describes the target's positional change pattern along the time axis, exhibiting continuity, smoothness, and differentiability. Motion trend parameters are extracted from the fitted function, including key indicators such as average velocity, rate of change of velocity, and motion direction angle. These parameters reflect the basic characteristics and trends of the target's motion. When extrapolating the future position of a person's motion trajectory function using a trajectory prediction model, the model predicts the target's possible position within a future time period based on the current trajectory function and motion trend parameters. The prediction model employs mathematical extrapolation to extend the established trajectory function to the future time range, while considering the impact of changes in motion trends on the prediction results. The future position extrapolation process first determines the prediction time window, which is set according to the needs of security monitoring and the system's response time, typically covering a time span from a few seconds to tens of seconds into the future. The extrapolation algorithm sets multiple time nodes within the prediction time window, calculating the corresponding predicted position coordinates for each time node. The calculation process considers the target's current position, velocity, and acceleration information. The prediction process also includes uncertainty analysis, calculating the confidence interval of the prediction results using error propagation theory to reflect the reliability of the prediction. The sequence of personnel positions within the prediction time window is a series of predicted position coordinates arranged chronologically, with each coordinate corresponding to a specific future moment. These coordinates constitute a discrete representation of the target's predicted motion trajectory.

[0042] When performing collision detection on a sequence of personnel positions against the boundary of a safety zone for medical equipment, collision detection is a crucial problem in computational geometry, used to determine whether a moving object will come into contact with a static obstacle. The safety zone boundary for medical equipment is typically defined as the geometry surrounding the equipment, which may be circular, rectangular, or a more complex polygonal region, each corresponding to different safety level requirements. The collision detection algorithm examines each coordinate point in the predicted position sequence, determining whether the point is inside the safety zone boundary. If the predicted position enters the safety zone, a potential collision is considered to have occurred. The algorithm also calculates the time required for the target to move from its current position to the safety zone boundary; this time value is the potential collision time point, which is crucial for early warning by the safety system. The collision probability calculation considers the uncertainty of the prediction and the randomness of the target's movement, using statistical analysis methods to assess the likelihood of a collision. Collision detection processing also includes collision point localization and collision angle analysis to determine the specific location and mode of contact between the target and the safety zone boundary. This information helps in developing appropriate safety protection measures.

[0043] When comprehensively scoring collision probability values ​​and motion trend parameters based on a threat assessment matrix, the threat assessment matrix is ​​a multi-dimensional evaluation system that comprehensively considers multiple factors to quantify the threat level of a target. The rows of the assessment matrix correspond to different threat factors, including collision probability, approach speed, direction of movement, and target size, while the columns correspond to different threat levels, ranging from low to high threat. The motion trend parameters specifically include three sub-parameters: speed magnitude, acceleration direction, and motion duration. Approach speed is the primary manifestation of the speed magnitude sub-parameter, reflecting how quickly the target approaches the medical equipment; acceleration direction reflects the changing trend of the target's motion state; and motion duration reflects the length of time the target continues to move towards the equipment.

[0044] The comprehensive scoring process first standardizes each threat factor, converting parameters of different dimensions into dimensionless scores between 0 and 1. Then, it assigns weight coefficients based on the importance of each factor. Collision probability, as the most important threat indicator, receives 50% weight; approach speed receives 30%; movement direction receives 15%; and target size receives 5%. Within the movement trend parameters, speed magnitude accounts for 60% weight, acceleration direction for 30%, and movement persistence for 10%. The score calculation uses a weighted summation method, multiplying the standardized scores of each factor by their corresponding weights and summing the results to obtain a comprehensive threat score within the range of 0 to 1. Personnel movement trajectory information is used in threat assessment to determine the target's movement intentions and behavioral patterns. Trajectories that continuously approach medical equipment are considered more threatening than random movements; for continuously approaching trajectories, the system adds a 20% threat bonus to the comprehensive threat score.

[0045] The triggering conditions for medical security alarms are determined by comparing a comprehensive threat score with a preset threshold. When the score exceeds the threshold for the corresponding level, the system triggers the corresponding level of security alarm, achieving tiered early warning and response. Specifically, when the comprehensive threat score is below 0.3, it is considered low threat and no alarm is triggered; when the score is between 0.3 and 0.6, it is considered medium threat and triggers a Level 1 alarm; when the score is between 0.6 and 0.8, it is considered high threat and triggers a Level 2 alarm; and when the score exceeds 0.8, it is considered extremely high threat and triggers a Level 3 alarm. Different alarm levels correspond to different intensities of audible and visual prompts and notification methods.

[0046] In one specific embodiment, the process of executing step S105 may specifically include the following steps: Based on the trajectory prediction algorithm, the time step segmentation process of the personnel movement trajectory function is performed to obtain multiple predicted time nodes within a preset time interval. The motion trend parameters are input into the position extrapolation module for recursive calculation to obtain the predicted distance and predicted angle values ​​corresponding to each prediction time node. Based on the medical equipment coordinate positioning algorithm, the predicted distance and predicted angle values ​​are processed by spatial coordinate transformation to obtain the predicted three-dimensional position coordinates relative to the medical equipment. The predicted three-dimensional position coordinates are verified by a trajectory continuity test algorithm to obtain a corrected set of predicted position points. The corrected set of predicted location points is arranged and combined according to the time series to obtain the personnel location sequence within the predicted time window.

[0047] Specifically, when performing time-step segmentation on the trajectory function of personnel movement based on trajectory prediction algorithms, time-step segmentation is a numerical processing method that divides a continuous time interval into discrete time points. The trajectory prediction algorithm first determines the total time span for prediction, which is set according to the needs of medical safety monitoring, typically covering a time range of three to ten seconds in the future. The algorithm divides the total time span by a preset time-step interval to obtain the number of time segments. The selection of the time step needs to balance prediction accuracy and computational efficiency; a step that is too small will increase the computational burden, while a step that is too large will reduce the temporal resolution of the prediction. The segmentation process starts from the current moment and sequentially sets future time nodes at fixed time intervals. Each time node represents a moment when position prediction is required. The preset time interval is typically set to tens to hundreds of milliseconds, taking into account the typical speed of personnel movement and the response requirements of medical equipment safety systems. Multiple prediction time nodes constitute discrete sampling points on the time axis. These sampling points provide a time reference for subsequent position prediction calculations, ensuring that the prediction results cover the entire time window of interest.

[0048] When motion trend parameters are input into the position extrapolation module for recursive calculation, the position extrapolation module is a dedicated computational unit for calculating future positions based on these motion parameters. Motion trend parameters include the current position, current velocity, current acceleration, and motion pattern information extracted from trajectory fitting; these parameters constitute the input data for position extrapolation. The recursive calculation process employs numerical integration, progressively calculating the position information for each predicted time point starting from the current moment. The calculation process is based on the equations of motion in classical mechanics, considering the differential relationships between position, velocity, and acceleration. Velocity information is converted into position changes, and acceleration information is converted into velocity changes, through time integration. The recursive algorithm assumes that the motion parameters remain constant within each time step, then calculates the position change within that time step, accumulating the changes to the position at the previous moment to obtain the new position. The predicted distance value represents the predicted distance between the target and the radar at each time point, and the predicted angle value represents the angular azimuth of the target relative to the radar coordinate system; these two parameters constitute a position description in polar coordinates.

[0049] When performing spatial coordinate transformation on predicted distance and angle values ​​using the medical device coordinate positioning algorithm, the coordinate transformation is a mathematical transformation process that converts position information represented in polar coordinates to Cartesian coordinates. The medical device coordinate positioning algorithm establishes a three-dimensional coordinate system with the medical device as the origin. The X, Y, and Z axes of this coordinate system correspond to the east-west, north-south, and up-down directions in the space surrounding the device, respectively. The coordinate transformation process converts the distance and angle values ​​in polar coordinates into X, Y, and Z coordinate values ​​in Cartesian coordinates. This transformation process needs to consider the radar's installation position and attitude angle relative to the medical device. The algorithm first converts the distance and azimuth angle in polar coordinates into two-dimensional coordinates in the radar coordinate system, and then converts the radar coordinates to coordinates in the device coordinate system through coordinate system rotation and translation transformations. Spatial coordinate transformation also needs to consider the target's height information. If the radar has elevation angle measurement capabilities, the three-dimensional position information can be obtained directly; otherwise, it is necessary to assume that the target is moving within a specific height plane. The predicted three-dimensional position coordinates relative to the medical device provide a precise description of the target's position in the space surrounding the device. This description facilitates subsequent safe zone determination and collision detection analysis.

[0050] When performing smoothness verification on predicted 3D position coordinates using a trajectory continuity check algorithm, continuity checking is a crucial step to ensure the predicted trajectory is physically reasonable. The algorithm first checks whether the distance changes between adjacent predicted position points conform to the physical constraints of the target's motion. If the distance changes between adjacent points exceed the theoretical limit calculated based on the target's maximum possible velocity, the prediction result is considered discontinuous. Smoothness verification assesses the smoothness of the trajectory by calculating the curvature change; excessive curvature changes indicate unreasonable sharp turns or jumps in the trajectory. The algorithm also checks the velocity continuity of the trajectory, ensuring that the predicted velocity changes do not exceed the target's maximum acceleration limit. For personnel targets, the algorithm sets reasonable maximum velocity and maximum acceleration constraints. When discontinuous or unsmooth prediction points are detected, the algorithm uses interpolation smoothing to correct them, eliminating anomalies by recalculating the position coordinates near the problem points. Correction may employ linear interpolation, spline interpolation, or other smoothing algorithms to ensure that the corrected trajectory satisfies the continuity requirement while remaining as close as possible to the original prediction result. The revised set of predicted location points is reliable prediction data that has been verified for continuity and smoothness, and this data provides a physically reasonable trajectory basis for subsequent safety analysis.

[0051] When the corrected set of predicted location points is arranged and combined according to a time series, this process involves reorganizing the discrete location points according to chronological order. The algorithm first sorts the time labels of all predicted location points to ensure the location sequence is strictly arranged in chronological order, avoiding any impact from time misordering on subsequent analysis. The arrangement process also includes time interval standardization to ensure consistency between adjacent location points. If uneven time intervals are found, the algorithm uses interpolation to supplement missing time points or delete redundant ones. The combination process associates the location coordinates with corresponding time, velocity, and acceleration information to form a trajectory data structure. Each data element contains a timestamp, three-dimensional location coordinates, and a velocity vector. The algorithm also assigns a unique sequence number to each location point for easy data indexing and retrieval. The personnel location sequence within the predicted time window is the final trajectory prediction result. This sequence, indexed by time, records the expected positions of the target at various future moments. This location information provides detailed predictive data support for threat assessment and alarm decision-making in security monitoring systems.

[0052] In one specific embodiment, the process of executing step S106 may specifically include the following steps: The personnel movement trajectory and medical safety alarm trigger conditions are input into the threat cascade decision-making device for multi-level decision processing, resulting in primary decision results, intermediate decision results and advanced decision results. Based on the medical device safety grading standard, the primary, intermediate, and advanced judgment results are classified into threat levels to obtain classification labels for extremely high threat level, high threat level, and medium threat level. The safety threshold of medical equipment is adaptively corrected based on the dynamic threshold adjustment algorithm to obtain the real-time safety threshold parameters under the current environment. The threat level classification identifier is compared and matched with the real-time security threshold parameter to obtain the corresponding alarm output control command and alarm intensity parameter; The hierarchical alarm output control module generates and processes alarm output control commands and alarm intensity parameters to obtain a hierarchical medical safety alarm output signal based on radio ranging.

[0053] Specifically, when personnel movement trajectories and medical safety alarm triggering conditions are input into the threat cascade decision-making unit for multi-level decision processing, the threat cascade decision-making unit is a hierarchical decision-making system that ensures the accuracy and reliability of alarm decisions through multiple decision levels. The decision-making unit first receives personnel movement trajectory data, which includes the target's historical location information, current location information, and predicted location information, as well as key parameters such as movement speed and direction. Medical safety alarm triggering conditions serve as the standard input for decision-making, including multiple judgment conditions such as distance threshold, speed threshold, and approach angle threshold. The primary decision-making process performs rapid evaluation based on real-time distance and speed data. Specifically, when the target distance is less than 2 meters and the approach speed exceeds 1 meter per second, the primary decision outputs a threat warning signal. The judgment result is marked as 1 indicating the presence of a threat, and marked as 0 indicating no threat. The intermediate decision-making process considers the consistency of data across multiple consecutive measurement cycles. By analyzing the trajectory data of the most recent 5 to 10 measurement cycles, it verifies the persistence and stability of the threat. Only when at least 80% of the consecutive measurement cycles show the presence of a threat does the intermediate decision confirm the authenticity of the threat and output a confirmation signal. Advanced decision processing is based on long-term motion trend analysis of more than 30 seconds, comprehensively considering the target's movement intentions, behavioral patterns, and environmental factors. It uses complex logical reasoning to determine the severity and urgency of the threat. When the target's trajectory shows a continuous approach with positive acceleration, advanced decision processing outputs a high-threat confirmation signal. The three decision levels correspond to different confidence levels: primary decision has a confidence level of 60%, intermediate decision has a confidence level of 80%, and advanced decision has a confidence level of 95%, providing multi-dimensional decision-making basis for subsequent threat classification.

[0054] When classifying threat levels based on the medical device safety classification standards (primary, intermediate, and advanced), the standards are pre-defined rules for categorizing threat levels, determined by the importance of the medical device, the safety requirements of the surrounding environment, and the severity of potential accident consequences. The classification algorithm comprehensively analyzes the three judgment results and uses a decision fusion method to determine the final threat level. The criteria for an extremely high threat level are: all three judgment results indicate the presence of a threat and the target is expected to contact the device within 3 seconds; in this case, the primary judgment result is 1, the intermediate judgment result is 1, and the advanced judgment result is 1. This level requires immediate emergency response measures. The criteria for a high threat level are: at least two judgment results indicate the presence of a threat and the target continues to approach the device; for example, both primary and intermediate judgments are 1, or both intermediate and advanced judgments are 1. This level requires close monitoring and preparation for contingency measures. The criteria for a moderate threat level are: only one or two judgment results indicate a minor threat and the target's movement trend is unclear; for example, only the primary judgment is 1 while the other judgments are 0. This level requires vigilance but does not require immediate action. The threat level classification also considers target type identification. Personnel targets and object targets are classified using different standards. Large objects with a reflective cross-section exceeding 5 square meters are classified as having a higher threat level. The classification identifier is output in the form of a numerical code: 3 for extremely high threat level, 2 for high threat level, and 1 for medium threat level, facilitating subsequent automated processing and manual identification.

[0055] When applying a dynamic threshold adjustment algorithm to adaptively correct the safety thresholds of medical equipment, the dynamic threshold adjustment is an adaptive mechanism designed to adapt to constantly changing environmental conditions. The algorithm's input parameters include: background noise level quantized by signal-to-noise ratio (SNR), personnel density quantified by the number of targets per unit area, ambient temperature and relative humidity directly measured by temperature and humidity sensors, and equipment operating status indicators obtained through status monitoring. The algorithm's output parameters include: adjusted real-time distance thresholds and real-time speed thresholds. The algorithm continuously monitors changes in environmental parameters. Background noise level is quantified by calculating the SNR, which is equal to the ratio of signal power to noise power and converted to decibels. Electromagnetic interference intensity is quantified by measuring the energy integral of the interference spectrum. Temperature and humidity are directly measured by sensors, and personnel density is quantified by dividing the number of targets detected per unit time by the monitored area. The adaptive environmental correction process analyzes the impact of these environmental parameters on radar performance and dynamically adjusts the safety threshold parameters to maintain the system's detection performance. The specific adjustment rules are as follows: When the signal-to-noise ratio (SNR) corresponding to the background noise is higher than 20 dB, it is considered a low-noise environment. In this case, the algorithm adjusts the real-time distance threshold to a base distance threshold multiplied by a factor of 0.9, and the real-time speed threshold is also adjusted to a base speed threshold multiplied by a factor of 0.9. This improves system sensitivity, triggering alarms at greater distances or lower speeds to avoid missed detections. When the SNR is between 10 and 20 dB, it is considered a normal environment. The real-time distance threshold and the real-time speed threshold remain the same as the base distance threshold and do not require adjustment. When the SNR is lower than 10 dB, it is considered a high-noise environment. The algorithm adjusts the real-time distance threshold to a base distance threshold multiplied by a factor of 1.2, and the real-time speed threshold is also adjusted to a base speed threshold multiplied by a factor of 1.2. This appropriately raises the alarm threshold to avoid false alarms caused by noise. When the detected personnel density exceeds 5 people per 10 square meters, regardless of the current noise level, the algorithm multiplies the adjusted thresholds by a factor of 0.85 to further lower the alarm threshold and increase the alert level, preventing missed detections in high-density environments. Changes in temperature and humidity affect the propagation characteristics of electromagnetic waves. When the ambient temperature deviates from the standard temperature by 20 degrees Celsius, the propagation speed of electromagnetic waves changes. The algorithm calculates a correction factor based on the temperature difference at a rate of 0.01% per degree Celsius. The final distance measurement result needs to be multiplied by this correction factor, and the corresponding safety distance threshold is adjusted proportionally. The resulting real-time safety threshold parameters are a dynamic set of thresholds after environmental correction. These parameters can adapt to current environmental conditions, ensuring that the safety monitoring system maintains stable performance under various environments.

[0056] When comparing threat level classification identifiers with real-time security threshold parameters, this comparison and matching is a crucial step in determining specific alarm response measures. The process first compares the numerical code of the threat level classification identifier with the actual distance to the current target to determine if the current threat level exceeds the threshold requirement for the corresponding level. The matching algorithm employs multiple judgment conditions, comparing not only the numerical magnitude of the threat level but also additional factors such as the duration, trend, and scope of the threat. Specifically, the matching rules are as follows: When the threat level classification identifier is 3 (extremely high threat level), the algorithm checks if the target distance is less than 50% of the real-time distance threshold. If so, it generates the highest-level alarm output control command, coded as Command 3, requiring the immediate activation of all available alarm devices and emergency response procedures. Simultaneously, the alarm intensity parameter is set to 100% of full scale, representing maximum intensity. When the threat level classification identifier is 2 (high threat level), the algorithm checks if the target distance is less than 70% of the real-time distance threshold. If so, it generates a medium-level alarm output control command, coded as Command 2, requiring the activation of major alarm devices and notification of relevant personnel. The alarm intensity parameter is set to 60% of full scale, representing medium intensity. When the threat level classification is 1, i.e., medium threat level, the algorithm checks if the target distance is less than 100% of the real-time distance threshold (i.e., the full threshold). If so, a low-level alarm output control command is generated. This command is encoded as Command 1, requesting the activation of a basic alarm notification. The alarm strength parameter is set to 30% of the full scale, indicating a low intensity. The matching process also includes alarm time control, determining the alarm duration and repetition frequency based on the urgency of the threat. Alarms for extremely high threats are continuously output until the threat is resolved; alarms for high threats repeat every 5 seconds; and alarms for medium threats only output a single notification. The comparison and matching results directly affect the final alarm output effect, ensuring that different threat levels receive alarm responses of corresponding strengths.

[0057] When the hierarchical alarm output control module generates and processes signals based on alarm output control commands and alarm intensity parameters, it serves as the output interface for the entire alarm system, responsible for converting digitized control commands into actual alarm signals. The signal generation and processing selects appropriate output channels and devices based on the level requirements of the alarm output control commands, including various output methods such as audible alarms, visual indicators, vibration alerts, and data communication interfaces. The alarm intensity parameters control the operating intensity of various output devices. The volume of the audible alarm is linearly adjusted according to the intensity parameter: 90 dB for 100% intensity, 75 dB for 60% intensity, and 60 dB for 30% intensity. The frequency of the audible alarm also varies according to the threat level: 3000 Hz for extremely high threats, 2000 Hz for high threats, and 1000 Hz for medium threats. The brightness, color, and flashing pattern of the light signal indicator also change accordingly. Extremely high threat corresponds to a strong red light signal flashing continuously at a frequency of 5 times per second; high threat corresponds to a medium-strong orange light signal flashing intermittently at a frequency of 2 times per second; and medium threat corresponds to a constant, non-flashing yellow light signal. Signal generation also includes a multimedia alarm function, which uses voice synthesis technology to play specific threat information and response suggestions, such as "Warning, personnel are detected approaching rapidly. Please stop the equipment immediately." The display screen shows the location coordinates and movement trajectory of the threat target. Data communication alarms send alarm messages to the monitoring center and relevant personnel's mobile devices via a network interface. The messages include detailed information such as threat level, target location, approach speed, and estimated contact time. The medical safety classification alarm output signal based on radio ranging is the final system output. This signal integrates multiple forms such as sound, light, vibration, and data, enabling timely and accurate communication of security threat information to relevant personnel, providing reliable early warning protection for medical equipment.

[0058] The above describes the intelligent alarm method based on millimeter-wave radar ranging in the embodiments of this application. The following describes the intelligent alarm system based on millimeter-wave radar ranging in the embodiments of this application. Please refer to [link / reference]. Figure 2 One embodiment of the intelligent alarm system based on millimeter-wave radar ranging in this application includes: The ranging module is used to perform radio wave ranging scans of the safety area around medical equipment using frequency-modulated continuous wave millimeter-wave radar to obtain radio reflection signal data of the layered safety detection area; The superposition module is used to coherently superimpose the radio reflection signal data according to the medical safety ranging signal processing algorithm to obtain the equipment safety frequency domain signal; The calculation module is used to perform distance calculation processing on the safe frequency domain signal of the equipment through the medical equipment ranging frequency estimation algorithm to obtain the distance measurement results and approach speed data between the personnel and the equipment; The prediction module is used to perform security threat prediction processing on the distance measurement results based on the medical safety trajectory tracking algorithm to obtain the personnel movement trajectory and medical safety alarm triggering conditions. The judgment module is used to intelligently judge the movement trajectory of the person and the safety threshold of the medical equipment to obtain a medical safety graded alarm output signal based on radio ranging.

[0059] above Figure 2 The intelligent alarm system based on millimeter-wave radar ranging in the embodiments of the present invention will be described in detail from the perspective of modular functional entities. The intelligent alarm device based on millimeter-wave radar ranging in the embodiments of the present invention will be described in detail from the perspective of hardware processing.

[0060] Reference Figure 3 This invention also provides an intelligent alarm device based on millimeter-wave radar ranging. This intelligent alarm device can be a server, and its internal structure can be as follows: Figure 3 As shown, the intelligent alarm device based on millimeter-wave radar ranging includes a processor, memory, display screen, input device, network interface, and database connected via a system bus. The processor, designed as a computer, provides computational and control capabilities. The memory of the intelligent alarm device based on millimeter-wave radar ranging includes a non-volatile storage medium and internal memory. The non-volatile storage medium stores the operating system, computer programs, and database. The internal memory provides an environment for the operation of the operating system and computer programs in the non-volatile storage medium. The database of the intelligent alarm device based on millimeter-wave radar ranging is used to store the data corresponding to this embodiment. The network interface of the intelligent alarm device based on millimeter-wave radar ranging is used for communication with external terminals via a network connection. When the computer program is executed by the processor, it implements the above-described method.

[0061] Those skilled in the art will understand that Figure 3 The structure shown is merely a block diagram of a portion of the structure related to the present invention and does not constitute a limitation on the intelligent alarm device based on millimeter-wave radar ranging applied thereto.

[0062] The present invention also provides a computer-readable storage medium, which can be a non-volatile computer-readable storage medium or a volatile computer-readable storage medium, wherein the computer-readable storage medium stores instructions that, when the instructions are executed on a computer, cause the computer to perform the steps of the intelligent alarm method based on millimeter-wave radar ranging.

[0063] Those skilled in the art will clearly understand that, for the sake of convenience and brevity, the specific working processes of the systems and units described above can be referred to the corresponding processes in the foregoing method embodiments, and will not be repeated here.

[0064] If the integrated unit is implemented as a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, in essence, or the part that contributes to the prior art, or all or part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause an intelligent alarm device based on millimeter-wave radar ranging (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of the present invention. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0065] The above embodiments are only used to illustrate the technical solutions of the present invention, and are not intended to limit it. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. An intelligent alarm method based on millimeter-wave radar ranging, characterized in that, The method includes: Radio wave ranging scans of the safety area around medical equipment are performed using frequency-modulated continuous wave millimeter-wave radar to obtain radio reflection signal data of the layered safety detection area; The radio reflection signal data is coherently superimposed using a medical safety ranging signal processing algorithm to obtain a device safety frequency domain signal; The distance calculation and processing of the safe frequency domain signal of the equipment is performed by the medical equipment ranging frequency estimation algorithm to obtain the distance measurement results and approach speed data between the personnel and the equipment. The distance measurement results are processed for security threat prediction based on the medical safety trajectory tracking algorithm to obtain the personnel movement trajectory and medical safety alarm triggering conditions. The movement trajectory of the person is intelligently compared with the safety threshold of the medical equipment to obtain a medical safety graded alarm output signal based on radio ranging.

2. The intelligent alarm method based on millimeter-wave radar ranging according to claim 1, characterized in that, The method involves using frequency-modulated continuous wave millimeter-wave radar to perform radio wave ranging scans of the safety area surrounding the medical equipment, obtaining radio reflection signal data of the layered safety detection area, including: A linear frequency modulated signal is generated by a frequency modulated continuous wave millimeter-wave radar transmitter and transmitted, resulting in a millimeter-wave signal with sweep bandwidth and sweep period parameters. The millimeter-wave signal is transmitted directionally around the medical device through a multi-antenna array to obtain radio waves covering the near-field danger zone, the mid-field warning zone, and the far-field monitoring zone. The radio waves are reflected by the target object to obtain a reflected radio wave signal carrying distance and speed information. The reflected radio wave signal is input into the receiving antenna array for signal acquisition processing to obtain an intermediate frequency signal containing beat frequency information; The intermediate frequency signal is subjected to analog-to-digital conversion and digitization processing to obtain the radio reflection signal data of the layered security detection area.

3. The intelligent alarm method based on millimeter-wave radar ranging according to claim 1, characterized in that, The step of performing coherent superposition processing on the radio reflection signal data according to the medical safety ranging signal processing algorithm to obtain the device safety frequency domain signal includes: The radio reflection signal data is input into an I / Q demodulator for mixing to obtain a beat frequency signal containing target distance information; Based on the medical safety zone division rules, the beat frequency signal is processed for frequency band region identification. According to the frequency band where the beat frequency is located, the corresponding near-field danger zone, mid-field warning zone or far-field monitoring zone is identified. The beat frequency is proportional to the distance. The near-field danger zone corresponds to the low-frequency beat frequency, the mid-field warning zone corresponds to the mid-frequency beat frequency, and the far-field monitoring zone corresponds to the high-frequency beat frequency. The beat frequency signals of each region are subjected to phase compensation and amplitude superposition processing according to the multi-antenna coherent superposition formula to obtain the composite beat frequency signal after coherent superposition. The composite beat frequency signal is input into a fast Fourier transform (FFT) for frequency domain conversion to obtain the frequency domain amplitude spectrum and the frequency domain phase spectrum. The frequency domain amplitude spectrum and frequency domain phase spectrum are filtered and enhanced based on the medical safety frequency domain optimization algorithm to obtain the device safety frequency domain signal.

4. The intelligent alarm method based on millimeter-wave radar ranging according to claim 1, characterized in that, The process of performing distance calculation on the safe frequency domain signal of the device using a medical device ranging frequency estimation algorithm to obtain the distance measurement results and approach speed data between the person and the device includes: Based on the peak detection algorithm, the device's safe frequency domain signal is subjected to spectrum peak search processing to obtain the target frequency peak position and the corresponding frequency amplitude value; The target frequency peak position is input into a Chirp-Z converter for frequency refinement to obtain a high-resolution, accurate target frequency value. The precise target frequency value is converted into a distance value based on the millimeter-wave ranging formula to obtain the instantaneous distance between the personnel and the medical equipment. Based on the Doppler frequency shift extraction algorithm, the velocity component separation process of the device's safety frequency domain signal is performed to obtain the target radial velocity component and tangential velocity component. The instantaneous distance value and radial velocity component are fused to obtain the distance measurement result and approach speed data between the personnel and the equipment.

5. The intelligent alarm method based on millimeter-wave radar ranging according to claim 1, characterized in that, The process of performing security threat prediction on the distance measurement results based on the medical safety trajectory tracking algorithm to obtain the personnel movement trajectory and medical safety alarm triggering conditions includes: The distance measurement results and approach velocity data are input into a Kalman filter for state estimation processing to obtain a target state vector containing position, velocity and acceleration. Based on the cubic spline interpolation algorithm, the target state vector of multiple consecutive measurement cycles is subjected to trajectory fitting processing to obtain the personnel motion trajectory function and motion trend parameters; The trajectory prediction model is used to extrapolate the future positions of the personnel movement trajectory function to obtain the personnel position sequence within the prediction time window. The personnel location sequence is subjected to collision detection processing with the boundary of the medical equipment safety area to obtain potential collision time points and collision probability values; The collision probability value and motion trend parameters are comprehensively scored based on the threat assessment matrix to obtain the personnel movement trajectory and medical safety alarm triggering conditions.

6. The intelligent alarm method based on millimeter-wave radar ranging according to claim 5, characterized in that, The step of performing future position extrapolation processing on the personnel movement trajectory function based on the trajectory prediction model to obtain the personnel position sequence within the prediction time window includes: The trajectory prediction algorithm is used to perform time step segmentation on the personnel movement trajectory function to obtain multiple predicted time nodes within a preset time interval. The motion trend parameters are input into the position deduction module for recursive calculation to obtain the predicted distance and predicted angle values ​​corresponding to each predicted time node. The predicted distance and predicted angle values ​​are transformed into spatial coordinates based on the medical device coordinate positioning algorithm to obtain the predicted three-dimensional position coordinates relative to the medical device. The predicted three-dimensional position coordinates are smoothed using a trajectory continuity test algorithm to obtain a corrected set of predicted position points. The corrected set of predicted location points is arranged and combined according to the time series to obtain the personnel location sequence within the predicted time window.

7. The intelligent alarm method based on millimeter-wave radar ranging according to claim 1, characterized in that, The step of intelligently judging the movement trajectory of the personnel against the safety threshold of the medical equipment to obtain a medical safety graded alarm output signal based on radio ranging includes: The personnel movement trajectory and medical safety alarm triggering conditions are input into the threat cascade decision-making device for multi-level decision processing to obtain primary decision results, intermediate decision results and advanced decision results; Based on the medical device safety grading standard, the primary, intermediate, and advanced judgment results are classified into threat levels to obtain classification labels for extremely high threat level, high threat level, and medium threat level. The safety threshold of the medical device is adaptively corrected based on the environment according to the dynamic threshold adjustment algorithm to obtain the real-time safety threshold parameters under the current environment; The threat level classification identifier is compared and matched with the real-time security threshold parameter to obtain the corresponding alarm output control command and alarm intensity parameter; The hierarchical alarm output control module performs signal generation and processing on the alarm output control command and alarm intensity parameters to obtain the medical safety hierarchical alarm output signal based on radio ranging.

8. An intelligent alarm system based on millimeter-wave radar ranging, characterized in that, For implementing the intelligent alarm method based on millimeter-wave radar ranging as described in any one of claims 1-7, the intelligent alarm system based on millimeter-wave radar ranging comprises: The ranging module is used to perform radio wave ranging scans of the safety area around medical equipment using frequency-modulated continuous wave millimeter-wave radar to obtain radio reflection signal data of the layered safety detection area; The superposition module is used to coherently superimpose the radio reflection signal data according to the medical safety ranging signal processing algorithm to obtain the equipment safety frequency domain signal; The calculation module is used to perform distance calculation processing on the safe frequency domain signal of the equipment through the medical equipment ranging frequency estimation algorithm to obtain the distance measurement results and approach speed data between the personnel and the equipment; The prediction module is used to perform security threat prediction processing on the distance measurement results based on the medical safety trajectory tracking algorithm to obtain the personnel movement trajectory and medical safety alarm triggering conditions. The judgment module is used to intelligently judge the movement trajectory of the person and the safety threshold of the medical equipment to obtain a medical safety graded alarm output signal based on radio ranging.

9. An intelligent alarm device based on millimeter-wave radar ranging, characterized in that, The system includes a memory and a processor, wherein the memory stores a computer program that can run on the processor, and the processor executes the computer program to implement the intelligent alarm method based on millimeter-wave radar ranging as described in any one of claims 1 to 7.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is run by the processor, it causes the processor to execute the intelligent alarm method based on millimeter-wave radar ranging as described in any one of claims 1 to 7.

Images