Calibration method and system for portable respiratory flow measuring instrument

Abstract

The invention provides a portable respiratory flow measuring instrument calibration method and system, and the method comprises the steps: sequentially connecting a gas generation device, a to-be-calibrated portable respiratory flow measuring instrument and a high-precision gas flow calibration instrument in series, constructing a stable gas flow channel, and ensuring that gas flow sequentially flows through each instrument; starting a gas generation device, and synchronously acquiring data after the gas flow is stable, namely an original measurement physical quantity vector x, a theoretical flow value and actual flow data; repeating the acquisition process to obtain a data set covering the full scale; the collected data are input into terminal equipment, a proper machine learning algorithm is selected for training according to the instrument principle, the target is to learn the relation between x and x, and a flow function which is applicable to the full scale and finally forms correction is obtained; in actual operation, the instrument calculates theoretical flow through x collected in real time, the trained flow is called for correction, and a high-precision flow value is output. The technical problems that in the prior art, equipment is complex, cost is high, operation is tedious, efficiency is low, intelligence and precision are insufficient, and scene adaptability is single are solved.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of calibration of respiratory flow measurement instruments, and in particular to a portable respiratory flow measurement instrument calibration method and system. BACKGROUND

[0002] In the medical field, lung function detection is crucial, and portable respiratory flow measurement instruments have become a new demand for primary medical care. There are three types of mainstream calibration schemes at present:

[0003] Calibration cylinder calibration method: use a standard volume calibration cylinder to generate a known volume of air flow for comparison calibration. Although the principle is intuitive and the repeatability is strong, the calibration cylinder needs to be replaced frequently, only discrete calibration of fixed volume points can be achieved, it depends on multiple specifications of calibration cylinders, and the device is large in volume and high in storage cost.

[0004] Hot wire sensor neural network calibration method: based on neural network model, collect data within a limited flow rate range to train the curve and extrapolate. It can expand the flow rate range using existing data to provide more continuous calibration results, but it is only suitable for hot wire sensors and has no compatibility for devices based on pressure difference principle, and the extrapolation accuracy of small flow is insufficient.

[0005] Ultrasonic wave direction tidal breathing calibration method: simulate air flow by manually pushing and pulling a standard syringe to calibrate ultrasonic wave sensors. It can simulate variable flow air flow in real breathing scenarios and verify sensor performance, but it is only suitable for specific breathing mode sensors, has large manual operation error, and cannot meet the demand for automated calibration.

[0006] Deficiency of prior art

[0007] (1) Complex equipment and high cost

[0008] The traditional calibration cylinder calibration method requires multiple specifications of standard cylinders, and other schemes rely on large laboratory stable flow field or expensive gas flow control equipment (such as high-precision regulating valve + air blower combination), which is more than 60% higher in cost than the serial calibration system proposed in this scheme.

[0009] (2) Complicated operation and low efficiency

[0010] The traditional calibration cylinder needs to replace the volume points frequently, and the ultrasonic wave method relies on manual pushing and pulling of the syringe, which requires professional personnel to operate in a specific environment and takes a long time for single calibration.

[0011] (3) Lack of intelligence and precision

[0012] The traditional method relies on manual operation or simple data fitting (such as calibration cylinder method only for discrete point comparison), and cannot handle the non-linear relationship of complex flow field, with a calibration error generally higher than 5%.

[0013] (4) Single scene adaptability

[0014] The existing solutions are mostly designed for fixed devices or specific sensors, and cannot be compatible with the related measurement principles of various portable respiratory flow measurement instruments, and are difficult to cover the full flow range. SUMMARY

[0015] The present application aims to overcome the above technical deficiencies, and provides a portable respiratory flow measurement instrument calibration method and system, which solves the technical problems of complex equipment, high cost, tedious operation, low efficiency, insufficient intelligence and precision, and single scene adaptability in the prior art.

[0016] To achieve the above technical purpose, the technical scheme of the present application provides a portable respiratory flow measurement instrument calibration method, comprising the steps of:

[0017] The gas generating device, the portable respiratory flow measurement instrument to be calibrated, and the high-precision gas flow calibration instrument are connected in series to build a stable airflow channel, ensuring that the airflow flows through each instrument in order;

[0018] Adjust the gas generating device to output stable airflow in different flow ranges according to the physical principles and adaptive flow ranges of the portable respiratory flow measurement instrument;

[0019] Start the gas generating device, and after the airflow is stable, synchronously collect the following three groups of key data:

[0020] Calibration instrument: collect the original measurement physical quantity vector through the built-in sensor , and at the same time calculate the theoretical flow value through the theoretical flow formula ; High-precision gas flow calibration instrument: real-time monitoring and accurate acquisition of actual gas flow data as a calibration reference;

[0021] Adjust the gear of the gas generating device to change the output flow, and repeat the data collection process to obtain a data set covering the full range ;

[0022] Input the collected data into the terminal device, and select appropriate machine learning algorithms according to the instrument principle for training, the goal being to learn the relationship between x, and , and obtain a full-range applicable . Finally, form the corrected flow function .

[0023] In actual operation, the instrument calculates the theoretical flow through real-time collection of x, and calls the trained for correction, and outputs the high-precision flow value .

[0024] Compared with the prior art, the beneficial effects of the present application include:

[0025] The calibration method and system based on a portable respiratory flow measurement instrument innovatively use machine learning and other methods to construct a modified flow function based on a multi-dimensional physical quantity vector x . The method performs nonlinear correction on the full range through , effectively solving the nonlinear error problem in the low flow and high flow regions and comprehensively ensuring the accuracy and reliability of the calibration results. Through case testing, the error can be stably controlled within 5%, compared with the manual operation error of the calibration standard method and the ultrasonic laser method, effectively improving the measurement accuracy and stability of the instrument. Common gas generating devices and standard flow meters are used, without the need for large professional laboratory environment, and the equipment purchase and maintenance costs are greatly reduced, with a 60% reduction in calibration costs compared with traditional solutions. The system structure is simple, easy to operate, and can be completed by simply connecting the equipment, saving time and labor costs and significantly improving the calibration efficiency. The universality of this solution is reflected in two aspects: first, the calibration model is based on a multi-dimensional physical quantity vector x, which can be compatible with instruments of different physical principles (pressure difference, thermal, ultrasonic, etc.); second, the design of the modified coefficient function enables it to focus on solving specific measurement accuracy problems in the full range. Ultimately, it realizes accurate measurement of the instrument in different scenarios and improves its reliability and universality. Depending on the characteristics of different models and specifications of portable respiratory flow measurement instruments, the input combination of the physical quantity vector x can be flexibly adjusted, or the architecture and parameters of the machine learning model can be optimized, providing a standardized paradigm for calibration work.

[0026] According to some embodiments of the present application, the gas generating device serves as the gas flow source of the system and is used to generate stable and adjustable gas flow, and the output flow range can be flexibly set according to the range requirement of the instrument to be calibrated, covering the low, medium and high flow range.

[0027] According to some embodiments of the present application, the original measurement physical quantity vector x, the calculated flow value , and the flow value measured by the high-precision gas flow calibration instrument are transmitted to the terminal device through serial communication.

[0028] In the second aspect, the present application provides a portable respiratory flow measurement instrument calibration system, which comprises a gas generating device, a portable respiratory flow measurement instrument to be calibrated, and a high-precision gas flow calibration instrument connected in sequence, and applies the portable respiratory flow measurement instrument calibration method of any one of the first aspect.

[0029] Additional aspects and advantages of the application will be set forth in part in the description which follows, and in part will become apparent to those skilled in the art upon examination of the following and the accompanying drawings or can be learned by practice of the application. BRIEF DESCRIPTION OF DRAWINGS

[0030] The foregoing and / or additional aspects and advantages of the present application are achieved by providing a portable respiratory flow measuring instrument calibration method, which comprises the following steps.

[0031] Figure 1 A flow chart of a portable respiratory flow measuring instrument calibration method provided by the present application;

[0032] Figure 2 A system structure diagram of a portable respiratory flow measuring instrument calibration provided by the present application. DETAILED DESCRIPTION

[0033] In order to make the objects, technical solutions and advantages of the present application clearer, the present application will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present application and should not be used to limit the present application.

[0034] It should be noted that although the functional modules are divided in the system schematic diagram and the logical order is shown in the flow chart, in some cases, the steps shown or described can be performed in a manner different from the module division in the system or the order in the flow chart. The terms "first", "second", etc. in the specification and claims and the above drawings are used to distinguish similar objects, and do not necessarily describe a specific order or sequence.

[0035] Reference Figure 1 and Figure 2 , Figure 1 A flow chart of a portable respiratory flow measuring instrument calibration method provided by the present application; Figure 2 A system structure diagram of a portable respiratory flow measuring instrument calibration provided by the present application.

[0036] In an embodiment, the portable respiratory flow measuring instrument calibration method comprises the following steps: connecting a gas generating device, a portable respiratory flow measuring instrument to be calibrated and a high-precision gas flow calibration instrument in series to build a stable airflow channel and ensure that the airflow flows through each instrument in order; adjusting the gas generating device to make the gas generating device output stable airflow in different flow ranges according to the physical principle of the portable respiratory flow measuring instrument and the adaptive flow range; starting the gas generating device, and after the airflow is stable, synchronously collecting the following three groups of key data: the instrument to be calibrated: collecting the original measured physical quantity vector through the built-in sensor , and calculating the theoretical flow value through the theoretical flow formula at the same time High-precision gas flow calibration instrument: Real-time monitoring and accurate acquisition of actual gas flow data. This serves as a reference for calibration; the gas generator's setting is adjusted to change the output flow rate, and the data acquisition process is repeated to obtain data covering the entire range. Dataset; the collected data The dataset is transmitted to the terminal device. The collected data is input into the terminal device, and an appropriate machine learning algorithm is selected for training based on the instrument's principles. The goal is to learn x. and The relationship between them yields a full-range applicable value. The final result is the corrected flow function. .

[0037] In actual operation, the instrument calculates the theoretical flow rate by collecting x data in real time. and call the trained Make corrections and output high-precision flow rates. .

[0038] This patent addresses the following issues: This solution overcomes the shortcomings of existing technologies by addressing four core problems related to cost, efficiency, accuracy, and adaptability through the following technical means:

[0039] (1) Reduce hardware costs

[0040] This solution uses a series design of a gas generator, a calibration instrument, and a high-precision gas flow calibration instrument to build a stable airflow channel without the need for complex equipment, thus significantly reducing hardware costs.

[0041] (2) Improve calibration efficiency

[0042] This solution adopts a standardized calibration implementation process. Through system integration, parameter setting, and real-time automatic data collection and analysis, it automates the process, significantly improves calibration efficiency, and lowers the application threshold at the grassroots level.

[0043] (3) Achieve high-precision calibration

[0044] This scheme utilizes machine learning algorithms to construct a multidimensional physical quantity vector-based system. The corrected flow function. This model can be adapted to the physical principles of different instruments, and is achieved through a correction coefficient function. Solve the nonlinear relationship of complex flow fields across the entire range to achieve high-precision calibration.

[0045] (4) Enhance scene adaptability

[0046] This scheme combines a serial calibration system with a multidimensional physical quantity vector-based system. This is combined with the modified flow function. The model adapts to different physical quantity vectors. To be compatible with the measurement principles of various portable instruments; and at the same time, to obtain the correction coefficient function through machine learning. It flexibly covers the full range of requirements, completely solving the problem of the limitations of traditional solutions in terms of scenarios.

[0047] (I) Composition of the calibration system

[0048] (1) Gas generating device

[0049] As the system's airflow source, it generates a stable and adjustable airflow. Its output flow range can be flexibly set according to the measurement range requirements of the instrument to be calibrated, covering the entire flow range from low to high.

[0050] (2) Portable respiratory flow measurement instrument to be calibrated

[0051] The instrument is a calibration target. It has a built-in sensor that can measure the vector of raw physical quantities generated when airflow passes through it, based on specific physical principles (such as differential pressure, thermal, or ultrasonic methods). .

[0052] (3) High-precision gas flow calibration instrument

[0053] It serves as a calibration reference. Connected in series after the instrument to be calibrated, it is used to accurately measure the actual gas flow rate through the system. Its measurement accuracy is far higher than that of the instrument to be calibrated, providing a "true value" reference for subsequent model correction.

[0054] (II) Generalized Model Correction and Calibration Method

[0055] This patent proposes a generalized calibration method based on theoretical model correction. It uses machine learning technology to correct the theoretical flow formula established by the instrument based on its physical principles, so as to achieve high-precision calibration across the entire range.

[0056] (1) Theoretical Model

[0057] Let the multiple physical parameters measured by the instrument to be calibrated form a vector. Based on its physical principles, a theoretical flow rate formula can be established. The composition of the vector x varies depending on the instrument's principle. For example, in a differential pressure principle, it may include pressure or pressure difference, while in a thermal principle, it may include temperature or temperature difference.

[0058] (2) Correcting the model construction

[0059] To improve accuracy, a correction coefficient function related to x is introduced into the theoretical formula. The corrected flow formula is obtained as follows:

[0060]

[0061] in: This represents the actual flow rate measured by a high-precision calibration device.

[0062] (3) Method for obtaining the correction coefficient function

[0063] Using machine learning algorithms, based on multi-dimensional calibration data covering the entire measurement range, we learn the physical quantity vector x theoretical flow rate. and actual traffic The mapping relationship between them is used to fit the correction coefficient function. , making As close as possible to the true value .

[0064] (4) Data Acquisition

[0065] To train the model, the following data needs to be collected synchronously at different traffic points to construct a system containing... Training dataset for the correspondence:

[0066] x: The vector of the original physical quantity measured by the instrument to be calibrated;

[0067] From theoretical formulas The calculated theoretical flow rate;

[0068] The actual flow rate reference value measured by high-precision calibration equipment.

[0069] (5) Model Training and Application

[0070] The collected data is input into the terminal device, and an appropriate machine learning algorithm is selected for training based on the instrument's principles. The goal is to learn x. and The relationship between them yields a full-range applicable value. The final result is the corrected flow function. .

[0071] In actual operation, the instrument calculates the theoretical flow rate by collecting x data in real time. and call the trained Make corrections and output high-precision flow rates. .

[0072] (III) Calibration Process

[0073] (1) System Connection

[0074] Connect the gas generator, the portable respiratory flow measuring instrument to be calibrated, and the high-precision gas flow calibration instrument in series to create a stable airflow channel, ensuring that the airflow flows through each instrument in sequence. See details... Figure 1 .

[0075] (2) Parameter settings

[0076] Adjust the gas generator to output a stable airflow within different flow ranges based on the physical principles of different portable respiratory flow measuring instruments (such as differential pressure type, thermal type, ultrasonic type, etc.) and their respective applicable flow ranges, thus creating specific flow conditions for calibration work.

[0077] (3) Single-point data acquisition

[0078] Start the gas generator and, after the gas flow stabilizes, simultaneously collect the following three sets of key data:

[0079] Instrument to be calibrated: Acquires raw measurement physical quantity vectors through built-in sensors. And simultaneously through its theoretical flow formula Calculate the theoretical flow rate.

[0080] High-precision gas flow calibration instrument: Utilizing existing high-precision measurement technology, it monitors and accurately acquires actual gas flow data in real time. , as a reference standard for calibration.

[0081] (4) Repeated single-point data collection

[0082] Adjust the speed setting of the gas generator to change the output flow rate, repeat the data acquisition process, and obtain data covering the entire range. Dataset.

[0083] (5) Data processing and correction

[0084] The collected The dataset is transmitted to the terminal device and processed using a full-range correction method based on machine learning.

[0085] Based on the physical principles of the instrument to be calibrated, a suitable machine learning algorithm is selected, and all available resources are utilized. The dataset is used for training to obtain the corrected flow function covering the entire range as defined in the (II) generalized model correction calibration method. This enables accurate calibration across the entire range.

[0086] V. Results Achieved

[0087] (1) High precision and reliability

[0088] Based on the calibration method and system of portable respiratory flow measurement instruments, this paper innovatively applies machine learning and other methods to construct a corrected flow function based on a multidimensional physical quantity vector x. This method is achieved through... The function performs nonlinear correction across the entire measurement range, effectively resolving nonlinear error issues in low and high flow rates, and comprehensively ensuring the accuracy and reliability of calibration results. Case studies have shown that the error can be stably controlled within 5%, effectively improving the instrument's measurement accuracy and stability compared to the manual operation errors inherent in calibration standard methods and ultrasonic / laser methods.

[0089] (2) Low cost and high cost performance

[0090] Using common gas generators and standard flow meters, no large professional laboratory environment is required, significantly reducing equipment purchase and maintenance costs, and reducing calibration costs by 60% compared to traditional solutions.

[0091] (3) Convenience and efficiency

[0092] The system has a simple structure and is easy and quick to operate. Calibration can be completed by simply connecting devices in series, saving time and labor costs and significantly improving the efficiency of calibration work.

[0093] (4) High applicability

[0094] The universality of this solution is reflected on two levels:

[0095] First, the calibration model is based on a multidimensional physical quantity vector x, making it compatible with instruments based on different physical principles (pressure difference, thermal, ultrasonic, etc.); second, the correction coefficient function... Its design allows it to focus on solving specific measurement accuracy problems across the entire measurement range. Ultimately, this enables the instrument to perform accurate measurements in different scenarios, improving its reliability and versatility.

[0096] (5) Good scalability

[0097] According to the characteristics of portable respiratory flow measurement instruments of different models and specifications, the input combination of physical quantity vector x can be flexibly adjusted, or the architecture and parameters of machine learning models can be optimized, providing a standardized paradigm for calibration work.

[0098] Six specific case studies for implementation analysis

[0099] To more clearly demonstrate the practical application effect of this calibration scheme, we will take the calibration of a portable respiratory flow measuring instrument based on the Venturi measurement principle as an example to explain the relevant case implementation process in detail.

[0100] (1) System Connection

[0101] Connect the outlet of the gas generator to the inlet of the instrument to be calibrated, and connect the outlet of the instrument to be calibrated to the inlet of the high-precision gas flow calibrator, forming a series airflow channel to ensure that the airflow passes through the instrument to be calibrated and the high-precision gas flow calibrator in sequence.

[0102] The instrument has an internal pipe structure with openings at both ends serving as inlets and outlets. The inner diameter of both inlet and outlet pipes is 20mm, and the neck diameter is 12mm. Mounting holes are provided at the inlet and neck of the pipe for installing pressure sensors. Based on the Venturi flow measurement principle, the gas flow rate is measured by combining the data collected by the sensors.

[0103] (2) Parameter settings:

[0104] Set the initial setting of the gas generator to produce a stable airflow within a known flow range.

[0105] (3) Data collection:

[0106] Start the gas generator and wait for the airflow to stabilize. Then, the two pressure sensors of the instrument to be calibrated collect the pressure values ​​at two different locations in the pipeline (i.e., the pressure at the gas inlet of the pipeline). and pressure at the neck of the pipe The data is then transmitted to the STM32F103C8T6 microcontroller in the instrument to be calibrated for processing.

[0107] The microcontroller calculates the flow rate Q in the pipe based on the collected pressure data, using the principle of the Venturi flowmeter and a correction formula. The Venturi flowmeter principle is based on Bernoulli's equation and the continuity equation; its core is to calculate the flow velocity and flow rate by measuring the pressure difference between the fluid at the inlet and the neck. When the fluid flows from the inlet of the Venturi tube to the neck, the flow velocity increases and the pressure decreases; the flow velocity and flow rate can be derived from the pressure difference between the two measuring points.

[0108] The correction process is as follows:

[0109] Original theoretical formula: (Based on the ideal fluid assumption, without considering factors such as viscosity and turbulence of actual fluids)

[0110] Correction formula: (Introducing a correction function) correction)

[0111] Explanation of key variables:

[0112] Actual flow rate (volume flow rate), unit: m³ / s, the actual flow rate value after correction factor.

[0113] Theoretical flow rate, unit: m³ / s, is the theoretical flow rate calculated based on the ideal fluid assumption.

[0114] Coefficient functions, dimensionless, are used to correct the difference between theoretical and actual values, taking into account practical factors such as fluid viscosity and turbulence.

[0115] Cross-sectional area at the neck of the pipe, unit: m²

[0116] Cross-sectional area at the air inlet of the pipe, unit: m²

[0117] Flow velocity at the neck of a Venturi tube, unit: m / s; the velocity of the fluid in the contraction section.

[0118] Pressure at the air inlet of the pipeline, unit: Pa

[0119] Pressure at the neck of the pipe, unit: Pa

[0120] Density, unit: kg / m³, the density of the fluid being measured.

[0121] For the simulation results, a correction coefficient function was added based on the original pressure-transformer flow rate formula. This solution uses machine learning to fit this... The function corrects for the entire range.

[0122] Meanwhile, the high-precision gas flow calibration instrument measures the actual gas flow rate in real time. , which serves as the standard gas flow rate for calibration.

[0123] (4) Repeated single-point data collection:

[0124] The gas generator has three speed settings: low, medium, and high. We selected the low speed setting (0-200 L / min) for this test; the data acquisition and processing methods for the other speed settings are the same. By adjusting the low speed setting of the gas generator to change the output flow rate, we repeatedly executed the data acquisition process to obtain multiple sets of measurement data and standard data at different flow rates.

[0125] (5) Data processing and correction:

[0126] The data x collected by the microcontroller (including...) is transmitted via serial communication. (etc.), the calculated flow rate value and the flow rate measured by a high-precision gas flow rate calibration instrument. The data is then transferred to a computer. Next, machine learning algorithms are used to perform in-depth analysis of the data to construct a correction function. The specific analysis and adjustment methods are as follows:

[0127] Error characteristics of different flow ranges (basis for machine learning fitting):

[0128] Low flow rate region: Susceptible to noise interference, with large fluctuations. Medium flow rate region: Good linearity. High flow rate region: Increased turbulence, prominent nonlinear characteristics.

[0129] 1. The dynamic adjustment method is as follows:

[0130] Will contain The full-range dataset is uniformly input into machine learning algorithms (FNN, GNN) for training, fitting a correction coefficient function that covers the entire range. .

[0131] Should Functions are learned (Right now , )and The relationship enables it to automatically and dynamically correct the above error characteristics according to different inputs x (corresponding to different flow zones), thereby strengthening the suppression of noise in the low zone and compensating for turbulence deviation in the high zone.

[0132] 2. Achieve accurate calibration across the entire measurement range:

[0133] Finally, the correction coefficient function obtained from training is... Substitute into the formula:

[0134] ,

[0135] After correction, the error in the low range can be stably controlled within 5%.

[0136] The final correction function is constructed in this way, achieving accurate calibration of the full-range flow rate. Specifically, two types of neural networks are used: one is a feedforward neural network (FNN) as... The first method is to use a fitter to fit the complex nonlinear relationship between flow velocity and pressure; the second method is to use a graph neural network (GNN) to predict sensor measurements based on the structural features of the pressure field, thereby improving the accuracy and reliability of the data.

[0137] The gas generator serves as the system's airflow source, producing a stable and adjustable flow rate. Its output flow rate range can be flexibly set according to the measurement requirements of the instrument to be calibrated, covering the low, medium, and high flow rate ranges. The built-in sensor acquires the original measured physical quantity vector via serial communication. Calculated flow rate values and the flow rate measured by a high-precision gas flow rate calibration instrument. Transmitted to the terminal device.

[0138] Among them, the piecewise correction method based on machine learning uses two types of neural networks: one is a feedforward neural network (FNN) to fit the complex nonlinear relationship between flow velocity and pressure, and solve the deviation problem under extreme flow rates; the other is a graph neural network (GNN) to predict sensor measurements based on the structural features of the pressure field, thereby improving data accuracy and reliability.

[0139] In one embodiment, a portable respiratory flow measurement instrument calibration system includes a gas generator, a portable respiratory flow measurement instrument to be calibrated, and a high-precision gas flow calibration instrument connected in sequence, and applies the portable respiratory flow measurement instrument calibration method described above.

[0140] The above is a detailed description of the preferred embodiments of the present invention. However, the present invention is not limited to the above embodiments. Those skilled in the art can make various equivalent modifications or substitutions without departing from the spirit of the present invention. All such equivalent modifications or substitutions are included within the scope defined by the claims of the present invention.

[0141] The specific embodiments of the present invention described above do not constitute a limitation on the scope of protection of the present invention. Any other corresponding changes and modifications made in accordance with the technical concept of the present invention should be included within the scope of protection of the claims of the present invention.

Claims

1. A method of calibrating a portable respiratory flow measuring instrument, characterized in that, The method comprises the steps of: connecting a gas generating device, a portable respiratory flow measurement instrument to be calibrated, and a high-precision gas flow calibration instrument in sequence to construct a stable airflow channel and ensure that the airflow flows through each instrument in sequence; adjusting the gas generating device to make the gas generating device output stable airflow in different flow ranges according to the physical principle and adaptive flow range of the portable respiratory flow measurement instrument; starting the gas generating device, and synchronously collecting the following three groups of key data after the airflow is stable: Instrument to be calibrated: acquisition of the raw measurement physical quantity vector through the built-in sensor and at the same time the theoretical flow value is calculated through the theoretical flow formula ; High-precision gas flow calibration instrument: real-time monitoring and accurate acquisition of actual gas flow data as a calibration reference Adjust the gear of the gas generating device to change the output flow, repeat the data acquisition process to obtain a data set covering the full range ​ The collected data is input into a terminal device, and a suitable machine learning algorithm is selected according to the principle of the instrument for training, the goal being to learn the relationship between x, and y, obtain a full-range applicable , and finally form a corrected flow function ; In actual operation, the instrument corrects the theoretical flow calculated by the real-time collected x , calls the trained , and outputs high-precision flow values .

2. The method of calibrating a portable respiratory flow measurement instrument of claim 1, wherein, The gas generating device serves as the airflow source of the system, is used to generate stable and adjustable airflow, and the output flow range can be flexibly set according to the range requirement of the instrument to be calibrated, covering the low, medium and high flow range.

3. The method of calibrating a portable respiratory flow measurement instrument of claim 1, wherein, The original measurement physical quantity vector collected by the built-in sensor is transmitted to the terminal device through serial communication. The calculated flow value , and the flow value measured by the high-precision gas flow calibration instrument are transmitted to the terminal device.

4. A portable respiratory flow measuring instrument calibration system comprising, in sequence, a gas generating device, a portable respiratory flow measuring instrument to be calibrated, and a high-precision gas flow calibration instrument, characterized in that, The portable respiratory flow measurement instrument calibration method of any one of claims 1-3 is applied.

Images