A method for calibrating and controlling a mixed oxygen proportional valve

Abstract

This application provides a method for calibrating and controlling a mixed-oxygen proportional valve, comprising: consulting a current-flow rate correspondence table of the mixed-oxygen proportional valve based on the target oxygen flow rate to obtain the feedforward opening current of the mixed-oxygen proportional valve; employing an active disturbance rejection controller to establish a control loop with an oxygen concentration sensor as the closed loop, and performing real-time control of the input current of the mixed-oxygen proportional valve; the current-flow rate correspondence table of the mixed-oxygen proportional valve is obtained by performing dual-thread calibration on the mixed-oxygen proportional valve. The advantages of this application are: dual-thread calibration of the mixed-oxygen proportional valve allows for accurate control of both the increasing and decreasing oxygen concentration processes; the use of an active disturbance rejection control algorithm not only eliminates the experimental process of adjusting PID parameters but also meets the control and measurement accuracy requirements of the proportional valve-oxygen concentration control system, ensuring that both are within the error range.

Description

Technical Field

[0001] This application belongs to the field of respiratory humidification therapy devices, specifically relating to a method for calibrating and controlling a mixed oxygen proportional valve. Background Technology

[0002] High-flow humidified respiratory therapy devices, as respiratory treatment equipment, need to provide patients with a mixed gas with adjustable oxygen concentration to normalize blood oxygenation and achieve therapeutic effects. For high-flow humidified respiratory therapy devices that combine a proportional valve and a fan, both the proportional valve and the fan must be opened simultaneously during oxygen mixing. Since the oxygen source is a strong gas source, the opening degree and response time of the proportional valve directly affect the speed and stability of the mixed oxygen concentration, and also have a certain impact on the fan control. In existing technologies, oxygen concentration monitoring is generally carried out using oxygen concentration sensors, but these sensors have a slow response, resulting in a slow response in oxygen concentration control. Furthermore, there is a discrepancy between the current and flow rate curves of the mixed oxygen proportional valve, thus requiring calibration of the mixed oxygen proportional valve.

[0003] Currently, the calibration method for proportional oxygen valves primarily uses the current-flow rate curve. However, the calibration process typically only calibrates the current rise curve. Proportional valves often exhibit a hysteresis loop, meaning there may be a significant deviation between the current decrease and rise processes. This results in substantial differences in oxygen concentration as it decreases, leading to slow oxygen concentration control response. After flow rate calibration, it's necessary to convert oxygen concentration to oxygen flow rate, and then apply traditional PID control for oxygen concentration. This control method uses an oxygen concentration sensor as feedback, resulting in a slow response. Furthermore, the control parameter tuning relies heavily on engineer experience, making the process cumbersome and unable to cover all equipment operating conditions. Summary of the Invention

[0004] The purpose of this application is to overcome the shortcomings of existing oxygen mixing proportional valves, such as slow response, cumbersome process of control parameter tuning relying on engineers' experience, and inability to cover all equipment operating conditions.

[0005] To achieve the above objectives, this application proposes a method for calibrating and controlling a mixed-oxygen proportional valve, comprising:

[0006] Based on the target oxygen flow rate, consult the current-flow rate correspondence table of the oxygen mixing proportional valve to obtain the feedforward opening current of the oxygen mixing proportional valve.

[0007] An active disturbance rejection controller is used to implement a control loop with an oxygen concentration sensor as the closed loop, and the input current of the oxygen mixing proportional valve is controlled in real time.

[0008] The current-flow rate correspondence table of the oxygen mixing proportional valve is obtained by performing dual-thread calibration on the oxygen mixing proportional valve.

[0009] As an improvement to the above method, the oxygen mixing proportional valve is subjected to dual-thread calibration, which includes flow rate increase thread calibration and flow rate decrease thread calibration.

[0010] As an improvement to the above method, the process of calibrating the flow-increasing thread includes:

[0011] The opening of the oxygen mixing proportional valve gradually increases from 0. When the flow sensor detects that the flow rate is greater than the set threshold, the opening is recorded as the critical opening of the oxygen mixing proportional valve, Valve_Open.

[0012] Adjust the opening of the oxygen mixing proportional valve to the maximum value, then gradually reduce the opening and record the opening when the flow sensor is lower than the set maximum flow. This opening is the maximum opening of the oxygen mixing proportional valve, Valve_Max.

[0013] The oxygen mixing proportional valve is further divided into N equal parts between the critical opening and the maximum opening, that is, the interval between each opening is: Valve_Delta = (Valve Max -Valve Open The value is set to y1, y2, ..., yN, where N>0, and the opening is set according to this interval Valve_Delta. Data is recorded simultaneously. The opening is set as {Valve_Open+Value_Delta, Valve_Open+2*Value_Delta, ..., Valve_Open+N*Value_Delta}, and the corresponding flow rate is {y1, y2, ..., yN}. N}

[0014] As an improvement to the above method, the process of traffic reduction thread calibration includes:

[0015] The opening of the oxygen mixing proportional valve gradually decreases from its maximum value. When the flow sensor detects that the flow rate is less than the set threshold, the current is recorded as the critical opening of the oxygen mixing proportional valve, Valve_Open.

[0016] Set the opening of the oxygen mixing proportional valve to 0, then gradually increase the opening. Record the opening when the flow sensor is higher than the set maximum flow rate. This opening is the maximum opening of the oxygen mixing proportional valve, Valve_Max.

[0017] The oxygen mixing proportional valve is further divided into N equal parts between the critical opening and the maximum opening, that is, the interval between each opening is: Valve_Delta = (Value Max -Valve OpenThe value is set to ) / N, where N>0, and the opening is set according to this interval Valve_Delta. Data is recorded simultaneously. The opening is set as {Valve_Open+Value_Delta,Valve_Open+2*Value_Delta,…,Valve_Open+N*Value_Delta}, and the corresponding flow rate is {z1,z2,…z N}

[0018] As an improvement to the above method, the control loop with the oxygen concentration sensor as the closed loop includes the following control process:

[0019] The oxygen mixing proportional valve obtains its initial opening based on the feedforward opening current that has been limited by the limiting module;

[0020] The real-time oxygen concentration is measured by an oxygen concentration sensor, and the oxygen concentration error is obtained.

[0021] The active disturbance rejection controller is activated, and the oxygen concentration error is used as the control target. The error e1 is obtained by comparing it with the output value of the linear extended state observer. After passing through the linear state error feedback module, the control quantity u0 is output. After compensation by the total disturbance estimate z2 obtained by the linear extended state observer, the final control quantity u is obtained by passing through the reciprocal of the controller gain. u is then passed through the limiting module to output the compensation control opening of the oxygen mixing proportional valve, and finally the oxygen concentration is controlled.

[0022] As an improvement to the above method, the linear state error feedback module adopts a PD controller, whose controller transfer function is:

[0023] G(s)=k p +k d s

[0024] Where, k p k is the proportionality coefficient. d is the differential coefficient; s is the complex plane independent variable; the proportional coefficient and differential coefficient are obtained by the ZN tuning method.

[0025] As an improvement to the above method, the linearly extended state observer is designed as follows:

[0026]

[0027] Where β1 and β2 represent the pole placement parameters of the linearly extended state observer, determining the system convergence speed; u represents the system control law; and z is the state variable output matrix of the linearly extended state observer, z = [z1z2]. TT represents transpose; z1 tracks the real-time oxygen concentration y output by the system. After the dynamic control process converges, z1 tends towards y; z2 tracks the total disturbance. After the dynamic control process converges, z2 tends towards the total disturbance; b0 = 2.18; This represents the rate of change of the output matrix z of the state variable; This represents the estimated real-time oxygen concentration y of the system.

[0028] Compared with existing technologies, the advantages of this application are:

[0029] 1. The oxygen mixing proportional valve is calibrated in two threads, which can accurately control the process of oxygen concentration increasing from small to large and decreasing from large to small.

[0030] 2. The adoption of active disturbance rejection control algorithm not only eliminates the experimental process of adjusting PID parameters, but also meets the control accuracy and measurement accuracy requirements of the proportional valve-oxygen concentration control system, keeping them within the error range. Attached Figure Description

[0031] Figure 1 The diagram shown is a flowchart of the calibration and control method for a mixed oxygen proportional valve.

[0032] Figure 2 The diagram shows the flowchart of the dual-thread calibration method for the oxygen mixing proportional valve.

[0033] Figure 3 The diagram shown is a block diagram for controlling oxygen concentration. Detailed Implementation

[0034] The technical solution of this application will be described in detail below with reference to the accompanying drawings.

[0035] like Figure 1 As shown, the oxygen mixing proportional valve calibration and control method provided in this application first performs dual-thread calibration of the oxygen mixing proportional valve. Using dual-thread calibration, high-flow humidified breathing therapy achieves more accurate oxygen flow control during oxygen concentration control, thereby improving the response time of oxygen concentration control. Furthermore, in the control loop with oxygen concentration sensor as feedback, oxygen concentration control is performed using the oxygen concentration sensor as a closed-loop control loop, avoiding the cumbersome parameter tuning process of PID control, while achieving rapid response and no overshoot in the dynamic process.

[0036] like Figure 2 As shown, the dual-thread calibration process of the oxygen mixing proportional valve includes flow rate increase thread calibration and flow rate decrease thread calibration; among which,

[0037] The flow rate increase calibration process is as follows: The proportional valve opening is gradually increased from 0. When the flow sensor detects a flow rate > 0.5 LPM, this opening is recorded as the critical opening of the proportional valve, Valve_Open. The proportional valve opening is then increased to its maximum value, and the opening is gradually decreased. The opening when the flow sensor detects a flow rate below the maximum flow rate of 0.5 LPM is recorded. This opening is the maximum opening of the oxygen mixing proportional valve, Valve_Max. The proportional valve is further divided into N equal parts between the critical opening and the maximum opening, i.e., the interval between each opening is: Valve_Dalta = (Value... Max -Valve Open The value is set to y1, y2, ..., yN, where N>0, and the opening is set according to this interval Valve_Delta. Data is recorded simultaneously. The opening is set as {Valve_Open+Value_Delta, Valve_Open+2*Value_Delta, ..., Valve_Open+N*Value_Delta}, and the corresponding flow rate is {y1, y2, ..., yN}. N}

[0038] The calibration process for the flow rate decrease thread is the reverse of the flow rate decrease thread described above, including: gradually decreasing the proportional valve opening from its maximum value; when the flow sensor detects a flow rate < 0.5 LPM, recording this current as the critical opening of the proportional valve, Valve_Open; adjusting the proportional valve opening to 0; then gradually increasing the opening; recording the opening when the flow sensor detects a flow rate higher than the maximum flow rate by 0.5 LPM; this opening is the maximum opening of the oxygen mixing proportional valve, Valve_Max; further dividing the proportional valve between the critical opening and the maximum opening into N equal parts, i.e., the interval between each opening is: Valve_Delta = (Value Max -Valve Open The value is set to ) / N, where N>0, and the opening is set according to this interval Valve_Delta. Data is recorded simultaneously. The opening is set as {Valve_Open+Value_Delta,Valve_Open+2*Value_Delta,…,Valve_Open+N*Value_Delta}, and the corresponding flow rate is {z1,z2,…z N}

[0039] The opening degree of each proportional valve corresponds to the amount of current allocated to it.

[0040] After calibration, the dual-thread data is subjected to three-point linear smoothing, and then piecewise linear processing is performed to obtain the current magnitude that should be given under different flow rates during the rising and falling processes.

[0041] After the flow rate calibration is completed, the current-flow rate curve is used as a feedforward for oxygen concentration control and added to the output of the oxygen concentration controller to improve the speed of oxygen concentration control.

[0042] Based on the user-defined oxygen concentration SetO2, the following can be obtained from the formula for converting oxygen concentration, total flow rate Flow, and oxygen flow rate O2Flow:

[0043]

[0044] Substitute the target oxygen flow rate Tar_O2Flow into the set oxygen concentration value in equation (1), and use the linearized lookup table method to look up the current-flow correspondence table obtained from the dual-thread calibration process of the oxygen-mixing proportional valve to obtain the feedforward opening value for the oxygen concentration control of the proportional valve. The constants in equation 1 are set by the algorithm.

[0045] Based on this, to ensure error-free oxygen concentration control, a control loop with the oxygen concentration sensor as the closed loop is implemented. This control algorithm employs an ADRC (Active Disturbance Rejection Control) controller, avoiding the cumbersome parameter tuning process of PID control, and ensuring system overshoot-free operation while meeting the requirements for rapid system adjustment.

[0046] The control block diagram of the control loop with the oxygen concentration sensor as the closed loop is shown in Figure 3:

[0047] After the proportional valve provides the flow rate based on the feedforward opening, the flow rate passes through the limiting module to obtain the final proportional valve output opening. The real-time oxygen concentration is measured by the oxygen concentration sensor, yielding the oxygen concentration error O2Error. At this point, ADRC control is activated, using the oxygen concentration error as the control target. This error is compared with the output value of the Linear Extended State Observer (LESO) to obtain the error e1. This error is then fed back through the designed linear state error feedback module, which uses a classic PD controller with the controller transfer function G(s) = k p +k d s, where k p k is the proportionality coefficient. d The differential coefficient is obtained by the ZN (Ziegler-Nichols) tuning method, and s is the complex plane independent variable. Its output is the control quantity u0. After compensation by the total disturbance estimate z2 obtained by the linear extended state observer, it is then passed through the reciprocal of the controller gain to obtain the final control quantity u. u is then passed through the output proportional valve of the limiting module to compensate for the control opening, ultimately achieving fast and accurate control of oxygen concentration.

[0048] To design a linear expansion state observer, a rough mathematical model of the proportional valve needs to be established, and the step response method is selected here.

[0049] Based on the step response method, a proportional valve is designed to step from the critical opening to 10% of the maximum opening, denoted as Δu. The continuous response curve of the oxygen concentration is collected by an oxygen concentration sensor. The initial oxygen concentration value is denoted as y(0); after the system stabilizes, the oxygen concentration value is denoted as y(∞). The response curve resembles a typical first-order inertial element with a delay element. The system model can be established using the two-point method, that is, selecting the oxygen concentration y corresponding to different time points during the rising phase of the response curve. * (t) Establish the equations for the proportional valve-oxygen concentration system:

[0050]

[0051] Where t1 and t2 are two time points selected during the rising phase of the response curve; y(t) is the flow rate corresponding to time t; y * (t) represents the ratio of real-time flow rate to the oxygen concentration value after stabilization; τ represents the delay time; and T represents the time constant.

[0052] Typical value: y * (t1)=0.39,y * (2) = 0.63, therefore:

[0053]

[0054] After solving, we find that the controlled object gain K = 1.24, the controlled object time constant T = 0.57, and the controlled object time delay τ = 0.000107. Therefore, the proportional valve-oxygen concentration transfer function model is actually a typical first-order inertial element, and its transfer function expression is:

[0055]

[0056] In the formula, s is the complex plane independent variable, which is equivalent to a differential operator, transforming G(s) into the form of a differential equation:

[0057]

[0058] Corresponding to this with the self-disturbance paradigm, we get:

[0059]

[0060] By combining equations (5) and (6), we can see that b0 = 2.18 and the total disturbance f = -1.75y.

[0061] Select state variables x1 and x2, where, make The derivative of the total system disturbance, where y is the system oxygen concentration output, is converted into a state-space expression:

[0062]

[0063] In the formula: C = [1,0], D = [0]

[0064] The corresponding linear expansion state controller is designed as follows:

[0065]

[0066] In the formula: z is the state variable output matrix of the linearly extended state observer, z = [z1z2] T The output z1 tracks the oxygen concentration setpoint x1, which is the real-time oxygen concentration y output by the system. After the dynamic control process converges, z1 tends towards y. z2 tracks x2, which is the total disturbance f. After the dynamic control process converges, z2 tends towards f. L = [β1β2] T If parameters are configured for the two undetermined poles of the linearly extended state observer, then the linearly extended state observer is designed as follows:

[0067]

[0068] Where β1 and β2 represent the pole placement parameters of the linearly extended state observer, which determine the system convergence speed; u represents the system control law; This represents the rate of change of the output matrix z of the state variable; This represents the estimated real-time oxygen concentration y of the system.

[0069] This linearly extended state observer can obtain the total perturbation f,z2→f, and introduce the total perturbation f into u.

[0070]

[0071] Substituting equation (8) into equation (7), we get: The above model can then be transformed into a single-integral model, whose open-loop transfer function is: G1(s) = ω c / s, where ω c Let be the system bandwidth, s be a complex variable, s = σ + jω, σ ∈ R + Let ω be the real part and ω∈R be the imaginary part, and R and R + Let represent the real number field and the positive real number field, respectively. The closed-loop transfer function is then:

[0072]

[0073] Time constant T = 1 / ω c Based on the time-domain system response, the closed-loop output of the system can reach 99% of the oxygen concentration setpoint at 5T. According to the oxygen concentration control requirements, the setpoint is reached in 1 second. Therefore, ω... c= 5 rad / s.

[0074] Design the bandwidth ω of the linearly extended state observer. o Since the controlled object is set as a single integral model, the observer bandwidth needs to be larger than the controller bandwidth, so it can be set to (3~5)*ω. c The specific method depends on the system. In the proportional valve-oxygen concentration control process, a 5ω value is used. c According to the bandwidth method, we know that:

[0075]

[0076] To verify the above calibration and control methods, the ventilation flow rate of the high-flow humidification therapy device was set to 35 L / min, and three sets of oxygen concentration parameters were set: 30%, 60%, and 90%. The initial current value of the proportional valve of the high-flow humidification therapy device, the actual oxygen concentration RTO2 monitored by the oxygen concentration sensor, and the SensorO2 of the third-party monitoring device were recorded. The response time of each parameter was also recorded. The specific data are shown in the table below.

[0077] Table 1 Calibration and Control Algorithm Verification of Oxygen Mixing Proportional Valve

[0078]

[0079] Therefore, it can be seen that the oxygen mixing proportional valve is accurately calibrated, with an absolute error of ±1L / min, which can meet the feedforward requirements of the oxygen mixing control system. The adoption of the active disturbance rejection control algorithm not only eliminates the experimental process of adjusting PID parameters, but also meets the control accuracy and measurement accuracy of the proportional valve-oxygen concentration control system, keeping them both within the error range.

[0080] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of this application and are not intended to limit it. Although this application has been described in detail with reference to the embodiments, those skilled in the art should understand that modifications or equivalent substitutions to the technical solutions of this application do not depart from the spirit and scope of the technical solutions of this application, and should all be covered within the scope of the claims of this application.

Claims

1. A calibration and control system for a mixed oxygen proportional valve, comprising: Based on the target oxygen flow rate, consult the current-flow rate correspondence table of the oxygen mixing proportional valve to obtain the feedforward opening current of the oxygen mixing proportional valve. An active disturbance rejection controller is used to implement a control loop with an oxygen concentration sensor as the closed loop, and the input current of the oxygen mixing proportional valve is controlled in real time. The current-flow rate correspondence table of the oxygen mixing proportional valve is obtained by performing dual-thread calibration on the oxygen mixing proportional valve. The oxygen mixing proportional valve undergoes dual-thread calibration, including flow rate increase thread calibration and flow rate decrease thread calibration. After calibration, the dual-thread data is subjected to three-point linear smoothing, and then piecewise linear processing is performed to obtain the current magnitude that should be given under different flow rates during the rising and falling processes. The process of calibrating the thread with increasing traffic includes: The opening of the oxygen mixing proportional valve gradually increases from 0. When the flow sensor detects that the flow rate exceeds the set threshold, this opening is recorded as the critical opening of the oxygen mixing proportional valve, Valve1. Open ; Adjust the opening of the oxygen mixing proportional valve to its maximum value, then gradually decrease the opening. Record the opening at which the flow sensor reading falls below the set maximum flow rate; this opening is the maximum opening of the oxygen mixing proportional valve, Valve1. Max ; The oxygen mixing proportional valve is further divided into N1 equal parts between the critical opening and the maximum opening, that is, the interval between each opening is: Valve1 Delta =(Valve1) Max -Valve1 Open ) / N1, N1>0, and according to this interval Valve1 Delta Set the corresponding opening degree and record the data simultaneously, where the opening degree is set to {Valve1}. Open +Valve1 Delta Valve1 Open +2*Valve1 Delta Valve1 Open +N1*Valve1 Delta The one-to-one flow rate is {y1, y2, ... y}. N }; The process of calibrating the flow reduction thread includes: The opening of the oxygen mixing proportional valve gradually decreases from its maximum value. When the flow sensor detects that the flow rate is less than the set threshold, this opening is recorded as the critical opening of the oxygen mixing proportional valve, Valve2. Open ; Set the oxygen mixing proportional valve opening to 0, then gradually increase the opening. Record the opening when the flow sensor reading exceeds the set maximum flow rate; this opening is the maximum opening of the oxygen mixing proportional valve, Valve2. Max ; The oxygen mixing proportioning valve is further divided into N2 equal parts between the critical opening and the maximum opening, that is, the interval between each opening is: Valve2 Delta =(Valve2) Max -Valv2e Open ) / N2, N2>0, and follow this interval Valve2 Delta Set the appropriate opening degree and record the data, where the opening degree is set to {Valve2}. Open +Valve2 Delta Valve2 Open +2*Valve2 Delta Valve2 Open +N2*Valve2 Delta The one-to-one flow rate is {z1, z2, ... z}. N } 2. The oxygen mixing proportioning valve calibration and control system according to claim 1, characterized in that: The control loop, which uses an oxygen concentration sensor as a closed loop, includes the following control process: The oxygen mixing proportional valve obtains its initial opening based on the feedforward opening current that has been limited by the limiting module; The real-time oxygen concentration is measured by an oxygen concentration sensor, and the oxygen concentration error is obtained. The active disturbance rejection controller is activated, and the oxygen concentration error is used as the control target. The error e1 is obtained by comparing it with the output value of the linear extended state observer. After passing through the linear state error feedback module, the control quantity u0 is output. After compensation by the total disturbance estimate z2 obtained from the linear extended state observer, the final control quantity u is obtained by passing through the reciprocal of the controller gain. u is then passed through the limiting module to output the compensation control opening of the oxygen mixing proportional valve, and finally the oxygen concentration is controlled.

3. The oxygen mixing proportioning valve calibration and control system according to claim 2, characterized in that: The linear state error feedback module uses a PD controller, whose controller transfer function is: G(s)=k p +k d s Where, k p k is the proportionality coefficient. d is the differential coefficient; s is the complex plane independent variable; the proportional coefficient and differential coefficient are obtained by the ZN tuning method.

4. The oxygen mixing proportioning valve calibration and control system according to claim 2, characterized in that: The linearly extended state observer is designed as follows: Where β1 and β2 represent the pole placement parameters of the linearly extended state observer, determining the system convergence speed; u represents the system control law; and z is the state variable output matrix of the linearly extended state observer, z = [z1z2]. T T represents transpose; z1 tracks the real-time oxygen concentration y output by the system. After the dynamic control process converges, z1 tends towards y; z2 tracks the total disturbance. After the dynamic control process converges, z2 tends towards the total disturbance; b0 = 2.18; This represents the rate of change of the output matrix z of the state variable; This represents the estimated real-time oxygen concentration y of the system.

Images