An oxygen detection method, device and storage medium based on an oxygen generation device

Abstract

The present invention discloses an oxygen detection method, device and storage medium based on an oxygen generation device, relating to the technical field of oxygen detection. By calculating the real-time torsion angle of the dumbbell ball and comparing it with a threshold value, when it exceeds, the swing amplitude is estimated using the Kalman filtering algorithm in combination with the air flow disturbance data, and then input into a support vector machine model to classify the influence degree of the air flow disturbance and generate a judgment rule, thus solving the problem of measurement accuracy caused by gas flow; at the same time, a temperature and pressure sensor is used to collect data and the swing amplitude data is corrected by combining a compensation algorithm, solving the problem of interference caused by temperature and pressure changes; in addition, according to the corrected torsion angle data, it is judged whether it exceeds the linear response range of the photoelectric sensor. When it exceeds, a sensor array is used for piecewise fitting, and the oxygen concentration is calculated by combining a fitting algorithm, solving the problem of sensor response limitation, and effectively improving the accuracy, reliability and measurement range of oxygen concentration measurement.

Description

Technical Field

[0001] The present invention relates to the technical field of oxygen detection, and specifically, to an oxygen detection method, device, and storage medium based on an oxygen generation device. Background Art

[0002] In the medical field, the accurate measurement of oxygen concentration is crucial. Currently, the commonly used method for measuring oxygen concentration is to utilize the paramagnetic principle of oxygen and determine its concentration by measuring the magnetic field change caused by oxygen. However, in practical applications, due to the influence of gas flow in the detection chamber, the suspended dumbbell-shaped quartz glass balls will generate a certain swing, resulting in a change in the position of the reflected light spot on the photoelectric sensor, thereby affecting the measurement accuracy. Summary of the Invention

[0003] The purpose of the present invention is to provide an oxygen detection method, device, and storage medium based on an oxygen generation device, which realizes high-precision, high-stability, and real-time monitoring of the oxygen concentration in the detection chamber, significantly improves the reliability and safety of oxygen detection in the oxygen generation device, and provides a strong guarantee for the accurate measurement of oxygen concentration in medical and other fields.

[0004] The purpose of the present invention can be achieved through the following technical solutions:

[0005] The present application provides an oxygen detection method based on an oxygen generation device, including the following steps:

[0006] S1. Obtain the magnetic field change data caused by the oxygen concentration in the detection chamber, reduce external interference through a high-stability magnetic field source to obtain a magnetic field change signal, and then calculate the torsion angle of the dumbbell ball. When the torsion angle exceeds a preset threshold, judge the influence of air flow disturbance on the swing of the dumbbell ball to obtain the swing amplitude data;

[0007] S2. According to the swing amplitude data, use a temperature sensor and a pressure sensor to collect the temperature and pressure fluctuation data in the detection chamber in real time to obtain temperature and pressure compensation parameters, and combine a preset compensation algorithm to correct the swing amplitude data to obtain the corrected torsion angle data of the dumbbell ball;

[0008] S3. According to the corrected torsion angle data of the dumbbell ball, judge whether it exceeds the linear response range of the photoelectric sensor. When it exceeds the range, use multiple photoelectric sensors to form an array to obtain piecewise linear fitting data;

[0009] S4. According to the piecewise linear fitting data, combine a preset fitting algorithm to calculate the corresponding relationship between the oxygen concentration and the torsion angle to obtain the oxygen concentration measurement result, and judge whether there is non-linear distortion. When there is distortion, recalculate through the piecewise fitting data of the sensor array to obtain the corrected oxygen concentration value.

[0010] Further, obtain the magnetic field change data caused by the oxygen concentration in the detection chamber, reduce external interference through a high-stability magnetic field source, and obtain the magnetic field change signal, specifically including:

[0011] Adopt a high-stability magnetic field source to generate a uniform and stable magnetic field through the magnetic field source, reduce the influence of external environmental magnetic field interference on the indoor magnetic field change, arrange multiple magnetic field sensors indoors, obtain the magnetic field change data at different positions, and improve the spatial resolution and coverage of the magnetic field change data through multi-point sampling;

[0012] Preprocess the obtained magnetic field change data, and according to the characteristics of the magnetic field change data, adopt the wavelet transform signal processing method to extract the characteristic information in the magnetic field change signal as an indicator of the oxygen concentration change;

[0013] Then establish a mapping relationship model between the magnetic field change signal and the oxygen concentration change, and train the mapping model through machine learning algorithms such as support vector machines or neural networks;

[0014] In the mapping relationship model, by designing a real-time monitoring algorithm for the oxygen concentration change, continuously collect and analyze the magnetic field change signal, judge whether the current oxygen concentration is within the normal range, and give a timely warning when it is lower than the preset threshold.

[0015] Further, after obtaining the magnetic field change signal, it also includes: obtaining the magnetic field change signal of the dumbbell ball, calculating the real-time torsion angle of the dumbbell ball through the magnetic field change signal, comparing the calculated real-time torsion angle of the dumbbell ball with the preset torsion angle threshold, and judging whether the torsion angle of the dumbbell ball exceeds the preset threshold;

[0016] When the torsion angle of the dumbbell ball exceeds the preset threshold, obtain the current airflow disturbance data, analyze the influence degree of the airflow disturbance on the swing of the dumbbell ball through the airflow disturbance data, and then according to the airflow disturbance data and the dumbbell ball torsion angle data, adopt the Kalman filter algorithm to evaluate the swing amplitude of the dumbbell ball and obtain an estimated value of the swing amplitude of the dumbbell ball;

[0017] Compare the estimated value of the swing amplitude of the dumbbell ball with the preset swing amplitude threshold, and judge whether the swing amplitude of the dumbbell ball exceeds the preset threshold;

[0018] When the swing amplitude of the dumbbell ball exceeds the preset threshold, input the torsion angle data and swing amplitude data of the dumbbell ball into a pre-trained support vector machine model, and classify the influence degree of the airflow disturbance through the support vector machine model to obtain a classification result of the influence degree of the airflow disturbance;

[0019] According to the classification results of the influence degree of airflow disturbance, a decision tree algorithm is used to generate the judgment rules for the influence degree of airflow disturbance, and the judgment rules are applied to the monitoring of the swing of the dumbbell ball to make real-time judgment and early warning of the airflow disturbance.

[0020] Furthermore, according to the swing amplitude data, a temperature sensor and a pressure sensor are used to collect the temperature and pressure fluctuation data in the detection room in real time to obtain the temperature and pressure compensation parameters, specifically including:

[0021] Through the combination of a temperature sensor and a pressure sensor, the temperature and pressure fluctuation data in the room are collected in real time, and then the original temperature and pressure data collected by the sensors are preprocessed to obtain effective temperature and pressure data;

[0022] According to the historical temperature and pressure data, a multiple regression model of temperature-pressure-swing amplitude is established, the model parameters are trained by a machine learning algorithm, and the temperature and pressure data collected in real time are input into the trained regression model to predict the estimated value of the swing amplitude in the current environment;

[0023] Then calculate the temperature and pressure compensation parameters, and by comparing the actually measured swing amplitude with the model prediction value, obtain the deviation amount caused by temperature and pressure.

[0024] Furthermore, in combination with a preset compensation algorithm, the swing amplitude data is corrected to obtain the corrected dumbbell ball torsion angle data, specifically including:

[0025] The Kalman filtering algorithm is used to filter the swing amplitude data, establish the state equation and observation equation of the dumbbell ball movement, and use recursive calculation to obtain the optimal estimated value to obtain the filtered swing amplitude data;

[0026] According to the temperature and pressure compensation parameters and in combination with a preset compensation algorithm, non-linear compensation is performed on the filtered swing amplitude data. By establishing a non-linear mapping relationship between the dumbbell ball swing amplitude and the torsion angle, the corresponding compensation value is determined by using the look-up table method or the interpolation method to obtain the compensated torsion angle data;

[0027] The least squares method is used to perform curve fitting on the compensated torsion angle data. By selecting a suitable fitting function and using the principle of minimizing the sum of squared residuals to determine the coefficients of the fitting function, the curve of the fitted torsion angle data is obtained;

[0028] According to the curve of the fitted torsion angle data, the torsion angle data is smoothed. By setting the smoothing window and the weighting coefficient, the data is smoothed by using methods such as moving average or exponential smoothing to obtain the smoothed torsion angle data;

[0029] The adaptive threshold method is used to perform threshold processing on the smoothed torsional angle data. By dynamically adjusting the threshold size, binary processing is performed on the data using the threshold judgment condition to obtain the torsional angle data after threshold processing;

[0030] According to the torsional angle data after threshold processing, the corrected torsional angle data of the dumbbell ball is obtained by using the weighted fusion or decision tree method.

[0031] Furthermore, according to the corrected torsional angle data of the dumbbell ball, it is judged whether it exceeds the linear response range of the photoelectric sensor. When it exceeds the range, an array composed of multiple photoelectric sensors is used to obtain piecewise linear fitting data, specifically including:

[0032] Obtain the corrected torsional angle data of the dumbbell ball, input the data into the judgment module, and the judgment module judges whether the torsional angle data exceeds the linear response range of a single photoelectric sensor according to the preset linear response range threshold;

[0033] When the torsional angle data does not exceed the linear response range, a single photoelectric sensor is used to obtain the complete torsional angle data. When the torsional angle data exceeds the linear response range, the sensor array module is triggered, and according to the preset sensor layout scheme, the matching number of photoelectric sensors is selected to form a sensor array;

[0034] The sensor array module controls each photoelectric sensor to obtain the torsional angle data within different angle ranges respectively, forming piecewise data, and then inputs the obtained piecewise torsional angle data into the data fitting module, and uses the piecewise linear fitting algorithm to fit each segment of data to obtain the complete torsional angle fitting curve.

[0035] Furthermore, according to the piecewise linear fitting data, combined with the preset fitting algorithm, the corresponding relationship between the oxygen concentration and the torsional angle is calculated to obtain the oxygen concentration measurement result, specifically including:

[0036] Obtain the set of data points of the piecewise linear fitting, perform linear fitting on each data segment using the least squares method to obtain the linear function expression of this segment, and then according to the preset fitting algorithm, determine the position of the segmentation point of the piecewise linear function, and connect the linear functions of adjacent segments to obtain the complete piecewise linear function;

[0037] Obtain the currently measured torsional angle value, calculate the theoretical value of the oxygen concentration corresponding to this torsional angle through the piecewise linear function, obtain the actual measured value of the oxygen concentration sensor, and compare it with the theoretical calculated value. When the deviation exceeds the preset threshold, it is judged as abnormal and an alarm prompt is given;

[0038] According to the oxygen concentration values measured multiple times, the Kalman filtering algorithm is used for noise filtering to obtain a more accurate and stable measurement result of the oxygen concentration, and the corresponding relationship between the measured oxygen concentration result and the torsion angle is stored in the database.

[0039] According to the corrected oxygen concentration value, the final measurement result of the oxygen concentration is output to complete the measurement process.

[0040] An oxygen detection device based on an oxygen generation device, including a processor, a memory, and computer program instructions stored in the memory, which implement the above-mentioned oxygen detection method based on an oxygen generation device when the computer program instructions are executed by the processor.

[0041] A storage medium stores computer program instructions, which implement the above-mentioned oxygen detection method based on an oxygen generation device when the computer program instructions are executed by the processor.

[0042] The beneficial effects of the present invention are as follows:

[0043] The present invention calculates the real-time torsion angle by obtaining the magnetic field change signal of the dumbbell ball and compares it with a preset threshold. When the threshold is exceeded, the airflow disturbance data is obtained, and the Kalman filtering algorithm is used to estimate the swing amplitude. Then, the relevant data is input into the support vector machine model to classify the influence degree of the airflow disturbance and generate judgment rules for application in monitoring, thereby effectively solving the problem that the swing of the dumbbell-shaped quartz glass ball caused by the indoor gas flow affects the measurement accuracy, and improving the accuracy of the oxygen concentration measurement;

[0044] The temperature sensor and the pressure sensor are used to collect the temperature and pressure fluctuation data in the detection room in real time to obtain compensation parameters, and combined with a preset compensation algorithm, a series of correction processes are performed on the swing amplitude data, including Kalman filtering, nonlinear compensation, curve fitting, smoothing processing, and threshold processing, etc. Finally, the corrected torsion angle data of the dumbbell ball is obtained, thereby solving the interference problem of temperature and pressure changes on the oxygen concentration measurement result and making the measurement data more accurate and reliable;

[0045] According to the corrected torsion angle data of the dumbbell ball, it is judged whether it exceeds the linear response range of the photoelectric sensor. When it exceeds, an array composed of multiple photoelectric sensors is used to obtain segmented data and perform segmented linear fitting to obtain a complete torsion angle fitting curve. Then, combined with a preset fitting algorithm, the corresponding relationship between the oxygen concentration and the torsion angle is calculated to obtain the oxygen concentration measurement result, and recalculation and correction are performed when there is nonlinear distortion, thereby effectively expanding the measurement range and solving the problem of inaccurate measurement caused by the limited linear response range of a single photoelectric sensor, and ensuring accurate measurement of the oxygen concentration at different torsion angles. Description of the Drawings

[0046] For better understanding and implementation, the technical solution of the present application will be described in detail below with reference to the accompanying drawings.

[0047] Figure 1 It is a schematic flowchart of an oxygen detection method based on an oxygen generation device provided by the present application;

[0048] Figure 2 It is a schematic flowchart of obtaining a magnetic field change signal by an oxygen detection method based on an oxygen generation device provided by the present application;

[0049] Figure 3 It is a schematic flowchart of obtaining temperature and pressure compensation parameters by an oxygen detection method based on an oxygen generation device provided by the present application. Detailed implementation manners

[0050] To further elaborate on the technical means and effects adopted by the present invention to achieve the predetermined invention purpose, exemplary embodiments will be described in detail here, and the examples are shown in the accompanying drawings. When the following description refers to the accompanying drawings, unless otherwise indicated, the same numbers in different drawings represent the same or similar elements. The implementation manners described in the following exemplary embodiments do not represent all implementation manners consistent with the present application. On the contrary, they are only examples of methods and systems consistent with some aspects of the present application as detailed in the appended claims.

[0051] The terms used in the present application are only for the purpose of describing specific embodiments and are not intended to limit the present application. The singular forms "a", "the", and "said" used in the present application and the appended claims are also intended to include the plural forms unless the context clearly indicates otherwise. It should also be understood that the term "and / or" used herein refers to and includes any or all possible combinations of one or more of the associated listed items.

[0052] The following will describe in detail the specific implementation manners, features, and effects of the present invention with reference to the accompanying drawings and preferred embodiments.

[0053] Please refer to Figures 1 - 3 , this embodiment provides an oxygen detection method based on an oxygen generation device, including the following steps:

[0054] S1. Obtain the magnetic field change data caused by the oxygen concentration in the detection chamber, reduce external interference through a high-stability magnetic field source to obtain a magnetic field change signal, and then calculate the torsion angle of the dumbbell ball. When the torsion angle exceeds the preset threshold, judge the influence of the airflow disturbance on the swing of the dumbbell ball to obtain the swing amplitude data;

[0055] Further, obtaining the magnetic field change data caused by the oxygen concentration in the detection chamber and reducing external interference through a high-stability magnetic field source to obtain a magnetic field change signal specifically includes:

[0056] S11. Adopt a high-stability magnetic field source to generate a uniform and stable magnetic field through the magnetic field source, reduce the influence of external environmental magnetic field interference on the indoor magnetic field change, improve the accuracy of the magnetic field change data, arrange multiple magnetic field sensors indoors to obtain the magnetic field change data at different positions, and improve the spatial resolution and coverage of the magnetic field change data through multi-point sampling;

[0057] S12. Preprocess the obtained magnetic field change data, including removing outliers, smoothing noise, etc., improve the signal-to-noise ratio and quality of the magnetic field change data, and adopt signal processing methods such as wavelet transform according to the characteristics of the magnetic field change data to extract the characteristic information in the magnetic field change signal, such as signal amplitude, frequency, phase, etc., as the indication quantity of the oxygen concentration change;

[0058] S13. Then establish a mapping relationship model between the magnetic field change signal and the oxygen concentration change, and train the mapping model through machine learning algorithms such as support vector machines or neural networks to realize the function of accurately estimating the indoor oxygen concentration from the magnetic field change signal;

[0059] S14. In the mapping relationship model, by designing a real-time monitoring algorithm for the oxygen concentration change, continuously collect and analyze the magnetic field change signal, and judge whether the current oxygen concentration is within the normal range. When it is lower than the preset threshold, give an early warning in time.

[0060] Specifically, through a series of methods such as adopting a high-stability magnetic field source, arranging magnetic field sensors at multiple points, preprocessing the magnetic field change data, and establishing a mapping relationship model, the high-precision, high-stability and real-time monitoring of the indoor oxygen concentration are realized. Specifically, the high-stability magnetic field source reduces external interference, multi-point sampling improves the spatial resolution and coverage of the magnetic field change data, the preprocessing and signal processing methods improve the data quality and the accuracy of feature extraction, and the mapping model trained by machine learning algorithms realizes the accurate estimation from the magnetic field change signal to the oxygen concentration. The real-time monitoring algorithm can continuously collect and analyze the magnetic field change signal, timely judge whether the oxygen concentration is normal and give an early warning. This effectively solves the problems such as large external interference, low data accuracy, and untimely monitoring that may exist in traditional oxygen detection methods, improves the reliability and safety of oxygen detection in oxygen production equipment, and provides more accurate oxygen concentration monitoring guarantee for relevant application scenarios.

[0061] Furthermore, after obtaining the magnetic field change signal, it also includes: obtaining the magnetic field change signal of the dumbbell ball, calculating the real-time torsion angle of the dumbbell ball through the magnetic field change signal, comparing the calculated real-time torsion angle of the dumbbell ball with the preset torsion angle threshold, and judging whether the torsion angle of the dumbbell ball exceeds the preset threshold;

[0062] When the torsion angle of the dumbbell ball exceeds the preset threshold, the current airflow disturbance data is obtained, and the influence degree of the airflow disturbance on the swing of the dumbbell ball is analyzed through the airflow disturbance data. Then, according to the airflow disturbance data and the dumbbell ball torsion angle data, the Kalman filter algorithm is used to evaluate the swing amplitude of the dumbbell ball, and an estimated value of the swing amplitude of the dumbbell ball is obtained;

[0063] The estimated value of the swing amplitude of the dumbbell ball is compared with the preset swing amplitude threshold to determine whether the swing amplitude of the dumbbell ball exceeds the preset threshold;

[0064] When the swing amplitude of the dumbbell ball exceeds the preset threshold, the torsion angle data and swing amplitude data of the dumbbell ball are input into a pre-trained support vector machine model, and the influence degree of the airflow disturbance is classified through the support vector machine model to obtain a classification result of the influence degree of the airflow disturbance;

[0065] According to the classification result of the influence degree of the airflow disturbance, a decision tree algorithm is used to generate a judgment rule for the influence degree of the airflow disturbance, and the judgment rule is applied to the subsequent monitoring of the swing of the dumbbell ball to realize the real-time judgment and early warning of the influence of the airflow disturbance.

[0066] Specifically, by obtaining the magnetic field change signal of the dumbbell ball and calculating its real-time torsion angle, combining with the airflow disturbance data, using the Kalman filter algorithm to estimate the swing amplitude of the dumbbell ball, and using the support vector machine model to classify the influence degree of the airflow disturbance, and then using the decision tree algorithm to generate a judgment rule, the accurate monitoring and real-time early warning of the influence of the airflow disturbance are realized, effectively improving the accuracy and reliability of oxygen detection, and solving the detection error problem caused by the airflow disturbance to the swing of the dumbbell ball.

[0067] S2. According to the swing amplitude data, a temperature sensor and a pressure sensor are used to collect the temperature and pressure fluctuation data in the detection room in real time, obtain the temperature and pressure compensation parameters, and combine with the preset compensation algorithm to correct the swing amplitude data to obtain the corrected dumbbell ball torsion angle data;

[0068] Furthermore, according to the swing amplitude data, a temperature sensor and a pressure sensor are used to collect the temperature and pressure fluctuation data in the detection room in real time, and the temperature and pressure compensation parameters are obtained, specifically including:

[0069] S21. Through the combination of a temperature sensor and a pressure sensor, the temperature and pressure fluctuation data in the room are collected in real time, and the original temperature and pressure data collected by the sensor are preprocessed, including data cleaning, outlier filtering, etc., to obtain effective temperature and pressure data;

[0070] S22. Establish a multiple regression model of temperature - pressure - swing amplitude based on historical temperature and pressure data, train the model parameters through machine learning algorithms such as support vector machines or random forests, input the real - time collected temperature and pressure data into the trained regression model, and predict the estimated value of the swing amplitude in the current environment;

[0071] S23. Then calculate the temperature and pressure compensation parameters. By comparing the actually measured swing amplitude with the model prediction value, obtain the deviation amount caused by temperature and pressure.

[0072] Specifically, by collecting the indoor temperature and pressure fluctuation data in real - time, combining the multiple regression model and machine learning algorithms, calculate the temperature and pressure compensation parameters, and then correct the swing amplitude data to obtain more accurate dumbbell ball torsion angle data, effectively improving the oxygen detection accuracy and solving the problem of the influence of temperature and pressure changes on the detection results.

[0073] Furthermore, combine the preset compensation algorithm to correct the swing amplitude data to obtain the corrected dumbbell ball torsion angle data, specifically including:

[0074] Use the Kalman filter algorithm to filter the swing amplitude data, establish the state equation and observation equation of the dumbbell ball movement, and obtain the filtered swing amplitude data by using recursive calculation to get the optimal estimated value;

[0075] According to the temperature and pressure compensation parameters and the preset compensation algorithm, perform non - linear compensation on the filtered swing amplitude data. By establishing the non - linear mapping relationship between the dumbbell ball swing amplitude and the torsion angle, use the look - up table method or interpolation method to determine the corresponding compensation value to obtain the compensated torsion angle data;

[0076] Use the least - squares method to perform curve fitting on the compensated torsion angle data. By selecting an appropriate fitting function and using the principle of minimizing the sum of squared residuals to determine the coefficients of the fitting function, obtain the fitted torsion angle data curve;

[0077] According to the fitted torsion angle data curve, perform smoothing processing on the torsion angle data. By setting the smoothing window and weighting coefficient, use methods such as moving average or exponential smoothing to smooth the data to obtain the smoothed torsion angle data;

[0078] Use the adaptive threshold method to perform threshold processing on the smoothed torsion angle data. By dynamically adjusting the threshold size and using the threshold judgment condition to perform binary processing on the data, obtain the threshold - processed torsion angle data;

[0079] According to the threshold - processed torsion angle data, use methods such as weighted fusion or decision tree to obtain the corrected dumbbell ball torsion angle data.

[0080] Specifically, the swing amplitude data is filtered by the Kalman filter algorithm, and then non-linear compensation is performed in combination with temperature and pressure compensation parameters. After that, curve fitting is carried out using the least squares method, and the smoothed data is processed. Then, the adaptive threshold method is used for threshold processing. Finally, the corrected dumbbell ball torsion angle data is obtained through methods such as weighted fusion or decision tree. This series of operations effectively improves the accuracy and stability of oxygen detection, making the detection results more accurate and reliable, and better reflecting the actual oxygen concentration situation.

[0081] S3. According to the corrected dumbbell ball torsion angle data, determine whether it exceeds the linear response range of the photoelectric sensor. When it exceeds the range, an array composed of multiple photoelectric sensors is used to obtain piecewise linear fitting data.

[0082] Furthermore, according to the corrected dumbbell ball torsion angle data, determine whether it exceeds the linear response range of the photoelectric sensor. When it exceeds the range, an array composed of multiple photoelectric sensors is used to obtain piecewise linear fitting data, which specifically includes:

[0083] Obtain the corrected dumbbell ball torsion angle data and input the data into the judgment module. The judgment module determines whether the torsion angle data exceeds the linear response range of a single photoelectric sensor according to the preset linear response range threshold.

[0084] When the torsion angle data does not exceed the linear response range, the complete torsion angle data is directly obtained by using a single photoelectric sensor. When the torsion angle data exceeds the linear response range, the sensor array module is triggered, and according to the preset sensor layout scheme, a matching number of photoelectric sensors are selected to form a sensor array.

[0085] The sensor array module controls each photoelectric sensor to obtain the torsion angle data within different angle ranges respectively, forming piecewise data. Then, the obtained piecewise torsion angle data is input into the data fitting module, and the piecewise linear fitting algorithm is used to fit each piece of data to obtain a complete torsion angle fitting curve.

[0086] Among them, according to the torsion angle fitting curve, the angle change data of the dumbbell ball during the entire torsion process is determined as the input for subsequent motion analysis.

[0087] The rule for determining the number of matches is as follows: when the corrected torsional angle data of the dumbbell ball does not exceed the linear response range of a single photoelectric sensor, only 1 photoelectric sensor is required; when it exceeds this range, it is calculated based on the ratio of the actual torsional angle range of the dumbbell ball to the linear response range of a single photoelectric sensor. Specifically, in the two-way measurement scenario, the number of matches is the result of rounding up this ratio multiplied by 2 to ensure coverage of the entire range from the maximum negative torsional angle to the maximum positive torsional angle; in the one-way measurement scenario, the number of matches is the result of rounding up this ratio, thus ensuring measurement accuracy in the corresponding direction.

[0088] Specifically, by judging whether the corrected torsional angle data of the dumbbell ball exceeds the linear response range of the photoelectric sensor, and when it exceeds the range, using multiple photoelectric sensors to form an array for piecewise linear fitting, the detection range is effectively expanded, the detection accuracy is improved, and the angle change data of the dumbbell ball during the entire torsional process can be accurately obtained, providing more reliable data support for subsequent oxygen concentration analysis.

[0089] S4. According to the piecewise linear fitting data, combined with a preset fitting algorithm, calculate the corresponding relationship between the oxygen concentration and the torsional angle, obtain the oxygen concentration measurement result, and judge whether there is non-linear distortion. When there is distortion, recalculate through the piecewise fitting data of the sensor array to obtain the corrected oxygen concentration value;

[0090] Furthermore, according to the piecewise linear fitting data, combined with a preset fitting algorithm, calculate the corresponding relationship between the oxygen concentration and the torsional angle, obtain the oxygen concentration measurement result, specifically including:

[0091] Obtain the set of data points for piecewise linear fitting, use the least squares method to perform linear fitting on each data segment to obtain the linear function expression of this segment, and then according to the preset fitting algorithm, determine the position of the segmentation point of the piecewise linear function, and connect the linear functions of adjacent segments to obtain a complete piecewise linear function;

[0092] Obtain the currently measured torsional angle value, calculate the theoretical value of the oxygen concentration corresponding to this torsional angle through the piecewise linear function, obtain the actual measured value of the oxygen concentration sensor, compare it with the theoretically calculated value, and when the deviation exceeds the preset threshold, judge it as abnormal and give an alarm prompt;

[0093] According to the oxygen concentration values measured multiple times, use the Kalman filtering algorithm to filter out noise, obtain a more accurate and stable oxygen concentration measurement result, and store the corresponding relationship between the measured oxygen concentration result and the torsional angle in the database for subsequent data analysis and model optimization.

[0094] Specifically, the corresponding relationship between the oxygen concentration and the torsion angle is calculated by piecewise linear fitting of data combined with a preset fitting algorithm to obtain the oxygen concentration measurement result, and it can be judged whether there is non-linear distortion. When distortion exists, the piecewise fitting data of the sensor array is used to recalculate the corrected oxygen concentration value, thereby improving the accuracy and reliability of the oxygen concentration measurement, providing more accurate data for the oxygen detection of the oxygen generation equipment, and ensuring the stability and effectiveness of the oxygen concentration measurement.

[0095] S5. Output the final oxygen concentration measurement result according to the corrected oxygen concentration value to complete the measurement process.

[0096] An oxygen detection device based on an oxygen generation device includes a processor, a memory, and computer program instructions stored in the memory. When the computer program instructions are executed by the processor, the above-mentioned oxygen detection method based on an oxygen generation device is implemented.

[0097] A storage medium stores computer program instructions, and when the computer program instructions are executed by a processor, the above-mentioned oxygen detection method based on an oxygen generation device is implemented.

[0098] The oxygen detection method based on an oxygen generation device in this embodiment first generates a uniform and stable magnetic field through a high-stability magnetic field source to reduce external interference, and arranges multiple magnetic field sensors indoors to obtain magnetic field change data at different positions, improving the spatial resolution and coverage of the magnetic field change data. Then, the magnetic field change data is preprocessed, including removing outliers, smoothing noise, etc. Next, signal processing methods such as wavelet transform are used to extract the characteristic information in the magnetic field change signal, establish a mapping relationship model between the magnetic field change signal and the oxygen concentration change, and train the model through machine learning algorithms such as support vector machines or neural networks to achieve the function of accurately estimating the indoor oxygen concentration from the magnetic field change signal. And a real-time monitoring algorithm is designed to continuously collect and analyze the magnetic field change signal to judge whether the oxygen concentration is normal and give an early warning in time;

[0099] After obtaining the magnetic field change signal, further obtain the magnetic field change signal of the dumbbell ball to calculate the real-time torsion angle, compare it with a preset threshold. When it exceeds the threshold, obtain the air flow disturbance data, use the Kalman filter algorithm to estimate the swing amplitude of the dumbbell ball, then input the relevant data into the support vector machine model to classify the influence degree of the air flow disturbance, and generate a judgment rule for application in monitoring to achieve real-time judgment and early warning of the influence of the air flow disturbance;

[0100] At the same time, a temperature sensor and a pressure sensor are used to collect the temperature and pressure fluctuation data in the detection room in real time to obtain compensation parameters, and combined with a preset compensation algorithm, the swing amplitude data is corrected, including Kalman filtering, non-linear compensation, curve fitting, smoothing processing, and threshold processing, etc., to obtain the corrected torsion angle data of the dumbbell ball and improve the oxygen detection accuracy;

[0101] Then, based on the corrected dumbbell ball torsion angle data, it is determined whether it exceeds the linear response range of the photoelectric sensor. When it exceeds the range, an array composed of multiple photoelectric sensors is used for piecewise linear fitting to obtain a complete torsion angle fitting curve, and the angle change data of the dumbbell ball during the entire torsion process is determined, providing reliable support for subsequent analysis;

[0102] Finally, based on the piecewise linear fitting data, the corresponding relationship between the oxygen concentration and the torsion angle is calculated by combining the preset fitting algorithm to obtain the oxygen concentration measurement result, and it is determined whether there is non-linear distortion. When there is distortion, the piecewise fitting data of the sensor array is used to recalculate the corrected oxygen concentration value to improve the measurement accuracy and reliability, and finally the corrected oxygen concentration measurement result is output to complete the measurement process.

[0103] Through a series of delicate steps and advanced algorithms, the whole method effectively solves problems such as external interference, airflow disturbance, temperature and pressure changes, and sensor response limitations, realizes high-precision, high-stability and real-time monitoring of the oxygen concentration in the detection chamber, provides more accurate and reliable data for the oxygen detection of oxygen production equipment, ensures the stability and effectiveness of the oxygen concentration measurement, improves the reliability and safety of the oxygen detection of oxygen production equipment, and provides a strong guarantee for relevant application scenarios.

[0104] The above is only a preferred embodiment of the present invention, and does not impose any form of limitation on the present invention. Although the present invention has been disclosed as above with the preferred embodiment, it is not intended to limit the present invention. Any person skilled in the art can make some changes or modifications to the equivalent embodiments by using the disclosed technical content within the scope of the technical solution of the present invention. However, as long as it does not depart from the content of the technical solution of the present invention, any brief modifications, equivalent changes and modifications made to the above embodiments according to the technical essence of the present invention still fall within the scope of the technical solution of the present invention.

Claims

1. An oxygen detection method based on an oxygen generation device, characterized in that: It includes the following steps: Obtain the magnetic field change data caused by the oxygen concentration in the detection chamber, reduce external interference through a high-stability magnetic field source to obtain a magnetic field change signal, then calculate the torsion angle of the dumbbell ball. When the torsion angle exceeds the preset threshold, judge the influence of air flow disturbance on the swing of the dumbbell ball, analyze the influence degree of air flow disturbance on the swing of the dumbbell ball through air flow disturbance data, and then according to the air flow disturbance data and the dumbbell ball torsion angle data, use the Kalman filter algorithm to evaluate the swing amplitude of the dumbbell ball to obtain an estimated value of the swing amplitude of the dumbbell ball, and obtain the swing amplitude data; According to the swing amplitude data, use temperature sensors and pressure sensors to collect the temperature and pressure fluctuation data in the detection chamber in real time. Establish a multiple regression model based on the historical temperature, pressure and swing amplitude data. Input the real-time temperature and pressure data into the model to predict the estimated value of the swing amplitude. Obtain the deviation amount caused by temperature and pressure by comparing the actual measurement value with the predicted value, obtain the temperature and pressure compensation parameters, and combine with the preset compensation algorithm to perform non-linear compensation on the filtered swing amplitude data. By establishing a non-linear mapping relationship between the swing amplitude of the dumbbell ball and the torsion angle, use the look-up table method or interpolation method to determine the corresponding compensation value to correct the swing amplitude data and obtain the corrected dumbbell ball torsion angle data; According to the corrected dumbbell ball torsion angle data, judge whether it exceeds the linear response range of the photoelectric sensor. When it exceeds the range, use multiple photoelectric sensors to form an array to obtain piecewise linear fitting data; According to the piecewise linear fitting data, combine with the preset fitting algorithm to calculate the corresponding relationship between the oxygen concentration and the torsion angle, obtain the oxygen concentration measurement result, and judge whether there is non-linear distortion. When there is distortion, recalculate through the piecewise fitting data of the sensor array to obtain the corrected oxygen concentration value.

2. The oxygen detection method based on an oxygen generation device according to claim 1, wherein: Obtain the magnetic field change data caused by the oxygen concentration in the detection chamber, reduce external interference through a high-stability magnetic field source to obtain a magnetic field change signal, specifically including: According to the high-stability magnetic field source, generate a uniform and stable magnetic field through the magnetic field source to reduce the influence of external environmental magnetic field interference on the indoor magnetic field change. Arrange multiple magnetic field sensors in the room to obtain magnetic field change data at different positions, and improve the spatial resolution and coverage of the magnetic field change data through multi-point sampling; Preprocess the obtained magnetic field change data. According to the characteristics of the magnetic field change data, use the wavelet transform signal processing method to extract the characteristic information in the magnetic field change signal as an indicator of the oxygen concentration change; Then establish a mapping relationship model between the magnetic field change signal and the oxygen concentration change, and train the mapping model through machine learning algorithms such as support vector machines or neural networks; In the mapping relationship model, through the design of a real-time monitoring algorithm for oxygen concentration change, continuously collect and analyze the magnetic field change signal, judge whether the current oxygen concentration is within the normal range, and give a timely warning when it is lower than the preset threshold.

3. The oxygen detection method based on an oxygen generation device according to claim 2, wherein: After obtaining the magnetic field change signal, it further includes: acquiring the magnetic field change signal of the dumbbell ball, calculating the real-time torsion angle of the dumbbell ball through the magnetic field change signal, comparing the calculated real-time torsion angle of the dumbbell ball with a preset torsion angle threshold, and determining whether the torsion angle of the dumbbell ball exceeds the preset threshold; When the torsion angle of the dumbbell ball exceeds the preset threshold, acquire the current airflow disturbance data, analyze the influence degree of the airflow disturbance on the swing of the dumbbell ball through the airflow disturbance data, and then evaluate the swing amplitude of the dumbbell ball by using the Kalman filter algorithm based on the airflow disturbance data and the dumbbell ball torsion angle data to obtain an estimated value of the swing amplitude of the dumbbell ball; Compare the estimated value of the swing amplitude of the dumbbell ball with a preset swing amplitude threshold to determine whether the swing amplitude of the dumbbell ball exceeds the preset threshold; When the swing amplitude of the dumbbell ball exceeds the preset threshold, input the torsion angle data and swing amplitude data of the dumbbell ball into a pre-trained support vector machine model, classify the influence degree of the airflow disturbance through the support vector machine model to obtain a classification result of the influence degree of the airflow disturbance; According to the classification result of the influence degree of the airflow disturbance, generate a judgment rule for the influence degree of the airflow disturbance by using the decision tree algorithm, and apply the judgment rule to the monitoring of the dumbbell ball swing to perform real-time judgment and early warning on the airflow disturbance.

4. The oxygen detection method based on an oxygen generation device according to claim 1, wherein: According to the swing amplitude data, use a temperature sensor and a pressure sensor to collect the indoor temperature and pressure fluctuation data in real time to obtain temperature and pressure compensation parameters, specifically including: Through the combination of a temperature sensor and a pressure sensor, collect the indoor temperature and pressure fluctuation data in real time, and then preprocess the original temperature and pressure data collected by the sensor to obtain effective temperature and pressure data; According to the historical temperature and pressure data, establish a multiple regression model of temperature-pressure-swing amplitude, train the model parameters through a machine learning algorithm, and input the real-time collected temperature and pressure data into the trained regression model to predict the estimated value of the swing amplitude in the current environment; Then calculate the temperature and pressure compensation parameters, and obtain the deviation amount caused by temperature and pressure by comparing the actually measured swing amplitude with the model prediction value.

5. The oxygen detection method based on an oxygen generation device according to claim 1, characterized in that: Combine a preset compensation algorithm to correct the swing amplitude data to obtain the corrected torsion angle data of the dumbbell ball, specifically including: Use the Kalman filter algorithm to filter the swing amplitude data, establish the state equation and observation equation of the dumbbell ball movement, and use recursive calculation to obtain the optimal estimated value to obtain the filtered swing amplitude data; According to the temperature and pressure compensation parameters and combine the preset compensation algorithm to perform non-linear compensation on the filtered swing amplitude data. By establishing a non-linear mapping relationship between the swing amplitude and torsion angle of the dumbbell ball, use the look-up table method or interpolation method to determine the corresponding compensation value to obtain the compensated torsion angle data; Use the least squares method to perform curve fitting on the compensated torsion angle data. By selecting a fitting function and using the principle of minimizing the sum of squared residuals to determine the coefficients of the fitting function, obtain the fitted torsion angle data curve; According to the fitted torsional angle data curve, smooth the torsional angle data. By setting a smoothing window and a weighting coefficient, use the moving average or exponential smoothing method to smooth the data and obtain the smoothed torsional angle data; Use the adaptive threshold method to perform threshold processing on the smoothed torsional angle data. By dynamically adjusting the threshold size, use the threshold judgment condition to perform binary processing on the data and obtain the torsional angle data after threshold processing; According to the torsional angle data after threshold processing, use the weighted fusion or decision tree method to obtain the corrected torsional angle data of the dumbbell ball.

6. The oxygen detection method based on an oxygen generation device according to claim 1, characterized in that: According to the corrected torsional angle data of the dumbbell ball, judge whether it exceeds the linear response range of the photoelectric sensor. When it exceeds the range, use multiple photoelectric sensors to form an array to obtain piecewise linear fitting data, specifically including: Obtain the corrected torsional angle data of the dumbbell ball and input the data into the judgment module. The judgment module judges whether the torsional angle data exceeds the linear response range of a single photoelectric sensor according to the preset linear response range threshold; When the torsional angle data does not exceed the linear response range, use a single photoelectric sensor to obtain the complete torsional angle data. When the torsional angle data exceeds the linear response range, trigger the sensor array module, and select a matching number of photoelectric sensors to form a sensor array according to the preset sensor layout scheme; The sensor array module controls each photoelectric sensor to obtain the torsional angle data in different angle ranges respectively to form segmented data, and then input the obtained segmented torsional angle data into the data fitting module. Use the piecewise linear fitting algorithm to fit each segment of data to obtain the complete torsional angle fitting curve.

7. The oxygen detection method based on an oxygen generation device according to claim 1, wherein: According to the piecewise linear fitting data, combine the preset fitting algorithm to calculate the corresponding relationship between the oxygen concentration and the torsional angle to obtain the oxygen concentration measurement result, specifically including: Obtain the set of data points of the piecewise linear fitting, use the least squares method to perform linear fitting on each data segment to obtain the linear function expression of this segment, and then determine the position of the segmentation point of the piecewise linear function according to the preset fitting algorithm. Connect the linear functions of adjacent segments to obtain the complete piecewise linear function; Obtain the currently measured torsional angle value, calculate the theoretical value of the oxygen concentration corresponding to this torsional angle through the piecewise linear function, obtain the actual measured value of the oxygen concentration sensor, and compare it with the theoretically calculated value. When the deviation exceeds the preset threshold, it is judged as abnormal and an alarm prompt is given; According to the oxygen concentration values measured multiple times, use the Kalman filter algorithm to filter out noise to obtain a more accurate and stable oxygen concentration measurement result, and store the corresponding relationship between the measured oxygen concentration result and the torsional angle in the database.

8. A method for detecting oxygen based on an oxygen generation device according to claim 1, characterized in that: It also includes: Output the final oxygen concentration measurement result according to the corrected oxygen concentration value to complete the measurement process.

9. An oxygen detection device based on an oxygen generation device, characterized in that: It includes a processor, a memory, and computer program instructions stored in the memory. When the computer program instructions are executed by the processor, it implements an oxygen detection method based on an oxygen generation device described in any one of claims 1-8 above.

10. A storage medium having computer program instructions stored thereon, characterized in that, When the computer program instructions are executed by a processor, an oxygen detection method based on an oxygen generation device described in any one of claims 1-8 above is implemented.

Images