CN111529847A - Breathing machine system with breathing phase identification control applied to intelligent medical system and control method - Google Patents

Breathing machine system with breathing phase identification control applied to intelligent medical system and control method Download PDF

Info

Publication number
CN111529847A
CN111529847A CN202010384776.3A CN202010384776A CN111529847A CN 111529847 A CN111529847 A CN 111529847A CN 202010384776 A CN202010384776 A CN 202010384776A CN 111529847 A CN111529847 A CN 111529847A
Authority
CN
China
Prior art keywords
pressure
time
real
sequence
air passage
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Withdrawn
Application number
CN202010384776.3A
Other languages
Chinese (zh)
Inventor
冯叶
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Shandong Kaixin Hongye Biotechnology Co ltd
Original Assignee
Shandong Kaixin Hongye Biotechnology Co ltd
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Shandong Kaixin Hongye Biotechnology Co ltd filed Critical Shandong Kaixin Hongye Biotechnology Co ltd
Priority to CN202010384776.3A priority Critical patent/CN111529847A/en
Publication of CN111529847A publication Critical patent/CN111529847A/en
Withdrawn legal-status Critical Current

Links

Images

Classifications

    • A—HUMAN NECESSITIES
    • A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
    • A61M—DEVICES FOR INTRODUCING MEDIA INTO, OR ONTO, THE BODY; DEVICES FOR TRANSDUCING BODY MEDIA OR FOR TAKING MEDIA FROM THE BODY; DEVICES FOR PRODUCING OR ENDING SLEEP OR STUPOR
    • A61M16/00—Devices for influencing the respiratory system of patients by gas treatment, e.g. mouth-to-mouth respiration; Tracheal tubes
    • A—HUMAN NECESSITIES
    • A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
    • A61M—DEVICES FOR INTRODUCING MEDIA INTO, OR ONTO, THE BODY; DEVICES FOR TRANSDUCING BODY MEDIA OR FOR TAKING MEDIA FROM THE BODY; DEVICES FOR PRODUCING OR ENDING SLEEP OR STUPOR
    • A61M16/00—Devices for influencing the respiratory system of patients by gas treatment, e.g. mouth-to-mouth respiration; Tracheal tubes
    • A61M16/0057—Pumps therefor
    • A61M16/0066—Blowers or centrifugal pumps
    • A—HUMAN NECESSITIES
    • A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
    • A61M—DEVICES FOR INTRODUCING MEDIA INTO, OR ONTO, THE BODY; DEVICES FOR TRANSDUCING BODY MEDIA OR FOR TAKING MEDIA FROM THE BODY; DEVICES FOR PRODUCING OR ENDING SLEEP OR STUPOR
    • A61M16/00—Devices for influencing the respiratory system of patients by gas treatment, e.g. mouth-to-mouth respiration; Tracheal tubes
    • A61M16/0003—Accessories therefor, e.g. sensors, vibrators, negative pressure
    • A61M16/0009—Accessories therefor, e.g. sensors, vibrators, negative pressure with sub-atmospheric pressure, e.g. during expiration
    • A—HUMAN NECESSITIES
    • A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
    • A61M—DEVICES FOR INTRODUCING MEDIA INTO, OR ONTO, THE BODY; DEVICES FOR TRANSDUCING BODY MEDIA OR FOR TAKING MEDIA FROM THE BODY; DEVICES FOR PRODUCING OR ENDING SLEEP OR STUPOR
    • A61M16/00—Devices for influencing the respiratory system of patients by gas treatment, e.g. mouth-to-mouth respiration; Tracheal tubes
    • A61M16/0003—Accessories therefor, e.g. sensors, vibrators, negative pressure
    • A61M2016/0015—Accessories therefor, e.g. sensors, vibrators, negative pressure inhalation detectors
    • A—HUMAN NECESSITIES
    • A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
    • A61M—DEVICES FOR INTRODUCING MEDIA INTO, OR ONTO, THE BODY; DEVICES FOR TRANSDUCING BODY MEDIA OR FOR TAKING MEDIA FROM THE BODY; DEVICES FOR PRODUCING OR ENDING SLEEP OR STUPOR
    • A61M16/00—Devices for influencing the respiratory system of patients by gas treatment, e.g. mouth-to-mouth respiration; Tracheal tubes
    • A61M16/0003—Accessories therefor, e.g. sensors, vibrators, negative pressure
    • A61M2016/0027—Accessories therefor, e.g. sensors, vibrators, negative pressure pressure meter
    • A—HUMAN NECESSITIES
    • A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
    • A61M—DEVICES FOR INTRODUCING MEDIA INTO, OR ONTO, THE BODY; DEVICES FOR TRANSDUCING BODY MEDIA OR FOR TAKING MEDIA FROM THE BODY; DEVICES FOR PRODUCING OR ENDING SLEEP OR STUPOR
    • A61M16/00—Devices for influencing the respiratory system of patients by gas treatment, e.g. mouth-to-mouth respiration; Tracheal tubes
    • A61M16/0003—Accessories therefor, e.g. sensors, vibrators, negative pressure
    • A61M2016/003—Accessories therefor, e.g. sensors, vibrators, negative pressure with a flowmeter
    • A—HUMAN NECESSITIES
    • A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
    • A61M—DEVICES FOR INTRODUCING MEDIA INTO, OR ONTO, THE BODY; DEVICES FOR TRANSDUCING BODY MEDIA OR FOR TAKING MEDIA FROM THE BODY; DEVICES FOR PRODUCING OR ENDING SLEEP OR STUPOR
    • A61M2205/00—General characteristics of the apparatus
    • A61M2205/18—General characteristics of the apparatus with alarm
    • A—HUMAN NECESSITIES
    • A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
    • A61M—DEVICES FOR INTRODUCING MEDIA INTO, OR ONTO, THE BODY; DEVICES FOR TRANSDUCING BODY MEDIA OR FOR TAKING MEDIA FROM THE BODY; DEVICES FOR PRODUCING OR ENDING SLEEP OR STUPOR
    • A61M2205/00—General characteristics of the apparatus
    • A61M2205/33—Controlling, regulating or measuring
    • A—HUMAN NECESSITIES
    • A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
    • A61M—DEVICES FOR INTRODUCING MEDIA INTO, OR ONTO, THE BODY; DEVICES FOR TRANSDUCING BODY MEDIA OR FOR TAKING MEDIA FROM THE BODY; DEVICES FOR PRODUCING OR ENDING SLEEP OR STUPOR
    • A61M2205/00—General characteristics of the apparatus
    • A61M2205/33—Controlling, regulating or measuring
    • A61M2205/3327—Measuring
    • A—HUMAN NECESSITIES
    • A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
    • A61M—DEVICES FOR INTRODUCING MEDIA INTO, OR ONTO, THE BODY; DEVICES FOR TRANSDUCING BODY MEDIA OR FOR TAKING MEDIA FROM THE BODY; DEVICES FOR PRODUCING OR ENDING SLEEP OR STUPOR
    • A61M2205/00—General characteristics of the apparatus
    • A61M2205/33—Controlling, regulating or measuring
    • A61M2205/3331—Pressure; Flow
    • A61M2205/3334—Measuring or controlling the flow rate

Abstract

An intelligent breathing machine system with breathing phase recognition control comprises a controller, a fan, a pipeline and a sensor, wherein the fan outputs rotating speed under set working voltage, the rotating speed is converted into wind pressure through a blade and then is transmitted to the pipeline, and the sensor detects real-time flow of an air passage, real-time pressure of the air passage and concentration of real-time carbon dioxide and feeds the real-time flow, the real-time pressure of the air passage and the concentration of the real-time carbon dioxide back to the controller; the controller realizes the control of the breathing machine according to the real-time flow of the air passage and the real-time pressure of the air passage, and the control comprises breathing phase identification control.

Description

Breathing machine system with breathing phase identification control applied to intelligent medical system and control method
Technical Field
The invention belongs to the field of breathing machines, and particularly relates to an intelligent breathing machine system with breathing phase identification control and a control method thereof.
Background
The breathing machine is a mechanical ventilation device which can be used for replacing, controlling or changing the spontaneous respiration of a human body so as to increase the lung ventilation volume, improve the respiratory function and relieve the fatigue of respiratory muscles, is widely applied to the conditions of clinical respiratory failure, respiratory support treatment, anesthesia respiratory management during major operations, emergency resuscitation and the like, works through the pressure difference between alveoli and an airway, and realizes the mechanical ventilation in the respiratory process of a user by regulating the pressure difference. With the rapid development of computer technology and the gradual improvement of intelligent control theory, the performance of the breathing machine is more improved, and the breathing machine is developing towards networking and intellectualization. The conventional ventilator has the following problems:
(1) the control method of the breathing machine is easy to be interfered by the outside when in work to cause deviation between the output pressure and the set pressure, and for a nonlinear time-varying system, the conventional feedback control has obvious defects in robustness and stability;
(2) the estimation of breathing parameters is lacked, so that the breathing condition of a patient is estimated by evaluating the breathing parameters, and the possible danger of the patient is known by medical staff in advance;
(3) the traditional ventilator has the disadvantages of difficult real-time online adjustment, weak self-adaptive capacity, slow corresponding speed and incapability of timely adjusting complex conditions;
(4) the traditional breathing machine lacks of judgment on an expiratory phase and an inspiratory phase, and cannot perform corresponding adjustment aiming at different breathing stages, so that initial judgment is performed in time;
(5) lack of accurate measurement of gas concentration within the ventilator;
(6) there is a lack of accurate estimation of whether or not fluid accumulation is present in the ventilator.
Disclosure of Invention
The invention aims to provide an intelligent breathing machine system with breathing phase identification control, which can quickly and accurately realize breathing phase identification and corresponding control.
In order to achieve the purpose, the technical scheme of the invention is as follows:
an intelligent breathing machine system with breathing phase recognition control comprises a controller, a fan, a pipeline and a sensor, wherein the fan outputs rotating speed under set working voltage, the rotating speed is converted into wind pressure through a blade and then is transmitted to the pipeline, and the sensor detects real-time flow of an air passage, real-time pressure of the air passage and concentration of real-time carbon dioxide and feeds the real-time flow, the real-time pressure of the air passage and the concentration of the real-time carbon dioxide back to the controller; the controller realizes the control of the breathing machine according to the real-time flow of the air passage and the real-time pressure of the air passage, and the control comprises breathing phase identification control.
A control method of a breathing machine with a breathing phase recognition control comprises the following steps:
step 1, initializing a system, and when a singlechip is electrified, starting to power on, wherein the system comprises a timer and an I2C, communicating with the SPI, initializing A/D conversion and initializing related variables;
step 2, configuring a processor interface;
step 3, initializing a display;
step 4, data acquisition and data processing, wherein the acquisition is executed once every 50ms, the real-time pressure P and the real-time flow f of the air passage are acquired, and the concentration of carbon dioxide gas in the air passage is measured by using two channels;
step 5, packaging the respiratory data and performing wireless transmission;
step 6, key scanning, namely traversing every 10ms, judging whether a key instruction exists, changing program parameters according to the instruction and refreshing an interface of the display;
step 7, setting breathing parameters;
step 8, judging whether the fan is started, if so, entering step 9, otherwise, returning to step 6;
step 9, judging the state of the fan, and inquiring once every 50 ms;
step 10, respiratory phase identification and fan control;
step 11, whether the alarm condition is met or not is judged, the alarm is executed once every 30ms, if yes, the alarm is given, and the step 12 is carried out; otherwise, directly entering step 12;
step 12, displaying the working parameters by a display;
step 13, storing the breathing data, executing the storage task once every 200ms, and storing the breathing machine airway acquisition parameters in a proper format;
step 14, predicting airway pressure according to the stored respiratory data, and further predicting the respiratory state of the patient;
step 15, monitoring according to the stored respiratory data, judging whether accumulated liquid exists in the airway according to the real-time flow and the real-time pressure of the airway, if so, returning to the step 11, and triggering alarm;
wherein, the step 10 breathing phase identification specifically comprises:
step 10.1, judging whether the real-time pressure P of the air passage is larger than the pressure P when the mask is not worn0If yes, wearing the mask, otherwise, entering step 10.2;
step 10.2, starting inspiration, and synchronously providing an inspiratory positive pressure IPAP by the respirator;
step 10.3, judging whether the airway real-time pressure P maintaining time t is larger than the time defined by the respiratory arrestTdIf yes, entering step 10.10, otherwise, entering step 10.4;
step 10.4, judging whether the real-time pressure P of the air passage is greater than the high trigger air passage pressure limit value HTP and the real-time flow f of the air passage is less than the low trigger air passage flow limit value LTF, if so, entering step 10.5, otherwise, returning to step 10.2;
step 10.5, entering an expiratory phase, and enabling the patient to enter an expiratory state;
step 10.6, starting expiration, and synchronously providing positive expiratory pressure (EPAP) by the respirator;
step 10.7, judging whether the airway real-time pressure P maintaining time T is larger than the time T defined by the respiratory arrestdIf yes, entering step 10.10, otherwise, entering step 10.8;
step 10.8, judging whether the real-time pressure P of the air passage is smaller than the low trigger air passage pressure limit LTP and the real-time flow f of the air passage is larger than the high trigger air passage flow limit HTF, if so, entering step 10.9, otherwise, returning to step 10.6;
step 10.9, entering an inspiration phase, enabling the patient to enter an inspiration state, and returning to the step 10.2;
and step 10.10, judging apnea, alarming and taking off the mask.
The invention has the beneficial effects that:
1) the self-tuning processing mode reduces pressure fluctuation generated by the respiratory disturbance of the patient, eliminates the uncertainty of the model, has good self-adaptation and self-learning capabilities, has simple structure and high response speed, and can effectively adjust the control parameters in a complex process and a time-varying system;
2) the air pressure output by the breathing machine in a stable operation state can be quickly, accurately and comfortably switched between an expiration phase and an inspiration phase, and the air pressure is adjusted by the feedback control of the fan according to the error between a given pressure input value and a real-time monitored pressure value, so that the optimal control effect is obtained;
3) the breathing machine provides positive pressure of different levels to carry out breathing treatment in the processes of inspiration and expiration of a patient, and the switching of a breathing phase and the control of the output pressure of the fan are judged through feedback values of the pressure difference sensor and the pressure sensor in the working process of the breathing machine so as to achieve an ideal treatment effect;
4) the bi-level respirator system can generate a large amount of breathing data during working, and the current physical condition of a user can be reflected and predicted to a certain degree by researching and analyzing the data, so that the bi-level respirator system has important guiding significance on respiratory therapy; the method can well fit the fluctuation of the observed value sequence and analyze the prediction result, and the model has certain applicability to the prediction of the breathing machine airway pressure waveform;
5) the concentration measurement method is improved, the linearity degree of the measurement result is increased, uncontrollable errors are compensated, the robustness of the measurement result is increased, and low-frequency noise generated by concentration is eliminated;
6) the automatic detection algorithm for the effusion in the pipeline of the breathing machine based on the wavelet transform realizes the informatization and the digitization of the breathing machine and the automation of the effusion detection in the pipeline of the breathing machine.
Drawings
FIG. 1 is a flow chart of a method of controlling a ventilator in accordance with the present invention;
FIG. 2 is a flow chart of breath phase identification control according to the present invention;
FIG. 3 is a flow chart of the fan control of the present invention;
FIG. 4 is a flow chart of the present invention for estimating a respiratory state of a patient.
Detailed Description
The invention is further described with reference to the following figures and examples.
Embodiments of the present invention are illustrated with reference to fig. 1-4.
An intelligent respirator system, the respirator includes controller, blower, pipeline and sensor, the blower outputs the rotational speed under setting for the operating voltage, convert into the wind pressure to the pipeline through the blade, the sensor detects real-time flow, real-time pressure and concentration of real-time carbon dioxide of the air drain, and feedback to the controller; the controller realizes control of the breathing machine according to real-time flow of the air passage and real-time pressure of the air passage, the control comprises breathing phase identification control and self-setting processing of a stable running state of the fan, the breathing state of a patient is estimated according to collected real-time pressure data of the air passage, the concentration of carbon dioxide gas in the air passage is measured by using double channels, and whether effusion exists in the air passage is judged according to the real-time flow of the air passage and the real-time pressure of the air passage.
A method of controlling a ventilator, comprising the steps of:
step 1, initializing a system, and when a singlechip is electrified, starting to power on, wherein the system comprises a timer and an I2C, communicating with the SPI, initializing A/D conversion and initializing related variables;
step 2, configuring a processor interface;
step 3, initializing a display;
step 4, data acquisition and data processing, wherein the acquisition is executed once every 50ms, the real-time pressure P and the real-time flow f of the air passage are acquired, and the concentration of carbon dioxide gas in the air passage is measured by using two channels;
step 5, packaging the respiratory data and performing wireless transmission;
step 6, key scanning, namely traversing every 10ms, judging whether a key instruction exists, changing program parameters according to the instruction and refreshing an interface of the display;
step 7, setting breathing parameters;
step 8, judging whether the fan is started, if so, entering step 9, otherwise, returning to step 6;
step 9, judging the state of the fan, and inquiring once every 50 ms;
step 10, respiratory phase identification and fan control;
step 11, whether the alarm condition is met or not is judged, the alarm is executed once every 30ms, if yes, the alarm is given, and the step 12 is carried out; otherwise, directly entering step 12;
step 12, displaying the working parameters by a display;
step 13, storing the breathing data, executing the storage task once every 200ms, and storing the breathing machine airway acquisition parameters in a proper format;
step 14, predicting airway pressure according to the stored respiratory data, and further predicting the respiratory state of the patient;
and 15, judging whether accumulated liquid exists in the air passage or not according to the real-time flow and the real-time pressure of the air passage, if so, returning to the step 11, and triggering to alarm.
Wherein, the step 10 breathing phase identification specifically comprises:
step 10.1, judging whether the real-time pressure P of the air passage is larger than the pressure P when the mask is not worn0If yes, wearing the mask, otherwise, entering step 10.2;
step 10.2, starting inspiration, and synchronously providing an inspiratory positive pressure IPAP by the respirator;
step 10.3, judging whether the airway real-time pressure P maintaining time T is larger than the time T defined by the respiratory arrestdIf yes, entering step 10.10, otherwise, entering step 10.4;
step 10.4, judging whether the real-time pressure P of the air passage is greater than the high trigger air passage pressure limit value HTP and the real-time flow f of the air passage is less than the low trigger air passage flow limit value LTF, if so, entering step 10.5, otherwise, returning to step 10.2;
step 10.5, entering an expiratory phase, and enabling the patient to enter an expiratory state;
step 10.6, starting expiration, and synchronously providing positive expiratory pressure (EPAP) by the respirator;
step 10.7, judging whether the airway real-time pressure P maintaining time T is larger than the time T defined by the respiratory arrestdIf yes, entering step 10.10, otherwise, entering step 10.8;
step 10.8, judging whether the real-time pressure P of the air passage is smaller than the low trigger air passage pressure limit LTP and the real-time flow f of the air passage is larger than the high trigger air passage flow limit HTF, if so, entering step 10.9, otherwise, returning to step 10.6;
step 10.9, entering an inspiration phase, enabling the patient to enter an inspiration state, and returning to the step 10.2;
and step 10.10, judging apnea, alarming and taking off the mask.
The fan control process of step 10 is as follows:
step 10A.1, acquiring real-time pressure P and real-time flow f of an air passage;
step 10A.2, if the real-time pressure P of the air passage is smaller than the set working initial pressure, the system is in a boosting operation state;
if the real-time pressure P of the airway is greater than the set working initial pressure and less than the inspiratory positive pressure IPAP, the step 10A.3 is carried out;
if the real-time airway pressure P is greater than the inspiratory positive pressure IPAP and the hold time exceeds the high pressure threshold time ThThe system enters a pressure release state;
step 10A.3, if the real-time airway pressure P is less than the positive expiratory pressure EPAP and the maintenance time exceeds the air leakage threshold time TleakEntering a gas leakage compensation state, otherwise entering a step 10 A.4;
step 10A.4, if the real-time airway pressure P is greater than the inspiratory positive pressure IPAP and the hold time exceeds the high pressure threshold time ThIf not, the system enters a stable operation state;
and step 10A.5, judging whether the user stops treatment, if so, returning to the step 10A.1, otherwise, entering a running stop state.
The fan switches working states in real time according to the breathing condition of a user, and a finite state machine model of the fan is established, wherein the finite state machine model comprises a stable operation state, a pressure release state, a boosting operation state, an air leakage compensation state and a stop operation state;
(1) the pressure-boosting operation state is that the fan is accelerated uniformly and the real-time pressure P of the air passage is increased uniformly in the period from the mask wearing to the time when the real-time pressure P of the air passage reaches the set working initial pressure;
(2) the ventilator is in a normal working stage in a stable running state, the ventilator performs self-setting processing on sensor data in an expiratory positive pressure EPAP and an inspiratory positive pressure TPAP, and the output voltage controls the rotating speed of the fan;
(3) in the pressure release state, the respirator is in a high-pressure processing stage, after a high-pressure alarm value is identified in the state, the pressure is uniformly reduced to the suction phase positive pressure TPAP, and in the time period, the blower uniformly decelerates, and the real-time pressure P of the air passage is uniformly reduced;
(4) the air leakage compensation state, the air leakage treatment stage of the respirator, when entering the state, the rotating speed of the motor is adjusted, the output pressure of the fan is increased to be used as the pressure compensation of air leakage, and an alarm is given until entering the stable operation stage;
(5) and in the running stop state, the fan stops running, the breathing machine is in the non-working state, and the user instruction is waited in the state.
The stable operation state is a normal treatment stage of the respirator, the stable output of treatment pressure must be ensured, the control quantity output of the blower system is obtained through self-tuning treatment, the automatic adjustment of the blower speed is realized, and the effect of pressure self-adaptive control is achieved.
The finite-state machine model is a mathematical model of the operation state of the system in normal operation and the actions and the transfer among the states, and after the breathing condition of the user is obtained, the state of the fan is judged, if the finite-state machine model is in the state of the fan, the specific state and the fan task are further judged, and then data processing is carried out and the action of the fan is triggered to be executed;
the self-tuning process of the stable operation state in step 10a.4 is as follows:
step 10A.4.1, establishing a relation function G (t) between the angular speed omega (t) of the fan and the input voltage U (t),
Figure BDA0002480969220000071
wherein, L' is phase inductance; t is the time; r is resistance; j is the rotational inertia of the fan blade; z is a damping coefficient; k is a radical oftIs a torque coefficient; k is a radical ofeIs the back electromotive force coefficient;
step 10A.4.2, establishing a pipeline model,
when the fan works, the impeller drives the air to rotate, the air is gathered to the spiral shell along with the inertia effect, at the moment, a transient vacuum state is formed in the shell, the air suction port flows in fresh air under the action of atmospheric pressure, so that the effect of air delivery is achieved, in the air delivery channel of the ventilator fan, the total pressure at a certain position of the air channel at the same horizontal height can be kept unchanged, and the value of the total pressure is equal to the static pressure value P at a certain position of the air channelD(t) and dynamic pressure P at a certain position of the airwayJ(t) composition:
PF(t)＝PD(t)+PJ(t)；
PFthe output pressure of the fan is;
Figure BDA0002480969220000081
ρ is the air density; r is the radius of the fan impeller; λ is tip speed ratio;
static pressure value P at certain position of air passageD(t) measured by a pressure sensor;
dynamic pressure value P at certain position of air passageJ(t) is:
Figure BDA0002480969220000082
Figure BDA0002480969220000083
a is the sectional area of the air passage; qv(t) is the volume flow, c is the outflow coefficient, d is the orifice diameter of the orifice, β is the diameter ratio, Δ P (t) is the pressure difference;
step 10A.4.3, inputting the initial angular velocity omega of the fan0Converted into the initial voltage U by the relation function G (t) of step 10A.4.10Driving the fan to rotate;
step 10A.4.4, collecting static pressure value P of the air passage measured by the pressure sensorD(t), differential pressure Δ P (t) measured by the differential pressure sensor and volumetric flow Q measured by the flow sensorv(t) calculating the angular velocity ω (t) of the fan through the pipeline model of step 10 a.4.2;
step 10a.4.5, performing feedback control, obtaining an input voltage u (t) as a control quantity through the relation function g (t) in the step 10a.4.1, discretizing u (t) into u (k), and establishing a discretization control expression:
Figure BDA0002480969220000084
k is the sampling number, e (K) is the deviation of K times of sampling, U (K) is the voltage control quantity output in K times of sampling, KPIs a proportionality coefficient, KLIs an integral coefficient，KDIs a differential coefficient;
calculates an input control voltage increment deltau (k),
ΔU(k)＝KP(e(k)-e(k-1))+KLe(k)+KD[e(k)-2e(k-1)+e(k+2)]
and step 10A.4.6, adding the calculated voltage increment and the previous voltage control quantity to obtain the current voltage control quantity to drive the fan to rotate, judging whether the voltage control is finished, jumping out if yes, and returning to the step 10A.4.4 if not.
In consideration of various factors such as functions, performance and economy of the breathing machine, in order to reduce pressure fluctuation generated by respiratory disturbance of a patient and eliminate uncertainty, the invention adopts self-setting processing to control the output pressure of the fan when a breathing machine system stably runs, has good self-adaption and self-learning capabilities, simple structure and high response speed, well solves the problem that parameters of the traditional control method are not easy to adjust on line in real time, and can effectively adjust the control parameters in a complex process and a time-varying system.
Step 14, predicting airway pressure according to the stored respiratory data, and further predicting the respiratory state of the patient specifically comprises:
step 14.1, the respiratory data is preprocessed,
if the respiratory data sequence is in a linear trend, the respiratory data sequence is stabilized through first-order difference; if the respiratory data sequence is in a curve trend, performing second-order or third-order differential processing to remove the influence caused by the curve trend; if the respiratory data sequence is a fixed period variation trend, adopting a differential operation with the step length as the period length;
step 14.2, judging the sequence type,
let xtFor the value of the processed stationary sequence at time t, the covariance of two random variables at time k will be separated, i.e. the left-lagging autocovariance sequence gammakReferred to as the auto-covariance function,
γk＝Cov(xt，xt-k)，
γ0＝Var(xt)，
var () is a variance function, Cov () is a covariance function,
with the phase-separated time k as lag phase, xtHas an autocorrelation function of pk(k＝0，1，2，...)：
Figure BDA0002480969220000091
The error is measured using a criterion that minimizes the sum of squared errors, and the selected coefficient psi is then determinedkj(j ═ 1, 2.. times, k) is such that
Figure BDA0002480969220000092
At minimum, it is called psikj(j ═ 1, 2.. k.) is xtWherein E () represents a mathematical expectation;
a determination is made as to the type of smoothing sequence generated at step 14.1,
(1) if the autocorrelation function of the stationary sequence generated in step 14.1 has a tail characteristic and the partial autocorrelation function has a tail-cut characteristic, then the sequence is an autoregressive sequence;
(2) if the autocorrelation function of the stationary sequence generated in step 14.1 has a tail-biting characteristic and the partial autocorrelation function has a tailing characteristic, then the sequence is a moving average sequence;
(3) if the autocorrelation function and the partial autocorrelation function of the stationary sequence generated in step 14.1 both have a tail characteristic;
step 14.3, prediction is carried out according to the sequence type,
(1) autoregressive model
Figure BDA0002480969220000101
φiIs the regression coefficient of the autoregressive model, p is the regression order of the autoregressive model,
Figure BDA0002480969220000102
the variance of the random white noise is such that,tandsrepresenting random white noise at different timesSequence, xtAnd xsRespectively representing the values at the t moment and the s moment, and calculating to obtain a data value at the t moment according to p-order data before the t moment through an autoregressive model;
(2) moving average model
Figure BDA0002480969220000103
θiIf the moving average model regression coefficient is adopted, and q is the regression order of the moving average model, calculating to obtain a data value at the t moment according to the q order data before the t moment through the moving average model;
(3) the model of the auto-regressive moving average,
Figure BDA0002480969220000104
calculating to obtain a data value at the t moment according to q-order data before the t moment by an autoregressive moving average model;
and step 14.4, when t is t +1, judging whether the cycle number reaches the preset number n, if not, returning to the step 14.2, if so, ending the prediction, and generating a prediction sequence (x)1，x2，...，xn) By the prediction sequence (x)1，x2，...，xn) The respiratory state of the patient is estimated.
Wherein the step 14.4 is performed by the prediction sequence (x)1，x2，...，xn) The specific process of estimating the respiratory state of the patient is as follows:
at step 14.4.1, the estimation of the derivative,
the derivative of the estimated point is expressed by using half of the slope of the line connecting adjacent points on the left and right sides of the estimated point
Figure BDA0002480969220000111
Generating derivative estimation sequences
Figure BDA0002480969220000112
Figure BDA0002480969220000113
xhIs (x)1，x2，...，xn) The data of (1);
at step 14.4.2, the data segments,
setting a threshold parameter to judge the change amplitude of the derivative estimation sequence, if the derivative value of one point is less than the threshold value, considering that the change amplitude of the point is small compared with the previous point and basically has no change, the two points are divided into the same segment and have the same change sign, otherwise, considering that the point has obvious change, the adjacent two points are represented by different signs, and the derivative estimation sequence is represented by
Figure BDA0002480969220000114
Conversion into a sequence of symbolic representations (R)1，R2，...，Rn) The sequence notation is expressed as follows:
Figure BDA0002480969220000115
step 14.4.3, transforming the signature sequence
Will (R)1，R2，...，Rn) Conversion into a sequence of features
Figure BDA0002480969220000116
Wherein the content of the first and second substances,
Figure BDA0002480969220000117
Sjthe jth sub-sequence of the symbol sequence is shown. k is a radical ofjIs SjThe number of symbols of (a) is,
Figure BDA0002480969220000118
the symbols show that most adjacent points in the sequence have the same variation trend and are represented by the same trend symbols, adjacent points with the same symbols are combined, and the original sequence is segmented by combining the adjacent points with the same symbols. In order not to lose the information of the sequence length, each segment is composed ofTwo values constitute: symbol RjAnd the number of symbols kjIs represented by (R)j，kj) The division point from segment to segment is the sequence (R)1，R2，...，Rn) Where the symbol changes, T is expressed as
Figure BDA0002480969220000121
The symbol value in the symbol sequence is approximately represented by the change trend of the time sequence, and the number of the symbols represents the duration of the change trend;
14.4.4, obtaining characteristic sequences of the predicted data and the typical sample data through steps 14.4.1-14.4.3 respectively
Figure BDA0002480969220000122
And
Figure BDA0002480969220000123
wherein, typical sample data is breathing data under various breathing abnormal conditions;
step 14.4.5, calculate feature sequence
Figure BDA0002480969220000124
And
Figure BDA0002480969220000125
the distance of (2) is obtained, the segment matching is performed,
if the characteristic signs of the segments are different, the variation trends of the subsequences represented by the two segments are different, and the subsequences are dissimilar, the distance function is:
D(S1i，S2j)＝1，if R1i≠R2j，
wherein, S1iAnd S2jRespectively represent sequences
Figure BDA0002480969220000126
And sequence
Figure BDA0002480969220000127
Segment of (5), R1iAnd R2jRespectively, represent the segment S1iAnd S2jA corresponding symbolic representation;
if the characteristic symbols of the segments are the same, the characteristic symbols representing the two segments are the same and represent the same variation trend, when similarity measurement is carried out, the segments with the same symbols are matched and aligned, two subsequences can be considered to be similar, and the distance function is:
Figure BDA0002480969220000128
where the parameters ζ, ζ ∈ [0, 1) are the proportionality coefficients k1 and k2jRespectively, represent the segment S1iAnd S2jThe corresponding symbol number, max { } and min { } respectively represent a maximum function and a minimum function;
if the symbol sequences are not of equal length, no similar segment can match for the part beyond the longer sequence, so that it should be handled according to the first case of different symbols, the distance function is:
D(S1i，-)＝1，
step 14.4.6, calculate separately through step 14.4.5
Figure BDA0002480969220000129
And
Figure BDA00024809692200001210
each segment S1iAnd S2jAccumulating the distance between the two data blocks, if the accumulated value is within the threshold range, determining that the predicted data is matched with the typical sample data, outputting the name of the typical sample data and finishing matching; if the typical sample data are completely matched, the fact that proper typical sample data are not found is considered, the predicted data are saved, and matching is finished; if the accumulated value is outside the threshold range, the two are considered to be not matched and typical sample data is replaced, returning to step 14.4.3.
The specific steps of using the double channels to measure the concentration of the carbon dioxide gas in the air passage in the step 4 are as follows:
step 4.1, obtainingTaking the transmitted light intensity I (lambda) of the measurement channel and the reference channel1) And I (lambda)2)，λ1And λ2Respectively the wavelength of the measurement signal and the wavelength of the reference signal, lambda1Wavelength at carbon dioxide absorption peak, λ2Is a wavelength near the carbon dioxide absorption peak;
the dual-wavelength measurement method can eliminate the influence of instability and temperature drift of a time drift light source of the thermoelectric device on signals.
Step 4.2, establishing a concentration measurement formula of carbon dioxide:
Figure BDA0002480969220000131
wherein C is the concentration of carbon dioxide; l is the optical path length of the adapter; mu (lambda)1) Wavelength lambda1Absorption coefficient of carbon dioxide
The measuring scheme of the two channels can effectively convert logarithmic operation into linear operation, improve the linearity of measurement, effectively reduce the generation of errors and simultaneously improve the precision of measurement
4.3, filtering and denoising the image,
the measured carbon dioxide concentration is sampled to obtain n sampling data (C)1，C2，C3，...，Cn)，
Figure BDA0002480969220000132
Figure BDA0002480969220000133
The final data is represented by the data of the display,
Figure BDA0002480969220000134
to be the average of n-1 sample data,
Figure BDA0002480969220000135
is the average of n sampled data.
The algorithm for determining the effusion in the ventilator circuit is designed on the basis of each breathing cycle. The mechanical ventilation provided by the ventilator is performed through an exhalation tube and an inhalation tube, and thus, during use of the ventilator, an inhalation phase or a effusion of the respiratory phase may occur. In addition, the inspiratory waveform and the expiratory waveform of the patient are different, and if the difference between the expiratory phase and the inspiratory phase is ignored and the design of the effusion algorithm is uniformly carried out, the inspiratory phase and the expiratory phase in each respiratory cycle are respectively designed.
Wherein, the step 15 comprises the following steps:
step 15.1, the respiratory cycle is separated,
step 15.2, detecting the switching point of inspiration and expiration,
the breathing cycle starts with an inspiration phase and ends with an expiration phase, the effusion detection is respectively detected aiming at the inspiration phase and the expiration phase, the inspiration phase and the expiration phase are divided according to a first point less than zero in the flow rate time waveform, and when F (i) is equal to 0, the ith sampling point is an inspiration and expiration switching point;
step 15.3, wavelet decomposition and reconstruction of the pressure-time waveform,
step 15.3.1, intercepting the pressure-time waveforms of the inhalation phase and the exhalation phase from the pressure-time waveforms according to the breathing cycle and the inhalation-exhalation switching point determined in the step 1-2;
step 15.3.2, performing wavelet transformation on the pressure-time waveforms of the inhalation phase and the exhalation phase, as follows:
Figure BDA0002480969220000141
f (t) is a function of the pressure-time waveform, WTf() For discrete wavelet transform, k is the order, n is the decomposition level, ψ () is the mother function of the wavelet, 2kIs a scaling factor;
step 15.3.3, wavelet decomposition,
four-stage decomposition is carried out on the pressure time waveform respectively to obtain an approximate signal and a detail signal which are used for observing the characteristic signal centralized distribution condition of effusion in the breathing machine pipeline, a3Approximation signal representing a 3-level decomposition, di(i ═ 1, 2, 3) represents the detail signal of the ith stage;
step 15.3.4, the signal is reconstructed and,
S＝K(d1+d2+d3)，
wherein S is a reconstruction signal, K is a constant, preferably K ═ 0.05, 0.1, 0.5, 1, 2;
step 15.4, the threshold value is calculated,
the reconstructed signal S contains characteristic information capable of distinguishing pressure time waveforms in normal and effusion states, the effusion waveform is detected through setting a threshold value,
because the respiratory states of different patients and even each respiratory cycle of the same patient have differences, the threshold value is selected by considering the differences among individuals and the differences among individuals, and in order to avoid uncertainty caused by the differences, the threshold value is calculated by taking the mean value and the standard deviation calculated by the detail signal in the current reconstruction signal as the basis, as follows:
Figure BDA0002480969220000151
where μ is the mean of the detail signal, σ is the standard deviation of the detail signal, and T is the threshold.
And step 15.5, judging the effusion condition according to the reconstruction signals of the inspiratory phase and the expiratory phase obtained in the step 3 and the step 4 and a threshold value, and when the reconstruction signals exceed the threshold value, giving an alarm if the effusion exists in the surface airway.
The above-described embodiment merely represents one embodiment of the present invention, but is not to be construed as limiting the scope of the present invention. It should be noted that, for a person skilled in the art, several variations and modifications can be made without departing from the inventive concept, which falls within the scope of the present invention.

Claims (7)

1. An intelligent breathing machine system with breathing phase recognition control comprises a controller, a fan, a pipeline and a sensor, wherein the fan outputs rotating speed under set working voltage, the rotating speed is converted into wind pressure through a blade and then is transmitted to the pipeline, and the sensor detects real-time flow of an air passage, real-time pressure of the air passage and concentration of real-time carbon dioxide and feeds the real-time flow, the real-time pressure of the air passage and the concentration of the real-time carbon dioxide back to the controller; the controller realizes the control of the breathing machine according to the real-time flow of the air passage and the real-time pressure of the air passage, and the control comprises breathing phase identification control.
2. A ventilator control method for an intelligent ventilator system with respiratory phase recognition control as claimed in claim 1, comprising the steps of:
step 1, initializing a system, and when a singlechip is electrified, starting to power on, wherein the system comprises a timer and an I2C, communicating with the SPI, initializing A/D conversion and initializing related variables;
step 2, configuring a processor interface;
step 3, initializing a display;
step 4, data acquisition and data processing, wherein the acquisition is executed once every 50ms, the real-time pressure P and the real-time flow f of the air passage are acquired, and the concentration of carbon dioxide gas in the air passage is measured by using two channels;
step 5, packaging the respiratory data and performing wireless transmission;
step 6, key scanning, namely traversing every 10ms, judging whether a key instruction exists, changing program parameters according to the instruction and refreshing an interface of the display;
step 7, setting breathing parameters;
step 8, judging whether the fan is started, if so, entering step 9, otherwise, returning to step 6;
step 9, judging the state of the fan, and inquiring once every 50 ms;
step 10, respiratory phase identification and fan control;
step 11, whether the alarm condition is met or not is judged, the alarm is executed once every 30ms, if yes, the alarm is given, and the step 12 is carried out, otherwise, the step 12 is directly carried out;
step 12, displaying the working parameters by a display;
step 13, storing the breathing data, executing the storage task once every 200ms, and storing the breathing machine airway acquisition parameters in a proper format;
step 14, predicting airway pressure according to the stored respiratory data, and further predicting the respiratory state of the patient;
step 15, monitoring according to the stored respiratory data, judging whether accumulated liquid exists in the airway according to the real-time flow and the real-time pressure of the airway, if so, returning to the step 11, and triggering alarm;
wherein, the step 10 breathing phase identification specifically comprises:
step 10.1, judging whether the real-time pressure P of the air passage is larger than the pressure P when the mask is not worn0If yes, wearing the mask, otherwise, entering step 10.2;
step 10.2, starting inspiration, and synchronously providing an inspiratory positive pressure IPAP by the respirator;
step 10.3, judging whether the airway real-time pressure P maintaining time T is larger than the time T defined by the respiratory arrestdIf yes, entering step 10.10, otherwise, entering step 10.4;
step 10.4, judging whether the real-time pressure P of the air passage is greater than the high trigger air passage pressure limit value HTP and the real-time flow f of the air passage is less than the low trigger air passage flow limit value LTF, if so, entering step 10.5, otherwise, returning to step 10.2:
step 10.5, entering an expiratory phase, and enabling the patient to enter an expiratory state;
step 10.6, starting expiration, and synchronously providing positive expiratory pressure (PAP) by a breathing machine;
step 10.7, judging whether the airway real-time pressure P maintaining time T is larger than the time T defined by the respiratory arrestdIf yes, entering step 10.10, otherwise, entering step 10.8;
step 10.8, judging whether the real-time pressure P of the air passage is smaller than the low trigger air passage pressure limit LTP and the real-time flow f of the air passage is larger than the high trigger air passage flow limit HTF, if so, entering step 10.9, otherwise, returning to step 10.6;
step 10.9, entering an inspiration phase, enabling the patient to enter an inspiration state, and returning to the step 10.2;
and step 10.10, judging apnea, alarming and taking off the mask.
3. The control method of the ventilator according to claim 2, wherein the blower control process of step 10 is as follows:
step 10A.1, acquiring real-time pressure P and real-time flow f of an air passage;
step 10A.2, if the real-time pressure P of the air passage is smaller than the set working initial pressure, the system is in a boosting operation state;
if the real-time pressure P of the airway is greater than the set working initial pressure and less than the inspiratory positive pressure IPAP, the step 10A.3 is carried out;
if the real-time airway pressure P is greater than the inspiratory positive PAP and the hold time exceeds the high pressure threshold time ThThe system enters a pressure release state;
step 10A.3, if the real-time airway pressure P is less than the positive expiratory pressure EPAP and the maintenance time exceeds the air leakage threshold time TleakEntering a gas leakage compensation state, otherwise entering a step 10 A.4;
step 10A.4, if the real-time airway pressure P is greater than the inspiratory positive pressure IPAP and the hold time exceeds the high pressure threshold time ThIf not, the system enters a stable operation state;
and step 10A.5, judging whether the user stops treatment, if so, returning to the step 10A.1, otherwise, entering a running stop state.
4. The control method of the ventilator according to claim 3, wherein the fan switches the working state in real time according to the breathing condition of the user, and establishes a finite state machine model of the fan, which includes a stable operation state, a pressure release state, a boost operation state, an air leakage compensation state and a stop operation state;
(1) the pressure-boosting operation state is that the fan is accelerated uniformly and the real-time pressure P of the air passage is increased uniformly in the period from the mask wearing to the time when the real-time pressure P of the air passage reaches the set working initial pressure;
(2) the ventilator is in a normal working stage in a stable running state, the ventilator is in an expiratory positive pressure EPAP and an inspiratory positive pressure IPAP, the sensor data is subjected to self-setting processing, and the output voltage control quantity controls the rotating speed of the fan;
(3) in a pressure release state, the respirator is in a high-pressure processing stage, after a high-pressure alarm value is identified in the state, the pressure is uniformly reduced to an inspiratory positive pressure IPAP (initial pressure of inhalation), and in the time period, the blower uniformly decelerates, and the real-time pressure P of an air passage is uniformly reduced;
(4) the air leakage compensation state, the air leakage treatment stage of the respirator, when entering the state, the rotating speed of the motor is adjusted, the output pressure of the fan is increased to be used as the pressure compensation of air leakage, and an alarm is given until entering the stable operation stage;
(5) and in the running stop state, the fan stops running, the breathing machine is in the non-working state, and the user instruction is waited in the state.
5. The ventilator control method according to claim 4, wherein the steady operation state self-tuning process in step 10a.4 is as follows:
step 10A.4.1, establishing a relation function G (t) between the angular speed omega (t) of the fan and the input voltage U (t),
Figure FDA0002480969210000031
wherein, L' is phase inductance; t is the time; r is resistance; j is the rotational inertia of the fan blade; z is a damping coefficient; k is a radical oftIs a torque coefficient; k is a radical ofeIs the back electromotive force coefficient;
step 10A.4.2, establishing a pipeline model,
in the ventilator fan air delivery channel, the total pressure at a certain position of the air channel under the same horizontal height can be kept unchanged, and the value is represented by the static pressure value P at a certain position of the air channelD(t) and dynamic pressure P at a certain position of the airwayJ(t) composition: pF(t)＝PD(t)+PJ(t)；
PFThe output pressure of the fan is;
Figure FDA0002480969210000041
ρ is the air density; r is the radius of the fan impeller; λ is tip speed ratio;
a certain of the air passageAt a static pressure value PD(t) measured by a pressure sensor;
dynamic pressure value P at certain position of air passageJ(t) is:
Figure FDA0002480969210000042
Figure FDA0002480969210000043
a is the sectional area of the air passage; qv(t) is the volume flow, c is the outflow coefficient, d is the orifice diameter of the orifice, β is the diameter ratio, Δ P (t) is the pressure difference;
step 10A.4.3, inputting the initial angular velocity omega of the fan0Converted into the initial voltage U by the relation function G (t) of step 10A.4.10Driving the fan to rotate;
step 10A.4.4, collecting static pressure value P of the air passage measured by the pressure sensorD(t), differential pressure Δ P (t) measured by the differential pressure sensor and volumetric flow Q measured by the flow sensorv(t) calculating the angular velocity ω (t) of the fan through the pipeline model of step 10 a.4.2;
step 10a.4.5, performing feedback control, obtaining an input voltage u (t) as a control quantity through the relation function g (t) in the step 10a.4.1, discretizing u (t) into u (k), and establishing a discretization control expression:
Figure FDA0002480969210000044
k is the sampling number, e (K) is the deviation of K times of sampling, U (K) is the voltage control quantity output in K times of sampling, KPIs a proportionality coefficient, KLIs the integral coefficient, KDIs a differential coefficient;
calculates an input control voltage increment deltau (k),
ΔU(k)＝KP(e(k)-e(k-1))+KLe(k)+KD[e(k)-2e(k-1)+e(k+2)]
and step 10A.4.6, adding the calculated voltage increment and the previous voltage control quantity to obtain the current voltage control quantity to drive the fan to rotate, judging whether the voltage control is finished, jumping out if yes, and returning to the step 10A.4.4 if not.
6. The method according to claim 2, wherein step 14 predicts the airway pressure according to the stored breathing data, and further predicts the breathing state of the patient as:
step 14.1, the respiratory data is preprocessed,
if the respiratory data sequence is in a linear trend, the respiratory data sequence is stabilized through first-order difference; if the respiratory data sequence is in a curve trend, performing second-order or third-order differential processing to remove the influence caused by the curve trend; if the respiratory data sequence is a fixed period variation trend, adopting a differential operation with the step length as the period length;
step 14.2, judging the sequence type,
let xtFor the value of the processed stationary sequence at time t, the covariance of two random variables at time k will be separated, i.e. the left-lagging autocovariance sequence gammakReferred to as the auto-covariance function,
γk＝Cov(xt，xt-k)，
γ0＝Var(xt)，
var () is a variance function, Cov () is a covariance function,
with the phase-separated time k as lag phase, xtHas an autocorrelation function of pk(k＝0，1，2，...)：
Figure FDA0002480969210000051
The error is measured using a criterion that minimizes the sum of squared errors, and the selected coefficient psi is then determinedkj(j ═ 1, 2.. times, k) is such that
Figure FDA0002480969210000052
At minimum, it is called psikj(j ═ 1, 2.. k.) is xtWherein E () represents a mathematical expectation;
a determination is made as to the type of smoothing sequence generated at step 14.1,
(1) if the autocorrelation function of the stationary sequence generated in step 14.1 has a tail characteristic and the partial autocorrelation function has a tail-cut characteristic, then the sequence is an autoregressive sequence;
(2) if the autocorrelation function of the stationary sequence generated in step 14.1 has a tail-biting characteristic and the partial autocorrelation function has a tailing characteristic, then the sequence is a moving average sequence;
(3) if the autocorrelation function and the partial autocorrelation function of the stationary sequence generated in step 14.1 both have a tail characteristic;
step 14.3, prediction is carried out according to the sequence type,
(1) autoregressive model
Figure FDA0002480969210000061
φiIs the regression coefficient of the autoregressive model, p is the regression order of the autoregressive model,
Figure FDA0002480969210000064
the variance of the random white noise is such that,tandsrepresenting a random white noise sequence, x, at different timestAnd xsRespectively representing the values at the t moment and the s moment, and calculating to obtain a data value at the t moment according to p-order data before the t moment through an autoregressive model;
(2) moving average model
Figure FDA0002480969210000062
θiIf the moving average model regression coefficient is adopted, and q is the regression order of the moving average model, calculating to obtain a data value at the t moment according to the q order data before the t moment through the moving average model;
(3) the model of the auto-regressive moving average,
Figure FDA0002480969210000063
calculating to obtain a data value at the t moment according to q-order data before the t moment by an autoregressive moving average model;
and step 14.4, when t is t +1, judging whether the cycle number reaches the preset number n, if not, returning to the step 14.2, if so, ending the prediction, and generating a prediction sequence (x)1，x2，...，xn) By the prediction sequence (x)1，x2，...，xn) The respiratory state of the patient is estimated.
7. Method according to claim 6, characterized in that in step 14.4 the sequence (x) is predicted by means of the prediction1，x2，...，xn) Estimating the respiratory state of the patient, which comprises the following steps:
at step 14.4.1, the estimation of the derivative,
the derivative of the estimated point is expressed by using half of the slope of the line connecting adjacent points on the left and right sides of the estimated point
Figure FDA0002480969210000071
Generating derivative estimation sequences
Figure FDA0002480969210000072
Figure FDA0002480969210000073
xhIs (x)1，x2，...，xn) The data of (1);
at step 14.4.2, the data segments,
setting a threshold parameter to judge the change amplitude of the derivative estimation sequence, if the derivative value of one point is smaller than that of the previous point, considering that the change amplitude of the point is small and basically unchanged compared with the previous point, the two points are divided into the same segment and have the same change sign, otherwise, considering that the point is obviously changed,two adjacent points are represented by different signs, and the derivative estimation sequence is obtained
Figure FDA0002480969210000074
Conversion into a sequence of symbolic representations (R)1，R2，...，Rn) The sequence notation is expressed as follows:
Figure FDA0002480969210000075
at step 14.4.3, the signature sequences are transformed,
will (R)1，R2，...，Rn) Conversion into a sequence of features
Figure FDA0002480969210000076
Wherein the content of the first and second substances,
Figure FDA0002480969210000077
Sjdenoted is the jth subsequence, k, of the symbol sequencejIs SjThe number of symbols of (a) is,
Figure FDA0002480969210000078
for symbolic representation, most adjacent points in the sequence have the same variation trend and are represented by the same trend symbol, adjacent points with the same symbol are merged, and the original sequence is segmented by merging adjacent points with the same symbol. In order not to lose information of the sequence length, each segment consists of two values: symbol RjAnd the number of symbols kjIs represented by (R)j，kj) The division point from segment to segment is the sequence (R)1，R2，...，Rn) The point of change of the middle sign is
Figure FDA0002480969210000081
Is shown as
Figure FDA0002480969210000082
The symbol value in the symbol sequence is approximately represented by the change trend of the time sequence, and the number of the symbols represents the duration of the change trend;
14.4.4, obtaining characteristic sequences of the predicted data and the typical sample data through steps 14.4.1-14.4.3 respectively
Figure FDA0002480969210000086
And
Figure FDA0002480969210000087
wherein, typical sample data is breathing data under various breathing abnormal conditions;
step 14.4.5, calculate feature sequence
Figure FDA0002480969210000088
And
Figure FDA0002480969210000089
the distance of (2) is obtained, the segment matching is performed,
if the characteristic signs of the segments are different, the variation trends of the subsequences represented by the two segments are different, and the subsequences are dissimilar, the distance function is:
D(S1i，S2j)＝1，if R1i≠R2j，
wherein, S1iAnd S2jRespectively represent sequences
Figure FDA00024809692100000811
And sequence
Figure FDA00024809692100000810
Segment of (5), R1iAnd R2jRespectively, represent the segment S1iAnd S2jA corresponding symbolic representation;
if the characteristic symbols of the segments are the same, the characteristic symbols representing the two segments are the same and represent the same variation trend, when similarity measurement is carried out, the segments with the same symbols are matched and aligned, two subsequences can be considered to be similar, and the distance function is:
Figure FDA0002480969210000083
if R1i＝R2j
where the parameters ζ, ζ ∈ [0, 1) are the proportionality coefficients k1 and k2jRespectively, represent the segment S1iAnd S2jThe corresponding symbol number, max { } and min { } respectively represent a maximum function and a minimum function;
if the symbol sequences are not of equal length, no similar segment can match for the part beyond the longer sequence, so that it should be handled according to the first case of different symbols, the distance function is:
D(S1i，-)＝1，
step 14.4.6, calculate separately through step 14.4.5
Figure FDA0002480969210000084
And
Figure FDA0002480969210000085
each segment S1iAnd S2jAccumulating the distance between the two data blocks, if the accumulated value is within the threshold range, determining that the predicted data is matched with the typical sample data, outputting the name of the typical sample data and finishing matching; if the typical sample data are completely matched, the fact that proper typical sample data are not found is considered, the predicted data are saved, and matching is finished; if the accumulated value is outside the threshold range, the two are considered to be not matched and typical sample data is replaced, returning to step 14.4.3.
CN202010384776.3A 2020-05-07 2020-05-07 Breathing machine system with breathing phase identification control applied to intelligent medical system and control method Withdrawn CN111529847A (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202010384776.3A CN111529847A (en) 2020-05-07 2020-05-07 Breathing machine system with breathing phase identification control applied to intelligent medical system and control method

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202010384776.3A CN111529847A (en) 2020-05-07 2020-05-07 Breathing machine system with breathing phase identification control applied to intelligent medical system and control method

Publications (1)

Publication Number Publication Date
CN111529847A true CN111529847A (en) 2020-08-14

Family

ID=71970634

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202010384776.3A Withdrawn CN111529847A (en) 2020-05-07 2020-05-07 Breathing machine system with breathing phase identification control applied to intelligent medical system and control method

Country Status (1)

Country Link
CN (1) CN111529847A (en)

Cited By (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN113219042A (en) * 2020-12-03 2021-08-06 深圳市步锐生物科技有限公司 Device and method for analyzing and detecting components in human body exhaled air
CN113952566A (en) * 2021-10-22 2022-01-21 广西开蒙医疗科技有限公司 Breathing machine pressure control method, system, terminal and storage medium
CN114028677A (en) * 2021-12-01 2022-02-11 广东健奥科技有限公司 Breathing machine air pressure adjusting and monitoring system and application thereof
CN115995282A (en) * 2023-03-23 2023-04-21 山东纬横数据科技有限公司 Expiratory flow data processing system based on knowledge graph

Cited By (7)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN113219042A (en) * 2020-12-03 2021-08-06 深圳市步锐生物科技有限公司 Device and method for analyzing and detecting components in human body exhaled air
CN113952566A (en) * 2021-10-22 2022-01-21 广西开蒙医疗科技有限公司 Breathing machine pressure control method, system, terminal and storage medium
CN113952566B (en) * 2021-10-22 2024-04-09 广西开蒙医疗科技有限公司 Breathing machine pressure control method, system, terminal and storage medium
CN114028677A (en) * 2021-12-01 2022-02-11 广东健奥科技有限公司 Breathing machine air pressure adjusting and monitoring system and application thereof
CN114028677B (en) * 2021-12-01 2023-10-20 广东健奥科技有限公司 Breathing machine air pressure adjusting and monitoring system and application thereof
CN115995282A (en) * 2023-03-23 2023-04-21 山东纬横数据科技有限公司 Expiratory flow data processing system based on knowledge graph
CN115995282B (en) * 2023-03-23 2023-06-02 山东纬横数据科技有限公司 Expiratory flow data processing system based on knowledge graph

Similar Documents

Publication Publication Date Title
CN111529847A (en) Breathing machine system with breathing phase identification control applied to intelligent medical system and control method
US11027080B2 (en) System and method for determining ventilator leakage during stable periods within a breath
CN111657950A (en) Breathing machine system applied to intelligent medical treatment and capable of predicting breathing state of patient and control method thereof
US9254369B2 (en) Leak determination in a breathing assistance system
EP1996267B1 (en) Method and apparatus for improved flow limitation detection of obstructive sleep apnea
JP3477200B2 (en) In particular, a respiratory assist device for treating sleep apnea
JP4429599B2 (en) Device for supplying variable positive airway pressure
US5645053A (en) Auto CPAP system and method for preventing patient disturbance using airflow profile information
JP4918040B2 (en) Method and apparatus for treating Chainstokes respiratory disease
CN111658925A (en) Ventilator system with fan control in intelligent medical system and control method thereof
WO2021062737A1 (en) Method for evaluating volume responsiveness and medical device
CN111529849A (en) Breathing machine system with gas concentration measurement and applied to intelligent medical treatment and control method
CN107690310B (en) Non-invasive method for monitoring the respiratory state of a patient via continuous parameter estimation
CN112754465A (en) Method for estimating quasi-static compliance of lung under pressure-controlled mechanical ventilation
CN111529848A (en) Breathing machine system with effusion detection function in intelligent medical system and control method
AU2012258376B2 (en) Method and Apparatus for Improved Flow Limitation Detection of Obstructive Sleep Apnea
AU2015204360A1 (en) Method and Apparatus for Improved Flow Limitation Detection of Obstructive Sleep Apnea

Legal Events

Date Code Title Description
PB01 Publication
PB01 Publication
SE01 Entry into force of request for substantive examination
SE01 Entry into force of request for substantive examination
WW01 Invention patent application withdrawn after publication

Application publication date: 20200814

WW01 Invention patent application withdrawn after publication