Aircraft starter control valve dynamic response test method

Abstract

The invention discloses a dynamic response test method for an aircraft starter control valve, the valve is driven by an electromagnetic valve, the test method comprises the following steps: collecting an upstream air pressure value Pupstream of the valve in real time, and dynamically calculating an opening determination threshold 0.9 * Pupstream of the valve; when the valve is opened, the excitation triggering moment T0 of the electromagnetic valve is automatically captured, then the downstream air pressure value, obtained in real time, of the valve is compared with the opening judgment threshold value, the moment T1 when the downstream air pressure value reaches the opening judgment threshold value is automatically captured, and the opening response time delta T of the valve is calculated according to the formula that delta T = T1-T0. An upstream air pressure value Pupstream of a valve is collected in real time, a dynamic threshold value 0.9 * Pupstream used for valve opening judgment is calculated according to the air pressure value, then an excitation triggering moment T0 of an electromagnetic valve is automatically captured, the downstream air pressure is compared with a real-time value of the dynamic threshold value, and a moment T1 when the downstream air pressure dynamically reaches 90% of the upstream air pressure is automatically captured; and the valve opening response time delta T is automatically and accurately measured through T1-T0.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of aircraft component testing, in particular to a method for testing the dynamic response of an aircraft starter control valve, the main objective of the method is to obtain the relevant dynamic response parameters during the opening stage of the valve, so as to determine whether the performance of the valve meets the requirements of the manufacturer's technical document approved by the aviation authority. BACKGROUND

[0002] The response speed of the aircraft starter control valve directly affects the engine starting efficiency. The principle of aircraft engine starting is as follows: the air source upstream of the aircraft starter control valve comes from the Apu (auxiliary power unit) or the ground air source or the bleed air of another engine (all three air sources have pressure fluctuations), the valve downstream is connected to the starter of the engine, after the electromagnetic valve of the starter control valve is excited, the pressure downstream reaches the rated pressure within a specified time to drive the turbine blades of the starter to complete the starting of the engine. According to the manufacturer's technical document, the above-mentioned specified time refers to the opening time of the starter control valve ≤ 500 ms, that is, the time for the pressure downstream to reach 90% of the upstream pressure after the electromagnetic valve is excited ≤ 500 ms, and the rated pressure refers to the pressure rise rate (dP / dt) of the downstream air pressure during this period ≥ 30 psi / s.

[0003] The aircraft starter control valve is composed of an electromagnetic valve, an actuator, a flow shell (pipeline) and a butterfly valve piece. When the electromagnetic valve is not excited, the butterfly valve piece is closed, and when the electromagnetic valve is excited, the opening cavity of the actuator flows into the air pressure to make the connecting rod of the actuator move linearly, and the connecting rod is connected to the butterfly valve piece in the flow shell, so that the butterfly valve piece is tilted to a corresponding angle, allowing the downstream gas of the flow shell to reach the rated pressure within a specified time.

[0004] Problems existing in the prior art:

[0005] 1) Unable to automatically and accurately obtain the opening time ΔT of the valve. In the prior art, the opening time of the valve is obtained by manually judging when the downstream pressure reaches a fixed threshold, which cannot adapt to the pressure fluctuations (usually the upstream pressure fluctuates in the range of 54-60 psi, and during the opening of the valve, the upstream pressure will also fluctuate downward due to the inflow of upstream gas to the downstream of the valve), resulting in a large subjective error in the recorded ΔT;

[0006] 2) The pressure rise rate needs to be calculated manually, which is low in efficiency and prone to errors;

[0007] 3) Sensor noise leads to large dispersion of measurement results (error ≥ ± 15 ms), and the data utilization value is reduced;

[0008] 4) unable to synchronize the dynamic correlation between the air pressure change curve and the time parameter, lack of integrated analysis of the air pressure rising rate and time joint criterion, difficult to meet the multi-dimensional performance verification requirements in the aviation airworthiness standard. SUMMARY

[0009] The purpose of the present application is to provide a method for obtaining or further expanding the evaluation of the opening stage of the aircraft starter control valve to the important dynamic response parameters.

[0010] The purpose of the present application is achieved by the following technical scheme: a method for testing the dynamic response of an aircraft starter control valve, the valve is driven by an electromagnetic valve, the testing method comprises the following steps:

[0011] Real-time acquisition of the upstream air pressure value Pupstream of the valve, dynamic calculation of the opening judgment threshold value 0.9*Pupstream of the valve;

[0012] Automatic capture of the excitation trigger time T0 of the electromagnetic valve, then compare the real-time acquisition of the downstream air pressure value of the valve with the opening judgment threshold value, automatically capture the time T1 when the downstream air pressure value reaches the opening judgment threshold value, and calculate the opening response time ΔT=T1-T0 of the valve.

[0013] The present application acquires the upstream air pressure value Pupstream of the valve in real time, and calculates the dynamic threshold value 0.9*Pupstream for valve opening judgment according to the air pressure value, then captures the excitation trigger time T0 of the electromagnetic valve automatically, compares the real-time value of the downstream air pressure with the dynamic threshold value, automatically captures the time T1 when the downstream air pressure reaches 90% of the upstream air pressure, and realizes the automatic and accurate measurement of the opening response time ΔT of the valve through T1-T0.

[0014] As an optimization of the above scheme, the present application adopts the following preferred means:

[0015] The capture of the time T0 and T1 is realized by triggering FPGA. FPGA-level hardware is used to capture time stamp, which can obtain a time resolution significantly better than traditional microcontrollers (usually 1ms level).

[0016] The present application uses interpolation method to compensate the sampling interval to more accurately determine T1, thereby getting rid of the limitation of sensor sampling interval, and making the time resolution match the FPGA.

[0017] The present application sets a hysteresis interval, i.e. sets the upper and lower limits of the threshold value, and only when the downstream air pressure value continuously exceeds the upper limit of the threshold value for a period of time, it is determined as a valid trigger (trigger capture T1). It is recommended to set the upper and lower limits of the threshold value as 0.9*Pupstream±0.5psi.

[0018] The meaning of the hysteresis interval is that when the air pressure value fluctuates within this interval 0.9*Pupstream±0.5psi, the system will not make a valid trigger determination. The advantage of this is that it can avoid the system from being triggered by noise caused by air pressure fluctuation, and can improve the stability and accuracy of the system.

[0019] The present application adopts a software and hardware cooperative filtering method to solve the sensor noise problem:

[0020] The original collected signal is filtered by a hardware filter two-order RC low-pass filter circuit and converted into a digital signal, and then filtered by software moving average filtering or Kalman filtering according to the reliability of the data to output the upstream air pressure value and the downstream air pressure value; the reliability of the data is determined according to the signal-to-noise ratio of the signal filtered by the hardware, and it is recommended to switch the filtering mode from moving average to Kalman when the signal-to-noise ratio is less than 40DB. The present application adopts a software and hardware cooperative filtering method, which can improve the signal-to-noise ratio to more than 60dB.

[0021] In order to eliminate the influence of filter delay, the T1 is compensated after filter delay, and then the ΔT is calculated.

[0022] The present application also includes the step of automatically calculating the downstream air pressure rising rate dP / dt, which includes:

[0023] According to the collected downstream air pressure value, a linear function is used to fit the air pressure rising curve, and the goodness of fit R 2 ≥0.95 is verified, and then dP / dt is calculated.

[0024] The dP / dt calculation method is as follows:

[0025]

[0026] Where K is dP / dt, n represents the number of data points, t i represents the i-th time data point, represents each observation time, p i represents the i-th air pressure data point, which is the air pressure value observed at t i .

[0027] The present application also includes a fault diagnosis step, which includes:

[0028] When ΔT>500ms, an overtime alarm is performed;

[0029] When R 2 ≧0.95, but dP / dt<30psi / s, a stagnation alarm is performed;

[0030] When R 2 <0.95, a jitter alarm is performed;

[0031] When the initial air pressure value P_initial is greater than 5 psi, an initial air pressure abnormality alarm is performed.

[0032] The present application fuses multi-dimensional information, can perform multi-stage fault diagnosis on the valve, and can more comprehensively and accurately evaluate the valve state.

[0033] Beneficial effects:

[0034] 1) The present application can automatically and accurately obtain the opening time of the valve and the air pressure rising rate of the downstream, and uses a dynamic threshold to determine the valve opening time, which can automatically adapt to the fluctuations of the upstream air pressure.

[0035] 2) The present application uses a software and hardware cooperative filtering method, which can improve the signal-to-noise ratio to more than 60 dB.

[0036] 3) The present application fuses multi-dimensional information, can perform multi-stage fault diagnosis on the valve, and can more comprehensively and accurately evaluate the valve state, which can well meet the requirements of multi-dimensional performance verification in the aviation airworthiness standards. BRIEF DESCRIPTION OF DRAWINGS

[0037] Figure 1 is a downstream air pressure dynamic response curve;

[0038] Figure 2 is a system overall architecture schematic diagram;

[0039] Figure 3 is an FPGA timing analysis implementation diagram;

[0040] Figure 4 is a software and hardware cooperative filtering schematic diagram;

[0041] Figure 5 is a test flowchart;

[0042] Figure 6 is a fault diagnosis diagram. DETAILED DESCRIPTION

[0043] The present application aims to disclose a kind of aircraft starter control valve dynamic response test method, the main goal of this method is to obtain the dynamic response parameter related to valve opening stage, to judge whether the performance of valve meets the requirements of the factory technical document approved by the department.

[0044] Firstly, it discloses an automatic accurate measurement scheme of aircraft starter control valve opening response time ΔT.

[0045] Apu's bleed air is located upstream of the aircraft starter control valve, downstream of which is connected to the engine starter, and the valve downstream needs to reach the rated air pressure within a specified time to drive the turbine blades of the starter to complete the start of the engine, which requires the valve to tilt to the corresponding angle within a specified time at a certain rate.

[0046] The valve is driven by a solenoid valve. The existing method for calculating the opening response time ΔT of the valve relies on manual calculation, which has large errors. On the other hand, the pressure upstream of the valve is provided by the Apu compressor, ground air source or bleed air of another engine, and the pressure value fluctuates. The opening judgment is made by using a fixed threshold value (the downstream pressure reaches the fixed threshold value to determine the opening of the valve), which cannot adapt to the fluctuation of air pressure. For example, if the initial pressure is high and the subsequent pressure decreases, the downstream pressure may not reach the set threshold value (the opening judgment condition of the valve specified in the manufacturer's technical document is that the downstream pressure reaches 90% of the upstream pressure), so the time ΔT cannot be obtained.

[0047] In view of these problems, the present application adopts the concept of dynamic threshold value, and uses FPGA to automatically record and calculate the opening response time ΔT of the valve. The specific scheme is as follows:

[0048] The upstream air pressure value Pupstream of the valve is collected in real time, and the opening judgment threshold value 0.9*Pupstream of the valve is dynamically calculated;

[0049] The excitation trigger time T0 of the solenoid valve is automatically captured, then the downstream air pressure value of the valve is compared with the opening judgment threshold value, the time T1 when the downstream air pressure value reaches the opening judgment threshold value is automatically captured, and the opening response time ΔT of the valve is calculated as ΔT=T1-T0. The effect diagram is shown in Figure 1 .

[0050] Figure 1 In the figure, the horizontal axis is time (ms), and the vertical axis is air pressure (psi). 301 represents the upstream pressure curve (after filtering, which will be described below), 303 represents the dynamic threshold value (after filtering), 302 represents the downstream pressure curve (after filtering), 304 represents the T0 trigger time vertical line (0ms), 305 represents the T1 trigger time vertical line, and 306 represents the linear fitting segment DP / DT (air pressure rising rate), which will be introduced below.

[0051] The hardware system architecture based on which the present application is shown in Figure 2 , which includes the following core components:

[0052] 101: excitation control module, configured to output a 10V / 300mA solenoid valve driving signal;

[0053] 102: Upstream pressure sensor (Honeywell 24PC, ±0.1% accuracy), real-time acquisition of 54-60 psi pressure signal;

[0054] 103: Downstream pressure sensor (TE MS5803-30BA), sampling rate 1 kHz;

[0055] 104: Data processing unit (NIcRIO-9049), integrated FPGA and real-time controller, for real-time filtering, dynamic threshold calculation, timing analysis (capture timestamp, etc.), pressure rise rate calculation, etc.

[0056] 105: Human-machine interaction terminal, displaying ΔT, dP / dt, and fault codes, etc.

[0057] The data processing unit of the application uses FPGA-level hardware to capture timestamps. The FPGA uses a 40MHz on-board clock source, and the timing period for generating timestamps is 25ns, with a theoretical time resolution of up to 0.1ms, significantly better than traditional microcontrollers (usually 1ms level), and a 64-bit timestamp counter is used to support long continuous timing and avoid overflow problems, ensuring the continuity and uniqueness of timestamps, as shown in Figure 3 .

[0058] The FPGA captures timestamps in the following way:

[0059] As shown in Figure 3 , the solenoid valve excitation trigger signal is sent to the FPGA's hardware interrupt controller, which can detect the solenoid valve's trigger signal (solenoid valve excitation, low level to high level, rising edge trigger) through the IO pin and perform operations such as recording timestamp T0 and synchronously acquiring downstream pressure values. Timestamp T0 is stored in the FIFO buffer. The dynamic threshold comparator in the figure is used to compare the downstream pressure with the dynamic threshold. When the downstream pressure is greater than or equal to the dynamic threshold P threshold = 0.9*Pupstream and lasts for 10ms, the valve opening time T1 is recorded and stored in the FIFO buffer.

[0060] To suppress noise interference, the system sets a hysteresis interval and sets upper and lower limits for the threshold. In this embodiment, 0.9*Pupstream ± 0.5psi is set. Only when the downstream pressure exceeds the upper limit (for 10ms) is it considered a valid trigger. When the pressure value fluctuates within this interval, 0.9*Pupstream ± 0.5psi, the system will not make a valid trigger judgment. The advantage of this is that it can avoid false triggering caused by noise-induced pressure fluctuations, and can improve the stability and accuracy of the system.

[0061] Timestamps T0 and T1 are transmitted to the FPGA's real-time controller via the DMA channel.

[0062] As an improvement, the present invention also uses an interpolation method to compensate for the sampling interval. In this embodiment, the details are as follows:

[0063] In this embodiment, the downstream air pressure signal is sampled at 1kHz with a sampling interval of 1ms. Direct sampling of the sampling points may not accurately capture the moment when the threshold is crossed. This invention uses the air pressure values ​​and timestamps of adjacent sampling points to accurately calculate the actual trigger moment, reducing the time error from ±0.5ms of the sampling interval to ±0.1ms.

[0064] The formula for calculating T1 by interpolation is:

[0065] T1=Tprev+(P threshold -Pprev) / (Pcurrent-Pprev)*(Tcurrent-Tprev)

[0066] P threshold is the calculated opening judgment threshold, Pprev is the value of reaching P threshold The downstream pressure of the previous sampling point, Pcurrent, is the pressure reaching P threshold The downstream pressure of the collection point is Tprev, Tcurrent is the time when Pprev is collected, and Tcurrent is the time when Pcurrent is collected. threshold When an event occurs where the value of Pupstream = 0.9*Pupstream+0.5psi lasting for 10ms, T1 is no longer determined by relying on the downstream sampling time. Instead, T1 is accurately determined using interpolation based on the value of T1 between Tprev and Tcurrent. This method can align the trigger recording event of T1 with the timestamp of the FPGA, which is also a manifestation of the high-precision timing analysis of the present invention.

[0067] The present invention solves the sensor noise problem by filtering and reducing noise. Specifically, in this embodiment, it adopts the means of software and hardware collaborative filtering, such as Figure 4 As shown:

[0068] Hardware part (left side):

[0069] 401: Second-order RC low-pass filter circuit (100Hz cutoff frequency);

[0070] The circuit parameters of the second-order RC low-pass filter circuit are R1=R2=1 kΩ, C1=C2=1.6 μF, and the transfer function is: H(s)=1 / (1+R1 C1 s)(1+R2 C2 s).

[0071] 402: sensor signal input interface;

[0072] 403: The filtered signal is output to the ADC.

[0073] Software part (right side) :

[0074] 404: Moving average filter (window width 10 ms) ;

[0075] 405: Kalman filter (Q=0.5, R=1.0) ;

[0076] 406: Output of the signal after noise reduction.

[0077] The relationship between 404 and 405 is selected, and the reliability of the data is judged by judging the signal-to-noise ratio (SNR) of the signal filtered by hardware, and the filter mode is switched from moving average to Kalman when SNR < 40 DB.

[0078] Due to the complex test conditions, such as vibration, high temperature of gas and electromagnetic interference, the data stability and reliability of the system in the complex noise environment can be significantly improved by the software and hardware cooperative filtering mode, and the signal-to-noise ratio can be improved to more than 60 dB by the software and hardware cooperative filtering mode.

[0079] In order to eliminate the influence of filter delay, the T1 compensation filter delay is calculated by interpolation method, and then ΔT is calculated. The filter delay can be calibrated by experiment.

[0080] The present application automatically calculates the downstream pressure rising rate dP / dt in the following way:

[0081] According to the collected downstream pressure value, the pressure rising curve is fitted, and dP / dt is calculated, and the fitting degree requirement R 2 ≥0.95. It is recommended to use least square method to fit dP / dt.

[0082] The valve is composed of electromagnetic valve, actuator, flow shell (pipeline) and butterfly valve piece. After the electromagnetic valve is excited, the opening cavity of the actuator flows into the gas pressure to make the connecting rod of the actuator do linear motion, the connecting rod connects the butterfly valve piece in the flow shell, and the butterfly valve piece is inclined to the corresponding angle within the specified time and rate. Under normal circumstances, the valve piece of the valve is smoothly opened, and will not swing, so linear function fitting is usually used.

[0083] The slope calculation method is as follows:

[0084]

[0085] Where K is dP / dt. n represents the number of data points, t i represents the i-th time data point, represents each observation time. p i represents the i-th pressure data point, is the pressure value observed at t i time.

[0086] If R 2 <0.95, usually indicates that the valve is shaking, for example, the butterfly valve reciprocating swing is large, causing the downstream pressure fluctuation is too large, which is usually the problem of the actuator of the actuator cylinder (verified that the transmission of the downstream sensor is normal).

[0087] The steps of the present application for judging whether the performance of the valve meets the requirements of the relevant technical documents by using the measured valve opening response time ΔT and the downstream air pressure rise rate dP / dt are as follows:

[0088] Figure 5 The test flowchart of the embodiment of the present application.

[0089] As Figure 5 shown, the test flow is as follows:

[0090] 201: system initialization;

[0091] 202: downstream initial air pressure judgment node whether less than or equal to 5psi;

[0092] 203: solenoid excitation trigger action;

[0093] 204: multi-channel data synchronous acquisition, including collecting upstream and downstream pressure sensor data, second-order RC low-pass filtering, sliding average or Kalman filtering processing;

[0094] 205: dynamic threshold determination;

[0095] 206: ΔT calculation, least square fitting of downstream air pressure rise curve;

[0096] 208: result output and report generation.

[0097] In step 202, when judging whether the downstream initial air pressure is less than or equal to 5psi, if it is not, 207 is performed: that is, the initial air pressure P_initial> 5psi, the initial air pressure (the solenoid is not excited, and the valve is in the closed stage) abnormal alarm is performed, which usually indicates that the sealing performance of the valve disc is decreased, the sealing is invalid, and the fault code F04 is output.

[0098] In addition, if the valve downstream pressure ≥ 90% P_up is not detected within 2 seconds, 207 is performed: indicating that the valve response is timed out, and the fault code F01 is output. The intention of setting 2 seconds instead of 500ms in the manufacturer's technical manual is to enlarge the data sampling window to two seconds, and provide two seconds of data in the opening stage of the valve for fault analysis. If the downstream pressure ≥ 90% P_up within two seconds, ΔT calculation is performed, the downstream air pressure rise curve is fitted by least square method, and subsequent Figure 6 related fault diagnosis is performed.

[0099] Figure 6 Fault diagnosis chart for specific embodiments of the present application.

[0100] As shown in Figure 6 , the following states can be diagnosed:

[0101] Initialization is complete, and the default is to enter 601: normal state.

[0102] 602: When ΔT>500ms, a timeout alarm is performed, which generally indicates that the response time of the valve exceeds the limit, and outputs fault code F01.

[0103] 603: When dP / dt<30psi / s (R 2 ≧0.95, under the premise of high fitting degree), a jam alarm is performed, which generally indicates that the valve is jammed, for example, the sealing ring of the butterfly valve plate is deformed and interferes with the flow shell, resulting in insufficient angle of tilting, insufficient efficiency of gas pressure transmission to the downstream, etc., and outputs fault code F02.

[0104] 604: When R 2 <0.95 (low fitting degree), a jitter alarm is performed, which generally indicates that the valve is jittering and the actuator of the actuating cylinder has a problem, and outputs fault code F03.

[0105] The existing precision-recognized test scheme is as follows:

[0106] The pressure sensors of the upstream and downstream are connected to an oscilloscope, the relationship between the voltage values of the pressure sensors of the upstream and downstream and time is obtained, and then the data is manually intercepted and the data waveform is exported to a graphic analysis software to complete data analysis.

[0107] Taking the values obtained by calculation using the high-precision oscilloscope and graphic analysis software as the benchmark, the ΔT error of the present embodiment is ≤±0.1ms, the dP / dt error is ≤±1.5psi / s, and the single test time is shortened from 5 minutes to 1 minute.

[0108] The present application is also characterized in that the present application integrates multi-dimensional information, can perform multi-level fault diagnosis on the valve, and can more comprehensively and accurately evaluate the state of the valve.

[0109] Application Embodiment

[0110] CFM56 engine starter valve test

[0111] 1. Test conditions:

[0112] Upstream gas pressure: 57psi, ambient temperature: 25℃;

[0113] Filtering parameters: sliding average window 10ms, Kalman Q=0.5, R=1.0.

[0114] 2. Test Results:

[0115] Delta T = 322.3 ± 0.1 ms;

[0116] dP / dt = 42.1 ± 1.3 psi / s;

[0117] Fault Injection Test: Fault injection cases and detection rates are shown in Table 1.

[0118] Table 1

[0119]

[0120] The application comprehensively covers the core technology of dynamic response test, has high precision, high reliability and wide applicability, and meets the aviation level test standard.

[0121] The application has been verified by Guangzhou Aircraft Maintenance Company Accessory Business Center, is suitable for the starter control valve of multiple engine types such as CFM56 and LEAP, can be integrated into a portable test equipment or an airborne health management system, significantly reduces the maintenance cost and improves the flight safety.

Claims

1. A method for testing the dynamic response of an aircraft starter control valve, wherein the valve is driven by a solenoid valve, characterized in that: The testing method comprises the following steps: The upstream air pressure value Pupstream of the valve is collected in real time, and the opening judgment threshold value of the valve is dynamically calculated as 0.9*Pupstream; Automatically capture the excitation triggering moment T0 of the solenoid valve, then compare the real-time downstream air pressure value of the valve with the opening judgment threshold, automatically capture the moment T1 when the downstream air pressure value reaches the opening judgment threshold, and calculate the opening response time ΔT=T1-T0 of the valve.

2. The testing method according to claim 1, wherein: The capture of the times T0 and T1 is achieved by triggering the FPGA.

3. The testing method according to claim 2, wherein: It uses interpolation to compensate for the sampling interval to more accurately determine T1.

4. The testing method according to claim 1, wherein: It sets a hysteresis interval, that is, sets the upper and lower limits of the threshold, and determines that it is effectively triggered only when the downstream air pressure value continues to exceed the upper limit of the threshold for a period of time.

5. The testing method according to claim 4, characterized in that: The upper and lower limits of the threshold are set to 0.9*Pupstream±0.5psi.

6. The testing method according to claim 1, wherein: It uses software and hardware collaborative filtering to solve the sensor noise problem: The original collected signal is filtered through a second-order RC low-pass filter circuit of a hardware filter and converted into a digital signal. After that, software sliding average filtering or Kalman filtering is selected according to the reliability of the data, and then the output is the upstream air pressure value and the downstream air pressure value; The reliability of the data is determined based on the signal-to-noise ratio of the signal after hardware filtering.

7. The testing method according to claim 6, characterized in that: When the signal-to-noise ratio is less than 40DB, the trigger filter mode switches from sliding average to Kalman.

8. The testing method according to claim 6, wherein: In order to eliminate the influence of filtering delay, the filtering delay is compensated for T1 before calculating the ΔT.

9. The testing method according to any one of claims 1 to 8, characterized in that: The method also includes the steps of automatically calculating the downstream pressure rise rate dP / dt, specifically: According to the collected downstream air pressure value, a linear function is used to fit the air pressure rise curve to verify the goodness of fit R 2 ≥0.95, then calculate dP / dt; dP / dt is calculated as follows: Where K is dP / dt, n is the number of data points, t i Represents the i-th time data point, representing each observation moment, p i Indicates the i-th air pressure data point, corresponding to t i The observed air pressure value at any moment.

10. The testing method according to claim 9, characterized in that: It also includes troubleshooting steps, specifically: When ΔT>500ms, a timeout alarm will be issued; In R 2 ≧0.95, but dP / dt<30psi / s, a jam alarm will be issued; In R 2 When <0.95, a jitter alarm will be issued; When the downstream initial air pressure value P_initial>5psi, an initial air pressure abnormality alarm is issued.

Images