CN108888271B

CN108888271B - Physiological parameter measuring system and intelligent seat with same

Info

Publication number: CN108888271B
Application number: CN201810833110.4A
Authority: CN
Inventors: 贾振华; 张午阳; 张智杰; 安宁; 林晓东; 理查德-霍华德; 许辰人; 张燕咏
Original assignee: Foshan Measure X Technology Co ltd
Current assignee: Foshan Measure X Technology Co ltd
Priority date: 2018-07-25
Filing date: 2018-07-25
Publication date: 2020-05-22
Anticipated expiration: 2038-07-25
Also published as: CN111493875A; CN111493874A; CN108888271A; CN111493874B; WO2020019358A1

Abstract

A physiological parameter measuring system is characterized in that a data acquisition unit acquires a first signal caused by human physiological activity in a non-contact mode, and a central processing unit performs operation processing on the first signal at least according to the following steps to obtain a respiratory frequency parameter of a human body: filtering the first signal according to a mode of filtering signals of a specific frequency band to obtain a second signal, performing operation processing on the second signal based on an autocorrelation function, and determining peak points in an operation result along a time axis according to a set sequence; and selecting n peak points closest to the origin of the time axis coordinate according to the mode that the proportion of the peak points in the total amount is f so as to calculate the average value of the respiratory frequency. An intelligent seat adopts the physiological parameter measuring system. The physiological parameter measuring system and the intelligent seat can acquire the respiratory frequency of a human body in a non-contact mode, and the use convenience is improved.

Description

Physiological parameter measuring system and intelligent seat with same

Technical Field

The invention belongs to the technical field of measurement, and particularly relates to a physiological parameter measuring system and an intelligent seat with the same.

Background

The respiration rate is an important human body physical sign parameter, and how to realize accurate measurement of the parameter through a non-contact measurement mode is one of the hot spots of researches in the fields of biomedical engineering and instruments. The imaging type photoplethysmography technology is a non-contact physiological parameter detection technology developed on the basis of PPG, and the technology utilizes imaging equipment to carry out video acquisition on information containing a detected part, and realizes a biomedical detection method for extracting physiological parameters such as heart rate, respiratory rate, blood oxygen saturation and the like by processing a sensitive area of a video image. The IPPG technology has the advantages of low cost, non-contact, safety, continuous measurement, simple operation and the like, and provides a new solution for the research of non-contact physiological signal measurement and remote medical monitoring.

In the prior art, in the research aspect of extracting human heart rate indexes based on the IPPG technology, the method for extracting the heart rate and the respiratory rate is mainly used for direct application or improvement and optimization of research ideas. The method mainly comprises the steps of extracting a heart rate from a G channel of an observation signal generated by a face video, or extracting the heart rate by three-channel blind source separation based on a JADE (JavaScript and XML) and other classical ICA (independent component analysis) algorithms, and further extracting a respiration rate from spectrum analysis of the heart rate. In addition, a face video tracking algorithm is integrated in part of the method to overcome the problem of noise interference. In most cases, the effect of using the G-channel method is really good when there is less interference noise, but the method has a drawback when there is more interference noise, and more noise sources affect the separation effect of the three-channel ICA algorithm. Although the above-mentioned algorithm based on face video tracking can solve the problem of the face movement of the subject to a certain extent, it cannot overcome the noise interference caused by the local subtle changes of the face and the weak changes of light. Meanwhile, the existing method also faces the interference problem of a noise source for the judgment of the source signal channel after ICA separation basically according to the power spectrum after FFT, and the judgment accuracy of the source channel is very important for the robustness of the algorithm under the condition of more separation channels. In addition, the existing method cannot realize synchronous extraction of heart rate and respiration signals, and further realize synchronous measurement of the heart rate and the respiration.

Patent document CN105147293A discloses a system and method for measuring respiratory rate, which includes a blood flow change collecting module for illuminating skin surface with white LED lamp and collecting blood flow change signal caused by heartbeat of human body; the respiratory wave signal acquisition module is used for processing the blood flow wave change signal to acquire a respiratory wave signal; and the respiratory frequency calculation module is used for processing the respiratory wave signal to obtain the respiratory frequency. The system and the method for realizing respiratory rate measurement by adopting the structure can apply the functional module to various positions of a human body suitable for measurement, and obtain real-time respiratory rate through white light optical irradiation and blood flow change monitoring. In the process of measuring the respiratory rate, the user needs to fix the data acquisition module at the designated position of the body in a wearing manner, and the operation is complex and is easily influenced by the environment.

Disclosure of Invention

The word "module" as used herein describes any type of hardware, software, or combination of hardware and software that is capable of performing the functions associated with the "module".

In view of the deficiencies of the prior art, the present invention provides a physiological parameter measuring system, at least comprising a data acquisition unit communicatively coupled to a central processing unit, the data acquisition unit acquiring a first signal caused by physiological activity of a human body based on at least one sensor in a non-contact manner, wherein the central processing unit performs an operation processing on the first signal to obtain a respiratory rate parameter of the human body at least according to the following steps: filtering the first signal according to a mode of filtering signals of a specific frequency band to obtain a second signal, performing operation processing on the second signal based on an autocorrelation function, and determining peak points in an operation result along a time axis according to a set sequence; selecting n peak points closest to the origin of the time axis coordinate in a manner that the percentage of the peak points in the total amount is f to calculate the average value of the respiratory frequency

Where D is the total distance between the first peak point and the nth peak point.

According to a preferred embodiment, the central processing unit is further configured to re-determine the operation mode of the remaining peak points with percentage 1-f in the operation result based on the average value of the breathing frequency, wherein the re-determining of the remaining peak points with percentage 1-f comprises at least the following steps:at a time range of [0, T]The point with the maximum operation result value is screened as the first peak point in the operation result of (1), and t is used₁Indicating the time of its occurrence. The occurrence time at which the h-th peak point has been confirmed and corresponds to the h-th peak point is t_hIn the case of being in the time range of

The point with the maximum operation result value is screened from the operation results to be used as the h +1 th peak value point, and t is used_h+1Represents the occurrence time thereof, wherein h is an integer of 2 or more.

According to a preferred embodiment, the autocorrelation function is defined for a time series x (t) by the following formula:

(x_t+h-x’)(x_t-x '), 0 ≦ h ≦ n, where n represents the number of sample points, h represents the time interval between sample points, and x' represents the mean of the sample points and is defined by the following formula:

according to a preferred embodiment, the first signal is a harmonic signal composed of at least a vibration wave signal generated by limb movement and a vibration wave signal generated by respiration, and a plurality of sensors are arranged in a manner of being spaced from each other to form a matrix form of at least three orders so as to acquire the first signal, wherein the specific frequency band is set to be 6Hz to 10Hz to filter the vibration wave signal generated by limb movement; or the specific frequency band is determined at least according to the following steps: and determining a first frequency domain, a second frequency domain and a third frequency domain in which the amplitudes of at least three first signals acquired by sensing devices on the same side of the matrix tend to increase along the extending direction parallel to the legs of the human body when the human body is seated, wherein when the first frequency domain, the second frequency domain and the third frequency domain all intersect with each other, the specific frequency band is determined by the maximum left endpoint and the minimum right endpoint of the endpoints of the first frequency domain, the second frequency domain and the third frequency domain.

According to a preferred embodiment, the step of sequentially determining the peak points in the operation result along the time axis in the sequence at least comprises the following steps: and performing initial peak value judgment on the sampled data subjected to smooth filtering, acquiring all sampled data between a rising edge starting point and a falling edge end point of the peak value when the sampled data meet the requirement of a preset peak value, and recording a peak value channel number and a peak value sequence. And performing weighted average calculation on all the acquired sampling data between the starting point of the rising edge and the end point of the falling edge, and acquiring and storing an accurate value of the peak position.

According to a preferred embodiment, the physiological parameter measuring system further comprises an ac amplifier for amplifying the signal, an analog-to-digital converter for converting an analog signal into a digital signal, and a filter for filtering out a signal of a specific frequency band, wherein the first signal is transmitted to the central processing unit in such a manner that the first signal is sequentially processed by the ac amplifier, the analog-to-digital converter, and the filter; the AC amplifier is configured to have an operation mode of a first-stage amplification circuit having a gain of 10 and a second-stage amplification circuit having a maximum gain of 20 and set to an operation mode in which the gain thereof can be increased or decreased.

According to a preferred embodiment, the physiological parameter measurement system is further configured to pre-warn of a forward fall of the seated person in a time-advanced manner by:

respectively determining the time T when the data signals respectively acquired by the first sensor and the second sensor are continuously equal to zero for the first time in a first set time period A₁And time T₂(ii) a At omega₁< 0 and T₂-T₁>α, the data signal collected at the third sensor is first less than the data collected at the second sensor for a time period T₂-T₁Time T of inner average₃And generating early warning information of forward leaning and falling.

The invention also provides a smart seat equipped with a physiological parameter measuring system as claimed in the preceding claims to collect at least respiratory rate data of a seated person, the smart seat comprising at least a cushion and a backrest, wherein a number of said sensors are arranged on the cushion and the backrest to collect said first signal.

According to a preferred embodiment, the smart seat further comprises at least a sitting position identification unit and an identity identification unit for identifying the identity information of the sitting person, both communicatively coupled to the central processing unit, wherein the identity identification unit is configured to be able to determine the working mode of the identity of the sitting person at least based on fingerprint recognition, weight recognition and/or sitting behavior recognition; wherein the sitting posture identifying unit is configured to be in a working mode of identifying the sitting posture of the sitting person according to the following steps: establishing a three-dimensional coordinate system by taking the geometric center of the cushion as an origin of coordinates; acceleration values a in the x-axis direction, the y-axis direction and the z-axis direction are respectively obtained_x、a_yAnd a_z(ii) a Respectively acquiring angular velocity values w along the x-axis direction, the y-axis direction and the z-axis direction_x、w_yAnd w_z(ii) a Determining a rotation angle θ based on the acceleration value and the angular velocity value₀Angle of pitch omega₀And roll angle

Based on the rotation angle theta₀Angle of pitch omega₀And roll angle

The sitting posture of the sitting person, such as forward leaning, backward leaning, left leaning, right leaning or rotation, is identified.

According to a preferred embodiment, the cushion is provided with a plurality of sensors which are arranged at intervals according to a matrix form, wherein the sensors at the edges of the matrix collect the first signals in a manner that the sampling frequency is lower than that of the sensors at the central part of the matrix; in the height direction along the backrest, a first sensor corresponding to the shoulder and back of the human body, a second sensor corresponding to the chest of the human body, and a third sensor corresponding to the waist and back of the human body are arranged at intervals, wherein the second sensor is arranged to acquire the first signal in a manner that the number of the second sensors is greater than that of the first sensors or the second sensors.

The invention has the beneficial technical effects that:

(1) in the process of analyzing the respiratory frequency, the data signals acquired by the sensor are very easily influenced by interference signals such as the movement of limbs of a human body and the vibration in the surrounding environment, and the vibration signals caused by respiration can be effectively obtained by filtering the interference signals with specific frequencies through the filter.

(2) In the process of determining the peak value of the autocorrelation function based on the peak searching algorithm to determine the respiratory frequency, the invention intercepts the first 20% of calculation results to calculate the average respiratory frequency, and determines the remaining 80% of peak value points in the initial data again by taking the calculated average respiratory frequency as the reference, thereby effectively avoiding the drift of the calculation results caused by the quasi-periodic characteristics of respiration.

(3) The intelligent seat analyzes and calculates the breathing frequency of the sitting person in a non-contact mode, inconvenience and rapidness caused by the fact that the sitting person collects the breathing frequency by wearing special equipment are avoided, meanwhile, the back and/or the neck of the sitting person can be measured without leaning against a backrest when the sitting person sits, and the intelligent seat is more in line with the usual use state of the seat.

Drawings

FIG. 1 is a modular schematic view of a preferred physiological parameter measurement system of the present invention;

FIG. 2 is a schematic diagram of a fast Fourier transform processed vibration wave signal acquired by tapping the mattress once in 30 seconds;

FIG. 3 is a schematic diagram of a fast Fourier transform processed vibration wave signal acquired by beating the mattress every 30 seconds;

FIG. 4 is a schematic diagram of a fast Fourier transform processed vibration wave signal acquired by continuous scratching of a mattress over a set period of time;

FIG. 5 is a schematic diagram of a preferred smart seat of the present invention;

FIG. 6 is a schematic diagram of the peak value captured based on the peak finding algorithm in the preferred embodiment of the present invention;

FIG. 7 is a schematic flow chart of the initial determination of the peak value according to the present invention;

FIG. 8 is a schematic diagram of the working principle of the preferred sitting posture identifying unit of the present invention; and

fig. 9 is a circuit schematic of a preferred ac amplifier of the present invention.

List of reference numerals

1: the sensor 2: and (3) a filter: cushion pad

4: the backrest 5: the sitting posture identifying unit 6: identity recognition unit

7: the storage unit 8: the central processing unit 9: fingerprint input unit

10: the alternating current amplifier 11: the analog-to-digital converter 12: mobile terminal

13: data acquisition unit

101: the first sensor 102: the second sensor 103: third sensor

Detailed Description

The following detailed description is made with reference to the accompanying drawings.

Example 1

The embodiment provides a method for measuring the respiratory rate of a human body, which at least comprises the following steps:

s1: collecting data signals related to a human body through a data collector;

s2: filtering the data signal to obtain sample data;

s3: determining a sample autocorrelation function based on the sample data;

s4: capturing and measuring the peak value of the sample autocorrelation function;

s5: the breathing frequency is extracted based on the peak value.

For ease of understanding, the following discussion is in detail based on the steps.

S1: and collecting data signals related to the human body through a data collector.

When a person inhales, the thoracic diaphragm contracts and causes expansion of the thoracic cavity. The natural expansion forces air into the two lungs to equalize the pressure inside and outside the body. Exhalation begins when inhalation ends. During exhalation, the thoracic diaphragm relaxes, which in turn causes contraction of the thoracic cavity, and air is expelled from both lungs. In the process of breathing and inhaling, the regular contraction of the thoracic diaphragm causes the human body to vibrate up and down slightly. The minute vibrations are hardly visible to the human eye but can be acquired based on sensor technology. For example, in the case where a human body sits on a seat, several sensors 1 may be provided on a cushion of the seat contacting with the buttocks of the human body to collect slight up-and-down vibration of the buttocks. The sensor may be at least one of a geophone, a pressure sensor or a vibration sensor for monitoring seismic waves. For example, in the case where the user lies down on the sofa or mattress at night for rest, the above-mentioned sensor may be provided in the interlayer of the sofa or mattress to collect the data signal related to the human body.

Preferably, the data signal related to the human body at least includes a vibration wave signal caused by respiration, a vibration wave signal caused by limb movement of the human body, and a vibration wave signal caused by an external object within a certain distance from the human body, wherein the vibration wave signal caused by respiration, the vibration wave signal caused by limb movement of the human body, and the vibration wave signal caused by the external object within the certain distance from the human body cannot be individually collected only by means of the sensor. The vibration signal collected by the sensor is a mixed signal of the three vibration wave signals. The sensor 1 may be arranged to acquire the vibration signal in a sampling period of a certain frequency.

S2: and filtering the data signal to obtain sample data.

Under the condition that the sensor 1 collects vibration data in a time period, the vibration data are filtered based on the filter 2 to filter out interference data in the vibration data. For example, the present invention is directed to determining the respiratory rate of a human body by analyzing minute vibrations caused during respiration of the human body, and thus, interference data are vibration wave signals caused by movement of limbs of the human body collected by a sensor and vibration wave signals caused by external objects within a certain range from the human body. A data signal generated entirely by the vibrations of the human breath is obtained by means of the filter 2.

Preferably, the filter may be configured to perform filtering processing in such a manner as to filter out a signal of a set frequency. The frequency of the signals to be filtered can be determined by carrying out comparison tests, for example, when vibration wave signals caused by the movement of the limbs of the human body need to be filtered, three sets of comparison tests are arranged and the vibration caused by the movement of the limbs of the human body is simulated in different modes. Specifically, the test group a is set to work mode according to which the mattress is tapped once within a certain time period and the vibration wave signal is collected by the sensor. The test group B is set to a working mode in which the mattress is tapped at intervals within a certain time period and the vibration wave signal is collected by the sensor. Test group C is set to the mode of operation according to the continuous scratching of the mattress and the acquisition of the vibration wave signal by the sensor over a certain period of time. And the acquired vibration wave signals are displayed after being processed by fast Fourier transform operation. For example, fig. 2 shows a fft-processed vibration wave signal obtained by tapping the mattress once in 30 seconds, fig. 3 shows a fft-processed vibration wave signal obtained by tapping the mattress once every 30 seconds, and fig. 4 shows a fft-processed vibration wave signal obtained by continuously tapping the mattress for a set period of time. Based on the vibration wave signals shown in fig. 2, 3 and 4, it is clear that the frequency of the vibration wave signals caused by the movement of the limbs of the human body is generally greater than 6hz, and a steep trend appears at 8 hz. The working mode of setting the filter to filter the vibration wave signals with the frequency of 6Hz to 10Hz is enough to filter the vibration wave signals caused by the limbs of the human body. Preferably, the set frequency is 8Hz to 8.4 Hz.

Preferably, the specific frequency band may be determined according to the following steps: and determining a first frequency domain, a second frequency domain and a third frequency domain in which the amplitudes of first signals acquired by at least three sensors on the same side of the matrix tend to increase along the extending direction parallel to the legs of the human body when the human body is seated, wherein when the ranges of the first frequency domain, the second frequency domain and the third frequency domain intersect with each other, the specific frequency band is determined by the maximum left endpoint and the minimum right endpoint of the endpoints of the first frequency domain, the second frequency domain and the third frequency domain. As shown in fig. 5, when 9 sensors are installed on the cushion of the seat in such a manner as to constitute a 3-step matrix, the buttocks of the sitting person are completely in contact with the 9 sensors at the same time when the sitting person is seated and in a normal sitting posture. The three sensors on the left and right sides of the matrix correspond to the left and right legs of the human body, respectively, in the direction along the thigh extension. At this time, most of the vibration wave signals caused by the limbs of the human body come from the legs, and the sensors on the left side and the right side of the matrix are most directly influenced by the legs, and can reflect the real frequency domain of the vibration wave signals caused by the legs. The sensor positioned in the middle of the matrix is most affected by human respiration, the collected data is used for calculating the respiratory frequency with higher accuracy, and the vibration wave signal of the specific frequency of the sensor positioned in the middle of the matrix can be filtered based on the specific frequency confirmed by the edge of the matrix. For example, after the data collected by the three sensors corresponding to the left leg of the human body on the left side of the matrix is processed to form a curve with the abscissa as frequency and the ordinate as amplitude as shown in fig. 2, frequency domains in which the amplitudes of the three sensors tend to increase are [10, 20], [15, 30], [25, 40], where the value 10, the value 15, and the value 25 represent the left end points of the three sensors, the value 20, the value 30, and the value 40 represent the right end points of the three sensors, respectively, and the specific frequency band can be determined to be [20, 25] by selecting the maximum left end point 25 and the minimum right end point 20. The filter is set to a working mode of filtering vibration wave signals with the frequency of 20Hz to 25 Hz. The specific frequency band determined in the above manner is always kept in the true frequency domain most representative of the vibration wave signal caused by the movement of the limb of the human body. For example, the frequency domains acquired by three sensors can all represent amplitude increase caused by human limb movement to a certain extent, the frequency domains [15, 30] can be determined to represent the frequency domain with the minimum error to a certain extent by adopting a method of taking an intermediate value, obviously, the frequency domains [10, 20] and [25, 40] have errors but can also reflect the real frequency of the vibration wave signal caused by limb movement, and a new frequency domain [20, 25] formed by further reducing the frequency domains [15, 30] by using the right end point of the frequency domain [10, 20] and the left end point of the frequency domain [25, 40] has higher accuracy.

Preferably, the time domain signal collected by the sensor is converted into a sinusoidal signal based on a frequency domain based on a fast fourier algorithm. For example, for an aperiodic continuous-time signal x (n), a continuous spectrum signal x (k) calculated by a fourier transform algorithm can be represented by the following formula:

wherein k is an integer of 0 to N-1,

x (N) is an input sequence of column length N, i.e. slice data in the time domain acquired by the sensor. x (k) is the output sequence of column length N, i.e. the data in the frequency domain after fast fourier transform.

S3: a sample autocorrelation function is determined based on the sample data.

The autocorrelation function of the samples is arranged to extract the periodicities from a time series. For a time series x (t), the sample autocorrelation function is defined by the following equation:

(x_t+h-x’)(x_t-x’)，0≤h≤n

where n represents the number of sample points, h represents the time interval, and x' represents the mean of the samples, defined by the following equation:

when the time interval h is zero, the vibration power signal is perfectly aligned with itself and the autocorrelation reaches a maximum. When the time delay interval starts to increase, the first vibration power signal remains the same, while the second vibration power signal shifts to the right. A mismatch between the two signals results in a reduction of the value of the sample autocorrelation function. In the case where the time interval is equal to an integer multiple of the detection interval of the respiratory oscillations, the oscillation pulses in the first oscillation power signal are well matched to the oscillation pulses in the second oscillation power signal to produce a larger autocorrelation function. The inference of the vibration frequency can be achieved by detecting a peak in the sample autocorrelation function calculation.

Preferably, the second signal is operated based on the autocorrelation function, and the peak points in the operation result are determined along the time axis according to a set sequence, where the set sequence may be according to a chronological sequence, so as to sequentially determine the peak points on the entire time axis. Preferably, the second signal with the set sequence can be divided into a plurality of parts along the time axis, the peak value of each divided part is obtained through parallel calculation by a plurality of processors, and the obtained peak values are restored to the whole time axis according to the time sequence, so that the running processing time of the data can be effectively shortened.

S4: peak capture and measurement of the sample autocorrelation function. And determining the position and the value of the peak value of the sample autocorrelation function based on a peak searching algorithm. Fig. 7 is a schematic flow chart illustrating initial determination of the peak value, and as shown in fig. 7, the initial determination of the peak value may be performed through the following steps:

s401: and performing initial peak value judgment on the data subjected to smooth filtering, acquiring all sampling data between a rising edge starting point and a falling edge end point of the peak value when the data meet the requirement of a preset peak value, and recording a peak value channel number and a peak value sequence. The initial determination of the peak value of the data can be realized, for example, as follows. When the sampling data of one channel is received, the maximum value of the sampling data in the channel can be found through a comparison mode. The peak value judgment runs with a clock, can be realized by adopting a flow mode, and has the following judgment principle: and setting the current value to be 0, assigning the current value when the current value is larger than the current value, and obtaining the maximum value when the running water is finished. And under the condition of determining the maximum value, extending the maximum value downwards by a preset value by taking the maximum value as a vertex, if the maximum value extends downwards by 15dB to obtain a reference value, recording the maximum value as a peak value if the maximum value is larger than the reference value, and making no judgment if the maximum value is smaller than the reference value. All sample data between the start of the rising edge and the end of the falling edge of the peak are acquired and the peak channel number and the peak sequence are recorded.

S402: and performing weighted average calculation on all the acquired sampling data between the starting point of the rising edge and the end point of the falling edge, and acquiring and storing an accurate value of the peak position. For example, a set of data from the beginning of a rising edge to the end of a falling edge of the same channel is read. And setting the abscissa of the starting point of the rising edge as zero, multiplying the read ordinate by the corresponding abscissa and accumulating to obtain data Y. And accumulating the read vertical coordinates to obtain data T. And calculating the ratio Y/T of the data Y and the data T, wherein the ratio Y/T is the accurate value of the peak position.

S5: the breathing frequency is extracted based on the peak value. The spacing between two adjacent peaks is the breathing frequency.

As shown in fig. 6, the peak value of the sample data after being processed can be obtained based on the peak finding algorithm. The first 20% of the total number of peak points was taken for averaging to obtain an average of the peak intervals to represent the breathing frequency. For example, as shown in fig. 6, in the case that there are 13 peak points in total, only the distance between the first 3 peak points is selected for average calculation, so that the situation that the calculation result is biased due to the quasi-periodic characteristic of the peak searching algorithm can be effectively avoided.

Preferably, the number of peak points determined based on the peak finding algorithm should theoretically be the same as the number of breaths in a given time period. The quasi-periodic characteristic of the peak-finding algorithm causes errors in the calculation results, so that deviations occur based on the determined positions of the peak points after a certain amount of calculation. The correct location of the remaining 80% of the peak points can be determined again by:

s501: if n peak points belong to the first 20% of the total peak points and the sum of the distances between the first peak point and the nth peak point is D, the method comprises the steps ofThe average respiratory rate can be used

Represents;

s502: based on the calculated and estimated respiratory frequency average value T, the time range of [0, T ] is found from the sample data in a mode of searching the maximum amplitude value]1 st peak point in, and using t₁Indicating the time of occurrence thereof;

s503: the occurrence time corresponding to the h-th peak point after h peak points have been confirmed is t_hIn the case of (2), the time range of (1) is found from the sample data by finding the maximum amplitude value

The h +1 th peak point;

s504: step S503 is repeated until all remaining 80% of the peak points are determined.

Example 2

This embodiment is a further improvement of embodiment 1, and repeated contents are not described again.

The invention also provides an intelligent seat which at least comprises a seat cushion 3 and a backrest 4, wherein a plurality of sensors 1 are arranged on the seat cushion and the backrest to collect data signals related to the human body. The sensor 1 is preferably a capacitance type sensor and is arranged on the cushion and the backrest in a mode of being connected with a resistor to form a resistance-capacitance circuit, so that the current use state of the intelligent seat can be identified by measuring the change of the capacitance of the sensor.

Preferably, as shown in fig. 5, the smart seat uses a total of 16 sensors 1 to form a data acquisition unit 13 for acquiring data signals related to a human body, wherein 9 sensors are disposed on the seat cushion and 7 sensors are disposed on the backrest. The 9 sensors are arranged on the cushion in a manner to form a shape similar to a third order matrix, wherein the sensors located at the edge of the matrix sample at a first sampling frequency and the sensors located at the middle of the matrix sample at a second sampling frequency. The first sampling frequency is less than the second sampling frequency to save energy consumption. Meanwhile, the sensor positioned in the middle of the matrix is influenced most by the vertical vibration caused by the respiration of the human body, so that the sensor acquires data at a higher sampling frequency and can improve the calculation precision of the respiratory frequency of the human body. In the height direction of the seat, 7 sensors are arranged on the backrest in such a manner as to be divided into three rows arranged in parallel at intervals with each other, wherein two sensors are arranged on a first row closest to the seat cushion and a third row farthest from the seat cushion, and three sensors are arranged on a second row located in the middle of the first row and the third row to correspond to the lung positions of the human body. In a standard sitting position, the second row can correspond to a human lung position. The more sensors of quantity on the second line can gather more comprehensive and back fluctuation state relevant data, and then can be better calculate human respiratory frequency through the periodic fluctuation of back.

Preferably, after the sensor is connected with the resistor to form the impedance circuit, whether the seat belongs to the user sitting state or the idle state can be determined based on the change of the capacitance value of the sensor. By the following formula

The discharge process of the sensor is modeled, where V (t) represents the voltage across the sensor at time t, V₀Is the voltage across the sensor at the initial moment, R represents a resistor with a fixed resistance value and may preferably have a resistance value of 2M omega. The calculation formula of the sensor capacitance can be obtained by reversely solving the modeling formula of the sensor discharge process

Preferably, the sitting posture change of the sitting person can be analyzed and monitored based on the capacitance signal acquired by the sensor 1. When the user sits on the intelligent seat in a stationary state, such as sleeping and resting, the gap between the skin of the human body and the sensor is equal to the thickness of the worn clothes. When the sitting posture of the sitting person changes, for example, the sitting person changes from a left-leaning posture to a right-leaning posture, the gap between the skin of the human body and the sensor changes remarkably, and the current sitting posture of the sitting person can be identified by matching the change rule of capacitance signals of the 9 sensors on the cushion.

Preferably, the smart seat further comprises a sitting posture identifying unit 5 communicatively coupled with the sensor, the sitting posture identifying unit 5 being configured to receive a capacitance signal of the sensor on the seat cushion and/or the backrest and to process a change in the capacitance signal to determine the current sitting posture of the seated person.

Preferably, the smart seat is configured with an operation mode of detecting the breathing frequency of the sitting person according to the method for measuring the breathing frequency of the human body in the embodiment 1.

For ease of understanding, the specific working principle of the sitting posture identifying unit 5 will be discussed in detail as follows.

Fig. 8 shows a schematic view of the working principle of the sitting posture identifying unit. As shown in fig. 8, the pressure data collected by the sensor is converted into a digital signal represented by a capacitance value, and then the digital signal is transmitted to the sitting posture identification unit 5 for processing, the sitting posture identification unit 5 establishes a coordinate system xyz by using the position of the sensor located at the center of the matrix on the seat cushion as the origin of coordinates, the sitting posture identification unit 5 is internally provided with an accelerometer and a gyroscope in an integrated manner, wherein the accelerometer measures acceleration values a in three axial directions of xyz_x、a_yAnd a_zCalculating angular velocity values w in directions of three axes of xyz by a gyroscope_x、w_yAnd w_z. By making a pair of w_zThe integral calculation is performed to obtain the rotation angle theta₀. Based on the acceleration value, the pitch angle omega can be respectively calculated by the following formula₀And roll angle

Wherein the angle of rotation theta₀Angle of pitch omega₀And roll angle

In the direction indicated by the arrow shown in FIG. 8A positive value. The sensor is arranged to sample the pressure data based on a sampling frequency, which may be performed once every time T. For example, the time interval T may be set to 1 minute, and the rotation angle, the pitch angle, and the flip angle detected at the moment before the time interval T are set to θ, respectively₁、ω₁And

the rotation angle, the pitch angle and the roll-over angle detected at a time after the time interval T are respectively theta₂、ω₂And

by setting the threshold values theta of the rotation angle, the pitch angle and the flip angle in advance respectively_max、ω_maxAnd

the magnitude of the inclination, e.g. sitting, can be determined, e.g. at

Under the condition (2), the user can be preliminarily judged to have a left-leaning trend

In this case, it can be determined that the user is greatly inclined to the left. At omega₁Under the condition of less than 0, the user can be preliminarily judged to have the forward tilting trend at omega₁＜-ω_maxIn this case, it can be judged that the user's sitting posture is greatly inclined forward. At theta₂＜θ₁In the case of (1), it can be judged that the user has a left rotation tendency at | θ₂-θ₁|＞θ_maxUnder the condition of (3), the sitting posture of the user can be judged to be rotated leftwards greatly.

Preferably, as shown in fig. 5, a first sensor 101, a second sensor 102 and a third sensor 103 are provided at intervals in the height direction of the seat, wherein the third sensor may be provided at a lower portion of the backrest corresponding to the lumbar and back of the human body, the second sensor may be provided at a middle portion of the backrest corresponding to the thoracic region of the human body, and the first sensor may be provided at an upper portion of the backrest corresponding to the shoulder and back of the human body. The sitting posture identifying unit 5 is also configured to early warn of a forward fall of the seated person with a time advance based on a sensor on the seat back. Specifically, the sitting posture identifying unit 5 performs early warning on forward leaning and falling of the sitting person according to a time advance mode at least according to the following steps:

s1: respectively determining the time T when the data signals respectively acquired by the first sensor 101 and the second sensor 102 are continuously equal to zero for the first time in the first set time period A₁And time T₂。

The data signal collected by the sensor is equal to zero, which indicates that the seated person is out of contact with the sensor and that the human body is not exerting pressure on the sensor. When the human body is not in the forward leaning state, the back of the sitting person is completely contacted with the third sensor, the second sensor and the first sensor. When the sitting person is tired to cause the human body to incline forward, the first sensor corresponding to the shoulder and the back of the human body, the second sensor corresponding to the chest cavity of the human body and the third sensor corresponding to the waist and the back of the human body are gradually separated from the human body according to the sequence. The first sensor and the second sensor are both continuously less than zero in the first set time period A, which is enough to indicate that the human body is in a forward tilting state. The sitting person is in a waking state only by conscious behaviors performed for adjusting sitting postures or treating special conditions by attaching the backrest with the human body after being separated from the backrest for a short time, and the mental state of the sitting person can be judged currently by setting the first set time period A. S2: at omega₁< 0 and T₂-T₁>α, the data signal acquired at the third sensor 103 is first less than the data acquired at the second sensor for a time period T₂-T₁Time T of inner average₃And generating early warning information of forward leaning and falling.

T₂-T₁>α indicate that the seated person is in a slow forward leaning process, e.g. the elderly are in the eveningIn the case of watching tv, forward leaning of the body due to fatigue is performed slowly. When T is₂-T₁>α and ω₁If the forward leaning degree is less than 0, the possibility that the forward leaning of the sitting person is an unconscious behavior caused by fatigue can be preliminarily judged, and the sitting person has a great risk of forward leaning and falling under the condition that the behavior is not prevented by early warning. When the data collected by the third sensor corresponding to the waist and the back of the human body is zero, the sitting person is completely separated from the backrest and is in a completely forward tilting state, and the early warning is obviously too late. The decreasing trend of the data collected by the second sensor during the unintentional forward tilting of the body of the sitting person caused by fatigue is approximately the same as the trend of the data collected by the third sensor by applying the data collected by the second sensor during the time period T₂-T₁The average value in the time interval is used as a standard for judging that the body of the sitting person is completely separated from the backrest, and early warning information is generated at the moment, so that the early warning has a certain time lead, and the forward leaning and falling caused by untimely early warning are avoided.

Preferably, at average respiratory rate

Less than β and T₂-T₁>α, the data signal acquired at the third sensor 103 is first less than the data acquired at the second sensor for a time period T₂-T₁Time T of inner average₃And generating early warning information of forward leaning and falling.

The human body has different breathing frequencies when in a waking state and a light sleeping state caused by fatigue, for example, a sitting person with snoring does not snore when in the waking state, the breathing frequency is in a normal value, when the sitting person snores, due to the difficulty in breathing through the nose, the sitting person can breathe through the mouth to obtain sufficient oxygen, so that the breathing frequency is reduced, and under the condition that the average breathing frequency is less than β, the sitting person can be preliminarily judged to be in an unconscious state.

Example 3

This embodiment is a further improvement on embodiments 1 and 2, and repeated details are not repeated.

The smart seat of the present invention further comprises an identification unit 6. The identity recognition unit is configured to learn an operational mode of a sitting behavior of the sitting person based on a machine learning algorithm. For example, in the process of sitting in contact with a seat and sitting perfectly steady, different persons may have the process last for different lengths of time based on different seating habits. For example, the sitting personnel can form distinct sitting steps due to different sitting habits, and for younger young people, the sitting is often performed in a direct sitting mode, so that the instantaneous impact strength on the seat is large, and the pressure data collected by the sensor can be displayed in a mode of being steeply increased at a certain moment through a time window after being processed. For the middle-aged and the elderly, the sitting is usually performed in a slow sitting mode, and the pressure data collected by the sensor is processed and then displayed through a time window, so that the pressure data can be continuously increased within a time range. For example, a seated person may experience the same sitting posture adjustment process throughout the sitting process in order to find a sitting posture that fits his or her hip shape, leg length, and/or spine configuration. The sitting person may take a form of sitting down and then standing up and pulling the seat back and forth to find a comfortable sitting posture, or the sitting person may take a form of sitting down and then rocking the body left and right or rocking the body back and forth to find a comfortable sitting posture. The identity recognition unit memorizes and learns different sitting behaviors of different sitting persons through a machine learning algorithm, and distinguishes identities of family members, for example, in the same family, based on the difference between the sitting behaviors.

Preferably, the identity recognizing unit 6 is further configured with an operation mode capable of recognizing the identity of the seated person based on the weight of the seated person. When the smart seat is used in the range of family members in units of parents and two children, the smart seat may be further configured to include a storage unit 7 for storing weight data and identity data of the family members. After the intelligent seat is purchased, identity information of family members is input in the storage unit 7 in a keyboard input or voice input mode, and the storage unit stores the identity information, such as 'son' and '40 Kg' in mass, in a corresponding and mutually-associated mode. When the person sits on the intelligent seat, the intelligent seat collects pressure data through the sensor and transmits the pressure data to the central processing unit 8 to be processed and calculated, so that the quality data of the person sits can be obtained, and the identity of the person sits can be obtained after the quality data is matched with the quality data stored in the storage unit. After the identification of the sitting person is completed, the quality data of the sitting person stored in the storage unit may be replaced with the quality data calculated by the central processing unit 8.

Preferably, the identification unit 6 is further configured to be an operation mode capable of identifying the identity of the sitting person based on fingerprint information of the sitting person. The smart seat further comprises a fingerprint entry unit 9. After the intelligent seat is purchased, the fingerprints of family members are recorded through the fingerprint recording unit and stored through the storage unit 7, the identity information corresponding to the fingerprints is recorded when the fingerprints are recorded, and the fingerprints and the identity information are stored in the storage unit in a related storage mode. When sitting personnel sit on intelligent seat, sitting personnel type the unit through the fingerprint and carry out the unblock to intelligent seat, and under the fingerprint that fingerprint and the memory cell storage of typeeing had the condition of matcing, intelligent seat circular telegram work and accomplish the discernment to sitting personnel's identity.

Example 4

This embodiment is a further improvement of the foregoing embodiment, and repeated contents are not described again.

The smart seat further comprises an ac amplifier 10 for amplifying the ac signal and an analog-to-digital converter 11 for converting the analog signal into a digital signal. The data signal acquired by the sensor is first transmitted to an ac amplifier 10 for amplification. As shown in fig. 9, the ac amplifier includes at least two stages of amplifying circuits, wherein the first stage amplifying circuit and the second stage amplifying circuit are both provided with RC band pass filters that allow only waves in a specific frequency band to pass through, and the specific frequency band of the RC band pass filters is set to 0.25Hz to 10 kHz. The gain of the first stage amplification circuit is 10 to reduce a part of the interference from the amplification circuit itself, the maximum gain of the second stage amplification circuit is 20 and the gain of the second stage amplification circuit is set to an operation mode capable of being dynamically adjusted by the resistor R7. The maximum overall gain of the entire ac amplifier is 200.

Preferably, the signal processed by the ac amplifier 10 is transmitted to an analog-to-digital converter to be converted into a digital signal for subsequent processing.

Example 5

As shown in fig. 1, the present invention further provides a physiological parameter measuring system, which at least includes a sensor 1, a filter 2, a storage unit 7, an ac amplifier 10, an analog-to-digital converter 11 and a mobile terminal 12, all of which are communicatively coupled with a central processing unit 8, wherein a data signal collected by the sensor 1 is sequentially subjected to a first-stage processing by the ac amplifier and the analog-to-digital converter, and then transmitted to the filter for processing to filter an interference signal with a specific frequency, and a first data processed by the filter is transmitted to the central processing unit for a second-stage processing.

Preferably, the central processing unit is configured to perform the second stage processing on the first data in the following manner.

A1: determining a sample autocorrelation function for the first data based on the first data, wherein for a time series x (t) the sample autocorrelation function is defined as:

(x_t+h-x’)(x_t-x’)，0≤h≤n

where n denotes the number of sample points, h denotes the time interval, x' denotes the mean of the samples, where,

a2: peak capture and measurement of the sample autocorrelation function. And determining the position and the value of the peak value of the sample autocorrelation function based on a peak searching algorithm. Wherein the peak finding algorithm is defined by at least the following steps:

a201: and performing initial peak value judgment on the data subjected to smooth filtering, acquiring all sampling data between a rising edge starting point and a falling edge end point of the peak value when the data meet the requirement of a preset peak value, and recording a peak value channel number and a peak value sequence. The initial determination of the peak value of the data can be realized, for example, as follows. When the sampling data of one channel is received, the maximum value of the sampling data in the channel can be found through a comparison mode. The peak value judgment runs with a clock, can be realized by adopting a flow mode, and has the following judgment principle: and setting the current value to be 0, assigning the current value when the current value is larger than the current value, and obtaining the maximum value when the running water is finished. And under the condition of determining the maximum value, extending the maximum value downwards by a preset value by taking the maximum value as a vertex, if the maximum value extends downwards by 15dB to obtain a reference value, recording the maximum value as a peak value if the maximum value is larger than the reference value, and making no judgment if the maximum value is smaller than the reference value. All sample data between the start of the rising edge and the end of the falling edge of the peak are acquired and the peak channel number and the peak sequence are recorded.

A202: and performing weighted average calculation on all the acquired sampling data between the starting point of the rising edge and the end point of the falling edge, and acquiring and storing an accurate value of the peak position. For example, a set of data from the beginning of a rising edge to the end of a falling edge of the same channel is read. And setting the abscissa of the starting point of the rising edge as zero, multiplying the read ordinate by the corresponding abscissa and accumulating to obtain data Y. And accumulating the read vertical coordinates to obtain data T. And calculating the ratio Y/T of the data Y and the data T, wherein the ratio Y/T is the accurate value of the peak position.

A3: the breathing frequency is extracted based on the peak value. The spacing between two adjacent peaks is the breathing frequency. The method of peak-based extraction of the breathing frequency is defined by at least the following steps:

a301: assume that n peak points belong to the first 20% of the total number of peak points, firstThe sum of the distances between the peak point and the nth peak point is D, the average respiratory frequency can be used

Represents;

a302: based on the calculated and estimated respiratory frequency average value T, the time range of [0, T ] is found from the sample data in a mode of searching the maximum amplitude value]1 st peak point in, and using t₁Indicating the time of occurrence thereof;

a303: the occurrence time corresponding to the h-th peak point after h peak points have been confirmed is t_hIn the case of (2), the time range of (1) is found from the sample data by finding the maximum amplitude value

The h +1 th peak point;

a304: step a303 is repeated until all remaining 80% of the peak points are determined.

Preferably, the detection system may further comprise a sitting posture identification unit 5 for identifying the sitting posture of the sitting person, an identity identification unit 6 for identifying the identity of the sitting person by the user and a fingerprint entry unit 9 for collecting the fingerprint of the sitting person, all communicatively coupled to the central processing unit.

Preferably, the storage unit 7 is configured to be capable of storing at least an operation mode of data information such as fingerprint data entered through the fingerprint entry unit, identity data recognized through the identity recognition unit, sitting posture data recognized through the sitting posture recognition unit, and respiratory rate processed by the central processing unit.

Preferably, the mobile terminal 12 is configured to be in an operation mode in which the stored data in the storage unit 7 is displayed for recall. Through the mobile terminal, the sitting person and/or the third party person can visually check the sitting data of the sitting person.

It should be noted that the above-mentioned embodiments are exemplary, and that those skilled in the art, having benefit of the present disclosure, may devise various arrangements that are within the scope of the present disclosure and that fall within the scope of the invention. It should be understood by those skilled in the art that the present specification and figures are illustrative only and are not limiting upon the claims. The scope of the invention is defined by the claims and their equivalents.

Claims

1. Physiological parameter measurement system comprising at least a data acquisition unit (13) communicatively coupled to a central processing unit (8), characterized in that said data acquisition unit (13) acquires, in a contactless manner, a first signal caused by a physiological activity of a human body based on at least one sensor (1), wherein said central processing unit (8) arithmetically processes said first signal to obtain a breathing frequency parameter of the human body at least according to the following steps:

filtering the first signal according to a mode of filtering signals of a specific frequency band to obtain a second signal, performing operation processing on the second signal based on an autocorrelation function, and determining peak points in an operation result along a time axis according to a set sequence;

selecting n peak points closest to the origin of the time axis coordinate in a manner that the percentage of the peak points in the total amount is f to calculate the average value of the respiratory frequency

Wherein D is the total distance between the first peak point and the nth peak point;

the central processing unit (8) is further configured to re-determine the operation mode of the remaining percentage 1-f peak points in the operation result based on the average value of the breathing frequency, wherein the re-determining of the remaining percentage 1-f peak points comprises at least the following steps:

at a time range of [0, T]The point with the maximum operation result value is screened as the first peak point in the operation result of (1), and t is used₁Indicating the time of occurrence thereof;

the occurrence time at which the h-th peak point has been confirmed and corresponds to the h-th peak point is t_hIn the case of being at timeIn the range of

2. A physiological parameter measurement system according to claim 1, wherein the autocorrelation function is defined for a time series x (t) by the formula:

where n represents the number of sample points, h represents the time interval between sample points, and x' represents the mean of the sample points and is defined by the following equation:

3. the physiological parameter measuring system according to claim 2, wherein said first signal is a harmonic signal composed of at least a vibration wave signal generated by limb movement and a vibration wave signal generated by respiration, a plurality of said sensors (1) are arranged in a matrix form of at least three orders in a manner of being spaced from each other to collect said first signal, wherein said specific frequency band is set to 6Hz to 10Hz to filter out said vibration wave signal generated by limb movement; or the specific frequency band is determined at least according to the following steps:

determining a first frequency domain, a second frequency domain and a third frequency domain in which the amplitudes of first signals acquired by at least three sensors (1) positioned on the same side of the matrix tend to increase in a direction parallel to the extension of the legs of a human body when sitting, wherein,

and when the first frequency domain, the second frequency domain and the third frequency domain all have an intersection with each other, determining the specific frequency band by using the maximum left endpoint and the minimum right endpoint of the endpoints of the first frequency domain, the second frequency domain and the third frequency domain.

4. The physiological parameter measuring system according to claim 2, wherein the determining of the peak point in the operation result sequentially along the time axis comprises at least the following steps:

carrying out initial peak value judgment on the sampled data subjected to smooth filtering, acquiring all sampled data between a rising edge starting point and a falling edge end point of a peak value when the sampled data meet the requirement of a preset peak value, and recording a peak value channel number and a peak value sequence;

and performing weighted average calculation on all the acquired sampling data between the starting point of the rising edge and the end point of the falling edge, and acquiring and storing an accurate value of the peak position.

5. The physiological parameter measuring system according to claim 4, further comprising an AC amplifier (10) for amplifying a signal, an analog-to-digital converter (11) for converting an analog signal into a digital signal, and a filter (2) for filtering a signal of a specific frequency band, wherein the first signal is transmitted to the central processing unit (8) in such a manner that the first signal is sequentially processed by the AC amplifier (10), the analog-to-digital converter (11), and the filter (2);

the AC amplifier (10) is configured to have an operation mode of a first stage amplification circuit having a gain of 10 and a second stage amplification circuit having a maximum gain of 20 and set to an operation mode in which the gain thereof can be increased or decreased.

6. The physiological parameter measurement system of claim 5, further configured to pre-warn of a forward fall of a seated person in a time-advanced manner by:

determining a first sensor (101) and a second sensor (101) respectivelyThe data signals respectively collected by the sensors (102) are firstly and continuously equal to zero at the time T within a first set time period A₁And time T₂；

At omega₁< 0 and T₂-T₁>α, the data signal acquired at the third sensor (103) is first less than the data acquired at the second sensor (102) for a time period T₂-T₁Time T of inner average₃And generating early warning information of forward leaning and falling.

7. A smart seat, characterized in that it is equipped with a physiological parameter measuring system according to one of the preceding claims to collect at least respiratory rate data of a seated person, said smart seat comprising at least a cushion (3) and a backrest (4), wherein several of said sensors (1) are arranged on the cushion (3) and the backrest (4) to collect said first signal.

8. The smart seat according to claim 7, further comprising at least a sitting position identification unit (5) and an identity identification unit (6) for identifying the identity information of the sitting person, both communicatively coupled to the central processing unit (8), wherein the identity identification unit (6) is configured to be able to determine an operational mode of the identity of the sitting person at least based on fingerprint identification, weight identification and/or sitting behavior identification; wherein the content of the first and second substances,

the sitting posture identifying unit (5) is configured to be in a working mode of identifying the sitting posture of a sitting person according to the following steps:

establishing a three-dimensional coordinate system by taking the geometric center of the cushion as an origin of coordinates;

acceleration values a in the x-axis direction, the y-axis direction and the z-axis direction are respectively obtained_x、a_yAnd a_z；

Respectively acquiring angular velocity values w along the x-axis direction, the y-axis direction and the z-axis direction_x、w_yAnd w_z；

Determining a rotation angle θ based on the acceleration value and the angular velocity value₀Angle of pitch omega₀And roll angle

Based on the rotation angle theta₀Angle of pitch omega₀And roll angle

9. The intelligent seat according to claim 8, characterized in that the seat cushion (3) is provided with a plurality of sensors (1) which are arranged at intervals according to a matrix form, wherein the sensors (1) at the edges of the matrix collect the first signals in a way that the sampling frequency is lower than that of the sensors (1) at the central part of the matrix;

in the height direction along the backrest (4), a first sensor (101) corresponding to the shoulder and back of the human body, a second sensor (102) corresponding to the chest of the human body and a third sensor (103) corresponding to the waist and back of the human body are arranged at intervals, wherein the second sensor (102) is arranged to acquire the first signal in a manner that the number of the second sensors is more than that of the first sensors or the second sensors.