Resuscitator dead space testing method and system

Abstract

The present invention relates to the field of dead space testing technology, specifically to a dead space testing method and system for a resuscitator, comprising the following steps: based on the readings of an airflow velocity sensor and a pressure sensor, calculating the flow-pressure ratio at each time point, determining the baseline data of the resuscitator under different operating conditions through the ratio, performing stability evaluation on the performance of the resuscitator, and obtaining preliminary stability diagnosis results. The present invention uses flow-pressure ratio calculation to more accurately identify and adjust the performance of the resuscitator under different operating conditions, predict the behavior of the resuscitator under future use conditions through real-time analysis and model optimization, and allow adjustment of its parameters to best suit current use needs, ensuring that the resuscitator can provide the most effective ventilation support at critical moments, and further performance deviation monitoring and abnormality identification capabilities, which can respond quickly when any deviation affecting the efficiency of the device is found, thereby reducing the risk of improper configuration of the resuscitator.

Description

Technical Field

[0001] The present invention relates to the technical field of dead space testing, and in particular to a method and system for testing dead space of a resuscitator. Background Art

[0002] The field of dead space testing technology involves measuring and evaluating the amount of air in the respiratory system that is not involved in gas exchange, the so-called "dead space". In medicine and bioengineering, this technology is mainly used to evaluate and optimize the performance of equipment such as ventilators and resuscitators. By measuring the dead space in the airway, medical personnel can better understand the efficiency of the equipment in actual use and the respiratory status of the patient. The determination of dead space capacity is crucial to ensure that patients receive appropriate ventilation, especially in clinical anesthesia and intensive care.

[0003] Among them, the resuscitator dead space test method focuses on the development and implementation of specific technologies for measuring dead space in resuscitators. A resuscitator is a medical device used to support breathing in emergency situations, usually when the patient is unable to breathe on their own. Dead space testing is extremely important for determining whether the resuscitator can effectively provide adequate ventilation to the patient and whether it helps to remove carbon dioxide, so as to ensure that the device can provide the expected support at critical moments and reduce the risk of potential complications caused by excessive dead space.

[0004] Traditional methods rely on static or preset measurement processes, which are insufficient to provide accurate performance evaluation in rapidly changing clinical operation scenarios. The lack of effective real-time data analysis capabilities leads to delays in existing methods in predicting the performance of the device under different operating conditions, limiting the application flexibility and practicality of the device. The inability to adjust and optimize device settings in a timely manner makes it impossible to accurately measure the dead space capacity in actual use, which in turn affects the overall device performance and safety. Summary of the invention

[0005] The purpose of the present invention is to solve the shortcomings of the prior art and to propose a resuscitator dead space testing method and system.

[0006] In order to achieve the above object, the present invention adopts the following technical solution: a resuscitator dead space testing method, comprising the following steps:

[0007] S1: Based on the readings of the airflow velocity sensor and the pressure sensor, the flow-pressure ratio at each time point is calculated, and the baseline data of the resuscitator under different operating conditions is determined by the ratio, and the stability of the resuscitator performance is evaluated to obtain preliminary stability diagnosis results;

[0008] S2: Based on the preliminary stability diagnosis results, using a gradient boosting tree, classifying the difference operation states, identifying the key operation variables that affect the performance, determining the key factors that affect the change in dead space capacity, and obtaining key factor identification information;

[0009] S3: Based on the key factor identification information, the data of the O2 concentration sensor is analyzed in time series, and the change of O2 concentration in the future state is predicted by moving average and exponential smoothing method, and the model parameters are adjusted according to the prediction results and historical data to obtain an optimized prediction model;

[0010] S4: Based on the optimized prediction model, the dead space volume is monitored, key time point data is recorded, the key time point data is analyzed, performance deviations and potential resuscitator abnormalities are identified, and abnormal test identification records are obtained.

[0011] The present invention is improved in that the calculation steps of the flow pressure ratio are specifically as follows:

[0012] S111: Based on the air velocity sensor reading and the pressure sensor reading, combined with the ambient temperature and humidity, the formula is used:

[0013]

[0014] Calculate each time point Flow pressure ratio ,in, Representing time point The air flow velocity, Representing time point The air flow pressure, Representing time point The ambient temperature, Representing time point The ambient humidity, is the temperature adjustment factor, is the adjustment factor for humidity;

[0015] S112: Using the collected flow pressure ratio , combined with the real-time calibration coefficient, using the formula:

[0016]

[0017] Adjust the flow-pressure ratio at each time point to obtain the calibrated flow-pressure ratio ,in, represents the calibration factor;

[0018] S113: Calculate the standard deviation of the calibrated flow-pressure ratio , the data consistency is evaluated using the formula:

[0019]

[0020] in, represents the sample size, yes The average value of .

[0021] The present invention is improved in that the steps of evaluating the performance stability of the resuscitator are specifically as follows:

[0022] S121: Using the flow pressure ratio and baseline data, the formula is:

[0023]

[0024] Calculate the difference percentage ,in, is the calibrated flow pressure ratio, is the baseline flow-pressure ratio, To optimize the numerical stability of a small constant, To avoid small positive values ​​with zero denominator;

[0025] S122: Comparison and stability threshold, using logical judgment: if < stability threshold, the resuscitator performance is stable. ≥ the stability threshold, the performance is unstable, and a preliminary stability diagnosis result is obtained based on the judgment result.

[0026] The present invention is improved in that the step of identifying the key operating variables is specifically as follows:

[0027] S211: Based on the preliminary stability diagnosis result, a gradient boosting tree model is applied to classify the difference operation states using the formula:

[0028]

[0029] in, Represents the classification result, is the splitting function in the decision tree, is an auxiliary analysis function, For the operating state, is the operating variable, For the The weight of a tree, represents the number of trees, and are weight adjustment coefficients, which are used to adjust and Contribution

[0030] S212: Based on the obtained classification results, calculate the criticality score and stability index of each variable in the classification results, using the formula:

[0031]

[0032] Generate key operating variables that affect performance , and is the weight coefficient, Function is used to calculate the given classification result Lower variable The key score, Functions are used to calculate Stability score;

[0033] S213: Identification of key operational variables For analysis, the formula is:

[0034]

[0035] Calculate weighted correlations between key operating variables and performance data ,in, is the correlation function used to calculate the key operating variables and performance data points The relationship between is a performance data point, It is The weight of the data point, is the total number of data points.

[0036] The present invention is improved in that the step of obtaining the key factor identification information is specifically as follows:

[0037] S221: Statistical analysis is performed on the key operating variables using the formula:

[0038]

[0039] in, is the influencing factor for initial identification, is the key operating variable, As the influencing factors, is the weight of the influencing factor, and Represents statistical analysis functions and auxiliary analysis functions, respectively, used to measure and The relationship between To adjust the coefficient, adjust the influence of the auxiliary analysis function in the overall analysis, is the total number of influencing factors;

[0040] S222: Yes To conduct factor analysis, the formula is:

[0041]

[0042] in, For the refined influencing factors, is the factor analysis function, It is The weight of the factor, is the number of factors;

[0043] S223: Based on the refined influencing factors , identify the key factors and apply the formula:

[0044]

[0045] in, Identify information for key factors, represents the weighting coefficient, is the number of weighting factors.

[0046] The present invention is improved in that the steps of obtaining the optimized prediction model are specifically as follows:

[0047] S311: Based on the key factor identification information, a time series analysis is performed on the data of the O2 concentration sensor, and a moving average method is applied, using the formula:

[0048]

[0049] Calculation time The weighted moving average , generate moving average data, where is the time window size, is the weight coefficient, For in time O2 concentration;

[0050] S312: Use exponential smoothing to iteratively smooth the data, using the formula:

[0051]

[0052] Calculate the time Exponential smoothing results for , generates exponential smoothing data, where Representative time O2 concentration, represents the smoothing coefficient, represents the adjustment item, It's time Exponential smoothing value of ;

[0053] S313: Predict the change of O2 concentration in the future state based on the moving average and exponential smoothing results, and adjust the model parameters in combination with historical data, using the formula:

[0054]

[0055] Get the optimized prediction model ,in, and are the moving average and exponential smoothing results respectively, is the adjustment factor, Adjust for bias.

[0056] The present invention is improved in that the step of recording the key time point data is specifically as follows:

[0057] S411: Obtain O2 concentration data from the optimized prediction model, perform monitoring activities, and regularly record O2 concentration using the formula:

[0058]

[0059] in, For time The calculated O2 concentration value, is the prediction model function, They are model parameters, which are used to adjust and match environmental changes and sensor properties;

[0060] S412: Obtained O2 concentration value Perform time series analysis to identify abnormal fluctuations or key changes using the formula:

[0061]

[0062] Get smoothed data ,in, For time The weighted nonlinear average O2 concentration, is the time window size, represents the time weight factor, is the data conversion parameter;

[0063] S413: Based on smoothed data , analyze key time points, identify key change points in the data, and use the formula:

[0064]

[0065] in, Represents the rate of change of data at key time points, and represent the sensitivity and exponential conversion parameters respectively.

[0066] The present invention is improved in that the step of obtaining the abnormal test identification record is specifically as follows:

[0067] S421: Using the key time point data, evaluate whether the data point indicates a performance deviation, using the formula:

[0068]

[0069] Generate preliminary anomaly identification results ,in, is the data change rate at the key time point, , and It is used to adjust the sensitivity and response curve of the model;

[0070] S422: Based on the preliminary abnormality identification results , analyze potential resuscitator anomalies using the formula:

[0071]

[0072] Generate exception records ,in, is the multiple threshold, for The standard deviation of Classified as abnormal;

[0073] S423: For the Keep records and establish abnormal test identification records , using the formula:

[0074]

[0075] in, is the total number of time points within the evaluation time range, It's time point The weight of .

[0076] A resuscitator dead space testing system, the system comprising:

[0077] The flow-pressure ratio module calculates the flow-pressure ratio at each time point based on the readings of the airflow velocity sensor and the pressure sensor, and evaluates the performance of the resuscitator under multiple operating states according to the flow-pressure ratio to obtain a preliminary stability diagnosis record;

[0078] The key variable analysis module classifies the operation status based on the preliminary stability diagnosis record using a gradient boosting tree, identifies key operation variables that affect performance, uses the key operation variables to determine key factors that affect the change in dead space capacity, and establishes key factor identification information;

[0079] The concentration prediction and model optimization module uses the key factor identification information to perform time series analysis on the O2 concentration sensor data, uses moving average and exponential smoothing methods to predict the O2 concentration information, and adjusts the model parameters according to the prediction results and historical data to obtain an optimized prediction model;

[0080] The dead space monitoring module uses the optimized prediction model to continuously monitor the dead space capacity of the resuscitator, record key time point data, analyze the data and identify performance deviations, and obtain preliminary abnormal test records;

[0081] The abnormality analysis module performs iterative analysis of performance deviations based on the preliminary abnormality test record, identifies and verifies potential resuscitator abnormalities by comparing historical data, and generates an abnormality test identification record.

[0082] Compared with the prior art, the advantages and positive effects of the present invention are:

[0083] In the present invention, the flow-pressure ratio calculation is used to more accurately identify and adjust the performance of the resuscitator in different operating states. Through real-time analysis and model optimization, the behavior of the resuscitator in future usage states can be predicted, allowing its parameters to be adjusted to best suit current usage needs, ensuring that the resuscitator can provide the most effective ventilation support at critical moments. Further performance deviation monitoring and abnormality identification capabilities can respond quickly when any deviation that affects the efficiency of the equipment is found, thereby reducing the risks of the resuscitator due to improper configuration. BRIEF DESCRIPTION OF THE DRAWINGS

[0084] Figure 1 A flow chart of a dead space testing method for a resuscitator proposed by the present invention;

[0085] Figure 2 A flow chart for calculating the flow-pressure ratio in the dead space test method of a resuscitator proposed by the present invention;

[0086] Figure 3 A flow chart for evaluating the stability of resuscitator performance in a resuscitator dead space testing method proposed by the present invention;

[0087] Figure 4 A flow chart for identifying key operating variables in a dead space test method for a resuscitator is provided for the present invention;

[0088] Figure 5 A flowchart for obtaining key factor identification information in a dead space test method for a resuscitator proposed by the present invention;

[0089] Figure 6 A flowchart for obtaining the optimized prediction model in the dead space testing method of the resuscitator proposed by the present invention;

[0090] Figure 7A flow chart for recording data at key time points in a dead space test method for a resuscitator is provided for the present invention;

[0091] Figure 8 The present invention proposes a flow chart for obtaining abnormal test identification records in a resuscitator dead space testing method. DETAILED DESCRIPTION

[0092] In order to make the purpose, technical solution and advantages of the present invention more clearly understood, the present invention is further described in detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not intended to limit the present invention.

[0093] In the description of the present invention, it should be understood that the terms "length", "width", "up", "down", "front", "back", "left", "right", "vertical", "horizontal", "top", "bottom", "inside", "outside" and the like indicate positions or positional relationships based on the positions or positional relationships shown in the accompanying drawings, and are only for the convenience of describing the present invention and simplifying the description, rather than indicating or implying that the device or element referred to must have a specific orientation, be constructed and operated in a specific orientation, and therefore cannot be understood as limiting the present invention. In addition, in the description of the present invention, the meaning of "plurality" is two or more, unless otherwise clearly and specifically defined.

[0094] Example

[0095] See also Figure 1 The present invention provides a technical solution: a resuscitator dead space testing method, comprising the following steps:

[0096] S1: Based on the readings of the airflow velocity sensor and the pressure sensor, the flow-pressure ratio at each time point is calculated, and the baseline data of the resuscitator under different operating conditions is determined by the ratio. The stability of the resuscitator performance is evaluated using the baseline data to obtain preliminary stability diagnosis results;

[0097] S2: Based on the preliminary stability diagnosis results, use the gradient boosting tree to classify the differential operating states, identify the key operating variables that affect the performance, analyze the key operating variables, determine the key factors that affect the change in dead space capacity, and obtain key factor identification information;

[0098] S3: Based on the key factor identification information, the data of the O2 concentration sensor is analyzed in time series, and the change of O2 concentration in the future state is predicted by moving average and exponential smoothing method. The model parameters are adjusted according to the prediction results and historical data to obtain the optimized prediction model;

[0099] S4: Based on the optimized prediction model, the dead space volume is monitored, the key time point data is recorded, the key time point data is analyzed, the performance deviation and potential resuscitator abnormalities are identified, and the abnormal test identification record is obtained.

[0100] The preliminary stability diagnosis results include flow stability, pressure stability, and performance consistency indicators. The key factor identification information includes operation frequency impact information, pressure adjustment sensitivity, and flow variation factor. The optimized prediction model includes prediction accuracy, response time optimization results, and historical fitting. The abnormal test identification records include abnormal occurrence frequency, abnormal intensity level, and identification efficiency information.

[0101] Using 100% oxygen as the tracking gas, a resuscitator was used to ventilate a "bag in a bottle" container, and the total dead space capacity of the resuscitator was calculated using the ventilation volume and the oxygen concentration of the inhaled gas in the bag.

[0102] Install the dead space tester, turn off the switch to the oxygen analyzer, open the ball valve, connect the resuscitator, and ventilate until the air bag completely fills the container and presses the inner wall. Close the ball valve, turn on the switch of the oxygen analyzer, turn on the flow meter, and inflate the container with 100% oxygen by volume. When the pressure gauge reads about 1kPa (≈10cmH2O), turn off the oxygen flow meter and the switch of the oxygen analyzer;

[0103] Connect the 22mm / 15mm test joint to the 22mm internal cone sleeve, and introduce an appropriate amount of air into the side mastoid joint;

[0104] Open the ball valve and flush the exhalation channel with 100% oxygen by volume;

[0105] The lungs are ventilated by blocking and releasing a 10-mm diameter hole with a finger, and the tidal volume is kept constant using a tidal volume meter and pressure gauge;

[0106] Close the ball valve, turn on the switch of the oxygen analyzer, adjust the flow rate of 100% oxygen by volume to 5L / min, and record the oxygen concentration reading in the air bag. , when the pressure gauge reading returns to 1kPa (≈10cmH2O), turn off the oxygen flow;

[0107] For each test parameter combination, the internal dead space capacity of the test instrument is determined. The test instrument can now be used to test the resuscitator using the same steps as the test connector.

[0108] Calculate the system dead space (including the test joint) using the following formula:

[0109]

[0110] In the formula, The dead space of the system. is the oxygen concentration in the airbag, Should be 20mL or less;

[0111] The dead space of the resuscitator under test is calculated using the following formula:

[0112]

[0113]

[0114] The above table shows the test parameters for calculating the dead space volume of the resuscitator.

[0115] See also Figure 2 , the calculation steps of flow pressure ratio are as follows:

[0116] S111: Based on the air velocity sensor reading and the pressure sensor reading, combined with the ambient temperature and humidity, the formula is used:

[0117]

[0118] Calculate each time point Flow pressure ratio ,in, Representing time point The air flow velocity, Representing time point The air flow pressure, Representing time point The ambient temperature, Representing time point The ambient humidity, is the temperature adjustment factor, is the humidity adjustment factor, which adjusts the pressure reading to reflect the effects of ambient conditions;

[0119] S112: Using the collected flow pressure ratio , combined with the real-time calibration coefficient, using the formula:

[0120]

[0121] Adjust the flow-pressure ratio at each time point to obtain the calibrated flow-pressure ratio ,in, represents the calibration factor;

[0122] S113: Calculate the standard deviation of the calibrated flow-pressure ratio , the data consistency is evaluated using the formula:

[0123]

[0124] in, represents the sample size, yes The average value of .

[0125] Assume the following data:

[0126] Assume that at a certain point in time , the measured air velocity m / s, air flow pressure Pa, ambient temperature ℃, ambient humidity , set the temperature adjustment coefficient and humidity adjustment factor .

[0127] calculate Values:

[0128]

[0129]

[0130] Then calculate

[0131]

[0132] Setting the calibration factor .

[0133] Calculate the calibrated flow-pressure ratio:

[0134]

[0135] Suppose five calibrated ratio data are collected:

[0136] =[0.00525, 0.00520, 0.00530, 0.00525, 0.00510];

[0137] Calculate the average:

[0138]

[0139] Calculate the standard deviation:

[0140]

[0141]

[0142] See also Figure 3 , the specific steps for evaluating the performance stability of the resuscitator are:

[0143] S121: Using the flow-pressure ratio and baseline data, use the formula:

[0144]

[0145] Calculate the difference percentage ,in, is the calibrated flow pressure ratio, is the baseline flow-pressure ratio, To optimize the numerical stability of a small constant, To avoid small positive values ​​with zero denominator;

[0146] S122: Comparison and stability threshold, using logical judgment: if < stability threshold, the resuscitator performance is stable. ≥ the stability threshold, the performance is unstable, and a preliminary stability diagnosis result is obtained based on the judgment result.

[0147] Assume the following data:

[0148] Assume 0.006;

[0149] Assume 0.0055;

[0150] A small constant that increases numerical stability, assumed to be 0.00001;

[0151] Avoid small positive values ​​with zero denominator, say 0.0001.

[0152] Calculation process:

[0153] First the square of the difference is calculated and the stability constant is added :

[0154]

[0155] Add to

[0156]

[0157] Square root:

[0158]

[0159] Next calculate the denominator and add the small positive value

[0160]

[0161] Final calculation

[0162]

[0163] This gives 57.14%, assuming a stability threshold of 10%.

[0164] judge:

[0165] <stability threshold, while 57.14%>10%, resulting in a preliminary stability diagnosis of “unstable”.

[0166] See also Figure 4 , the identification steps of key operating variables are as follows:

[0167] S211: Based on the preliminary stability diagnosis results, the gradient boosting tree model is applied to classify the differential operation states using the formula:

[0168]

[0169] in, Represents the classification result, is the splitting function in the decision tree, is an auxiliary analysis function, For the operating state, is the operating variable, For the The weight of a tree, represents the number of trees, and are weight adjustment coefficients, which are used to adjust and Contribution

[0170] S212: Based on the obtained classification results, calculate the criticality score and stability index of each variable in the classification results, using the formula:

[0171]

[0172] Generate key operating variables that affect performance , and is the weight coefficient used to adjust the impact of importance and stability indicators, Function is used to calculate the given classification result Lower variable The key score, Functions are used to calculate Stability score;

[0173] S213: Identification of key operational variables For analysis, the formula is:

[0174]

[0175] Calculate weighted correlations between key operating variables and performance data ,in, is the correlation function used to calculate the key operating variables and performance data points The relationship between is a performance data point, It is The weight of the data point, is the total number of data points.

[0176] Assume the following data:

[0177] and is the splitting function of the decision tree, which is used to split data based on different features and operation states;

[0178] is the operating state, assuming different machine settings (e.g., temperature, speed);

[0179] for operational variables, including different configuration parameters;

[0180] is the tree weight, assuming that each tree has equal weight, take =1 / m;

[0181] and is the weight adjustment coefficient, assuming =0.6 and =0.4;

[0182] is the number of trees, assuming there are 10 trees.

[0183] Assume there are three operating states and the corresponding operational variables , the splitting function in each state and The calculation results are as follows:

[0184]

[0185]

[0186]

[0187] calculate

[0188]

[0189]

[0190] Assume the following data:

[0191] Assumptions =0.7 and =0.3;

[0192] , calculated as:

[0193] =0.8, =0.6, importance =0.7;

[0194] Stability, calculated as:

[0195] stability =0.9, stability =0.85, stability =0.8.

[0196] calculate :

[0197]

[0198]

[0199] Assume the following data:

[0200] The correlation function assumes Generate random results;

[0201] The data point weights are assumed to be equal, i.e. ;

[0202] Performance data points, assuming there are five points .

[0203] calculate

[0204]

[0205] See also Figure 5 ,The specific steps for obtaining key factor identification information are:

[0206] S221: Perform statistical analysis on key operating variables using the formula:

[0207]

[0208] in, is the influencing factor for initial identification, is the key operating variable, As the influencing factors, is the weight of the influencing factor, and Represents statistical analysis functions and auxiliary analysis functions, respectively, used to measure and The relationship between To adjust the coefficient, adjust the influence of the auxiliary analysis function in the overall analysis, is the total number of influencing factors;

[0209] S222: Yes To conduct factor analysis, the formula is:

[0210]

[0211] in, For the refined influencing factors, is the factor analysis function, It is The weight of the factor, is the number of factors;

[0212] S223: Based on the refined influencing factors , identify the key factors and apply the formula:

[0213]

[0214] in, Identify information for key factors, represents the weighting coefficient, is the number of weighting factors.

[0215] Assume the following data:

[0216] Set them to equal weights, assuming each is 1;

[0217] To adjust the coefficient, adjust the influence of the auxiliary analysis function in the overall analysis, assuming it is 0.5;

[0218] The number of influencing factors, assumed to be 3;

[0219] ;

[0220] ;

[0221] .

[0222] calculate:

[0223]

[0224]

[0225]

[0226] Assume the following data:

[0227] Factor analysis function, to conduct in-depth analysis of initial factors;

[0228] is the factor weight, set to equal weight 1;

[0229] is the number of factors, assumed to be 2.

[0230] ;

[0231] .

[0232] calculate:

[0233]

[0234]

[0235] Assumptions =1, =1.

[0236] calculate:

[0237]

[0238]

[0239] See also Figure 6 , the specific steps for obtaining the optimized prediction model are:

[0240] S311: Based on the key factor identification information, the data of the O2 concentration sensor is analyzed in time series, and the moving average method is applied, using the formula:

[0241]

[0242] Calculation time The weighted moving average , generate moving average data, where is the time window size, is the weight coefficient, For in time O2 concentration;

[0243] S312: Use exponential smoothing to iteratively smooth the data, using the formula:

[0244]

[0245] Calculate the time Exponential smoothing results for , generates exponential smoothing data, where Representative time O2 concentration, represents the smoothing coefficient, represents the adjustment item, It's time Exponential smoothing value of ;

[0246] S313: Predict the change of O2 concentration in the future state based on the moving average and exponential smoothing results, and adjust the model parameters in combination with historical data, using the formula:

[0247]

[0248] Get the optimized prediction model ,in, and are the moving average and exponential smoothing results respectively, is the adjustment factor, Adjust for bias.

[0249] Assume the following data:

[0250] Defined as 5, it means considering the data of the current time point and the previous 4 time points;

[0251] Assume that the O2 concentrations at the last five time points are:

[0252] =400ppm;

[0253] =395ppm;

[0254] =390ppm;

[0255] =385ppm;

[0256] =380ppm.

[0257] Given a weight coefficient, the more recent the data, the greater the weight. For example:

[0258] =0.5;

[0259] =0.2;

[0260] =0.15;

[0261] =0.1;

[0262] =0.05.

[0263] calculate:

[0264]

[0265]

[0266]

[0267]

[0268]

[0269] Assume the following data:

[0270] It is set to 0.3 to define the influence of the current observation;

[0271] Set to +2 to correct the deviation;

[0272] ppm;

[0273] ppm (inherited from the previous cycle).

[0274] calculate:

[0275]

[0276]

[0277] Assume the following data:

[0278] It is set to 0.5 to balance the results of MA and ES;

[0279] Set to -3 for final output adjustment.

[0280] calculate:

[0281]

[0282]

[0283] See also Figure 7 , the specific steps for recording data at key time points are:

[0284] S411: Obtain O2 concentration data from the optimized prediction model, perform monitoring activities, and regularly record O2 concentration using the formula:

[0285]

[0286] in, For time The calculated O2 concentration value, is the prediction model function, They are model parameters, which are used to adjust and match environmental changes and sensor properties;

[0287] S412: Obtained O2 concentration value Perform time series analysis to identify abnormal fluctuations or key changes using the formula:

[0288]

[0289] Get smoothed data ,in, For time The weighted nonlinear average O2 concentration, is the time window size, represents the time weight factor, is the data conversion parameter;

[0290] S413: Based on smoothed data , analyze key time points, identify key change points in the data, and use the formula:

[0291]

[0292] in, Represents the rate of change of data at key time points, and represent the sensitivity and exponential conversion parameters respectively.

[0293] Assume the following data:

[0294] ;

[0295] time .

[0296] Assume the model is a linear regression model to predict O2 concentration:

[0297]

[0298]

[0299] Assume the following data:

[0300] Select 5;

[0301] Assumptions The data sequence is as follows:

[0302] ;

[0303] Weighting factors:

[0304] ;

[0305] .

[0306] Calculate the smoothed data:

[0307]

[0308]

[0309]

[0310]

[0311] Assume the following data:

[0312] Set to a fixed value of 0.05;

[0313] Set to 2.

[0314] Calculation process:

[0315] Assume that the smoothed data at the previous time point It is 40900 now is 40985.4:

[0316]

[0317]

[0318]

[0319]

[0320]

[0321] See also Figure 8 ,The specific steps for obtaining abnormal test identification records are:

[0322] S421: Using the key time point data, evaluate whether the data point indicates a performance deviation, using the formula:

[0323]

[0324] Generate preliminary anomaly identification results ,in, is the data change rate at the key time point, , and It is used to adjust the sensitivity and response curve of the model;

[0325] S422: Based on the preliminary abnormality identification results , analyze potential resuscitator anomalies using the formula:

[0326]

[0327] Generate exception records ,in, is the multiple threshold, for The standard deviation of Classified as abnormal;

[0328] S423: For the Keep records and establish abnormal test identification records , using the formula:

[0329]

[0330] in, is the total number of time points within the evaluation time range, It's time point The weight of .

[0331] Assume the following data:

[0332] , directly affects Contribution

[0333] , quadratic adjustment factor, enhanced The nonlinear effect of

[0334] , a logarithmic adjustment factor for smoothing Extreme values ​​of

[0335] The amount of data change at key time points, such as =1.

[0336] calculate:

[0337] Assumptions :

[0338]

[0339]

[0340]

[0341]

[0342]

[0343] Assume the following data:

[0344] , used to determine Whether it is abnormal, set to 2;

[0345] Assume it is 0.05.

[0346] calculate:

[0347] According to the above

[0348] ;

[0349] Not greater than 0.1;

[0350] (because did not exceed the set threshold and was therefore not marked as an anomaly).

[0351] Assume the following data:

[0352] Assume it is 10;

[0353] Assume that each =1.

[0354] calculate:

[0355] If only two of the 10 time points Marked as abnormal:

[0356] (other =0);

[0357] .

[0358] Resuscitator dead space test system, the system includes:

[0359] The flow-pressure ratio module calculates the flow-pressure ratio at each time point based on the readings of the airflow velocity sensor and the pressure sensor. Based on the flow-pressure ratio data, the performance of the resuscitator under multiple operating conditions is evaluated to obtain a preliminary stability diagnosis record.

[0360] The key variable analysis module uses a gradient boosting tree based on the preliminary stability diagnosis records to classify the operating status, identify the key operating variables that affect the performance, use the key operating variables to determine the key factors that affect the change in dead space capacity, and establish key factor identification information;

[0361] The concentration prediction and model optimization module uses key factor identification information to perform time series analysis on O2 concentration sensor data, uses moving average and exponential smoothing methods to predict O2 concentration information, and adjusts model parameters based on prediction results and historical data to obtain an optimized prediction model;

[0362] The dead space monitoring module uses an optimized prediction model to continuously monitor the dead space volume of the resuscitator, record data at key time points, analyze the data and identify performance deviations, and obtain preliminary abnormal test records;

[0363] The anomaly analysis module performs iterative analysis of performance deviations based on preliminary anomaly test records, identifies and verifies potential resuscitator anomalies by comparing historical data, and generates anomaly test identification records.

[0364] The above are only preferred embodiments of the present invention and are not intended to limit the present invention in other forms. Any technician familiar with the profession may use the technical contents disclosed above to change or modify them into equivalent embodiments with equivalent changes and apply them to other fields. However, any simple modification, equivalent change and modification made to the above embodiments based on the technical essence of the present invention without departing from the technical solution of the present invention still falls within the protection scope of the technical solution of the present invention.

Claims

1. A method for testing dead space of a resuscitator, characterized in that: The following steps are involved: Based on the readings of the airflow velocity sensor and the pressure sensor, the flow-pressure ratio at each time point is calculated, and the baseline data of the resuscitator under different operating conditions is determined by the ratio, and the stability of the resuscitator performance is evaluated to obtain preliminary stability diagnosis results; Based on the preliminary stability diagnosis results, using a gradient boosting tree, classifying the difference operating states, identifying key operating variables that affect performance, determining key factors that affect the change in dead space capacity, and obtaining key factor identification information; Based on the key factor identification information, the data of the O2 concentration sensor is analyzed in time series, and the change of O2 concentration in the future state is predicted by moving average and exponential smoothing method, and the model parameters are adjusted according to the prediction results and historical data to obtain an optimized prediction model; Based on the optimized prediction model, the dead space volume is monitored, key time point data is recorded, the key time point data is analyzed, performance deviations and potential resuscitator anomalies are identified, and an abnormal test identification record is obtained.

2. The resuscitator dead space testing method according to claim 1, characterized in that: The calculation steps of the flow pressure ratio are specifically as follows: Based on the air velocity sensor readings and the pressure sensor readings, combined with the effects of ambient temperature and humidity, the formula is used: Calculate each time point Flow pressure ratio ,in, Representing time point The air flow velocity, Representing time point The air flow pressure, Representing time point The ambient temperature, Representing time point The ambient humidity, is the temperature adjustment factor, is the adjustment factor for humidity; Using the collected flow pressure ratio , combined with the real-time calibration coefficient, using the formula: Adjust the flow-pressure ratio at each time point to obtain the calibrated flow-pressure ratio ,in, represents the calibration factor; Calculate the standard deviation of the calibrated flow-pressure ratio , the data consistency is evaluated using the formula: in, represents the sample size, yes The average value of .

3. The resuscitator dead space testing method according to claim 1, characterized in that: The steps for evaluating the performance stability of the resuscitator are specifically as follows: Using the flow pressure ratio and baseline data, the formula is used: Calculate the difference percentage ,in, is the calibrated flow pressure ratio, is the baseline flow-pressure ratio, To optimize the numerical stability of a small constant, To avoid small positive values ​​with zero denominator; contrast and stability threshold, using logical judgment: if < stability threshold, the resuscitator performance is stable. ≥ the stability threshold, the performance is unstable, and a preliminary stability diagnosis result is obtained based on the judgment result.

4. The resuscitator dead space testing method according to claim 1, characterized in that: The identification steps of the key operating variables are specifically as follows: Based on the preliminary stability diagnosis results, a gradient boosting tree model is applied to classify the differential operation states using the formula: in, Represents the classification result, is the splitting function in the decision tree, is an auxiliary analysis function, For the operating state, is the operating variable, For the The weight of a tree, represents the number of trees, and are weight adjustment coefficients, which are used to adjust and Contribution Based on the obtained classification results, the criticality score and stability index of each variable in the classification results are calculated using the formula: Generate key operating variables that affect performance , and is the weight coefficient, Function is used to calculate the given classification result Lower variable The key score, Functions are used to calculate Stability score; Identify key operational variables For analysis, the formula is: Calculate weighted correlations between key operating variables and performance data ,in, is the correlation function used to calculate the key operating variables and performance data points The relationship between is a performance data point, It is The weight of the data point, is the total number of data points.

5. The dead space testing method of a resuscitator according to claim 1, characterized in that: The steps for obtaining the key factor identification information are specifically as follows: The key operating variables were statistically analyzed using the formula: in, is the influencing factor for initial identification, is the key operating variable, As the influencing factors, is the weight of the influencing factor, and Represents statistical analysis functions and auxiliary analysis functions, respectively, used to measure and The relationship between To adjust the coefficient, adjust the influence of the auxiliary analysis function in the overall analysis, is the total number of influencing factors; right To conduct factor analysis, the formula is: in, For the refined influencing factors, is the factor analysis function, It is The weight of the factor, is the number of factors; Based on the refined influencing factors , identify the key factors and apply the formula: in, Identify information for key factors, represents the weighting coefficient, is the number of weighting factors.

6. The resuscitator dead space testing method according to claim 1, characterized in that: The steps for obtaining the optimized prediction model are specifically as follows: Based on the key factor identification information, the data of the O2 concentration sensor is analyzed in time series, and the moving average method is applied, using the formula: Calculation time The weighted moving average , generate moving average data, where is the time window size, is the weight coefficient, For in time O2 concentration; Use exponential smoothing to iteratively smooth the data using the formula: Calculate the time Exponential smoothing results for , generates exponential smoothing data, where Representative time O2 concentration, represents the smoothing coefficient, represents the adjustment item, It's time Exponential smoothing value of ; The changes in O2 concentration in the future are predicted based on the moving average and exponential smoothing results, and the model parameters are adjusted in combination with historical data. The formula is: Get the optimized prediction model ,in, and are the moving average and exponential smoothing results respectively, is the adjustment factor, Adjust for bias.

7. The resuscitator dead space testing method according to claim 1, characterized in that: The steps for recording the key time point data are specifically as follows: Obtain O2 concentration data from the optimized prediction model, perform monitoring activities, and regularly record O2 concentrations using the formula: in, For time The calculated O2 concentration value, is the prediction model function, They are model parameters, which are used to adjust and match environmental changes and sensor properties; The O2 concentration values ​​obtained Perform time series analysis to identify abnormal fluctuations or key changes using the formula: Get smoothed data ,in, For time The weighted nonlinear average O2 concentration, is the time window size, represents the time weight factor, is the data conversion parameter; Based on smoothed data , analyze key time points, identify key change points in the data, and use the formula: in, Represents the rate of change of data at key time points, and represent the sensitivity and exponential conversion parameters respectively.

8. The dead space testing method of a resuscitator according to claim 1, characterized in that: The steps for obtaining the abnormal test identification record are specifically as follows: Using the key time point data, evaluate whether the data point indicates a performance deviation using the formula: Generate preliminary anomaly identification results ,in, is the data change rate at the key time point, , and It is used to adjust the sensitivity and response curve of the model; Based on the preliminary abnormality identification results , analyze potential resuscitator anomalies using the formula: Generate exception records ,in, is the multiple threshold, for The standard deviation of Classified as abnormal; For those marked as abnormal Keep records and establish abnormal test identification records , using the formula: in, is the total number of time points within the evaluation time range, It's time point The weight of .

9. A resuscitator dead space test system, characterized in that: According to any one of claims 1 to 8, the resuscitator dead space testing method is implemented, and the system comprises: The flow-pressure ratio module calculates the flow-pressure ratio at each time point based on the readings of the airflow velocity sensor and the pressure sensor, and evaluates the performance of the resuscitator under multiple operating states according to the flow-pressure ratio to obtain a preliminary stability diagnosis record; The key variable analysis module classifies the operation status based on the preliminary stability diagnosis record using a gradient boosting tree, identifies key operation variables that affect performance, uses the key operation variables to determine key factors that affect the change in dead space capacity, and establishes key factor identification information; The concentration prediction and model optimization module uses the key factor identification information to perform time series analysis on the O2 concentration sensor data, uses moving average and exponential smoothing methods to predict the O2 concentration information, and adjusts the model parameters according to the prediction results and historical data to obtain an optimized prediction model; The dead space monitoring module uses the optimized prediction model to continuously monitor the dead space capacity of the resuscitator, record key time point data, analyze the data and identify performance deviations, and obtain preliminary abnormal test records; The abnormality analysis module performs iterative analysis of performance deviations based on the preliminary abnormality test record, identifies and verifies potential resuscitator abnormalities by comparing historical data, and generates an abnormality test identification record.

Images