Method for detecting the material level of an adsorption tower of a molecular sieve oxygen generator

Abstract

The application discloses a kind of molecular sieve oxygen generator's adsorption tower material level detection method, first in the compacting device in the adsorption tower of molecular sieve oxygen generator Set up several detection points, then obtain the pressure data of detection point by installing pressure detection element on detection point, and the relationship between pressure and adsorption tower material level depth is obtained by machine learning algorithm fitting, that is, the detection of adsorption tower material level depth can be indirectly realized by the pressure data of compacting device in the adsorption tower, and then feedback to host controller to realize early warning, alarm and shutdown control, detection method is simple and convenient and has higher accuracy.

Description

Technical Field

[0001] This invention belongs to the field of oxygen generation equipment testing technology, specifically relating to the design of a method for detecting the material level in the adsorption tower of a molecular sieve oxygen generator. Background Technology

[0002] Currently, molecular sieve oxygen generators are widely used in oxygen supply systems in hospitals or high-altitude areas, such as... Figure 1 As shown, it includes an air compressor unit, an adsorption tower, a storage tank, an oxygen compressor unit, and a nitrogen removal unit. The adsorption tower, as a key component of the molecular sieve oxygen generator, is connected to the air compressor unit, storage tank, and nitrogen removal unit. During the operation of the molecular sieve oxygen generator, the molecular sieves inside the adsorption tower gradually pulverize, causing a drop in the molecular sieve level within the tower. When the level drops and is not replenished in time, it exacerbates the pulverization of the molecular sieves, leading to a decrease in the overall oxygen concentration of the molecular sieve oxygen generator. Therefore, real-time monitoring of the adsorption tower level is crucial. However, because the adsorption tower is a relatively closed device, existing technologies lack a method for accurately monitoring the molecular sieve level in the adsorption tower in real time. Summary of the Invention

[0003] The purpose of this invention is to solve the problem that existing technologies cannot accurately detect the molecular sieve material level in the adsorption tower in real time, and to propose a method for detecting the material level in the adsorption tower of a molecular sieve oxygen generator.

[0004] The technical solution of this invention is: a method for detecting the material level in the adsorption tower of a molecular sieve oxygen generator, comprising the following steps:

[0005] S1. Set up several detection points on the compression device inside the adsorption tower of the molecular sieve oxygen generator.

[0006] S2. Install pressure detection elements at each detection point to detect the pressure at each detection point.

[0007] S3. Based on the pressure at each detection point, the adsorption tower material level depth corresponding to each detection point is obtained through machine learning algorithms.

[0008] S4. Send the adsorption tower material level depth corresponding to each detection point to the main controller of the molecular sieve oxygen generator for early warning, alarm or shutdown judgment.

[0009] S5. Determine the tilt and leakage of the adsorption tower based on the difference in material level depth between any two detection points.

[0010] Furthermore, the pressure sensing element in step S2 is a circular resistance strain gauge.

[0011] Furthermore, the formula for calculating the pressure at each detection point in step S2 is as follows:

[0012]

[0013] Where P represents the pressure at the detection point, U1 represents the input voltage of the circular resistance strain gauge, U2 represents the output voltage of the circular resistance strain gauge, h represents the thickness of the circular resistance strain gauge, E represents the Young's modulus of the circular resistance strain gauge material, K represents the longitudinal sensitivity coefficient of the circular resistance strain gauge, μ represents the Poisson's ratio of the circular resistance strain gauge material, R represents the radius of the circular resistance strain gauge, and l i This represents the distance from the i-th strain calculation point on the circular resistance strain gauge to the center of the circular resistance strain gauge, and n represents the number of strain calculation points.

[0014] Furthermore, step S3 includes the following sub-steps:

[0015] S31. Construct a fitting regression model of the adsorption tower material level depth H with respect to the pressure P at the detection point:

[0016]

[0017] Where ω represents the weight fitting parameter and b represents the bias fitting parameter.

[0018] S32. Set the initial value of the weight fitting parameter ω to ω0 and the initial value of the bias fitting parameter b to b0.

[0019] S33. Fit the weights of the current iteration to the parameters ω. j and bias fitting parameters b j Input the fitted regression model to obtain the predicted value H of the adsorption tower material level depth H corresponding to the pressure P at the detection point. y .

[0020] S34. Obtain the true value H of the adsorption tower material level depth corresponding to the same detection point pressure P. r Determine the predicted value H y and the true value H r If the difference is less than the preset threshold, proceed to step S35; otherwise, update the weight fitting parameter ω and the bias fitting parameter b, and return to step S33 to proceed to the next iteration.

[0021] S35. Fit the weights of the current iteration to the parameters ω. j and bias fitting parameters b j The final value is input into the fitted regression model, and the adsorption tower level depth corresponding to each detection point is obtained through the fitted regression model.

[0022] Furthermore, the formulas for updating the weight fitting parameter ω and the bias fitting parameter b in step S34 are as follows:

[0023]

[0024]

[0025] L = -[H r logH y +(1-H r log(1-H) y )]

[0026] Where ω j Let b represent the weight fitting parameters for the j-th iteration. j Let represent the bias fitting parameters for the j-th iteration, α represent the gradient descent parameters, and L represent the loss function.

[0027] Furthermore, step S4 includes the following sub-steps:

[0028] S41. Send the adsorption tower material level depth H corresponding to each detection point to the main controller of the molecular sieve oxygen generator. When the adsorption tower material level depth H corresponding to any detection point is lower than the preset first depth threshold H1, send a warning signal to the user periodically through the main controller.

[0029] S42. When the material level depth H of the adsorption tower corresponding to any detection point is lower than the preset second depth threshold H2, an alarm is periodically sent to the user through the host controller.

[0030] S43. When the material level depth H of the adsorption tower corresponding to any detection point is lower than the preset third depth threshold H3, the molecular sieve oxygen generator is stopped by the host controller.

[0031] Furthermore, H1>H2>H3. When H1>H>H2, the closer the adsorption tower level depth H is to the second depth threshold H2, the higher the frequency of sending early warning signals to users. When H2>H>H3, the closer the adsorption tower level depth H is to the third depth threshold H3, the higher the frequency of sending alarms to users.

[0032] Furthermore, step S5 includes the following sub-steps:

[0033] S51. When the difference in material level depth ΔH between any two detection points in the adsorption tower is greater than the preset material level depth difference threshold, it is determined that the clamping device of the adsorption tower is tilted, and an alarm is sent to the user through the host controller.

[0034] S52. When the tilt angle of the compression device of the adsorption tower is greater than the preset tilt angle threshold, it is determined that the adsorption tower is leaking, and the molecular sieve oxygen generator is stopped by controlling the main controller.

[0035] The beneficial effects of this invention are:

[0036] (1) This invention obtains pressure data at the detection point by installing a pressure detection element on the pressing device in the adsorption tower of the molecular sieve oxygen generator, and obtains the relationship between pressure and the material level depth in the adsorption tower by fitting the pressure through a machine learning algorithm. Thus, the material level depth in the adsorption tower can be indirectly detected by the pressure data of the pressing device in the adsorption tower, and then fed back to the host controller to realize early warning, alarm and shutdown control. The detection method is simple and convenient and has a high accuracy.

[0037] (2) The present invention can further determine the tilt and leakage of the adsorption tower based on the difference in material level depth of any two detection points, and has strong scalability and practicality.

[0038] (3) The present invention obtains the pressure at each detection point by means of a circular resistance strain gauge. The pressure value can be obtained indirectly by collecting the input voltage value and output voltage value of the circular resistance strain gauge. It is suitable for reading the pressure value of the detection point in a relatively closed device such as an adsorption tower. Attached Figure Description

[0039] Figure 1 The diagram shown is a structural block diagram of an existing molecular sieve oxygen generator.

[0040] Figure 2 The diagram shown is a flowchart of a method for detecting the material level in the adsorption tower of a molecular sieve oxygen generator according to an embodiment of the present invention. Detailed Implementation

[0041] Exemplary embodiments of the present invention will now be described in detail with reference to the accompanying drawings. It should be understood that the embodiments shown and described in the drawings are merely exemplary and are intended to illustrate the principles and spirit of the invention, and are not intended to limit the scope of the invention.

[0042] This invention provides a method for detecting the material level in the adsorption tower of a molecular sieve oxygen generator, such as... Figure 2 As shown, it includes the following steps S1 to S5:

[0043] S1. Set up several detection points on the compression device inside the adsorption tower of the molecular sieve oxygen generator.

[0044] In this embodiment of the invention, the clamping device inside the adsorption tower is a spring.

[0045] S2. Install pressure detection elements at each detection point to detect the pressure at each detection point.

[0046] In this embodiment of the invention, the pressure sensing element is a circular resistance strain gauge, and the formula for calculating the detected pressure is as follows:

[0047]

[0048] Where P represents the pressure at the detection point, U1 represents the input voltage of the circular resistance strain gauge, U2 represents the output voltage of the circular resistance strain gauge, h represents the thickness of the circular resistance strain gauge, E represents the Young's modulus of the circular resistance strain gauge material, K represents the longitudinal sensitivity coefficient of the circular resistance strain gauge, μ represents the Poisson's ratio of the circular resistance strain gauge material, R represents the radius of the circular resistance strain gauge, and l i This represents the distance from the i-th strain calculation point on the circular resistance strain gauge to the center of the circular resistance strain gauge, and n represents the number of strain calculation points.

[0049] S3. Based on the pressure at each detection point, the adsorption tower material level depth corresponding to each detection point is obtained through machine learning algorithms.

[0050] Step S3 includes the following sub-steps S31 to S35:

[0051] S31. Construct a fitting regression model of the adsorption tower material level depth H with respect to the pressure P at the detection point:

[0052]

[0053] Where ω represents the weight fitting parameter and b represents the bias fitting parameter.

[0054] S32. Set the initial value of the weight fitting parameter ω to ω0 and the initial value of the bias fitting parameter b to b0.

[0055] S33. Fit the weights of the current iteration to the parameters ω. j and bias fitting parameters b j Input the fitted regression model to obtain the predicted value H of the adsorption tower material level depth H corresponding to the pressure P at the detection point. y .

[0056] S34. Obtain the true value H of the adsorption tower material level depth corresponding to the same detection point pressure P. r Determine the predicted value H y and the true value H r If the difference is less than the preset threshold, proceed to step S35; otherwise, update the weight fitting parameter ω and the bias fitting parameter b, and return to step S33 to proceed to the next iteration.

[0057] In this embodiment of the invention, the formulas for updating the weight fitting parameter ω and the bias fitting parameter b are as follows:

[0058]

[0059]

[0060] L = -[H r logH y +(1-Hr log(1-H) y )]

[0061] Where ω j Let b represent the weight fitting parameters for the j-th iteration. j Let represent the bias fitting parameters for the j-th iteration, α represent the gradient descent parameters, and L represent the loss function.

[0062] S35. Fit the weights of the current iteration to the parameters ω. j and bias fitting parameters b j The final value is input into the fitted regression model, and the adsorption tower level depth corresponding to each detection point is obtained through the fitted regression model.

[0063] S4. Send the adsorption tower material level depth corresponding to each detection point to the main controller of the molecular sieve oxygen generator for early warning, alarm or shutdown judgment.

[0064] Step S4 includes the following sub-steps S41 to S45:

[0065] S41. Send the adsorption tower material level depth H corresponding to each detection point to the main controller of the molecular sieve oxygen generator. When the adsorption tower material level depth H corresponding to any detection point is lower than the preset first depth threshold H1, send a warning signal to the user periodically through the main controller.

[0066] In this embodiment of the invention, when H is lower than the preset first depth threshold H1, it indicates that the molecular sieve material level in the adsorption tower has decreased significantly, but it does not have a significant impact on the oxygen concentration of the entire molecular sieve oxygen generator. At this time, it is necessary to send a warning signal to the user through the host controller to remind the user to replenish the molecular sieve in time.

[0067] S42. When the material level depth H of the adsorption tower corresponding to any detection point is lower than the preset second depth threshold H2, an alarm is periodically sent to the user through the host controller.

[0068] In this embodiment of the invention, when H is lower than the preset second depth threshold H2, it indicates that the molecular sieve material level in the adsorption tower has decreased significantly and has affected the oxygen concentration of the molecular sieve oxygen generator. At this time, it is necessary to send an alarm to the user through the main controller to remind the user to take emergency measures to avoid the oxygen concentration of the molecular sieve oxygen generator from falling below the national standard (93% ± 3%).

[0069] S43. When the material level depth H of the adsorption tower corresponding to any detection point is lower than the preset third depth threshold H3, the molecular sieve oxygen generator is stopped by the host controller.

[0070] In this embodiment of the invention, when H is lower than the preset third depth threshold H3, it indicates that the molecular sieve material level in the adsorption tower has dropped below the critical value, causing the oxygen concentration of the molecular sieve oxygen generator to be lower than the national standard (93% ± 3%). At this time, it is necessary to control the molecular sieve oxygen generator to stop through the host controller.

[0071] Based on the above description, the depth thresholds in the embodiments of the present invention satisfy the following relationship: H1>H2>H3.

[0072] In this embodiment of the invention, when H1>H>H2, the closer the adsorption tower level depth H is to the second depth threshold H2 (i.e., the closer to the alarm), the higher the frequency of sending a warning signal to the user. When H2>H>H3, the closer the adsorption tower level depth H is to the third depth threshold H3 (i.e., the closer to shutdown), the higher the frequency of sending an alarm to the user.

[0073] S5. Determine the tilt and leakage of the adsorption tower based on the difference in material level depth between any two detection points.

[0074] Step S5 includes the following sub-steps S51 to S52:

[0075] S51. When the difference in material level depth ΔH between any two detection points in the adsorption tower is greater than the preset material level depth difference threshold, it is determined that the clamping device of the adsorption tower is tilted, and an alarm is sent to the user through the host controller.

[0076] S52. When the tilt angle of the compression device of the adsorption tower is greater than the preset tilt angle threshold, it is determined that the adsorption tower is leaking, and the molecular sieve oxygen generator is stopped by controlling the main controller.

[0077] Those skilled in the art will recognize that the embodiments described herein are intended to help the reader understand the principles of the invention, and should be understood that the scope of protection of the invention is not limited to such specific statements and embodiments. Those skilled in the art can make various other specific modifications and combinations based on the technical teachings disclosed in this invention without departing from the spirit of the invention, and these modifications and combinations are still within the scope of protection of this invention.

Claims

1. A method for detecting the material level in the adsorption tower of a molecular sieve oxygen generator, characterized in that, Includes the following steps: S1. Set up several detection points on the compression device inside the adsorption tower of the molecular sieve oxygen generator; S2. Install pressure detection elements at each detection point to detect the pressure at each detection point; S3. Based on the pressure at each detection point, the adsorption tower material level depth corresponding to each detection point is obtained through machine learning algorithms; S4. Send the material level depth of the adsorption tower corresponding to each detection point to the main controller of the molecular sieve oxygen generator for early warning, alarm or shutdown judgment; S5. Determine the tilt and leakage of the adsorption tower based on the difference in material level depth between any two detection points. The pressure detection element in step S2 is a circular resistance strain gauge. The formula for calculating the pressure at each detection point in step S2 is as follows: in Indicates the pressure at the detection point. This indicates the input voltage value of the circular resistance strain gauge. This indicates the output voltage value of the circular resistance strain gauge. This indicates the thickness of the circular resistance strain gauge. This represents the Young's modulus of the material used in circular resistance strain gauges. This represents the longitudinal sensitivity coefficient of a circular resistance strain gauge. The Poisson's ratio represents the material of a circular resistance strain gauge. This indicates the radius of the circular resistance strain gauge. Indicates the first on a circular resistance strain gauge The distance from each strain calculation point to the center of the circular resistance strain gauge This indicates the number of strain calculation points.

2. The method for detecting the material level in the adsorption tower of a molecular sieve oxygen generator according to claim 1, characterized in that, Step S3 includes the following sub-steps: S31, Constructing the adsorption tower material level depth H Regarding the pressure at the testing point P Fitting regression model: in Represents the weighted fitting parameters. b Indicates the bias fitting parameters; S32. Set weight fitting parameters initial value and bias fitting parameters b initial value ; S33. Fit parameters to the weights of the current iteration. and bias fitting parameters Input the fitted regression model to obtain the pressure at the detection point. P Corresponding adsorption tower material level depth H Predicted value ; S34. Obtain the pressure at the same detection point. P Corresponding adsorption tower material level depth H The true value Determine the predicted value and the true value If the difference is less than a preset threshold, proceed to step S35; otherwise, update the weight fitting parameters. and bias fitting parameters b Return to step S33 to proceed to the next iteration; S35. Fit parameters to the weights of the current iteration. and bias fitting parameters The final value is input into the fitted regression model, and the adsorption tower level depth corresponding to each detection point is obtained through the fitted regression model.

3. The method for detecting the material level in the adsorption tower of a molecular sieve oxygen generator according to claim 2, characterized in that, In step S34, the weight fitting parameters are updated. and bias fitting parameters b The formula is: in Indicates the first j The weighted fitting parameters for the next iteration. Indicates the first j The bias fitting parameters for the next iteration. Represents the gradient descent parameters. This represents the loss function.

4. The method for detecting the material level in the adsorption tower of a molecular sieve oxygen generator according to claim 1, characterized in that, Step S4 includes the following sub-steps: S41. Determine the adsorption tower material level depth corresponding to each detection point. H The data is sent to the main controller of the molecular sieve oxygen generator when the adsorption tower material level depth at any detection point is... H Below the preset first depth threshold At that time, the host controller periodically sends warning signals to the user; S42, When the adsorption tower material level depth corresponding to any detection point H Below the preset second depth threshold At that time, alarms are periodically sent to users through the host controller; S43, When the adsorption tower material level depth corresponding to any detection point H Below the preset third depth threshold At that time, the molecular sieve oxygen generator is stopped by controlling the main controller.

5. The method for detecting the material level in the adsorption tower of a molecular sieve oxygen generator according to claim 4, characterized in that, The ,when At that time, the material level depth in the adsorption tower H The closer to the second depth threshold The more frequently warning signals are sent to users, the more likely it is that the warning signal will be sent. At that time, the material level depth in the adsorption tower H The closer to the third depth threshold The more frequently alerts are sent to users, the better.

6. The method for detecting the material level in the adsorption tower of a molecular sieve oxygen generator according to claim 1, characterized in that, Step S5 includes the following sub-steps: S51, Difference in adsorption tower level depth between any two detection points When the difference in material level exceeds the preset threshold, it is determined that the clamping device of the adsorption tower is tilted, and an alarm is sent to the user through the main controller. S52. When the tilt angle of the compression device of the adsorption tower is greater than the preset tilt angle threshold, it is determined that the adsorption tower is leaking, and the molecular sieve oxygen generator is stopped by controlling the main controller.

Images