Method and device for evaluating recommended low-oxygen gas supply flow

Abstract

The invention discloses a method and a device for evaluating recommended supply flow of hypoxia gas, which are mainly used for supplying hypoxia air with proper concentration for a user to carry out hypoxia training. The method comprises the steps that the heart rate and the respiratory rate of a to-be-evaluated user are obtained; evaluating the exercise intensity condition of the user according to the heart rate and the respiratory rate; the heart rate variability rate and the blood oxygen concentration of the to-be-evaluated user are obtained; evaluating the adaptation degree of the user to the current altitude according to the heart rate variability and the blood oxygen concentration; and the low-oxygen gas supply flow is recommended according to the comprehensive adaptation degree condition. The low-oxygen gas is generated by recycling the exhaled gas, so that the cost is reduced, the weight is reduced, an intelligent decision-making and recommendation system is introduced, and training mode recommendation can be performed according to the personal condition of a user.

Description

Technical Field

[0001] The present invention relates to the technical field of hypoxia training, and in particular to a method and a device for evaluating a recommended hypoxic gas supply flow rate. Background Art

[0002] Residents from plain areas often experience a series of altitude sickness when they first enter the plateau, and the incidence of altitude sickness reaches 53.62% within a week. Hypoxia pre-adaptation training in plain areas is an effective means to prevent high-altitude diseases. In recent years, hypoxia generators have begun to be tried for hypoxia pre-adaptation training. However, existing hypoxia generators are expensive, bulky, and cannot recommend training modes based on the user's personal situation. Summary of the invention

[0003] The purpose of the present invention is to provide a method and device for evaluating the recommended air supply flow of hypoxic gas, which can reduce costs and weight by recycling and reusing exhaled gas to produce hypoxic gas, and introduce an intelligent decision-making and recommendation system to recommend training modes based on the user's personal situation.

[0004] To achieve the above object, the present invention provides the following technical solutions: In a first aspect, the present invention discloses a method for evaluating a recommended hypoxic gas supply flow rate, comprising the following steps: S1. Acquire the physiological parameters of the user in different motion states in the plain area, including heart rate HR, respiratory rate BR, heart rate variability HRv and blood oxygen saturation According to the heart rate HR and respiratory rate BR, the GWO-SVM classifier is trained to obtain the exercise intensity classification; according to the heart rate variability HRv and blood oxygen saturation Calculate the reference value of fitness level under different exercise intensity classifications and ;in, represents the mean minute heart rate variability under state i, Represents the mean minute blood oxygen drop in state i; S2, collecting the physiological parameters of the user in different exercise states in the plateau area, inputting the heart rate and breathing rate into the trained GWO-SVM classifier, and outputting the exercise intensity classification; S3. Calculate the user's average heart rate variability within one minute based on heart rate variability and blood sample saturation. and blood oxygen concentration drop And compare it with the corresponding adaptation degree reference value, and the root difference value determines the user's adaptation degree to the current motion state; S4. Recommend hypoxic gas supply flow rate according to the corresponding adaptation level.

[0005] Further solution: The training process of GWO-SVM classifier is: Get training data , all HR and BR are scaled to the interval [0,1]; Select RBF kernel function ,in is the kernel parameter; Set the initial penalty parameter of the SVR model and kernel parameters The initial value of Initialize the wolf pack: Randomly initialize the position of the wolf pack ; Defining wolf pack hierarchy: Determining the alpha wolf based on the pack's position , suboptimal wolf , the third best wolf ; Simulate hunting behavior: Update formula: ,in Ci , α Indicates The distance between a wolf and the leader.

[0006] Location Updates:

[0007] Update the parameters of the SVR model based on the location of the gray wolf and calculate the new potential solution; The algorithm terminates when the maximum number of iterations is reached or the optimal solution reaches the preset threshold; Use the optimized penalty parameter and kernel parameters Train the GWO-SVR model.

[0008] Further solution: Determine the user's adaptability to the current motion state as follows: Record heart rate variability HRv and blood oxygen saturation for multiple samples within 1 minute , calculate the mean heart rate variability and blood oxygen saturation decrease ; According to the exercise intensity classification determined in step S2, find the corresponding and , the fitness difference is calculated according to the following formula: , where α is the coefficient; like , then the adaptation level is recorded as Lv.1 fully adapted; if , then the adaptation level is recorded as Lv.2 basic adaptation; if , then the adaptation level is recorded as Lv.3, completely unadaptable.

[0009] Further solution: Step S4 is specifically: Record continuously for at least ten minutes ,calculate The proportion of each state in If the adaptation level of Lv.3 (completely unadapted) accounts for more than 50% of the total exercise time, it is recommended to increase the current air supply concentration, that is, lower the simulated altitude; if the adaptation level of Lv.2 (basic adaptation) and the adaptation level of Lv.1 (complete adaptation) account for more than 50%, it is recommended to maintain the current air supply concentration, that is, maintain the simulated altitude; otherwise, it is recommended to lower the current air supply concentration, that is, increase the simulated altitude.

[0010] Further solution: Exercise intensity classification includes resting, low-intensity exercise, moderate-intensity exercise, and high-intensity exercise.

[0011] Further plan: resting state is sitting or standing, METs of low-intensity exercise are less than 3.0, METs of moderate-intensity exercise are 3.0~7.0, and METs of high-intensity exercise are greater than 7.0.

[0012] Further solution: In the formula, the value of α is 0.5.

[0013] In a second aspect, the present invention discloses a computer-readable storage medium, on which are stored a number of acquisition classification programs, which are used to be called by a processor and execute the method for evaluating the recommended gas supply flow rate of hypoxic gas as described in any one of claims 1-7.

[0014] In a third aspect, the present invention discloses a hypoxic gas supply device for the above-mentioned method for evaluating the recommended hypoxic gas supply flow rate, comprising a breathing mask, the breathing mask is connected to a two-way three-way valve, a first exhaust port of the two-way three-way valve is sequentially connected to an exhaled gas processing module, a first gas component analysis module, a mixed gas chamber, and a second gas component analysis module, and the gas component analysis module is connected to the second exhaust port of the two-way three-way valve; The mixed gas chamber is also connected to a micro electric air pump, and the micro electric air pump is electrically connected to the control module, the intelligent recommendation module, and the physiological parameter monitoring module.

[0015] Compared with the prior art, the present invention has the following beneficial effects: The present invention generates hypoxic gas by recycling exhaled gas, thereby reducing costs and weight, and introduces an intelligent decision-making and recommendation system, which can recommend training modes according to the user's personal situation. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] Figure 1 It is a schematic diagram of the structure of the device in the present invention; Figure 2 A flowchart of determining the degree of adaptation of a user to a current motion state in the present invention; Figure 3 It is the algorithm flow chart of the GWO-SVR classifier in the present invention; Figure 4 Schematic diagram of the SVR algorithm in the present invention; In the figure: 1-breathing mask, 2-pressure sensor, 3-two-way three-way valve, 4-exhaled gas processing module, 5-first gas component analysis module, 6-physiological parameter monitoring module, 7-intelligent recommendation module, 8-control module, 9-micro electric air pump, 10-mixing gas chamber, 11-second gas component analysis module. DETAILED DESCRIPTION

[0017] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0018] In the description of the present invention, it should be noted that the directions or positional relationships indicated by the terms "upper", "lower", "left", "right", "inside", "outside", etc. are based on the directions or positional relationships shown in the accompanying drawings, or are the directions or positional relationships in which the product of the invention is usually placed when in use. They are only for the convenience of describing the present invention and simplifying the description, and do not indicate or imply that the device or element referred to must have a specific direction, be constructed and operated in a specific direction, and therefore cannot be understood as a limitation on the present invention.

[0019] See also Figure 1-3 In this embodiment, a method for evaluating the recommended supply flow rate of hypoxic gas includes the following steps: S0. Obtain the physiological parameters of the user to be evaluated when performing 4 different intensities of exercise in a plain area, train an artificial neural network classifier for exercise intensity classification based on heart rate and respiratory rate, and calculate the reference value of the degree of adaptation under different exercise states based on heart rate variability and blood oxygen saturation. Specifically, the following steps are included: S01. Collect physiological parameters of the user to be evaluated during 4 different intensity exercises in plain areas, each lasting 10 minutes; the exercise states include resting (Con1. Resting), 5km / h walking (Con2. Low-intensity exercise), 7km / h slow running (Con3. Medium-intensity exercise), and 10km / h fast running (Con4. High-intensity exercise). The physiological parameters include heart rate HR, respiratory rate BR, heart rate variability HRv, and blood oxygen saturation SpO2.

[0020] S02. Based on the heart rate HR and breathing rate BR, the support vector machine classification model (GWO-SVR) of the improved Grey Wolf algorithm is pre-trained to classify the exercise state into four types: Con1. Rest, Con2. Low-intensity exercise, Con3. Medium-intensity exercise, and Con4. High-intensity exercise, which are used for real-time exercise state classification in subsequent use. Specifically, the algorithm training process is: S021. Obtain training data , all HR and BR are scaled to the interval [0,1].

[0021] S022. Select RBF kernel function ,in is the kernel parameter.

[0022] S023. Set the initial penalty parameters of the SVR model and kernel parameters The initial value of .

[0023] S024. Initialize the wolf pack: Randomly initialize the position of the wolf pack .

[0024] S025. Define wolf pack levels: Determine the leader based on the pack's position , suboptimal wolf , the third best wolf .

[0025] S026. Simulate hunting behavior: Update formula: ,in Ci , α Indicates The distance between a wolf and the leader.

[0026] Location Updates:

[0027] S027. Update the parameters of the SVR model according to the position of the gray wolf and calculate the new potential solution.

[0028] S028. When the maximum number of iterations is reached or the optimal solution reaches a preset threshold, the algorithm terminates.

[0029] S029. Use optimized penalty parameters and kernel parameters Train the GWO-SVR model.

[0030] S03. Calculate the reference value of fitness under different exercise states based on heart rate variability HRv and blood oxygen saturation SpO2: and .in represents the mean minute heart rate variability under state i, represents the mean minute blood oxygen drop in state i. S1. Obtain the heart rate and respiratory rate of the user to be evaluated, and evaluate the user's exercise intensity according to the heart rate and respiratory rate. Specifically, the following steps are included: S11. Preprocess the heart rate HR and respiratory rate BR and scale them to the range of [0,1].

[0031] S12. Send the preprocessed data to the trained artificial neural network and output the classification results, which include Con1. Rest, Con2. Low-intensity exercise, Con3. Medium-intensity exercise, and Con4. High-intensity exercise.

[0032] S13, record the classification results and proceed to the next step of processing.

[0033] S2, obtaining the heart rate variability and blood oxygen concentration of the user to be evaluated, and evaluating the user's adaptation to the current altitude according to the heart rate variability and blood oxygen concentration. Specifically, the following steps are included: S21, record the heart rate variability HRv and blood oxygen saturation of multiple samples within 1 minute , calculate the mean heart rate variability and blood oxygen saturation decrease .

[0034] S22, according to the motion state determined in step S1, find the corresponding and .according to Calculate the fitness difference, where is the coefficient, which is usually 0.5.

[0035] S23. According to Determine the user's adaptability: If , then the record status is Lv.1 fully adapted; if , then the record status is Lv.2 basic adaptation; if , the recorded status is Lv.3: Totally unsuitable.

[0036] S3, recommending the hypoxic gas supply flow rate according to the comprehensive adaptability during the training process. Specifically, the following steps are included: S31. Record at least ten minutes ,calculate The proportion of each state in S32. If the proportion of Lv3. Completely unadapted exceeds 50% of the total time, it is recommended to increase the current air supply concentration, that is, to lower the simulated altitude; if the proportion of Lv.2 basic adaptation and Lv1. Complete adaptation exceeds 50%, it is recommended to maintain the current air supply concentration, that is, to maintain the simulated altitude; in other cases, it is recommended to lower the current air supply concentration, that is, to increase the simulated altitude.

[0037] Support vector regression (SVR) is a regression algorithm based on support vector machines. Compared with other traditional regression methods, SVR has more advantages in training higher-dimensional and nonlinear data and can achieve higher regression accuracy.

[0038] After the collected data is preprocessed, an SVR training data set consisting of n groups of data will be obtained. ,in For the By The input is composed of dimensional input vector, Estimate the oxygen consumption for the corresponding one-dimensional output, so, for indivual The input vector is composed of dimensional input array, while for The output energy consumption value is composed of dimensional output vector.

[0039] like Figure 4 As shown in the figure, each circle represents a set of data in the data set. The purpose of SVR is to obtain the optimal hyperplane by solving , so that the distance between each data and the hyperplane is less than the set value ,in and are weight vector and bias respectively. For energy consumption modeling, we essentially want to find a mapping , so that and As equal as possible, so for this regression model we have:

[0040] In the formula is a nonlinear function, which can make Mapped to a higher dimensional feature space, and Into a linear relationship.

[0041] In traditional regression models, the model output is usually The actual output The difference between the two is used to calculate the accuracy loss. In the SVR model, a special loss function is introduced. This loss function is less than is not sensitive to errors, that is and The difference between , the prediction is also considered correct and the loss is not included. This method can improve the robustness of the model. For this reason, the regression optimization problem can be written as:

[0042] The optimization problem described in formula (2) is based on the assumption that all data in the data set are in The center is centered, and the width of each side is However, in reality, some data are abnormal, and some points should be allowed to deviate to a certain extent. Therefore, the penalty parameter is introduced in formula (2) and slack variables and The slack variable refers to the positive or negative deviation outside the interval band, and the penalty parameter can control the weight of the deviation. At this point, the regression optimization problem can be written as:

[0043] In order to link the constraints to the objective function, a non-negative La-Grange multiplier is introduced: , , , .right , , and By finding the partial derivatives and setting them equal to 0, it can be transformed into the dual problem of SVR. According to the KKT conditions, the solution of SVR is as follows:

[0044] In formula (4) is called the kernel function. The actual calculation may be very complicated, so in SVR, it can be replaced by selecting a suitable kernel function. This paper adopts the RBF kernel function, and the expression is shown in formula (5):

[0045] In formula (5) is the kernel parameter.

[0046] When determining the penalty parameter and kernel parameters After finding the optimal value of and output variables The SVR-based exercise oxygen consumption model can be obtained by training with a known exercise oxygen consumption data set. ,When applying this model to predict energy consumption data, we only need to substitute the corresponding ,energy consumption related factors into the model as input variables to obtain the ,corresponding estimated oxygen consumption.

[0047] A device for evaluating a recommended hypoxic gas supply flow rate, comprising: Respiratory mask 1; A two-way three-way valve 3, a first end of which is connected to the breathing mask 1; An exhaled gas processing module 4, the air inlet of which is connected to the second end of the two-way three-way valve 3, and the third connection end of the two-way three-way valve 3 is connected to the exhaled gas processing module 4, the gas component analysis module 5, and the mixed gas chamber 10 in sequence, collects the exhaled gas in the breathing mask 1, processes and analyzes the exhaled gas, and then delivers it to the mixed gas chamber 10.

[0048] The micro electric air pump 9 is used to deliver air to the mixing chamber 10 ; the exhaled gas treated in the mixing chamber 10 is mixed with the air and outputs hypoxic gas to the breathing mask 1 .

[0049] The first gas component analysis module 5 and the second gas component analysis module 11 are used to monitor the oxygen content and carbon dioxide content of the flowing gas; Physiological parameter monitoring module 6, used to detect the user's heart rate, heart rate variability, respiratory rate, and blood oxygen saturation; The intelligent recommendation module 7 receives the data transmitted by the physiological parameter monitoring module 6, evaluates the training effect of the user, and recommends the training gas supply concentration to the user according to the evaluation effect.

[0050] And a control module 8 is used to control the working speed of the micro electric air pump 9.

[0051] Furthermore, it also includes a pressure sensor 2 for detecting the internal air pressure of the breathing mask 1, and the pressure sensor 2 is electrically connected to the control single module 8. When the pressure sensor 2 detects a positive pressure, the control single module 8 controls the second end (the first exhaust end) of the two-way three-way valve 3 to open and the third end (the second exhaust end) to close; when the pressure sensor 2 detects a negative pressure, the control single module 8 controls the second end of the two-way three-way valve 3 to close and the third end to open.

[0052] Furthermore, the exhaled gas processing module 4 has a built-in filter composed of a CO2 adsorbent and a desiccant, and the filter is surrounded by a refrigerant for removing CO2 and moisture from the exhaled gas and reducing the temperature of the exhaled gas.

[0053] Furthermore, the second end of the two-way three-way valve 3 and the exhaled gas processing module 4 are connected by a breathing air hose and a quick connector. The quick connector is self-locking, ensuring the stability of the connection and the continuity of the gas flow during system operation, while facilitating quick disconnection when replacing the filter and refrigerant.

[0054] Furthermore, the first gas component analysis module 5 and the second gas component analysis module 11 both include an oxygen sensor and a carbon dioxide sensor for detecting the oxygen concentration and the carbon dioxide concentration of the gas flowing therethrough and feeding back to the control module 8 .

[0055] Furthermore, the exhaled gas processing module 4 and the mixed gas chamber 10 are connected by a breathing air hose.

[0056] Although this specification is described according to implementation modes, not every implementation mode includes only one independent technical solution. This description of the specification is only for the sake of clarity. Those skilled in the art should regard the specification as a whole. The technical solutions in each embodiment may also be appropriately combined to form other implementation modes that can be understood by those skilled in the art.

[0057] Therefore, the above description is only a preferred embodiment of the present application and is not intended to limit the scope of implementation of the present application; that is, all equivalent changes made according to the scope of the claims of the present application are within the protection scope of the claims of the present application.

Claims

1. A method for evaluating the recommended supply flow rate of hypoxic gas, characterized in that: The following steps are involved: S1. Acquire the physiological parameters of the user in different motion states in the plain area, including heart rate HR, respiratory rate BR, heart rate variability HRv and blood oxygen saturation According to the heart rate HR and respiratory rate BR, the GWO-SVM classifier is trained to obtain the exercise intensity classification; according to the heart rate variability HRv and blood oxygen saturation Calculate the reference value of fitness level under different exercise intensity classifications and ;in, represents the mean minute heart rate variability under state i, Represents the mean minute blood oxygen drop in state i; S2, collecting the physiological parameters of the user in different exercise states in the plateau area, inputting the heart rate and breathing rate into the trained GWO-SVM classifier, and outputting the exercise intensity classification; S3. Calculate the user's average heart rate variability within one minute based on heart rate variability and blood sample saturation. and blood oxygen concentration drop And compare it with the corresponding adaptation degree reference value, and the root difference value determines the user's adaptation degree to the current motion state; S4. Recommend hypoxic gas supply flow rate according to the corresponding adaptation level.

2. The method for evaluating the recommended hypoxic gas supply flow rate according to claim 1, characterized in that: The training process of the GWO-SVM classifier is: Get training data , all HR and BR are scaled to the interval [0,1]; Select RBF kernel function ,in is the kernel parameter; Set the initial penalty parameter of the SVR model and kernel parameters The initial value of Initialize the wolf pack: Randomly initialize the position of the wolf pack ; Defining wolf pack hierarchy: Determining the alpha wolf based on the pack's position , suboptimal wolf , the third best wolf ; Simulate hunting behavior: Update formula: ,in Ci , α Indicates The distance between a wolf and the leader. Location Updates: Update the parameters of the SVR model based on the location of the gray wolf and calculate the new potential solution; The algorithm terminates when the maximum number of iterations is reached or the optimal solution reaches the preset threshold; Use the optimized penalty parameter and kernel parameters Train the GWO-SVR model.

3. The method for evaluating the recommended hypoxic gas supply flow rate according to claim 1, characterized in that: The degree of adaptability of the user to the current motion state is determined as follows: Record heart rate variability HRv and blood oxygen saturation for multiple samples within 1 minute , calculate the mean heart rate variability and blood oxygen saturation decrease ; According to the exercise intensity classification determined in step S2, find the corresponding and , the fitness difference is calculated according to the following formula: , where α is the coefficient; like , then the adaptation level is recorded as Lv.1 fully adapted; if , then the adaptation level is recorded as Lv.2 basic adaptation; if , then the adaptation level is recorded as Lv.3, completely unadaptable.

4. The method for evaluating the recommended hypoxic gas supply flow rate according to claim 3, characterized in that: Step S4 is specifically as follows: Record continuously for at least ten minutes ,calculate The proportion of each state in If the adaptation level of Lv.3 (completely unadapted) accounts for more than 50% of the total exercise time, it is recommended to increase the current air supply concentration, that is, lower the simulated altitude; if the adaptation level of Lv.2 (basic adaptation) and the adaptation level of Lv.1 (complete adaptation) account for more than 50%, it is recommended to maintain the current air supply concentration, that is, maintain the simulated altitude; otherwise, it is recommended to lower the current air supply concentration, that is, increase the simulated altitude.

5. The method for evaluating the recommended hypoxic gas supply flow rate according to claim 1, characterized in that: Exercise intensity classification includes resting, low-intensity exercise, moderate-intensity exercise, and high-intensity exercise.

6. The method for evaluating the recommended hypoxic gas supply flow rate according to claim 1, characterized in that: Resting state refers to sitting or standing, METs for low-intensity exercise are less than 3.0, METs for moderate-intensity exercise are 3.0~7.0, and METs for high-intensity exercise are greater than 7.

0.

7. The method for evaluating the recommended hypoxic gas supply flow rate according to claim 3, characterized in that: In the formula, the value of α is 0.

5.

8. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores a plurality of acquisition and classification programs, and the plurality of acquisition and classification programs are used to be called by a processor and execute the method for evaluating the recommended hypoxic gas supply flow rate as described in any one of claims 1-7.

9. A hypoxic gas supply device using the method for evaluating the recommended hypoxic gas supply flow rate according to any one of claims 1 to 7, characterized in that: The invention comprises a breathing mask (1), wherein the breathing mask (1) is connected to a two-way three-way valve (3), wherein a first exhaust port of the two-way three-way valve (3) is connected in sequence to an exhaled gas processing module (4), a first gas component analysis module (5), a mixed gas chamber (10), and a second gas component analysis module (11), and the gas component analysis module (11) is connected to the second exhaust port of the two-way three-way valve (3); The mixed gas chamber (10) is also connected to a micro electric air pump (9), and the micro electric air pump (9) is electrically connected to the control module (9), the intelligent recommendation module (7), and the physiological parameter monitoring module (6).

Images