A pressure control system and method for a turbo ventilator

Abstract

The application provides a pressure control system and method for a turbo ventilator, wherein the system comprises: a control module, which changes the rotating speed of the turbo according to the error between the airway pressure and the expected pressure signal, and makes the airway pressure track the expected pressure signal in the inhalation phase; a turbo module, which adjusts the airway pressure by changing the rotating speed of the turbo; characterized in that the pressure control system further comprises an observation module, which is used for estimating the total disturbance in the control process and feeding back the total disturbance to the control module, so as to compensate the rotating speed control of the turbo module and eliminate the total disturbance in the control process. Compared with the prior art, the application realizes shorter adjustment time and relatively smaller tracking error, improves the adaptability of the controller to different mechanical characteristics of the user's breathing system, thereby widening the parameter adaptation range, simplifies the design process of the control system, and improves the anti-interference ability of the control system.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of pressure control of a breathing machine, and in particular to a pressure control system and method for a turbo breathing machine. BACKGROUND

[0002] The working principle of a breathing machine is to realize mechanical ventilation by establishing a pressure difference between the airway and the alveoli. At present, most of the breathing machines on the market use a turbo device to generate the pressure difference to replace the external compressed air source. Controlling the pressure difference is the basis for the working of a turbo breathing machine. During the user's inhalation process, the prior art usually controls the turbo inside the breathing machine to change the user's airway pressure through algorithms such as proportional-integral-derivative (PID) control (the control logic of the prior art is shown in Figure 1 ), so that the airway pressure tracks the expected pressure signal (usually a square wave signal with a certain transition process) during the inhalation phase, but the existing control method cannot achieve satisfactory pressure tracking effect.

[0003] In order to solve the defects of the prior art in pressure tracking effect, the inventors have conducted a large amount of research and found that the reason why the prior art cannot achieve satisfactory pressure tracking effect is that the control of the turbo output pressure has the following two technical difficulties.

[0004] On the one hand, the operation of the turbo device involves quite complex fluid motion, and the relationship among its pressure, flow rate and rotational speed can be expressed in a simplified form as shown in the following equation:

[0005]

[0006] where P t is the pressure difference between the inlet and outlet of the turbo (used to represent the turbo output pressure), Q t is the output flow rate of the turbo outlet, N t is the rotational speed of the turbo; a1, a2, a3 are fitting coefficients which can be obtained by fitting the test data of the turbo. In the above relationship, P t , Q t and N t are coupled with each other, forming a complex nonlinear relationship, and the change of any one state will bring changes to the rest, so it is quite difficult to accurately control the output pressure of the turbo under different mechanical ventilation conditions, which also causes the problems of long adjustment time and large tracking error of the prior art.

[0007] On the other hand, the difference of the respiratory systems of different users also needs to be considered in the process of pressure control. The user group of the turbo breathing machine covers neonates, children and adults, and the mechanical characteristics of the respiratory systems of the users (reflected in the product of air resistance R and compliance C, denoted by symbol τRC The value of τ (expressed as τ = R x C, where R and C are the airway resistance and compliance, respectively) varies greatly with the age of the user or the disease suffered by the user, and the variation range usually exceeds 4 orders of magnitude (τ RC The wide variation range of the user characteristics increases the difficulty of pressure tracking control. In order to obtain relatively optimal control performance under different τ RC The existing technology usually needs to set the coefficients of the PID controller under different τ RC The design of such gain scheduling structure is very complex, and since the airway resistance and compliance of the user's respiratory system cannot be directly measured by sensors, the controller cannot obtain the changes in the characteristics of the controlled object, so the gain scheduling structure itself is difficult to establish and use.

[0008] In summary, the existing technology usually adopts a linear controller, and when facing the above-mentioned complex system (a series system composed of a turbine and a respiratory system) (including a nonlinear object, and the parameters of the controlled object will change greatly), it is difficult to achieve satisfactory results simply by relying on linear control. SUMMARY

[0009] To solve the above problems, the present application provides a pressure control system for a turbine ventilator, comprising: a control module that changes the speed of the turbine according to the error between the airway pressure and the desired pressure signal, and makes the airway pressure track the desired pressure signal during the inspiration phase; a turbine module that adjusts the airway pressure by changing the speed of the turbine; characterized in that the pressure control system further comprises an observation module, which is used to estimate the total disturbance in the control process and feed back the total disturbance to the control module to compensate for the speed control of the turbine module, so as to eliminate the total disturbance in the control process.

[0010] Specifically, the control module comprises a proportional control unit and a differential control unit (the control module can also comprise an integral control unit), wherein the proportional control unit and the differential control unit change the speed of the turbine in proportion to the error between the airway pressure and the desired pressure signal according to certain proportional and differential coefficients, and then realize the control of the airway pressure, wherein the proportional and differential coefficients are determined according to the mechanical characteristics of the respiratory system, and the mechanical characteristics of the respiratory system are the product of the airway resistance and the compliance.

[0011] Specifically, the process of "determining the proportional coefficient and differential coefficient based on the mechanical characteristics of the respiratory system" includes the following steps: (1) setting the initial mechanical characteristics of the respiratory system; the air resistance variation range of the respiratory system of general adult patients is R = 5 ~ 50 cmH2O·s / L, and the compliance variation range is C = 0.001 ~ 0.15 L / cmH2O. At the beginning of the controller design, the air resistance R = 20 cmH2O·s / L and the compliance C = 0.02 L / cmH2O are taken as the design origin within the above range. This set of parameters is close to the parameters of the commonly used simulated lung device "handheld splint lung", which is convenient for testing and verification; (2) taking the differential coefficient as zero, and taking a larger proportional coefficient until the airway pressure continues to oscillate or is close to diverging (the airway pressure overshoot situation is as follows). Figure 7 (as shown); (3) Reduce the proportional coefficient to half of the value taken in step (2) and observe the airway pressure overshoot: if the airway pressure still overshoots for several cycles, further reduce the proportional coefficient; if the airway pressure does not overshoot, gradually increase the proportional coefficient; adjust the proportional coefficient to make the pressure signal stabilize after one or two oscillations (the pressure signal at this time is as shown). Figure 8 (as shown), then fix and record the proportional coefficient at this time; (4) Gradually increase the differential coefficient. As the differential coefficient increases, the number of pressure signal oscillations and the overshoot will gradually decrease. When the oscillation period becomes less than 1 and the overshoot decreases to less than 10% (the pressure signal pattern at this time is as shown), Figure 9 (As shown), the differential coefficient is fixed and recorded at this point. Finally, the proportional and differential coefficients are fine-tuned to achieve better control quality.

[0012] Specifically, the "total disturbance" includes: the nonlinear relationship between the pressure, flow rate, and speed of the turbine module; large-scale changes in the mechanical characteristics of the breathing system; and possible external disturbances. The process by which the "observation module" estimates and eliminates the total disturbance includes the following steps:

[0013] (S1) The turbine module and breathing system are represented as the following nonlinear system:

[0014]

[0015] In the formula, x1 = P aw This refers to airway pressure; The rate of change of airway pressure; u t The parameters for turbine speed adjustment are derived from the pressure error; f(x1,x2) is a nonlinear function that includes the nonlinear relationship of the turbine, the uncertainty of the large-scale variation of the mechanical characteristics of the breathing system, and possible external disturbances; b0 is the parameter to be designed, used to match the approximate boost capacity of the turbine used.

[0016] (S2) Introduce an extended state variable x3, denoted as x3(t) = f(x1(t), x2(t)), and let The nonlinear system described in (S1) is extended as

[0017]

[0018] (S3) Establish an extended state observer for the extended system in (S2) as follows:

[0019]

[0020] wherein z1 is the airway pressure, z2 is the airway pressure rate of change, and z3 is the estimated value of the total disturbance; β1, β2, and β3 are three parameters to be designed for the observer; the bandwidth parameterization method is used to tune the parameters β1, β2, and β3 of the extended state observer; and e1 is the airway pressure observation error;

[0021] (S4) Discretize the extended state observer, denoted as The discretized structure is shown in the following formula:

[0022]

[0023] wherein k is the number of beats; h is the discretization step, which is selected according to the control interval of the controller; is the estimated value of the output signal; y(k) = P aw (k) is the measured value of the input signal;

[0024] (S5) Obtain the compensation parameter u eso for eliminating the total disturbance as follows:

[0025]

[0026] The turbine speed adjustment parameter n after eliminating the total disturbance is:

[0027]

[0028] The aforementioned "tuning the parameters β1, β2, and β3 of the extended state observer using the bandwidth parameterization method" in (S3) includes the following steps:

[0029] (S3-1) Let the expected characteristic equation of the third-order extended state observer be in the form of (s + ω0) 3 :

[0030] (s + ω0) 3 = s 3 + β1s 2 + β2s + β3, ​

[0031] The relationship between the observer parameter and the expected bandwidth ω0 is obtained as follows:

[0032]

[0033] (S3-2) determining the value of ω0 so that the extended state observer is not too sensitive to the sensor noise while ensuring the observation quality;

[0034] (S3-3) determining the to-be-designed parameters β1, β2, β3 of the extended state observer according to the value of ω0.

[0035] The process of determining the value of ω0 specifically includes the following steps:

[0036] (S3-2-1) approximating the closed-loop pressure control system as a first-order inertia link, and the corresponding transfer function G p (s) is as follows:

[0037]

[0038] In the formula, ω p is the system bandwidth, s = σ + jω is a complex variable, wherein the real part σ ∈ R + , the imaginary part ω ∈ R, R and R + represent the real number field and the positive real number field, respectively;

[0039] (S3-2-2) setting the expected regulation time T of the step response of the closed-loop system, and the regulation time T corresponds to the 0 rise to 95% of the steady-state amplitude, so that

[0040]

[0041] (S3-2-3) taking the observer bandwidth ω0 as at least two to three times of the expected bandwidth ω p of the closed-loop system, and ω0 can be further improved until the improvement of the value of ω0 is not beneficial to the control effect.

[0042] Correspondingly, the application also provides a pressure control method for a turbo ventilator, which comprises the following steps:

[0043] (SⅠ) control step: a control module changes the rotating speed of the turbo according to the error between the airway pressure and the expected pressure signal, so that the airway pressure tracks the expected pressure signal in the inspiration phase;

[0044] (SⅡ) observation step: an extended state observer is used to estimate the total disturbance in the control process to obtain a compensation parameter; the total disturbance includes the nonlinear relationship of the turbo, the uncertainty of the wide-range change of the mechanical characteristics of the respiratory system and possible external disturbances;

[0045] (SIII) Compensation step: the extended state observer feeds back the compensation parameter to the control module, and the control module compensates the control of the turbine speed to eliminate the total disturbance in the control process.

[0046] Specifically, the control module includes a proportional control link and a differential control link (which can also include an integral control link), wherein the proportional control link and the differential control link linearly change the speed of the turbine in proportion to the error between the airway pressure and the desired pressure signal according to certain proportional and differential coefficients, thereby achieving control of the airway pressure, wherein the proportional and differential coefficients are determined according to the mechanical characteristics of the respiratory system, and the mechanical characteristics of the respiratory system are the product of air resistance and compliance.

[0047] The technical effect of the present application is that, as can be seen from the examples below, compared with PID control or PD control alone, the present application on the one hand achieves a shorter adjustment time and a relatively smaller tracking error, and on the other hand improves the adaptability of the controller to different mechanical characteristics of the user's respiratory system, thereby widening the parameter adaptation range, avoiding the design difficulty and complexity caused by the gain scheduling structure of the prior art, and greatly simplifying the design process of the control system. In addition, since the ESO is introduced to consider the possible external disturbance, the present application can also improve the anti-interference ability of the control system.

[0048] The technical solutions of the present application are described in detail below in combination with the explanation of the principles:

[0049] In view of the problems existing in the prior art, the present application treats the nonlinear coupling of the turbine and the change of the mechanical characteristics of the user's respiratory system as the nonlinear dynamics and parameter uncertainty of the system respectively, designs an observer based on the idea of active disturbance rejection control, estimates the total disturbance of the nonlinear dynamics, parameter uncertainty and external disturbance using an extended state observer (ESO), compensates the closed-loop feedback control quantity on the basis of the PID control algorithm, makes the series system composed of the turbine and the respiratory system approach a double-integral system, reduces the control difficulty of the controlled object, makes the control parameters have a larger adaptation range, and at the same time improves the anti-interference ability of the system and reduces the tuning difficulty of the PID parameters.

[0050] 1. Pressure control process by the controller

[0051] As shown in Figure 1 , the prior art uses a PID controller alone to control the pressure of the turbine respirator. During the inhalation stage of the turbine respirator, the pressure tracking error aw (t) is obtained according to the measured pressure P pr (t) and the desired pressure signal P

[0052] e p (t) = P pr (t) - P aw (t)

[0053] Controlled pressure performance P aw is measured by sensors and fed back to the controller. The controller generates corresponding turbine speed adjustment parameter u p (t) according to the error e t (t),

[0054]

[0055] Where k p , k i , k d are proportional, integral and differential coefficients respectively. Assuming that the turbine has an ideal drive mechanism, the turbine drive mechanism will change the turbine speed N t in proportion to the turbine speed adjustment parameter u t (t), thereby achieving control of the output flow Q t and P aw .

[0056] In the embodiments of the present application, due to the introduction of ESO, the integral control (I) in the PID controller can be abandoned and replaced by a PD controller, so that the occurrence of integrator saturation problem can be avoided in actual engineering. The control effect comparison chart obtained by the inventor's comparison test of the different schemes of PID+ESO and PD+ESO is shown in Figure 6 .

[0057] The turbine speed control parameter after omitting integral control is

[0058]

[0059] The transfer function expression G c (s) of the controller is obtained by Laplace transform

[0060] G c (s) = sk d +k p (0.2)

[0061] 2. Introduce an extended state observer on the basis of PID or PD control algorithm

[0062] The principle of this invention lies in merging all unmodeled components of the system into a total disturbance within the framework of active disturbance rejection control (ADRC), and representing it using the system's extended state. An extended state observer (ESO) is then introduced to estimate the extended state based on the system's output. Once the extended state is estimated, the total disturbance can be eliminated in the control system, thereby eliminating nonlinear dynamics, parameter uncertainties, and external disturbances.

[0063] Specifically, considering that both the user's respiratory system and the turbine exhibit certain nonlinear dynamics, and that it is difficult to obtain accurate values ​​of the mechanical properties of the user's respiratory system, this invention will ( Figure 2-A The nonlinear dynamics of the series system consisting of the turbine and the user (in the B-type system), the uncertainty of the mechanical characteristics of the user's respiratory system, and external disturbances are combined into a total disturbance. An extended state observer is introduced to estimate the total disturbance, and real-time compensation is performed in the control input accordingly. This improves the adaptability and anti-interference ability to changes in the controlled object's parameters, and enhances the pressure control accuracy. In addition, this architecture can significantly reduce the difficulty of parameter tuning for the PD controller.

[0064] The pressure control principle block diagram of this invention based on ESO is shown below. Figure 2-A As shown in / B, where u t0 The original turbine speed adjustment parameters are output by the PD / PID controller. The turbine speed adjustment parameters are for compensation. The steps of the ESO-based pressure control process are described below:

[0065] (1) Establish an expansion system

[0066] Assumption Figure 2-A The system in / B, where the turbine and the user's breathing system are connected in series, can be simplified and linearized using feedback techniques to become a nonlinear system as shown below:

[0067]

[0068] Where, x1 = P aw For the user's airway pressure; The rate of change of airway pressure; u t0is the turbine speed adjustment parameter; f(x1, x2) is a nonlinear function containing the nonlinear relationship of the turbine described above, the uncertainty of the wide range of changes in the respiratory mechanics, and possible external disturbances, etc. unknown nonlinear link; b0 is a parameter to be designed, which is used to represent the boost capability brought by the unit turbine speed adjustment parameter, and matches the approximate boost capability of the turbine used (the selection of b0 is related to the boost capability of the turbine itself. It can be found from the above formula: ignoring the nonlinear term of the turbine, the airway pressure change rate is in linear relationship with the turbine speed adjustment parameter, wherein b0 is the proportional coefficient. Therefore, b0 represents the change of the airway pressure change rate that the turbine can bring under the unit turbine speed adjustment parameter. Through prior testing of the turbine characteristics, the approximate range of this value can be easily obtained, and one can be selected within the range. Generally, the selection of b0 is not sensitive, and b0 needs to be selected to be larger in the case of large disturbance and time delay).

[0069] An extended state variable x3 is introduced, denoted as x3(t) = f(x1(t), x2(t)), and denoted

[0070] Then the above system can be extended to a new system,

[0071]

[0072] (2) Establish an extended state observer

[0073] The extended state observer for the extended system is established as shown below:

[0074]

[0075] Wherein, z1, z2, z3 are the estimated values of the user's airway pressure, airway pressure change rate and total disturbance respectively; β1, β2, β3 are three parameters to be designed of the observer; e1 is the observation error of the user's airway pressure.

[0076] The characteristic equation corresponding to the extended state observer is:

[0077] s 3 +β1s 2 +β2s+β3,

[0078] In order to obtain ideal observation performance more simply, the bandwidth parameterization method is adopted to set the parameters to be designed β1, β2, β3 of the extended state observer. Specifically, all the closed-loop poles are distributed at -ω0, so that β1, β2, β3 become a function of ω0, and ω0 is the bandwidth of the observer. For a third-order ESO, the expected characteristic equation form is:

[0079] (s+ω0) 3

[0080] Let:

[0081] (s+ω0) 3 =s 3 +β1s 2 +β2s+β3,

[0082] The relationship between the observer parameters and the desired bandwidth ω0 is obtained as:

[0083]

[0084] Ideally, the higher the bandwidth parameter is set, the better the pressure tracking, disturbance rejection performance and sensitivity to parameter variations can be achieved, but in the actual system, the maximum bandwidth that can be achieved will be limited by sensor noise and dynamic uncertainty. Therefore, a certain trade-off needs to be made during parameter tuning. Specifically, the value of ω0 is usually related to the control performance requirements of pressure control. Assuming that the pressure regulation response time of the system needs to be within 150 ms (the step response is shown as Figure 10 p (s)

[0085]

[0086] Simulate the closed-loop pressure control system, then the closed-loop system bandwidth ω p = 25 rad / s is relatively close, then the bandwidth of the observer should be at least two to three times higher than the desired bandwidth of the closed-loop system, therefore, ω0 > 75 rad / s can meet the minimum requirement. On this basis, ω0 can be further increased until the increase in the value of ω0 is no longer beneficial to the control effect.

[0087] (3) Discretization of the state observer for continuous states

[0088] When implemented in engineering, the observer for continuous states needs to be discretized, let Then the discretization structure is shown in the following formula,

[0089]

[0090] Where k is the number of beats; β i , i = 1, 2, 3 are the observer pole placement parameters; h is the discretization step size; is the output signal estimate; y(k) = P aw (k) is the input signal measurement value.

[0091] 3. Obtain the turbine speed adjustment parameter that eliminates total disturbance

[0092] Turbine speed adjustment compensation parameter u​eso For

[0093]

[0094] So far, the total turbine speed adjustment parameter For

[0095] BRIEF DESCRIPTION OF DRAWINGS

[0096] Figure 1 The prior art PID pressure control principle diagram is shown in Fig. 1;

[0097] Figure 2-A The PD+ESO pressure control principle diagram of the present application is shown in Fig. 2;

[0098] Figure 2-B The PID+ESO pressure control principle diagram of the present application is shown in Fig. 3;

[0099] Figure 3 The effect comparison diagram of PID control alone and PID+ESO control of the present application is shown in Fig. 4;

[0100] Figure 4 The effect comparison diagram of PID control alone and PID+ESO control of the present application after changing the mechanical characteristics of the user's breathing system is shown in Fig. 5;

[0101] Figure 5 The effect comparison diagram of PID control alone and the present application after resetting the parameters for the changed mechanical characteristics of the user's breathing system is shown in Fig. 6;

[0102] Figure 6 The pressure control effect comparison diagram of the embodiment of the present application omitting the integral control link is shown in Fig. 7;

[0103] Figure 7 The airway pressure overshoot situation diagram caused by using a larger proportional coefficient when setting the proportional coefficient and the differential coefficient of the present application is shown in Fig. 8;

[0104] Figure 8 The diagram showing that the proportional coefficient of the present application reaches a suitable state is shown in Fig. 9;

[0105] Figure 9 The diagram showing that the proportional coefficient and the differential coefficient of the present application both reach a suitable state is shown in Fig. 10;

[0106] Figure 10 The diagram showing the expected closed-loop system step response in the embodiment of the present application is shown in Fig. 11. DETAILED DESCRIPTION

[0107] The following takes a pressure tracking numerical simulation of a turbine respirator as an example to give the specific implementation of the present technical method.

[0108] The user respiratory system resistance R = 20 cmH2O.s / L, and the compliance C = 0.02 L / cmH2O are set as the nominal parameter values of the user respiratory system mechanics.

[0109] First, in the numerical simulation environment, the parameters of the PID controller are tuned based on the nominal respiratory system mechanics parameters, and the relatively optimal control performance is obtained by adjusting the proportional, integral and differential coefficients. When it is found that the control effect cannot be improved by adjusting the parameters, the corresponding results are recorded as follows,

[0110] k p = 180, k i = 150, k d = 12.5,

[0111] Secondly, the extended state observer is established, and the bandwidth parameterization method is used to tune the observer parameters. In this embodiment, in order to quickly and accurately extract the disturbance and compensate, ω0 is set to 120 rad / s, which can ensure the observation quality and is not too sensitive to the sensor noise.

[0112] Thus, β i , i = 1, 2, 3 are obtained.

[0113]

[0114] The observer b0 represents the boost capability caused by the unit turbine speed adjustment parameter, and matches the approximate boost capability of the turbine used, and here b0 = 350 is selected. The discretization step is selected according to the control interval of the controller, and the control interval is 0.002 s in the simulation process, so h = 0.002. The parameters of the observer have great robustness, and generally a larger improvement can be observed as long as the approximate interval is taken; of course, more detailed parameter tuning is required for better control performance.

[0115] The designed parameters are substituted into equations (0.4) and (0.5) respectively to obtain the complete controller design results.

[0116] The following builds a pressure tracking mathematical model of the turbine respirator in the numerical simulation software, and uses specific embodiments to embody the technical advantages of the pressure control method based on the ESO proposed in this paper over the prior art.

[0117] First, under the above nominal respiratory system mechanics parameters, according to the controller design results, the pressure control effect comparison diagram between the present application and the conventional PID control can be obtained as shown in Figure 3 Figure 3 ​It can be obviously seen that the pressure control method based on the extended state observer proposed in the present application effectively shortens the regulation time compared with the PID control alone, so that the pressure rises faster and tracks to the steady state value.

[0118] Secondly, without changing any controller parameters, only the mechanical properties of the user's respiratory system are changed to air resistance R = 150 cmH2O·s / L and compliance C = 0.005 L / cmH2O, and the pressure control effect comparison diagram between the present application and the conventional PID control is shown in FIG. 3B. It can be seen that after changing the mechanical properties of the user's respiratory system, the pressure control method based on the extended state observer still has better regulation time and relatively smaller tracking error than the method of using the PID control alone. Figure 4

[0119] Thirdly, after changing the mechanical properties of the user's respiratory system, the pure PID control needs to be re-tuned to obtain better control effect. Therefore, the parameters of the pure PID controller are re-tuned to k p = 350, k i = 550, k d = 2, and the control parameters of the method proposed in the present application remain unchanged, and the effect comparison diagram between the re-tuned parameter PID control and the present application is shown in FIG. 3C. It can be seen that even after re-tuning the PID parameters, the pure PID still has a certain gap compared with the present application, in other words, compared with the PID control alone, the present application is relatively insensitive to the change of the mechanical properties of the user's respiratory system, and the parameter adaptation range is widened. Since the method proposed in the present application widens the parameter adaptation range, the design complexity of the gain scheduling structure can be greatly simplified. Figure 5

[0120] In addition, after introducing the ESO, the integral control (I) in the PID controller can be discarded and replaced by a PD controller. In this way, the occurrence of the integrator saturation problem can be avoided in actual engineering. For example, under the condition of air resistance R = 20 cmH2O·s / L and compliance C = 0.02 L / cmH2O, the separate PD control and the method of combining PD with ESO are used respectively, and the tracking effect is compared with that of the “PID+ESO” in the present application, and the details are shown in FIG. 4. It can be seen that the elimination of the integral control in the method proposed in the present application does not have too much influence on the tracking control effect, but the steady state error of the separate PD control is significantly increased. Figure 3 Figure 6

[0121] ​​​​From the above specific description of the present application, compared with the conventional PID control method, the pressure control method based on the extended state observer can improve the adaptability of the controller to different mechanical characteristics of the user's respiratory system, reduce the difficulty of PID parameter setting, and improve the anti-interference ability of the system.

[0122] Finally, it should be noted that the above examples are only used to illustrate the technical solutions of the present application and are not limiting. Although the present application has been described in detail with reference to the examples, those skilled in the art should understand that modifications or equivalent replacements of the technical solutions of the present application do not deviate from the spirit and scope of the present application, and they should be covered in the scope of the claims of the present application.

Claims

1. A pressure control system for a turbo ventilator, comprising: a control module, which changes the rotation speed of the turbo according to the error between the airway pressure and the desired pressure signal, and makes the airway pressure track the desired pressure signal during the inspiration phase; a turbo module, which adjusts the airway pressure by changing the rotation speed of the turbo; characterized in that the pressure control system further comprises an observation module, which estimates the total disturbance in the control process and feeds back the total disturbance to the control module to compensate for the rotation speed control of the turbo module so as to eliminate the total disturbance in the control process; the total disturbance includes the nonlinear relationship among the pressure, flow and rotation speed of the turbo module, the large-scale variation of the mechanical characteristics of the respiratory system, and possible external disturbances; the process of the observation module to estimate and eliminate the total disturbance comprises the following steps: (S1) representing the turbo module and the respiratory system as a nonlinear system as follows: where x1 = P aw is the airway pressure; is the airway pressure rate of change;u t is the turbine speed adjustment parameter derived from the pressure error; f(x1,x2) is a non-linear function including turbine non-linear relationship, uncertainty of wide range of respiratory system mechanics, and possible external disturbances; b0 is a design parameter to match the approximate pressure increasing capacity of the turbine used; (S2) Introduce an extended state variable x3, denoted as x3(t) = f(x1(t), x2(t)), and let The nonlinear system described in (S1) is extended as (S3) establishing an extended state observer for the extended system in (S2) as follows: wherein z1 is the airway pressure, z2 is the airway pressure change rate, and z3 is the estimated value of the total disturbance; β1, β2, β3 are three to-be-designed parameters of the observer, the to-be-designed parameters β1, β2, β3 of the extended state observer are tuned by using a bandwidth parameterization method; e1 is the airway pressure observation error; (S4) Discretize the extended state observer, denoted as The discretized structure is shown in the following equation. where k is the number of taps; h is the discretization step, chosen according to the control interval of the controller; is the output signal estimate; y(k) = P aw (k), is the input signal measurement; (S5) obtaining a compensation parameter u to eliminate the total disturbance eso is: then the turbine speed adjustment parameter after eliminating total disturbance is: the "tuning the to-be-designed parameters β1, β2, β3 of the extended state observer by using a bandwidth parameterization method" in (S3) comprises the following steps: (S3-1) Let the desired characteristic equation form of the third order extended state observer be (s + ω0) 3 is: (s + ω0) 3 = s 3 + β1s 2 + β2s + β3, the relationship between the observer parameters and the desired bandwidth ω0 is obtained as follows: (S3-2) determining the value of ω0 so that the extended state observer is not too sensitive to the sensor noise while ensuring the observation quality; (S3-3) determining the to-be-designed parameters β1, β2, β3 of the extended state observer according to the value of ω0; the process of "determining the value of ω0" in (S3-2) comprises the following steps: (S3-2-1) The closed-loop pressure control system is approximated as a first-order inertia link, and the corresponding transfer function G p (s) is: where ω p is the closed loop system bandwidth, s = σ + jω is a complex variable with real part σ ∈ R + and imaginary part ω ∈ R, R and R + denote the real and positive real number domains, respectively; (S3-2-2) Set the desired regulation time T for the step response in which the steady-state amplitude of the closed-loop system rises from 0 to 95%, then ω p The relationship with T is: (S3-2-3) Take the observer bandwidth ω0as the closed-loop system bandwidth ω corresponding to the desired regulation time T p two to three times, and further increase ω0on this basis until the increase in the value of ω0is no longer beneficial to the control effect.

2. The pressure control system according to claim 1, characterized in that the "control module" comprises a proportional control unit and a differential control unit, wherein the "proportional control unit and the differential control unit" linearly change the rotation speed of the turbo in proportion to the error between the airway pressure and the desired pressure signal according to certain proportional and differential coefficients, thereby realizing the control of the airway pressure, wherein the "proportional and differential coefficients" are determined according to the mechanical characteristics of the respiratory system, and the "mechanical characteristics of the respiratory system" is the product of the air resistance and the compliance.

3. The pressure control system of claim 2, wherein, the process of "determining the proportional and differential coefficients according to the mechanical characteristics of the respiratory system" comprises the following steps: (1) setting the initial mechanical characteristics of the respiratory system; (2) taking the differential coefficient as zero and taking a larger proportional coefficient until the airway pressure continuously oscillates or approaches divergence; (3) reducing the proportional coefficient to half of the value taken in step (2) and observing the airway pressure overshoot: if the airway pressure still overshoots for multiple cycles, further reduce the proportional coefficient; if the airway pressure does not overshoot, gradually increase the proportional coefficient; The proportional coefficient is fixed and recorded when the pressure signal is stabilized after one or two oscillations; (4) The differential coefficient is gradually increased. With the increase of the differential coefficient, the oscillation frequency and overshoot of the pressure signal gradually decrease. When the oscillation period is below 1 and the overshoot is reduced to below 10%, the differential coefficient at this time is fixed and recorded.

4. The pressure control system of claim 2, wherein, The control module further comprises an integral control unit.

5. A pressure control method for a turbo ventilator, which is executed by the pressure control system of claim 1, comprising the following steps: (SⅠ) Control step: the control module changes the speed of the turbine according to the error between the airway pressure and the expected pressure signal, and the airway pressure tracks the expected pressure signal in the inspiration phase; (SⅡ) Observation step: the total disturbance in the control process is estimated by the extended state observer to obtain compensation parameters; the total disturbance includes the uncertainty of the nonlinear relationship of the turbine, the wide range of changes in the mechanical properties of the respiratory system, and possible external disturbances; (SⅢ) Compensation step: the extended state observer feeds back the compensation parameters to the control module, and the control module compensates the speed control of the turbine to eliminate the total disturbance in the control process.

6. The pressure control method of claim 5, wherein The control module comprises a proportional control link and a differential control link, wherein The proportional control link and the differential control link change the speed of the turbine in proportion to the error between the airway pressure and the expected pressure signal according to a certain proportional coefficient and a differential coefficient, thereby realizing the control of the airway pressure, wherein The proportional coefficient and the differential coefficient are determined according to the mechanical properties of the respiratory system, and the mechanical properties of the respiratory system are the product of air resistance and compliance.

7. The pressure control method of claim 6, wherein The control step further comprises an integral control link.

Images