Respirator flow dynamic adjusting system based on airflow sensing

Abstract

The invention relates to the technical field of respirators, and discloses a respirator flow dynamic adjusting system based on airflow sensing, which is characterized in that after starting, a multi-mode sensing module acquires spatial flow velocity distribution and dynamic characteristics of respiratory airflow, and generates a composite data flow fused with flow field information; the composite data flow is subjected to nonlinear analysis to generate a state vector representing breathing stability, and the state vector is used for identifying fluid abnormal risks; the state vector generates an optimized execution instruction set by the self-adaptive control engine in combination with the energy state of thermoelectric conversion; the execution instruction set drives the piezoelectric valve and the shear response humidifying layer to cooperatively adjust the flow and the humidity, and outputs physiologically matched breathing gas; fusion collection of spatial flow velocity distribution and dynamic characteristics of respiratory airflow is achieved through the multi-mode sensing module, the limitation of traditional single-point measurement is broken through, and the sensing precision is remarkably improved; a breathing chaos index is constructed through nonlinear kinetic analysis, and early warning of the breathing rhythm stability and the fluid abnormal risk is achieved.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of respirators, in particular to a respiratory flow dynamic adjustment system based on airflow sensing. BACKGROUND

[0002] With the wide application of respiratory support technology in clinical first aid, chronic respiratory disease management and sleep disorder treatment, the flow adjustment performance of the respirator is directly related to the treatment effect and patient comfort. The existing respirators generally use thermal or differential pressure flow sensors to monitor the airflow in real time, and adjust the proportional valve or fan speed through feedback control to achieve basic adjustment of the output flow.

[0003] The existing respirator flow adjustment system generally relies on traditional thermal or differential pressure flow sensors for single-point measurement, and can only achieve passive response control of the airflow, with the problems of large response delay, asynchronization between man and machine, rigid humidification control, high energy consumption and lack of early abnormal warning capability. At the same time, the system uses fixed parameter control strategy, which is difficult to adapt to the dynamic changes of patient's respiratory rhythm, easy to cause airway discomfort, man-machine confrontation or condensate water risk, and the whole system depends on external power supply, which limits the portability and long-term use scene. In addition, the existing technology lacks comprehensive perception and intelligent control of airflow field spatial distribution, respiratory physiological stability and system self-learning ability, and cannot realize personalized, comfortable and intelligent respiratory support in a true sense. SUMMARY

[0004] In view of the above problems of the existing respiratory flow dynamic adjustment system based on airflow sensing, the present application is proposed.

[0005] Therefore, the purpose of the present application is to provide a respiratory flow dynamic adjustment system based on airflow sensing, which aims to: To solve the above technical problems, the present application provides the following technical scheme: after starting, the multi-modal perception module collects the spatial flow distribution and dynamic characteristics of the respiratory airflow, and generates a composite data stream fused with the flow field information; The composite data stream is analyzed by a nonlinear analysis to generate a state vector representing respiratory stability, and is used to identify fluid abnormality risk; The state vector is combined with the energy state of the thermoelectric conversion by the adaptive control engine to generate an optimized execution instruction set; The execution instruction set drives the piezoelectric valve and the shear response humidification layer to cooperatively adjust the flow and humidity, and outputs the physiological matching respiratory gas; The adjustment result is fed back for deviation evaluation and model updating, and triggers a safety response, completes the closed-loop regulation and control, and enters a low-power monitoring state, waiting for the next breath cycle to trigger the subsequent adjustment process.

[0006] As a preferred scheme of the respiratory flow dynamic adjustment system based on airflow sensing, the multi-modal sensing module comprises a 60GHz millimeter wave radar chip and a ring-shaped electrostatic induction electrode array, the millimeter wave radar is used to detect the Doppler shift caused by the movement of particles in the airflow to reconstruct the radial velocity distribution, and the electrostatic induction electrode is used to capture the time-varying charge signal generated by the airflow friction with the pipe wall, and the synchronous output realizes the fusion sensing of space and overall flow.

[0007] As a preferred scheme of the respiratory flow dynamic adjustment system based on airflow sensing, the non-linear analysis comprises phase space reconstruction on the airflow sequence, and calculation of the maximum Lyapunov index and sample entropy, and the respiratory chaos index RCI(t) is generated by weighting the normalized entropy change and Lyapunov offset, which is used to quantify the stability change trend of the respiratory rhythm.

[0008] As a preferred scheme of the respiratory flow dynamic adjustment system based on airflow sensing, the fluid abnormality risk comprises a turbulent flow starting position identified based on a Reynolds number, when the local Re>2300 and is located in the downstream area of the pipeline, it is determined as a potential airflow disturbance area, and is used as an input parameter for adjusting the control strategy.

[0009] As a preferred scheme of the respiratory flow dynamic adjustment system based on airflow sensing, the energy state of the thermoelectric conversion is provided by a flexible thermoelectric generator attached to the outer wall of the breathing pipeline, which generates electricity by using the temperature difference between the output gas and the environment, and supplies power to the sensing module and the edge computing unit through the MPPT circuit, realizing the self-powered operation of part of the functions.

[0010] As a preferred scheme of the respiratory flow dynamic adjustment system based on airflow sensing, the execution instruction set contains a flow rising slope, a valve opening time sequence and a humidification response parameter, when the state vector indicates that the respiration is unstable, the control engine increases the flow response speed by 15%-30% in advance by 0.1 seconds, and dynamically optimizes the piezoelectric valve control logic to suppress the development of turbulent flow.

[0011] As a preferred scheme of the respiratory flow dynamic adjustment system based on airflow sensing, the shear response humidification layer is composed of PVA / borax intelligent hydrogel, the permeability of which nonlinearly changes with the airflow shear stress, the water release rate is reduced at low flow, and the humidification capacity is automatically enhanced at high flow, realizing passive humidity matching without external control.

[0012] As a preferred scheme of the respiratory flow dynamic adjustment system based on airflow sensing, the deviation evaluation is achieved by calculating the weighted deviation index of the flow, humidity and pressure of the actual output airflow relative to the target value, triggering local data recording when the deviation continuously exceeds the limit, and uploading the model gradient to the cloud for federated learning to achieve continuous optimization of the control strategy.

[0013] As a preferred scheme of the respiratory flow dynamic adjustment system based on airflow sensing, the safety response includes determining that it is coughing or condensate accumulation when the static electricity sensing signal detects high-frequency abnormal pulses, and the system immediately starts the pulse purge program and reduces the peak flow by 10% to prevent aspiration or discomfort.

[0014] As a preferred scheme of the respiratory flow dynamic adjustment system based on airflow sensing, the low-power monitoring state only maintains the millimeter wave radar and static electricity sensing in a low-power detection mode, and the entire system is awakened as soon as airflow initiation characteristics are detected, re-entering the multi-modal perception and dynamic adjustment process to achieve a synergistic optimization of energy efficiency and response speed.

[0015] The present application has the following advantages: the multi-modal perception module is used to realize the fusion collection of the spatial flow velocity distribution and dynamic characteristics of respiratory airflow, breaking through the limitations of traditional single-point measurement and significantly improving the perception accuracy; the respiratory chaos index is constructed using nonlinear dynamics analysis to realize early warning of the stability of respiratory rhythm and the risk of fluid abnormalities, changing passive response to active prediction; the key modules are self-powered by using thermoelectric conversion technology, improving the energy efficiency and portability of the system; the piezoelectric valve and shear response humidification layer are driven by the adaptive control engine to work cooperatively, achieving precise and physiological matching output of flow and humidity; and the deviation feedback, federated learning and safety protection mechanism are introduced to form a closed-loop optimization and intelligent evolution capability. Overall, the problems of existing respirators such as response lag, asynchrony between man and machine, humidification rigidity, high energy consumption and insufficient intelligence are solved, and the comfort, safety and individualization level of respiratory support are significantly improved. BRIEF DESCRIPTION OF DRAWINGS

[0016] In order to more clearly illustrate the technical solutions of the embodiments of the present application, the drawings needed in the embodiment description will be briefly introduced. Obviously, the drawings in the following description are only some embodiments of the present application, and other drawings can be obtained by those skilled in the art without creative labor. Among them: Figure 1 The flow structure diagram of the present application. DETAILED DESCRIPTION

[0017] In order to make the above objectives, characteristics and advantages of the present application more apparent, a detailed description of the specific embodiments of the present application will be given below with reference to the accompanying drawings.

[0018] In the following description, numerous specific details are set forth in order to provide a thorough understanding of the present application. However, it will be apparent to one skilled in the art that the present application can be practiced without the specific details given herewith. In other instances, well-known methods have not been described in detail in order not to unnecessarily obscure aspects of the present application.

[0019] Secondly, the "one embodiment" or "embodiment" referred to herein means that a specific feature, structure or characteristic described can be included in at least one implementation of the present application. The "in one embodiment" appearing in different places in the specification does not mean the same embodiment, nor is it an embodiment that is independent of or mutually exclusive with other embodiments.

[0020] Thirdly, the present application is described in detail in conjunction with the schematic diagram. In the detailed description of the embodiments of the present application, the sectional view of the device structure is partially enlarged without the general proportion for the convenience of description, and the schematic diagram is only an example, which should not limit the scope of protection of the present application. In addition, the three-dimensional spatial dimensions of length, width and depth should be included in the actual manufacture.

[0021] Embodiments Reference Figure 1 For the embodiments of the present application, a respiratory flow dynamic adjustment system based on airflow sensing is provided, which comprises, After starting, the multi-modal sensing module collects the spatial flow velocity distribution and dynamic characteristics of the respiratory airflow, and generates a composite data stream that fuses flow field information; The composite data stream is analyzed by a nonlinear analysis to generate a state vector representing the stability of respiration, and is used to identify fluid abnormality risks; The state vector is combined with the energy state of thermoelectric conversion by the adaptive control engine to generate an optimized execution instruction set; The execution instruction set drives the piezoelectric valve and the shear response humidification layer to cooperatively adjust the flow and humidity, and outputs the physiological matching respiratory gas; The adjustment result is fed back for deviation evaluation and model updating, and triggers a safety response, completes the closed-loop regulation and control, and enters a low-power monitoring state, waiting for the next breathing cycle to trigger the subsequent adjustment process.

[0022] After the system is powered on, the multi-modal sensing module is first activated. The module integrates a 60GHz millimeter wave radar chip and a ring-shaped electrostatic induction electrode array. The millimeter wave radar emits high-frequency electromagnetic waves into the breathing pipeline. By detecting the Doppler shift caused by the movement of suspended particles in the airflow, the radial velocity distribution of the airflow is reconstructed in real time, and the spatial flow field information is obtained. At the same time, the electrostatic induction electrode captures the time-varying charge signal generated by the friction between the airflow and the pipe wall, reflecting the dynamic changes of the overall flow. The two signals are time-synchronized and data-fused to generate a composite data stream containing flow field structure characteristics and macro flow trends, providing high-dimensional input for subsequent intelligent analysis.

[0023] The composite data stream is sent to the edge computing unit for nonlinear dynamics analysis. The system first reconstructs the phase space of the airflow time series, determines the optimal time delay using the mutual information method, and constructs a high-dimensional state space trajectory with embedding dimension. On this basis, the maximum Lyapunov exponent is calculated to evaluate the chaotic characteristics of the system, and the sample entropy is calculated to quantify the complexity of the respiratory rhythm, and then the respiratory chaos index RCI(t) is generated by weighted generation, which is used to represent the change trend of respiratory stability. In addition, the local Reynolds number is calculated combined with the flow velocity distribution data provided by the millimeter wave radar to identify the starting position of the turbulent flow, and when Re>2300 and located in the downstream region of the pipeline, it is determined as a potential fluid disturbance risk area. Finally, the chaos index and flow field risk information are integrated into a multi-dimensional state vector as the core basis for control decision.

[0024] The state vector is input into the adaptive control engine to dynamically generate an optimized execution instruction set. The control engine determines whether there is a precursor to respiratory instability based on whether RCI(t) exceeds the individual baseline threshold. If there is, the flow rate rise slope is increased by 15%-30% in advance of 0.1 seconds to achieve feedforward compensation and reduce human-machine asynchronization. At the same time, the opening timing of the piezoelectric valve is optimized according to the identified turbulent flow position to suppress vortex development. In this process, the system dynamically adjusts the allocation of computing resources based on the energy state of the thermoelectric conversion module, i.e., the level of electrical energy generated by the flexible thermoelectric generator driven by the temperature difference of the pipeline, to ensure control accuracy while achieving energy efficiency optimization. Finally, a complete execution instruction set containing flow waveform, valve timing, and humidification parameters is generated.

[0025] The execution instruction set is issued to the execution layer to drive the piezoelectric ceramic micro valve to perform millisecond-level precise flow regulation and achieve stepless response in the range of 0-100 L / min; at the same time, the instruction synchronously activates the shearing response humidification layer, which is composed of PVA / borax intelligent hydrogel, and the micro network structure of the hydrogel changes adaptively with the airflow shear stress: at low flow rate, the network remains dense to reduce the water release rate and prevent excessive humidification; at high flow rate, the network expands and the permeability increases, automatically enhancing the humidification capacity; the whole process does not require external heating or pump control, and is completely driven by airflow power, realizing passive and on-demand humidity matching, and finally outputting respiratory gas with temperature, humidity and flow rate highly matched with the physiological needs of the patient.

[0026] After the adjustment is completed, the system samples the flow rate, humidity and pressure of the output airflow, compares them with the target set values, and calculates the weighted deviation index; if the deviation continues to exceed the limit, the local data recording mechanism is triggered, and the state vector and control parameters in this adjustment process are used to update the local model, and the model gradient is uploaded to the cloud for federated learning through a secure communication protocol, realizing continuous optimization of the global control strategy; in addition, the system continuously monitors the electrostatic induction signal, and once a high-frequency abnormal pulse (such as >100 Hz) is detected, it is determined to be a risk of choking or condensate accumulation, and the pulse purge program is immediately started and the peak flow rate is reduced by 10% to prevent aspiration or discomfort; after feedback and safety response, the system automatically enters a low-power listening state, maintaining only the millimeter wave radar and electrostatic sensing in a low-power detection mode, and monitoring the airflow starting characteristics in real time; once the next breathing action is detected, the whole system is awakened and the multi-modal sensing and dynamic adjustment process is re-entered, forming a complete closed-loop operation mechanism from sensing, decision-making, execution to learning evolution.

[0027] Specifically, the multi-modal sensing module includes a 60GHz millimeter wave radar chip and a ring-shaped electrostatic induction electrode array. The millimeter wave radar is used to detect the Doppler shift caused by the movement of particles in the airflow to reconstruct the radial velocity distribution, and the electrostatic induction electrode is used to capture the time-varying charge signal generated by the airflow friction with the pipe wall. The two synchronously output to realize the fusion sensing of space and overall flow.

[0028] Nonlinear analysis includes phase space reconstruction of airflow sequence and calculation of maximum Lyapunov index and sample entropy. The respiratory chaos index RCI(t) is generated by weighting the normalized entropy change and Lyapunov offset, which is used to quantify the stability change trend of respiratory rhythm.

[0029] Fluid anomaly risk includes the identification of turbulent flow initiation location based on Reynolds number. When the local Re is greater than 2300 and located in the downstream area of the pipeline, it is determined as a potential airflow disturbance zone and used as an input parameter for control strategy adjustment.

[0030] Further, the electrically converted energy state is provided by a flexible thermoelectric generator attached to the outer wall of the breathing circuit, which generates electricity using the temperature difference between the output gas and the environment, and provides power for the sensing module and edge computing unit through the MPPT circuit, realizing the self-powered operation of part of the functions.

[0031] Among them, the execution instruction set contains the flow rising slope, valve opening timing and humidification response parameters. When the state vector indicates that the breathing is unstable, the control engine increases the flow response speed by 15%-30% in advance 0.1 seconds, and dynamically optimizes the piezoelectric valve control logic to suppress the development of turbulence.

[0032] The shear-responsive humidification layer is composed of PVA / borax intelligent hydrogel, whose permeability changes nonlinearly with airflow shear stress, reducing the water release rate at low flow and automatically enhancing the humidification capacity at high flow, realizing passive humidity matching without external control.

[0033] Preferably, the deviation evaluation is calculated by calculating the weighted deviation index of the flow, humidity and pressure of the actual output airflow relative to the target value. When the deviation continues to exceed the limit, local data recording is triggered, and the model gradient is uploaded to the cloud for federated learning, realizing the continuous optimization of the control strategy.

[0034] The security response includes when the static induction signal detects high-frequency abnormal pulses, it is determined to be choking or condensate accumulation, and the system immediately starts the pulse purge program and reduces the peak flow by 10% to prevent aspiration or discomfort.

[0035] It should be noted that the low-power monitoring state only maintains the micro-power detection mode of the millimeter wave radar and the static sensor. Once the airflow starting characteristics are detected, the whole system is awakened, and the multi-modal sensing and dynamic adjustment process is re-entered, realizing the cooperative optimization of energy efficiency and response speed.

[0036] After the system is started, the multi-modal sensing module is activated first. The 60GHz millimeter wave radar detects the Doppler shift caused by the movement of particles in the airflow, reconstructs the radial velocity distribution in the breathing circuit, and obtains the spatial flow field information. At the same time, the annular static induction electrode captures the time-varying charge signal generated by the friction between the airflow and the pipe wall, reflecting the overall flow dynamics. The two signals are time-synchronized and fused to generate a composite data stream containing flow field structure and macro trend, providing high-dimensional input for subsequent analysis.

[0037] The composite data stream enters the edge computing unit for nonlinear dynamics analysis. The system reconstructs the phase space of the airflow time series, calculates the maximum Lyapunov exponent and sample entropy, and generates the respiratory chaos index RCI(t) by weighting, which is used to quantify the stability change trend of the respiratory rhythm; at the same time, the local Reynolds number is calculated combined with the flow velocity distribution data, and when Re>2300 and located in the downstream of the pipeline, it is determined as the starting area of turbulent flow, and the potential fluid disturbance risk is identified; finally, the chaos index and the flow field risk are integrated into a multi-dimensional state vector representing the respiratory state, which is the core basis for control decision.

[0038] The state vector is input into the adaptive control engine, combined with the energy state provided by the flexible thermoelectric generator using the temperature difference of the pipeline to dynamically generate an optimized execution instruction set; when the state vector indicates the precursor of respiratory instability, the control engine increases the flow rate rise slope by 15%-30% in advance 0.1 seconds, realizes feedforward compensation, and optimizes the piezoelectric valve opening timing according to the turbulent flow position to suppress vortex development, ensuring efficient and energy-saving adjustment process.

[0039] The execution instruction set drives the piezoelectric ceramic microvalve to perform millisecond-level precise flow regulation, while activating the shear-responsive humidification layer; the humidification layer is composed of PVA / borax intelligent hydrogel, whose permeability changes adaptively with airflow shear stress, suppressing water release at low flow rates and enhancing humidification capacity at high flow rates, achieving passive on-demand humidification without external heating or control, and finally outputting warm and humid gas that matches the physiological needs of patients.

[0040] After the adjustment is completed, the system samples the flow, humidity and pressure of the output airflow, and calculates the weighted deviation index relative to the target value; if the deviation continues to exceed the limit, local data recording is triggered and the model gradient is uploaded to the cloud for federated learning, realizing continuous optimization of the control strategy; at the same time, the system monitors high-frequency abnormal pulses through electrostatic induction signals, and once it detects the risk of choking or condensate accumulation, it immediately starts pulse purging and reduces the peak flow by 10%, preventing aspiration or discomfort.

[0041] After completing feedback and safety response, the system automatically enters low-power listening state, maintaining only millimeter wave radar and electrostatic sensor micro-power detection mode, real-time monitoring airflow starting characteristics; once the next respiratory action is detected, the whole system is immediately awakened, re-entering the multi-modal sensing and dynamic adjustment process, realizing a complete closed-loop operation from sensing, decision-making, execution to learning and energy saving, ensuring stability, intelligence and energy efficiency optimization in long-term use.

[0042] In use, after starting, the spatial flow velocity distribution and overall dynamic characteristics of the airflow are collected by the millimeter wave radar and the electrostatic induction electrode in cooperation to generate a composite data stream that fuses flow field information; through nonlinear dynamics analysis, the breathing chaos index is extracted to evaluate the rhythm stability, and the turbulent flow risk area is identified in combination with the Reynolds number, forming a multi-dimensional vector representing the physiological and fluid state; the state vector is input into the adaptive control engine, combined with the energy state provided by the thermoelectric generator driven by the pipeline temperature difference, to generate an optimized execution instruction set; the instruction set drives the piezoelectric valve to accurately adjust the flow, while activating the shear-responsive intelligent hydrogel humidification layer to make its permeability adaptively change with the airflow shear stress, achieving passive on-demand humidification; the adjustment result is fed back in real time for deviation evaluation and federal learning model updating, and safety protection is triggered when detecting abnormalities such as choking and water accumulation; after completing the closed-loop regulation, the system enters a low-power listening state, waiting for the next respiratory event to trigger a new round of regulation.

[0043] In summary, based on nonlinear dynamics analysis, the breathing chaos index is constructed to realize early warning of respiratory rhythm instability and change from passive response to active prediction; combined with thermoelectric conversion technology, the pipeline temperature difference is used to realize self-power supply of key modules, improving system energy efficiency and portability; through the adaptive control engine, the piezoelectric valve and the shear-responsive intelligent humidification layer work together to realize millisecond-level accurate regulation of flow and humidity and passive on-demand humidification, avoiding excessive or insufficient humidification; the closed-loop feedback mechanism integrates deviation evaluation, federal learning and safety protection, supporting continuous optimization of control strategies and real-time response to abnormal risks; finally, through the low-power listening mode, the energy efficiency and wake-up speed are balanced. The overall system significantly improves the comfort, synchronization, safety and intelligence level of respiratory support, effectively solving the core problems of human-machine asynchronization, response lag, humidification rigidity, high energy consumption and lack of intelligent evolution ability in existing technologies.

[0044] It is important to note that the construction and arrangement of the application shown in the various exemplary embodiments is illustrative only. Although only a few embodiments have been described in detail in this disclosure, those skilled in the art who review the present disclosure will readily appreciate that many modifications are possible (e.g., variations in sizes, dimensions, structures, shapes and proportions of the various elements, values of parameters, mounting arrangements, use of materials, colors, orientations, etc.) without materially departing from the novel teachings and advantages of the subject matter described herein. For example, elements shown as integrally formed can be constructed of multiple parts or elements, the position of elements can be reversed or otherwise varied, and the nature or number of elements can be altered or varied. Thus, the foregoing description is not intended to be taken in a limiting sense, but is made solely for illustration rather than limitation. All such modifications are intended to be included within the scope of the present application. The order or sequence of any process or method steps can be varied or re-sequenced without departing from the subject matter. Any "process" or "method" steps can be carried out simultaneously, concurrently, or in different order than as described. The application is to be limited only by the claims included herein and equivalents thereof. In the claims, any means-plus-function clause is intended to cover the structures described herein as performing the recited function and also material or equivalents equivalent thereto. Other substitutions, modifications, changes and omissions can be made in the design, operating conditions and arrangement of the exemplary embodiments without departing from the scope of the present application. Accordingly, the present application is not limited to the specific embodiments described herein, but only by the claims included herein and equivalents thereof.

[0045] Also, in an effort to provide a concise description of the exemplary embodiments, all features of an actual implementation can not be described (that is, those unrelated to the presently contemplated best mode of carrying out the application, or those unrelated to enabling the claimed application).

[0046] It should be noted that the above-mentioned embodiments are only used to illustrate the technical solutions of the present application, but not to limit the present application. Although the present application is described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical solutions of the present application can be modified or equivalently replaced without departing from the spirit and scope of the present application, which should be covered by the claims of the present application.

Claims

1. A respirator flow dynamic adjustment system based on airflow sensing, characterized in that: include, After startup, the multimodal sensing module collects the spatial velocity distribution and dynamic characteristics of respiratory airflow, and generates a composite data stream that integrates flow field information; The composite data stream is analyzed nonlinearly to generate a state vector characterizing respiratory stability, which is then used to identify the risk of fluid anomalies. This state vector is combined with the energy state of thermoelectric conversion by the adaptive control engine to generate an optimized set of execution instructions; The execution instruction set drives the piezoelectric valve and the shear-response humidification layer to coordinate the flow and humidity, and output physiologically matched breathing gas; The adjustment results are fed back for deviation assessment and model update, and trigger a safety response. After completing closed-loop control, it enters a low-power monitoring state, waiting for the next breathing cycle to trigger the subsequent adjustment process.

2. The respirator flow dynamic adjustment system based on airflow sensing according to claim 1, characterized in that: The multimodal sensing module includes a 60GHz millimeter-wave radar chip and a ring-shaped electrostatic induction electrode array. The millimeter-wave radar is used to detect the Doppler frequency shift caused by the movement of particles in the airflow to reconstruct the radial velocity distribution. The electrostatic induction electrode is used to capture the time-varying charge signal generated by the airflow rubbing against the pipe wall. The two are output synchronously to realize the fusion sensing of spatial and overall flow.

3. The respirator flow dynamic adjustment system based on airflow sensing according to claim 2, characterized in that: The nonlinear analysis includes phase space reconstruction of the airflow sequence and calculation of the maximum Lyapunov exponent and sample entropy. The respiratory chaos index RCI(t) is generated by weighting the normalized entropy change and Lyapunov shift, and is used to quantify the stability trend of respiratory rhythm.

4. The respirator flow dynamic adjustment system based on airflow sensing according to claim 3, characterized in that: The fluid anomaly risk includes the turbulence initiation location identified based on the Reynolds number. When the local Re > 2300 and is located in the downstream region of the pipeline, it is identified as a potential airflow disturbance zone and used as an input parameter for adjusting the control strategy.

5. The respirator flow dynamic adjustment system based on airflow sensing according to claim 4, characterized in that: The energy state of the thermoelectric conversion is provided by a flexible thermoelectric generator attached to the outer wall of the breathing tube. It generates electricity by utilizing the temperature difference between the output gas and the environment, and supplies power to the sensing module and edge computing unit through the MPPT circuit, realizing the self-powered operation of some functions.

6. The respirator flow dynamic adjustment system based on airflow sensing according to claim 5, characterized in that: The execution instruction set includes flow rate rise slope, valve opening timing and humidification response parameters. When the state vector indicates breathing instability, the control engine increases the flow response speed by 15%–30% 0.1 seconds in advance and dynamically optimizes the piezoelectric valve control logic to suppress turbulence development.

7. A respirator flow dynamic adjustment system based on airflow sensing according to claim 6, characterized in that: The shear-responsive humidification layer is composed of PVA / borax smart hydrogel, whose permeability changes nonlinearly with airflow shear stress. It reduces the moisture release rate at low flow rates and automatically enhances humidification capacity at high flow rates, achieving passive humidity matching without external control.

8. A respirator flow dynamic adjustment system based on airflow sensing according to claim 7, characterized in that: The deviation assessment calculates the weighted deviation index of the actual output airflow, humidity, and pressure relative to the target value. When the deviation continues to exceed the limit, local data recording is triggered, and the model gradient is uploaded to the cloud to participate in federated learning, thereby achieving continuous optimization of the control strategy.

9. A respirator flow dynamic adjustment system based on airflow sensing according to claim 8, characterized in that: The safety response includes determining that when the electrostatic induction signal detects a high-frequency abnormal pulse, indicating choking or condensation buildup, the system immediately initiates a pulse purging procedure and reduces the peak flow rate by 10% to prevent aspiration or discomfort.

10. A respirator flow dynamic adjustment system based on airflow sensing according to claim 9, characterized in that: The low-power listening state maintains only the low-power detection mode of millimeter-wave radar and electrostatic sensing. Once the initial characteristics of airflow are detected, the entire system is awakened and re-enters the multimodal sensing and dynamic adjustment process to achieve synergistic optimization of energy efficiency and response speed.

Images