An Adaptive Detection and Separation Method for Human Respiration and Heartbeat Signals in Fire Rescue

Abstract

This invention discloses an adaptive detection and separation method for human breathing and heartbeat signals in fire rescue. The invention includes software, hardware, and algorithms. It can automatically select the equipment's operating frequency band according to the needs of fire rescue, performing breathing detection after penetrating obstacles and when there are no obstacles or only a few obstacles. The detection function is implemented by controlling a USRP B210-based software-defined radar hardware through GNU Radio Companion software. Data preprocessing of the transmitted and received baseband signals yields a composite signal of superimposed human breathing and heartbeat signals. To address the difficulty in separating human breathing and heartbeat signals in software-defined radar detection, adaptive filtering and moving target display are applied to the composite signal to separate the breathing and heartbeat signals. The method includes: a human breathing and heartbeat signal detection method based on software-defined radar, and an adaptive human breathing and heartbeat signal separation method based on minimum mean square error moving target display.

Description

Technical Field

[0001] This invention relates to the field of radar systems and signal processing technology, specifically a software-defined radar-based adaptive detection and separation method for human respiratory and heartbeat signals used in fire rescue operations in the 1GHz to 6GHz frequency band. Background Technology

[0002] Human health can be reflected by many vital signs, among which the detection of respiration and heartbeat signals is extremely important. Traditional measurement methods are mostly contact-based. Contact-based methods have limitations; they cannot penetrate obstacles such as reinforced concrete walls, fires, and smoke. In recent years, non-contact vital sign detection technologies have developed rapidly, primarily based on radar detection.

[0003] In disaster relief environments, the penetration and environmental adaptability of detection technologies are of paramount importance. Targets requiring rescue may be buried several meters beneath obstacles, their vital signs extremely weak and difficult to detect. While millimeter-wave radar, with its short wavelength, can easily detect a human heartbeat, its poor penetration makes it unsuitable for effective firefighting and rescue applications. Classic radar equipment is limited by its hardware framework, resulting in high costs and complex structures. For example, once the hardware system of a traditional Doppler radar is built, it is difficult to adjust parameters such as operating frequency and waveform.

[0004] When detecting human vital signs, a composite signal of respiration and heartbeat is obtained. Therefore, the two signals should be separated during signal estimation. The respiration signal can usually be filtered out using a separate bandpass filter. However, the heartbeat signal is very weak, while the respiration signal is much stronger. The harmonics of the respiration signal and the heartbeat signal are prone to intermodulation, making the heartbeat signal difficult to separate. Summary of the Invention

[0005] To address the aforementioned problems, this invention provides an adaptive detection and separation method for human respiratory and heartbeat signals based on software-defined radar.

[0006] According to a first aspect of the present invention, a method for detecting human breathing and heartbeat signals based on software-defined radar is proposed. The method includes:

[0007] Step A: Firefighters assess the rescue environment to determine if there are any obstacles ahead. If there are obstacles, proceed to Step B; otherwise, proceed to Step D.

[0008] Step B: Utilize software-defined radar mode, i.e., the software GNU Radio Companion controls the hardware USRP B210 to realize radar functions; Step B1: Adjust the GNU Radio Companion to a suitable frequency band of 1GHz-2GHz for penetrating obstacles; Step B2: Control the USRP B210 to transmit and receive signals, during which the transmission carrier frequency band can be fine-tuned within the penetration frequency band as needed.

[0009] Step C: Determine whether there is a Doppler frequency that matches human physiological characteristics based on the echo of the software-defined radar. If the target is determined to exist, proceed to step D; otherwise, return to step A.

[0010] Step D: Adjust the GNU Radio Companion to a frequency band of 2GHz-6GHz suitable for human breathing and heartbeat detection. During this process, the transmission carrier frequency band can be fine-tuned as needed.

[0011] Step E: Perform data preprocessing on the transmitted baseband signal and the received baseband signal;

[0012] Step F involves preprocessing to obtain a composite signal that superimposes human respiratory and heartbeat signals.

[0013] The software-defined radar method (step B) of the above detection method is described in detail. The hardware includes a signal generation module and a signal transceiver module. The signal generation module, controlled by the GNU Radio Companion software, generates a digital baseband signal, which is transmitted to the signal transceiver module via a USB connection. The signal transceiver module receives data from the signal generation module using a USRP B210 software-defined radio device, up-converts the digital baseband signal, shifts it to radio frequency (RF), and transmits it through the transmitting antenna. The receiving antenna receives RF signals modulated by human respiration and heartbeat, down-converts them, and transmits them to the signal processing module via USB.

[0014] The above detection method step E is described in detail. The data preprocessing involves performing a series of signal processing steps on the transmitted baseband signal and the received baseband signal of the signal generation module to obtain a composite signal containing human breathing and heartbeat signals.

[0015] The preprocessing steps are as follows:

[0016] Step E1: Multiply the transmitted baseband signal and the received baseband signal.

[0017] Step E2 involves downsampling the multiplied signal.

[0018] Step E3: Perform DC compensation on the downsampled signal.

[0019] Step E4: Perform phase demodulation on the signal after DC compensation to obtain a composite signal containing human breathing and heartbeat signals.

[0020] According to a second aspect of the present invention, an adaptive method for separating human respiratory and heartbeat signals based on a minimum mean square error moving target display is proposed, which can effectively suppress the interference of human respiratory signals and their harmonics on heartbeat signals. The method steps are as follows:

[0021] Step one: The composite signal containing human breathing and heartbeat signals, which has been preprocessed by the first aspect of the present invention, is used as the signal input of the method.

[0022] Step two: Delay the input signal by one sampling interval to obtain the desired signal.

[0023] Step 3: The difference between the desired signal and the output signal when the filter has not reached its iterative optimum is used as the error signal. The filter weight vector is then iteratively updated using the least mean square algorithm.

[0024] Step four: Subtract the input signal from the iteratively optimal output signal.

[0025] Step 5: The subtracted signals are used to display the moving target and obtain the heartbeat signal.

[0026] Step six: After separating the human breathing and heartbeat signals, perform fast Fourier transform and independent bandpass filtering on the obtained breathing and heartbeat signals to complete the acquisition of time-domain and frequency-domain features of the measured breathing and heartbeat signals.

[0027] The advantages of this invention are:

[0028] 1. Software programming of the USRP B210 software-defined radio device can quickly realize a software-defined radar system, which can control the shape and frequency of the baseband signal waveform, as well as the gain of the transmitted and received signals, as needed, providing a low-cost and highly flexible solution.

[0029] 2. The proposed software-defined radar system operates in the 1GHz-6GHz frequency band. It can adjust the carrier frequency according to the rescue environment, automatically selecting low-frequency bands for penetrating wall-penetrating breathing detection and high-frequency bands for breathing and heartbeat detection. This can compensate for the poor penetration of current millimeter-wave radar.

[0030] 3. The proposed method for separating human respiratory and heartbeat signals is an adaptive method for separating human respiratory and heartbeat signals based on a minimum mean square error dynamic target display. The algorithm has low computational complexity and can effectively suppress the interference of human respiratory signals and their harmonics on heartbeat signals. Attached Figure Description

[0031] Figure 1 This is a block diagram of a method for detecting human breathing and heartbeat signals based on software-defined radar proposed in this invention.

[0032] Figure 2 Methods for defining radar hardware for the GNU Radio Companion software control software.

[0033] Figure 3 This is a schematic diagram of an adaptive human breathing and heartbeat signal separation method based on minimum mean square error dynamic target display proposed in this invention. Detailed Implementation

[0034] The first aspect of this invention proposes a method for detecting human breathing and heartbeat signals based on software-defined radar, such as... Figure 1 As shown, it specifically includes:

[0035] Step A: Firefighters assess the rescue environment to determine if there are any obstacles ahead. If there are obstacles, proceed to Step B; otherwise, proceed to Step D.

[0036] Step B utilizes software-defined radar mode, that is, the software GNU Radio Companion controls the hardware USRP B210 to implement radar functions. The processing method is as follows:

[0037] Step B1: Adjust the GNU Radio Companion to a frequency band of 70MHz-2GHz suitable for penetrating obstacles; within this frequency band, the wavelength range of the carrier signal is between 15 cm and 30 cm, which has strong obstacle penetration capability and can be used for wall detection.

[0038] Step B2: Control USRP B210 to transmit and receive signals. During this process, the transmission carrier frequency can be fine-tuned within the penetration band as needed.

[0039] Step C: Determine whether there is a Doppler frequency that matches human physiological characteristics based on the echo of the software-defined radar. If the target is determined to exist, proceed to step D; otherwise, return to step A.

[0040] Step D: Adjust the GNU Radio Companion to a frequency band of 2GHz-6GHz suitable for human breathing and heartbeat detection. During this process, the transmission carrier frequency band can be fine-tuned as needed.

[0041] Step E: Perform data preprocessing on the transmitted baseband signal and the received baseband signal;

[0042] Step F involves preprocessing to obtain a composite signal that superimposes human respiratory and heartbeat signals.

[0043] The method described above provides a detailed explanation of step B in the detection process. For example... Figure 2 As shown, the signal generation module control software GNU Radio Companion first generates the digital baseband signal for transmission:

[0044] T b (t)=cos(2πf b t) (1)

[0045] Where t is a time function, f b The frequency of the baseband signal.

[0046] Baseband data is transmitted to the USRP B210 transceiver module via USB 3.0. The FPGA-based transmit control module of the transceiver module controls the transmission of the baseband data. A digital upconversion (DUC) module upconverts the baseband data to an intermediate frequency (IF). The converted baseband signal is then converted to an analog signal by the USRP's digital-to-analog converter (DAC). After the DAC, a low-pass filter smooths the signal. The smoothed signal is then multiplied by a crystal oscillator. This ensures the signal is modulated to the set RF frequency, forming an RF signal. Finally, the RF signal is amplified by a power amplifier and transmitted via the transmitting antenna.

[0047] The signal emitted by the transmitting antenna can be represented as:

[0048]

[0049] Where f c Let φ(t) be the carrier frequency of the transmitted signal, and φ(t) represent the phase noise during data transmission. Because software-defined radio devices are affected by machine startup time when shifting frequencies, this introduces random phase interference, resulting in random phase noise.

[0050] Assume the distance between the human chest cavity and the antenna is:

[0051] x(t) = x0 + x b (t)+x h (t) (3)

[0052] Where x b (t) and x h (t) represent the minute displacements of the human chest cavity caused by respiration and heartbeat. x0 is the distance between the human chest cavity and the antenna, x b (t)+x h (t) << x0. Assume that the human body only performs breathing and heartbeat movements, and does not perform other limb movements.

[0053] According to the radar detection principle, the round-trip time of the signal between the human body and the antenna is:

[0054]

[0055] Where c is the speed of light. The echo signal is reflected after being modulated by the displacement of the human chest cavity. The echo signal received by the receiving antenna can be represented as...

[0056]

[0057] Where λ is the signal wavelength.

[0058] The echo signal is amplified by a low-noise amplifier and then multiplied by the USRP crystal oscillator. Following the same path as the transmitted signal, it is smoothed by a low-pass filter. Afterward, it undergoes analog-to-digital conversion (ADC) and is processed within the FPGA module. The processed data is then transmitted to the GNU Radio Companion via USB 3.0.

[0059] Input to GNU Radio Companion baseband signal R b (t) is represented as:

[0060]

[0061] Step E in the above detection method will be described in detail. For example... Figure 1 As shown, the data preprocessing steps are as follows:

[0062] Step E1, transmit baseband signal T b (t) and the received baseband signal R b (t) is multiplied to obtain information carrying human vital signs. The system baseband signal containing human respiratory and heartbeat signals can then be represented as:

[0063]

[0064] in, The phase shift is constant. Δφ(t) is the residual phase noise, which can be expressed as:

[0065]

[0066] The frequency modulation during USRP signal transmission and reception introduces random phase noise, so Δφ(t) is also random. However, human respiratory rate and heart rate are measured by detecting R... b (t) and T b The frequency of the phase change between (t) is used for calculation, and it is independent of the initial value of the phase of each signal. Therefore, the influence of Δφ(t) can be temporarily ignored.

[0067] Step E2 involves downsampling the multiplied signal. The signal processed in step E1 no longer contains T. b(t) and R b The frequency f of (t) b Therefore, the current sampling rate is oversampled relative to the frequencies of the breathing and heartbeat signals, and downsampling is required. Downsampling filters out high-frequency interference and reduces the total amount of data, which is beneficial for computation.

[0068] Further decompose B(t) into in-phase component B I (t) and orthogonal component B Q (t)

[0069]

[0070] Where b I and b Q This is a DC offset.

[0071] Step E3 involves DC compensation of the downsampled signal. The first step uses nonlinear least squares estimation (NLLS) to estimate the center and radius of the signal constellation. The second step, after algebraic simplification, transforms the problem into linear least squares estimation (LLSE) for DC compensation. DC compensation eliminates the fixed phase caused by the distance x0 between the human body and the radar, as well as other DC interference, improving the accuracy of subsequent phase demodulation.

[0072] Step E4 involves demodulating the DC-compensated signal to obtain a composite signal containing both respiratory and heartbeat signals. Ideally, the chest cavity displacement caused by respiration and heartbeat can be represented by phase demodulation.

[0073]

[0074] In a second aspect, the present invention proposes an adaptive method for separating human respiratory and heartbeat signals based on a moving target display with minimum mean square error.

[0075] like Figure 3 As shown, the method steps are as follows:

[0076] For ease of analysis, the time function t in formula (10) is converted to the sampling time n:

[0077] x(n)=x b (n)+x h (n)+x noise (n) (11)

[0078] Step one: The signal x(n) after signal preprocessing according to the first aspect of the present invention is used as the signal input of the method of the second aspect of the present invention. This signal also includes the respiratory signal x. b (n) and heartbeat signal x h(n).

[0079] Step 2: Delay the input signal x(n) by a sampling interval Tr to obtain the desired signal d(n).

[0080] d(n)=x(n+T r (12)

[0081] Step 3: The desired signal d(n) and the output signal when the filter has not reached its iterative optimum. The difference is taken as the error signal e(n).

[0082]

[0083] The filter weight vector is iteratively updated using the least mean square adaptive algorithm.

[0084]

[0085] Where μ is the step size.

[0086] In this invention, the number of iterations of the filter is set to the length of the input signal x(n) minus the number of filter taps. After iteration to the optimal value, the filter outputs the signal y(n).

[0087] Step four: Subtract the input signal x(n) from the iterative optimal output signal y(n), and denote the result as h(n):

[0088] h(n) = x(n) - y(n) (15)

[0089] Step 5: Further processing h(n) with moving target visualization yields the heartbeat signal x obtained by the algorithm. h (n)

[0090] x h (n) = h(n+1) - h(n) (16).

Claims

1. A method for detecting human breathing and heartbeat signals during fire rescue, characterized in that, The specific steps are as follows: Step A: Determine if there are obstacles ahead based on the rescue environment. If there are obstacles, proceed to Step B; otherwise, proceed to Step D. Step B utilizes a software-defined radar approach, where the software GNU Radio Companion controls the hardware USRP B210 to implement radar functionality: Step B1, adjusts the GNU Radio Companion to a suitable obstacle-penetrating frequency band of 1GHz-2GHz; Step B2, controls the USRP B210 to transmit and receive signals, during which the transmit carrier frequency band can be fine-tuned within the penetration frequency band as needed. Step C: Determine whether there is a Doppler frequency that matches human physiological characteristics based on the echo of the software-defined radar. If the target is determined to exist, proceed to step D; otherwise, return to step A. Step D: Adjust the GNU Radio Companion to a frequency band of 2GHz-6GHz suitable for human breathing and heartbeat detection. During this process, the transmission carrier frequency band can be fine-tuned as needed. Step E: Perform data preprocessing on the transmitted baseband signal and the received baseband signal; Step F: After preprocessing, a composite signal of human respiration and heartbeat signals is obtained; The hardware includes a signal generation module and a signal transceiver module; the signal generation module is controlled by the software GNU RadioCompanion to generate a digital baseband signal, represented as follows: ,in For baseband signal frequency, The baseband signal is a function of time. The signal is transmitted via USB cable to the signal transceiver module. The signal transceiver module receives data from the signal generation module using a USRP B210 software-defined radio device, performs up-conversion processing on the digital baseband signal, transfers it to radio frequency, and then transmits it via the transmitting antenna. This is represented as follows: ,in For transmitting signal carrier frequency, This represents phase noise during data transmission; the radio frequency signal modulated by human breathing and heartbeat received by the receiving antenna is represented as... ,in For the signal wavelength, This refers to the round-trip time of the signal between the human body and the antenna. The distance between the human body and the antenna. This is a breathing signal. This is a heartbeat signal. The speed is the speed of light; after down-conversion, the baseband signal is transmitted via USB to the GNU Radio Companion receiver, and is represented as follows: ; Step E specifically includes: Step E1: Multiply the transmitted baseband signal and the received baseband signal to obtain a signal containing human vital signs. ,Right now ,in, For constant phase shift; Step E2: Downsample the multiplied signal; Step E3: Perform DC compensation on the downsampled signal; Step E4: Perform phase demodulation on the signal after DC compensation to obtain a composite signal containing human breathing and heartbeat signals. ,in For noise, for In-phase components, for The orthogonal components.

2. A method for separating human breathing and heartbeat signals in fire rescue based on the detection method described in claim 1, characterized in that, The method steps are as follows: Step 1: Combine the signals containing human breathing and heartbeat signals. As a signal input to the method; Step two, input signal Delay one sampling interval As expected signal ,Right now ; Step 3, Expected Signal The output signal of the filter that did not reach the iterative optimum The difference serves as an error signal. ,Right now The filter weight vector is iteratively updated using the least mean square algorithm. ,in Step size, For the filter weight vector, Error signal transpose; Step 4, Input Signal With the iterative optimal output signal Subtract and record as ,Right now ; Step 5, for Heartbeat signal obtained from moving target display ,Right now .

Images