CN105816165A - Real-time dynamic heart rate monitoring device and monitoring method - Google Patents

Abstract

The invention discloses a real-time dynamic heart rate monitoring device and a monitoring method, wherein the device comprises a baseline drift elimination module, a band-pass filtering module, a frequency domain analysis module and a heart rate frequency point selection module; the baseline wander elimination module is used for eliminating interference signals causing wander of a baseline line of the pulse wave signals; the band-pass filtering module is used for acquiring frequency signal components belonging to a heart rate frequency band and eliminating noise signals outside the heart rate frequency band; the frequency domain analysis module is used for acquiring a signal frequency spectrum of the pulse wave signal in a preset specific frequency band interval; the heart rate frequency point selection module is used for acquiring a frequency point corresponding to the heart rate and outputting the frequency point. The invention eliminates the influence of human body movement and muscle activity on heart rate analysis, reduces hardware cost, can flexibly select the concerned frequency band interval, improves the frequency domain precision of signals only in the concerned frequency band interval, and avoids causing excessive calculation and storage expenses.

Description

A real-time dynamic heart rate monitoring device and monitoring method

technical field

The invention relates to the field of electronics, in particular to a real-time dynamic heart rate monitoring device and a monitoring method.

Background technique

With the application and development of mobile Internet technology in the medical and health field, a large number of different forms of smart watches, smart bracelets, and smart wristbands have appeared on the market, capable of measuring heart rate, blood pressure, blood oxygen concentration, respiratory rate, etc. Parameters of wearable mobile health products. Heart rate is defined as the number of heartbeats per minute of the human body, and is an important medical routine physiological parameter for evaluating the health status of a person. Heart rate monitoring is of great significance for disease risk early warning, disease diagnosis, and annual routine physical examination. In particular, sports such as fitness activities and outdoor running have extensive application requirements for real-time dynamic heart rate monitoring.

At present, the principle of heart rate monitoring technology adopted by most mobile health products is photoelectric transmission measurement. In the hardware design of the product, the sensor in contact with the human skin will emit a beam of light to hit the skin, and measure the light reflected or transmitted through the skin at the same time. Because the blood absorbs light of a specific wavelength, the process of heart pumping directly affects the change of the light signal intensity measured by the sensor. The hardware records the signal intensity change according to the set sampling rate and collects the original data, that is, the pulse wave signal. The data analysis software unit runs the heart rate monitoring algorithm to process the pulse wave signal, and outputs the heart rate value. Heart rate monitoring algorithm is the key core technology of heart rate measurement products, which determines the accuracy and reliability of heart rate measurement values. In practical applications, heart rate monitoring includes static heart rate monitoring and real-time dynamic heart rate monitoring. The latter has a wider application space, but also poses a great challenge to the existing technology.

Through actual tests, it is found that the current technical status of the dynamic heart rate monitoring algorithm of most mobile health products is traditional signal time-domain waveform analysis or signal frequency domain analysis combined with accelerometer readings to assist judgment, real-time monitoring of human heart rate in the state of exercise. The following disadvantages are exposed in an application scenario. First, due to noise interference, the feature points on the signal time-domain waveform are not obvious, which will cause the algorithm to fail to obtain complete input information; second, there are too many criteria for signal time-domain waveform matching, and it is difficult to set the specific value of the algorithm parameters ; Third, for embedded modules, the computational complexity of the waveform matching algorithm is large; Fourth, the traditional signal frequency domain analysis method will increase the calculation and data storage overhead when improving the frequency spectrum accuracy; Fifth, the accelerometer This increases hardware costs and also increases resource overhead in computing, storage, and energy consumption.

Therefore, the prior art still needs to be improved and developed.

Contents of the invention

In view of the deficiencies in the prior art, the purpose of the present invention is to provide a real-time dynamic heart rate monitoring device and a monitoring method, aiming at solving the problem of high computational complexity of the algorithm, high hardware cost, and computational complexity of the dynamic heart rate monitoring device in the prior art when monitoring heart rate , storage, and energy resource overhead are large drawbacks.

Technical scheme of the present invention is as follows:

A real-time dynamic heart rate monitoring device, wherein the device includes a baseline drift elimination module, a band-pass filter module, a frequency domain analysis module, and a heart rate frequency point selection module;

The baseline drift elimination module is used to eliminate interference signals that cause the baseline of the pulse wave signal to drift;

The band-pass filter module is used to obtain frequency signal components belonging to the heart rate frequency band while eliminating noise signals outside the heart rate frequency band;

The frequency domain analysis module is used to obtain the signal spectrum of the pulse wave signal in a preset specific frequency band interval;

The heart rate frequency point selection module is used to obtain the frequency point corresponding to the heart rate and output the frequency point;

The baseline drift elimination module is connected to the band-pass filter module, the band-pass filter module is connected to the frequency domain analysis module, and the frequency domain analysis module is also connected to the heart rate frequency point selection module.

The real-time dynamic heart rate monitoring device, wherein the baseline drift elimination module specifically includes a baseline drift trend item signal extraction unit and a signal linear superposition unit;

The baseline drift trend item signal extraction unit is used to obtain a baseline drift trend item signal of the same length as the original pulse wave signal;

The signal linear superposition unit is used to subtract the baseline drift trend item signal from the original pulse wave signal to obtain the pulse wave signal with the baseline drift removed;

The baseline drift trend item signal extraction unit is connected to the signal linear superposition unit.

The real-time dynamic heart rate monitoring device, wherein the bandpass filter module specifically includes a filter parameter setting unit and a filter unit,

The filter parameter setting unit is used to set the lower limit and upper limit frequency of the passband, the lower limit and upper limit frequency of the stopband, the attenuation coefficient in the passband, and the attenuation coefficient in the stopband;

The filter unit is used to filter the pulse wave signal after the baseline drift is eliminated by the baseline drift elimination module after obtaining the order and coefficient of the filter module, and obtain the pulse wave signal after the noise is removed;

The filtering parameter setting unit is connected to the filtering unit.

The real-time dynamic heart rate monitoring device, wherein the frequency domain analysis module specifically includes a specific frequency band interval setting unit, a frequency point power calculation unit and a signal spectrum acquisition unit;

The specific frequency band interval setting unit is used to set the start frequency point, the end frequency point, and the number of frequency point subdivisions of the specific frequency band interval;

The frequency point power calculation unit is used to calculate the real part of the frequency domain signal with the first transformation polynomial and the imaginary part of the frequency domain signal with the second transformation polynomial for each frequency point of the specific frequency band interval, according to the frequency domain signal Calculate the real and imaginary parts of the pulse wave signal at the frequency;

The signal spectrum acquisition unit is used to acquire the power of the pulse wave signal at each frequency point in a specific frequency band interval, and generate the signal spectrum of the pulse wave signal in the entire frequency band interval of interest after superposition;

The frequency point power calculation unit is respectively connected with the specific frequency band interval setting unit and the signal spectrum acquisition unit.

The real-time dynamic heart rate monitoring device, wherein the heart rate frequency point selection module specifically includes a heart rate frequency point selection unit and a heart rate frequency point dynamic tracking unit,

The heart rate frequency point selection unit is used to obtain the frequency point where the peak in the pulse wave signal spectrum is located as the heart rate frequency point in the initial state;

The heart rate frequency point dynamic tracking unit is used to perform high-precision frequency domain analysis in a spectrum range of a specific width around the center with the heart rate frequency point output from the previous tracking cycle as the center within the preset fixed tracking cycle, and select The frequency point where the spectrum peak is located is output as the heart rate frequency point;

The heart rate frequency point selection unit is connected with the heart rate frequency point dynamic tracking unit.

The real-time ambulatory heart rate monitoring device, wherein, the first transformation polynomial and the second transformation polynomial are orthogonal polynomials.

The real-time dynamic heart rate monitoring device, wherein the first transformation polynomial is one of Legendre polynomials, Jacobi polynomials, Laguerre polynomials, Chebyshev polynomials, and Hermite polynomials;

The second transformation polynomial is an orthogonal polynomial and is one of Legendre polynomials, Jacobi polynomials, Laguerre polynomials, Chebyshev polynomials, and Hermitian polynomials.

A monitoring method based on the real-time dynamic heart rate monitoring device, wherein the method comprises the steps of:

A. Obtain the pulse wave signal, and eliminate the baseline drift interference signal through the baseline drift elimination module;

B. The band-pass filter module filters the signal output by the baseline drift elimination module, and retains the signal components belonging to the heart rate frequency band;

C. The frequency domain analysis module acquires the signal spectrum of the pulse wave signal in the preset specific frequency band interval;

D. The heart rate frequency point selection module obtains and outputs the frequency point corresponding to the heart rate according to the signal frequency.

The real-time dynamic heart rate monitoring method, wherein, the step A specifically includes:

A1. Obtain the baseline drift trend item signal with the same length as the original pulse wave signal by signal filtering method or curve fitting method;

A2. Store the original pulse wave signal and the baseline drift trend item signal in a row vector or column vector, and then subtract the baseline drift trend item information from the original pulse wave signal according to the matrix addition rule to obtain the pulse wave signal with the baseline drift removed.

The real-time dynamic heart rate monitoring method, wherein, the step B specifically includes:

B1. Obtain the preset lower limit and upper limit frequency of the passband, the lower limit and upper limit frequency of the stopband, the attenuation coefficient in the passband, and the attenuation coefficient in the stopband;

B2. After obtaining the order and coefficient of the filtering module, filter the pulse wave signal after the baseline drift is eliminated by the baseline drift elimination module, and obtain the pulse wave signal after the noise is removed.

The present invention provides a real-time dynamic heart rate monitoring device and a monitoring method. The present invention can eliminate the influence of human body movement and muscle activity on heart rate analysis, reduce hardware costs, and can flexibly select the frequency range of interest, and only improve the heart rate within the frequency range of interest. The frequency domain accuracy of the signal avoids excessive calculation and storage overhead, and the technical route of signal spectrum analysis is selected to avoid the problems of finding the characteristic points of the time domain signal waveform and determining the specific value of the algorithm parameters.

Description of drawings

Fig. 1 is a functional principle block diagram of a preferred embodiment of a real-time dynamic heart rate monitoring device of the present invention.

Fig. 2 is a flowchart of a preferred embodiment of a monitoring method of a real-time dynamic heart rate monitoring device of the present invention.

detailed description

In order to make the object, technical solution and effect of the present invention more clear and definite, the present invention will be further described in detail below. It should be understood that the specific embodiments described here are only used to explain the present invention, not to limit the present invention.

The present invention also provides a functional schematic diagram of a preferred embodiment of a real-time dynamic heart rate monitoring device, as shown in Figure 1, wherein the device includes a baseline drift elimination module 100, a bandpass filter module 200, a frequency domain analysis module 300, Heart rate frequency selection module 400;

The baseline drift elimination module 100 is used to eliminate the interference signal that causes the baseline of the pulse wave signal to drift; the bandpass filter module 200 is used to obtain frequency signal components belonging to the heart rate frequency band while eliminating noise signals outside the heart rate frequency band; The frequency domain analysis module 300 is used to obtain the signal spectrum of the pulse wave signal in a preset specific frequency range; the heart rate frequency point selection module 400 is used to obtain the frequency point corresponding to the heart rate and output the frequency point; the baseline drift elimination The module 100 is connected to the bandpass filter module 200 , the bandpass filter module 200 is connected to the frequency domain analysis module 300 , and the frequency domain analysis module 300 is also connected to the heart rate frequency point selection module 400 .

During specific implementation, the baseline drift elimination module specifically includes a baseline drift trend item signal extraction unit and a signal linear superposition unit;

The baseline drift trend item signal extraction unit is used to obtain a baseline drift trend item signal of the same length as the original pulse wave signal;

The signal linear superposition unit is used to subtract the baseline drift trend item signal from the original pulse wave signal to obtain the pulse wave signal with the baseline drift removed;

The baseline drift trend item signal extraction unit is connected to the signal linear superposition unit.

During the specific implementation, the baseline of the pulse wave signal will drift up and down because of breathing, body movement or movement. Essentially, this baseline drift is a low-frequency signal that will interfere with the judgment of key information. The main function of the elimination of baseline drift module is to reduce the baseline drift interference caused by respiration and exercise on the pulse wave signal, and to avoid dependence on the auxiliary information of the accelerometer.

The module for eliminating baseline drift consists of two functional units: a baseline drift trend item signal extraction unit, and a signal linear superposition unit.

The baseline drift trend item signal extraction unit has two specific implementation methods, the first type is a signal filtering method, and the second type is a curve fitting method. The specific implementation of the signal filtering method is median filtering or mean filtering, using a sliding window to traverse the pulse wave signal, calculating the median or mean value of all signal values in the window, and finally obtaining a baseline drift trend item with the same length as the original pulse wave signal Signal. The specific implementation of the curve fitting method considers the baseline drift signal as a time function that can be expressed as a high-order polynomial, uses the pulse wave signal as an input data sample, and adopts a nonlinear fitting method to obtain the coefficients of the high-order polynomial, Finally, the baseline drift trend item signal with the same length as the original pulse wave signal is obtained through polynomial function calculation.

The signal linear superposition unit, the specific implementation method is to store the original pulse wave signal and the baseline drift trend item signal in a row vector or column vector, and then subtract the baseline drift trend item signal from the original pulse wave signal according to the matrix addition rule, and finally get the baseline drift signal Drifting pulse wave signal.

Further, the bandpass filtering module specifically includes a filtering parameter setting unit and a filtering unit,

The filter parameter setting unit is used to set the lower limit and upper limit frequency of the passband, the lower limit and upper limit frequency and the upper limit frequency of the stopband, the attenuation coefficient in the passband, and the attenuation coefficient in the stopband;

The filter unit is used to filter the pulse wave signal after the baseline drift is eliminated by the baseline drift elimination module after obtaining the order and coefficient of the filter module, and obtain the pulse wave signal after the noise is removed;

The filtering parameter setting unit is connected to the filtering unit.

During specific implementation, a large number of medical practices have shown that the range of the human heart rate is generally 40 beats/minute to 220 beats/minute, which indicates that the frequency range of the heart rate signal is 0.6 Hz to 3.7 Hz. Therefore, the signal frequency bands from 0 Hz to 0.6 Hz and above 3.7 Hz can be regarded as noise frequency bands. These noises include glitch noise caused by sensor circuits, noise introduced by human activities, and so on. The main function of the band-pass filter module is to retain only the signal components belonging to the heart rate frequency band, and at the same time eliminate the interference of out-of-band noise to the detection of the heart rate signal.

Band-pass filter module, the specific embodiment can adopt Butterworth filter, by setting passband lower limit and upper limit frequency, stopband lower limit and upper limit frequency, passband attenuation coefficient, stopband attenuation coefficient, refer to Butterworth The filter design process calculates the filter order and coefficients. The pulse wave signal from which the baseline drift has been removed is input into the band-pass filter module, and finally a pulse wave signal from which noise has been removed is obtained.

In the specific implementation, it is not ruled out that there are heart rate signals in extreme cases (disease, strenuous exercise). In order to ensure that heart rate changes are fully monitored, the dynamic range of the heart rate frequency band can be appropriately increased, for example, set to 0.5Hz to 4Hz; In addition to the Butterworth filter, Chebyshev filter and the like can also be selected for the implementation of the pass filter.

In a further embodiment, the frequency domain analysis module specifically includes a specific frequency band interval setting unit, a frequency point power calculation unit, and a signal spectrum acquisition unit;

The specific frequency band interval setting unit is used to set the start frequency point, the end frequency point, and the number of frequency point subdivisions of the specific frequency band interval;

The frequency point power calculation unit is used to calculate the real part of the frequency domain signal with the first transformation polynomial and the imaginary part of the frequency domain signal with the second transformation polynomial for each frequency point of the specific frequency band interval, according to the frequency domain signal Calculate the real and imaginary parts of the pulse wave signal at the frequency;

The signal spectrum acquisition unit is used to acquire the power of the pulse wave signal at each frequency point in a specific frequency band interval, and generate the signal spectrum of the pulse wave signal in the entire frequency band interval of interest after superposition;

The frequency point power calculation unit is respectively connected with the specific frequency band interval setting unit and the signal spectrum acquisition unit.

In specific implementation, the pulse wave signal from which the noise is removed is a time domain signal. In order to extract heart rate information more accurately, it is necessary to use a signal transformation method to obtain the frequency domain signal corresponding to the pulse wave signal, and to ensure that the signal spectrum has a sufficiently high frequency domain accuracy. The classic time-frequency signal transformation method is the fast Fourier transform or FFT. Its disadvantage is that in order to improve the frequency domain accuracy of the signal, it needs to pay the calculation and storage overhead that the embedded device cannot bear. The main function of the high-precision frequency domain analysis module is to flexibly select the frequency band interval of interest, and improve the frequency domain accuracy of the signal only in the frequency band interval of interest, avoiding excessive calculation and storage overhead.

The specific implementation of the high-precision frequency domain analysis module is to set the starting frequency point f1, the ending frequency point f2, and the number of subdivided frequency points Nf of the frequency band interval of interest.

Use q(n), n=0, 1, . . . , N-1 to represent a pulse wave signal sequence with a data length of N. For each frequency point f in the frequency band interval of interest, the real part of the frequency domain signal is calculated using the following transformation polynomial,

Qr=q(0)+q(1)*cos(f)+q(2)*T(2)+…+q(N-1)*T(N-1)

and compute the imaginary part of the frequency-domain signal using the following transformation polynomial,

Qi=-q(1)*sin(f)–q(2)*sin(f)*U(2)-…-q(N-1)*sin(f)*U(N-1)

T(n) is a first transformation polynomial, U(n) is a second transformation polynomial, and the first transformation polynomial and the second transformation polynomial are orthogonal polynomials. The first transformation polynomial is one of Legendre polynomials, Jacobi polynomials, Laguerre polynomials, Chebyshev polynomials, and Hermitian polynomials; the second transformation polynomial is an orthogonal polynomial that is Legendre One of polynomials, Jacobi polynomials, Laguerre polynomials, Chebyshev polynomials, and Hermite polynomials.

According to Qr and Qi, the power of the pulse wave signal at the frequency point f can be calculated.

After each frequency point in the frequency band of interest is calculated above, the signal spectrum of the pulse wave signal in the entire frequency band of interest can be obtained. The quantitative index of high-frequency domain accuracy is related to the number of frequency subdivisions Nf, and the greater the number of frequency subdivisions, the higher the frequency domain accuracy.

Further, the heart rate frequency point selection module specifically includes a heart rate frequency point selection unit and a heart rate frequency point dynamic tracking unit,

The heart rate frequency point selection unit is used to obtain the frequency point where the peak in the pulse wave signal spectrum is located as the heart rate frequency point in the initial state;

The heart rate frequency point dynamic tracking unit is used to perform high-precision frequency domain analysis in a spectrum range of a specific width around the center with the heart rate frequency point output from the previous tracking cycle as the center within the preset fixed tracking cycle, and select The frequency point where the spectrum peak is located is output as the heart rate frequency point;

The heart rate frequency point selection unit is connected with the heart rate frequency point dynamic tracking unit.

In the specific implementation, after obtaining the signal spectrum of the pulse wave signal in the attention frequency range, the heart rate value cannot be calculated directly, but the heart rate value needs to be found through analysis and judgment to find the frequency point representing the heart rate, which is the selection of the heart rate frequency point The role of the module.

The heart rate frequency point selection module is specifically implemented in two stages. The first stage is the precise selection of heart rate frequency points in the initial state; the second stage is dynamic tracking of heart rate frequency points.

Precise heart rate frequency point selection in the initial state. The specific implementation method is to ensure that the human body is in a quiet state, and obtain the pulse wave signal spectrum by eliminating the baseline drift module, band-pass filter module and high-precision frequency domain analysis module. The position is the frequency point representing the heart rate, and the heart rate value can be calculated.

Heart rate frequency point dynamic tracking, the specific implementation method is to cycle with a fixed tracking period, and in each tracking period, the pulse wave signal with noise removed is obtained by eliminating the baseline drift module and the band-pass filter module, and output in the previous tracking period The heart rate frequency point is the center, and high-precision frequency domain analysis is performed in a narrow width spectrum range around the center, and the frequency point where the spectrum peak is located is selected as the heart rate frequency point output.

The present invention also provides a preferred embodiment of a monitoring method based on the real-time dynamic heart rate monitoring device, as shown in Figure 2, wherein the method includes:

Step S100, acquire the pulse wave signal, and eliminate the baseline drift interference signal through the baseline drift elimination module;

Step S200, the bandpass filter module filters the signal output by the baseline drift elimination module, and retains the signal components belonging to the heart rate frequency band;

Step S300, the frequency domain analysis module acquires the signal spectrum of the pulse wave signal within a preset specific frequency band interval;

Step S400, the heart rate frequency point selection module obtains and outputs the frequency point corresponding to the heart rate according to the signal frequency.

In the specific implementation, the high-frequency noise mixed in the signal acquisition process is removed through a variety of filter techniques, the signal baseline drift caused by muscle shaking and breathing is removed, and a high-precision signal frequency domain analysis method with less computational complexity than FFT is adopted. , select and locate the frequency point representing the heart rate from the signal spectrum, and monitor the heart rate dynamically in real time. The hardware cost and resource overhead caused by the accelerometer is reduced. Second, it avoids the problems of finding the characteristic points of the time-domain signal waveform and determining the specific values of the algorithm parameters. Third, the computational complexity of the algorithm can be borne by ordinary embedded modules. The specific monitoring method is as described in the specific embodiment of the monitoring device above.

Further, the step S100 specifically includes:

Step S101. Obtain a baseline drift trend item signal with the same length as the original pulse wave signal by signal filtering method or curve fitting method;

Step S102: Store the original pulse wave signal and the baseline drift trend item signal in a row vector or column vector, and then subtract the baseline drift trend item information from the original pulse wave signal according to the matrix addition rule to obtain the pulse wave signal with the baseline drift removed.

During specific implementation, the specific implementation of the signal filtering method in the step S101 is median filtering or mean filtering, using a sliding window to traverse the pulse wave signal, calculating the median or mean value of all signal values in the window, and finally obtaining the same value as the original A baseline drift trend term signal equal in length to the pulse wave signal. The specific implementation of the curve fitting method considers the baseline drift signal as a time function that can be expressed as a high-order polynomial, uses the pulse wave signal as an input data sample, and adopts a nonlinear fitting method to obtain the coefficients of the high-order polynomial, Finally, the baseline drift trend item signal with the same length as the original pulse wave signal is obtained through polynomial function calculation.

In a further embodiment, the step S200 specifically includes:

Step S201, obtaining the preset passband lower limit and upper limit frequency, stopband lower limit and upper limit frequency, passband attenuation coefficient, and stopband attenuation coefficient;

Step S202, after obtaining the order and coefficient of the filtering module, filter the pulse wave signal after the baseline drift is eliminated by the baseline drift elimination module, and obtain the pulse wave signal after the noise is removed.

In the specific implementation, the Butterworth filter is used, by setting the lower limit and upper limit frequency of the pass band, the lower limit and upper limit frequency of the stop band, the attenuation coefficient in the pass band, and the attenuation coefficient in the stop band, and referring to the design process of the Butterworth filter to calculate the filter order and coefficients. The pulse wave signal from which the baseline drift has been removed is input into the band-pass filter module, and finally a pulse wave signal from which noise has been removed is obtained.

In the specific implementation, it is not ruled out that there are heart rate signals in extreme cases (disease, strenuous exercise). In order to ensure that heart rate changes are fully monitored, the dynamic range of the heart rate frequency band can be appropriately increased, for example, set to 0.5Hz to 4Hz; In addition to the Butterworth filter, Chebyshev filter and the like can also be selected for the implementation of the pass filter.

In summary, the present invention provides a real-time dynamic heart rate monitoring device and monitoring method. The device includes a baseline drift elimination module, a bandpass filter module, a frequency domain analysis module, and a heart rate frequency point selection module; the baseline drift elimination module uses It is used to eliminate the interference signal that causes the baseline of the pulse wave signal to drift; the bandpass filter module is used to obtain frequency signal components belonging to the heart rate frequency band while eliminating noise signals outside the heart rate frequency band; the frequency domain analysis module is used to obtain the pulse wave signal The signal spectrum of the wave signal in a preset specific frequency band interval; the heart rate frequency point selection module is used to obtain the frequency point corresponding to the heart rate and output the frequency point. The present invention eliminates the impact of human body movement and muscle activity on heart rate analysis, reduces hardware costs, can flexibly select the frequency range of interest, and only improves the frequency domain accuracy of the signal within the frequency range of interest, avoiding excessive calculation and storage overhead .

It should be understood that the application of the present invention is not limited to the above examples, and those skilled in the art can make improvements or transformations according to the above descriptions, and all these improvements and transformations should belong to the protection scope of the appended claims of the present invention.

Claims (10)

1. A real-time dynamic heart rate monitoring device, characterized in that the device includes a baseline drift elimination module, a bandpass filter module, a frequency domain analysis module, and a heart rate frequency point selection module; The baseline drift elimination module is used to eliminate interference signals that cause the baseline of the pulse wave signal to drift; The band-pass filter module is used to obtain frequency signal components belonging to the heart rate frequency band while eliminating noise signals outside the heart rate frequency band; The frequency domain analysis module is used to obtain the signal spectrum of the pulse wave signal in a preset specific frequency band interval; The heart rate frequency point selection module is used to obtain the frequency point corresponding to the heart rate and output the frequency point; The baseline drift elimination module is connected to the band-pass filter module, the band-pass filter module is connected to the frequency domain analysis module, and the frequency domain analysis module is also connected to the heart rate frequency point selection module. 2. The real-time dynamic heart rate monitoring device according to claim 1, wherein the baseline drift elimination module specifically includes a baseline drift trend item signal extraction unit and a signal linear superposition unit; The baseline drift trend item signal extraction unit is used to obtain a baseline drift trend item signal of the same length as the original pulse wave signal; The signal linear superposition unit is used to subtract the baseline drift trend item signal from the original pulse wave signal to obtain the pulse wave signal with the baseline drift removed; The baseline drift trend item signal extraction unit is connected to the signal linear superposition unit. 3. The real-time dynamic heart rate monitoring device according to claim 2, wherein the bandpass filtering module specifically includes a filtering parameter setting unit and a filtering unit, The filter parameter setting unit is used to set the lower limit and upper limit frequency of the passband, the lower limit and upper limit frequency and the upper limit frequency of the stopband, the attenuation coefficient in the passband, and the attenuation coefficient in the stopband; The filter unit is used to filter the pulse wave signal after the baseline drift is eliminated by the baseline drift elimination module after obtaining the order and coefficient of the filter module, and obtain the pulse wave signal after the noise is removed; The filtering parameter setting unit is connected to the filtering unit. 4. The real-time dynamic heart rate monitoring device according to claim 3, wherein the frequency domain analysis module specifically includes a specific frequency band interval setting unit, a frequency point power calculation unit and a signal spectrum acquisition unit; The specific frequency band interval setting unit is used to set the start frequency point, the end frequency point, and the number of frequency point subdivisions of the specific frequency band interval; The frequency point power calculation unit is used to calculate the real part of the frequency domain signal with the first transformation polynomial and the imaginary part of the frequency domain signal with the second transformation polynomial for each frequency point of the specific frequency band interval, according to the frequency domain signal Calculate the real and imaginary parts of the pulse wave signal at the frequency; The signal spectrum acquisition unit is used to acquire the power of the pulse wave signal at each frequency point in a specific frequency band interval, and generate the signal spectrum of the pulse wave signal in the entire frequency band interval of interest after superposition; The frequency point power calculation unit is respectively connected with the specific frequency band interval setting unit and the signal spectrum acquisition unit. 5. The real-time dynamic heart rate monitoring device according to claim 4, wherein the heart rate frequency point selection module specifically includes a heart rate frequency point selection unit and a heart rate frequency point dynamic tracking unit, The heart rate frequency point selection unit is used to obtain the frequency point where the peak in the pulse wave signal spectrum is located as the heart rate frequency point in the initial state; The heart rate frequency point dynamic tracking unit is used to perform high-precision frequency domain analysis in a spectrum range of a specific width around the center, centered on the heart rate frequency point output in the previous tracking cycle, within a preset fixed tracking period, and select The frequency point where the spectrum peak is located is output as the heart rate frequency point; The heart rate frequency point selection unit is connected with the heart rate frequency point dynamic tracking unit. 6. The real-time ambulatory heart rate monitoring device according to claim 5, characterized in that, the first transformation polynomial and the second transformation polynomial are orthogonal polynomials. 7. The real-time dynamic heart rate monitoring device according to claim 6, wherein the first transformation polynomial is Legendre polynomial, Jacobi polynomial, Laguerre polynomial, Chebyshev polynomial, Hermitian polynomial one of The second transformation polynomial is an orthogonal polynomial and is one of Legendre polynomials, Jacobi polynomials, Laguerre polynomials, Chebyshev polynomials, and Hermitian polynomials. 8. A monitoring method based on the real-time dynamic heart rate monitoring device according to claim 1, wherein the method comprises the steps of: A. Obtain the pulse wave signal, and eliminate the baseline drift interference signal through the baseline drift elimination module; B. The band-pass filter module filters the signal output by the baseline drift elimination module, and retains the signal components belonging to the heart rate frequency band; C. The frequency domain analysis module acquires the signal spectrum of the pulse wave signal in the preset specific frequency band interval; D. The heart rate frequency point selection module obtains and outputs the frequency point corresponding to the heart rate according to the signal frequency. 9. The real-time dynamic heart rate monitoring method according to claim 8, wherein said step A specifically comprises: A1. Obtain the baseline drift trend item signal with the same length as the original pulse wave signal by signal filtering method or curve fitting method; A2. Store the original pulse wave signal and the baseline drift trend item signal in a row vector or column vector, and then subtract the baseline drift trend item information from the original pulse wave signal according to the matrix addition rule to obtain the pulse wave signal with the baseline drift removed. 10. The real-time dynamic heart rate monitoring method according to claim 9, wherein said step B specifically comprises: B1. Obtain the preset lower limit and upper limit frequency of the passband, the lower limit and upper limit frequency of the stopband, the attenuation coefficient in the passband, and the attenuation coefficient in the stopband; B2. After obtaining the order and coefficient of the filtering module, filter the pulse wave signal after the baseline drift is eliminated by the baseline drift elimination module, and obtain the pulse wave signal after the noise is removed.