Intelligent multi-mode medical oxygen flow monitoring system, device and method

Abstract

The invention relates to the technical field of medical instruments, and discloses an intelligent multi-mode medical oxygen flow monitoring system, device and method. The system comprises a sensor module, a coupling modeling unit, a prediction unit, a control unit and an output module. Oxygen flow, temperature, humidity and pressure data are collected in real time through the sensor module, a dynamic coupling model of a flow field, a thermal field and a pressure field is established through the coupling modeling unit, and the prediction unit dynamically predicts the oxygen flow based on a time sequence and a state space model. The control unit optimizes sensor sampling and signal processing parameters through a feedback control algorithm, and the output module displays the oxygen flow and triggers abnormal alarm. According to the invention, high-precision monitoring, dynamic prediction and optimal control of the oxygen flow are realized, the problems of low monitoring precision, weak data analysis capability and poor interactivity in the prior art are solved, and the safety, real-time performance and intelligent level of oxygen supply are remarkably improved.

Description

Technical Field

[0001] The present invention relates to the technical field of medical devices, and in particular to an intelligent multi-modal medical oxygen flow monitoring system, device and method. Background Art

[0002] With the continuous advancement of modern medicine, oxygen therapy plays a key role in intensive care, emergency treatment, postoperative recovery and other fields, and puts forward higher accuracy and real-time requirements for patients' oxygen supply. However, traditional oxygen supply systems are usually only equipped with basic flow monitoring functions, lacking the ability to conduct real-time analysis and dynamic monitoring of multi-dimensional data such as oxygen flow, temperature, humidity and pressure. This limitation makes it difficult for medical staff to adjust the oxygen supply strategy in time according to the specific needs of patients, resulting in insufficient or overoxygenated oxygen supply, which in turn affects the treatment effect and the utilization efficiency of medical resources.

[0003] In existing technical solutions, oxygen flow monitoring equipment mostly uses a single sensor to collect data, usually relying on mechanical or low-sensitivity electronic flow sensors for monitoring. Its monitoring accuracy is poor under low flow rate conditions (such as ≤1L / min), and it cannot adapt to the dynamic changes of environmental factors such as temperature, humidity, and pressure. At the same time, these devices often lack in-depth data analysis and prediction capabilities, and can only provide real-time data display functions, making it difficult to predict the patient's oxygen demand trend through time series analysis. In terms of interaction methods, most devices only rely on a single display screen to display values, lack multimodal interaction functions, have insufficient remote monitoring capabilities, and have the risk of cross-infection during operation.

[0004] The existing technology also has problems with insufficient equipment durability and environmental adaptability. For example, the shell material of traditional equipment is not antibacterial, and the internal electronic components are easily affected by high humidity and corrosive environments and fail. At the same time, the lack of intelligent feedback control and parameter optimization design leads to poor adaptability of the equipment in complex medical scenarios, especially in high-pressure environments (such as emergency and ICU). It is difficult for the equipment to respond to the rapid changes in the patient's oxygen demand in a timely manner, which in turn affects the treatment effect. Therefore, there is an urgent need for a new monitoring system that can combine high-precision sensors, multi-physics field modeling, intelligent dynamic prediction and feedback control to solve the above technical bottlenecks. Summary of the invention

[0005] In view of the shortcomings of the prior art, the present invention provides an intelligent multimodal medical oxygen flow monitoring system, device and method to solve the problems of low monitoring accuracy, insufficient data analysis capability, backward interaction mode, poor adaptability and difficulty in meeting the needs of complex medical scenarios in the prior art.

[0006] To achieve the above objectives, the present invention is implemented through the following technical solutions: intelligent multi-modal medical oxygen flow monitoring system, device and method, the intelligent multi-modal medical oxygen flow monitoring system comprises the following steps: Multi-physics field sensing module for real-time collection of oxygen flow, temperature, humidity and pressure data; Central processing module, including: Multi-physics coupling modeling unit, used to establish dynamic coupling models between flow field, thermal field and pressure field; A dynamic flow prediction unit, used to predict the dynamic change of oxygen flow based on the state space model and time series data; An optimal control unit, used to dynamically adjust monitoring parameters according to the prediction results to optimize monitoring accuracy; The output and alarm module is used to display oxygen flow data in real time, trigger abnormal alarms, and send data to external terminals.

[0007] Preferably, the multi-physics field sensing module comprises: ; in, is the vortex frequency, is the Strouhal number, is the oxygen flow rate, is the characteristic size of the spoiler; The temperature and humidity sensor is used to collect temperature and humidity data in the pipeline and calculate the collected humidity signal through the following formula to obtain relative humidity: ; in, is the relative humidity, is the actual water vapor pressure, is the saturated water vapor pressure; Pressure sensor, used to measure static pressure and ambient pressure data in the pipeline in real time; A data preprocessing unit is used to preprocess the data sources from the flow sensor, the temperature and humidity sensor, and the pressure sensor, wherein the data preprocessing includes: The vortex frequency signal of the flow sensor is denoised, and the Kalman filter algorithm is used to optimize the collected signal. The flow velocity data after filtering is: ; in, is the estimated value of the state at the current moment, is the current measured value, is the Kalman gain, is the observation matrix; Perform linear interpolation on the sampling of temperature, humidity and pressure data to unify the time step; The flow and viscosity parameters are compensated based on temperature, and the compensation model is: ; in, For temperature The viscosity of oxygen under The base temperature The viscosity of is the coefficient of the effect of temperature change on viscosity.

[0008] Preferably, the multi-physics field coupling modeling unit establishes the dynamic coupling relationship among the flow field, the thermal field and the pressure field by the following modeling method: The flow field is established based on the fluid dynamics model, and the flow field characteristics are determined by the oxygen flow rate and pressure gradient; The thermal field is established based on the heat conduction model, and the thermal field characteristics are determined by the temperature distribution and the heat diffusion characteristics of the pipeline; The influence of thermal field on flow field is introduced into coupling modeling through the dynamic changes of oxygen density and viscosity.

[0009] Preferably, the dynamic traffic prediction unit uses a recursive neural network model to train time series data, and the recursive neural network is trained in the following manner: The input is the historical time series data of flow, temperature and pressure collected by the multi-physics field sensing module; The output is the predicted value of oxygen flow at the next moment; The network parameters are optimized using the loss function with the sum of squared errors as the target.

[0010] Preferably, the optimal control unit dynamically adjusts the following parameters through an optimal control algorithm: Design a cost function based on flow monitoring error and control cost; Adjust the sensor sampling frequency and signal filtering parameters in real time according to the feedback gain matrix; The control law uses input state error to feedback and correct the output signal.

[0011] Preferably, the output and alarm module includes: Real-time display subunit, used to display current oxygen flow, temperature, humidity and pressure; An alarm subunit triggers an alarm signal when the oxygen flow rate is lower than a preset threshold; The data transmission subunit is used to send the monitoring data to the remote terminal through wireless communication technology.

[0012] Preferably, the central processing module further comprises a calibration module, which is used to: Perform static zero flow detection regularly; Automatically adjust sensor sensitivity based on calibration data; When the monitoring error exceeds the preset threshold, the recalibration process is triggered.

[0013] Preferably, the flow sensor of the multi-physics field sensing module guides the flow of oxygen to form a vortex through a baffle, and calculates the oxygen flow rate by detecting the frequency and amplitude of the vortex, and the frequency of the vortex is determined by the flow rate.

[0014] Preferably, an intelligent multi-modal medical oxygen flow monitoring device is also provided, comprising: Sensor module for collecting oxygen flow, temperature, humidity and pressure data; A coupled modeling unit is used to establish a coupled model of oxygen flow field, thermal field and pressure field; A prediction unit, used for dynamically predicting oxygen flow based on a time series; a control unit for optimizing sensor sampling and signal processing parameters through feedback control; Output module, used to display oxygen flow and alarm information.

[0015] Preferably, an intelligent multi-modal medical oxygen flow monitoring method is also provided, comprising the following steps: Collect oxygen flow, temperature, humidity and pressure data in real time through sensor modules; According to the collected data, a coupled dynamic model of flow field, thermal field and pressure field is established; Based on time series data and state space model, recurrent neural network is used to predict the dynamic changes of oxygen flow rate; Optimize monitoring parameters through feedback control algorithms and adjust the sampling frequency and signal filtering strategy of sensors; Outputs real-time oxygen flow and triggers an alarm when an abnormality is detected.

[0016] The present invention provides an intelligent multi-modal medical oxygen flow monitoring system, device and method, which have the following beneficial effects: 1. The present invention adopts an integrated design of a highly sensitive oxygen flow sensor and a multi-modal interactive interface, achieving high-precision monitoring of oxygen flow and intuitive human-computer interaction, significantly improving user experience and reducing the probability of medical misjudgment due to monitoring data errors. Compared with the prior art design that only relies on low-precision flow sensors and a single interactive mode, the present invention solves the shortcomings of traditional monitoring systems in terms of insufficient sensitive capture of flow changes and inconvenient user operation, especially in low flow rates and dynamic environments, and can maintain high-precision monitoring.

[0017] 2. Since the central processing unit is equipped with an advanced data analysis module and uses AI algorithms to perform real-time analysis and dynamic prediction of oxygen flow data, the present invention has greatly improved the intelligence level of medical services and resource management efficiency. By accurately estimating the patient's future oxygen demand trend through a time series prediction model, the present invention significantly reduces resource waste and unnecessary manual intervention in the oxygen supply process. Compared with the existing solutions with limited data analysis capabilities and the inability to achieve intelligent prediction, the present invention solves the problems of inaccurate allocation of oxygen supply resources and low medical efficiency.

[0018] 3. The present invention introduces voice recognition and natural language processing technology to achieve a contactless user interactive query function. For example, patients or medical staff can use voice to inquire about the current oxygen concentration or remaining oxygen supply time of the device, thereby reducing the need to directly operate the device. The present invention effectively reduces the risk of cross-infection during the use of the device, especially in high-risk medical environments, which significantly protects the health and safety of medical staff and patients. Compared with the design scheme in the prior art that only supports touch screen or button operation, the present invention solves the problem that the traditional device interaction method has strong contact and high infection risk.

[0019] 4. The present invention has developed a supporting mobile application that supports medical staff to remotely monitor the oxygen usage of multiple patients through wireless data transmission technology. Medical staff can view the oxygen flow, cumulative oxygen usage and equipment status in real time through mobile phones or tablets, and adjust the oxygen supply plan in time according to patient needs. The present invention simplifies the daily oxygen supply management process, significantly saves time costs, and improves medical operation efficiency. Compared with the prior art solution that requires close-range operation of equipment and cannot be viewed remotely, the present invention solves the problems of low efficiency and cumbersome operation of traditional management models.

[0020] 5. The present invention uses antibacterial materials to manufacture the device housing, and by adding a moisture-proof design to the internal circuit board, the service life and operational stability of the device are enhanced. The antibacterial housing is treated with a nano-silver coating to effectively inhibit the growth of pathogenic microorganisms, and the moisture-proof design enables the device to adapt to high humidity and complex medical environments through a combination of a sealing structure and a protective coating. The present invention extends the equipment maintenance cycle and reduces long-term operating costs. Compared with the prior art solutions in which the equipment has poor durability and is easily affected by the environment, the present invention solves the problems of frequent failures and high maintenance costs of traditional equipment in high humidity environments.

[0021] 6. The present invention integrates a multimodal interactive interface, a data analysis module, an AI prediction model and a remote management function, and its overall design has reached a highly intelligent and automated level. Through unified data processing and high-precision control, the present invention not only improves the accuracy and efficiency of oxygen flow monitoring, but also provides a more intuitive interactive experience. Compared with the decentralized design and single-function system solutions in the prior art, the present invention solves the problem of insufficient functional integration of traditional systems and difficulty in forming an effective management closed loop. BRIEF DESCRIPTION OF THE DRAWINGS

[0022] Figure 1 It is a schematic diagram of the system architecture of the present invention; Figure 2 It is a schematic diagram of the structure of the device of the present invention; Figure 3 It is a schematic diagram of the method flow of the present invention. DETAILED DESCRIPTION

[0023] The following will be combined with the drawings in the specification 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.

[0024] Please see attached Figure 1 , an embodiment of the present invention provides an intelligent multi-modal medical oxygen flow monitoring system, comprising the following steps: Multi-physics field sensing module for real-time collection of oxygen flow, temperature, humidity and pressure data; First, in order to realize high-precision, multi-modal real-time monitoring of the intelligent multi-modal medical oxygen flow monitoring system, the present invention obtains multiple physical parameters such as oxygen flow, temperature, humidity and pressure through a multi-physical field sensing module, providing basic data input for subsequent multi-physical field modeling and dynamic prediction. The data acquisition results of the multi-physical field sensing module directly affect the accuracy of modeling, prediction and optimization control in the central processing module. Therefore, the multi-physical field sensing module needs to have high precision, high sensitivity and real-time performance.

[0025] In this embodiment, the multi-physics field sensing module includes a flow sensor, a temperature and humidity sensor, and a pressure sensor, which are used to collect multiple physical field data in real time and perform preliminary processing on the data for subsequent calculation and modeling. The specific implementation is as follows: Generally, the flow sensor in this embodiment adopts micro-electromechanical system (MEMS) technology, which has the characteristics of high sensitivity and fast response speed, and can accurately capture small changes in the oxygen flow in the pipeline.

[0026] Specifically, by setting a flow resistor in the pipeline, the flowing oxygen generates a periodic vortex signal, and the flow sensor uses this signal to measure the flow rate of oxygen. The oxygen flow rate and the vortex frequency generated by the flow resistor satisfy the following relationship: ; in, is the vortex frequency; is the Strouhal number, a dimensionless number determined by the spoiler characteristics; is the oxygen flow rate; is the characteristic size of the spoiler.

[0027] The vortex signal is captured by a high-frequency sensor, and the signal output by the sensor is amplified and filtered to reduce noise interference and improve data stability.

[0028] A noise elimination module can be added to the flow sensor, and the vortex frequency signal output by the sensor can be optimized through the Kalman filter algorithm. The filtered flow velocity signal can be expressed as: ; in, is the estimated value of the state at the current moment; is the predicted state of the previous moment; Kalman gain dynamically adjusts the weights of observed and predicted values; is the current observation value of the sensor; is the observation matrix, which is used to map the state variables to the measurement space.

[0029] The filtered signal is further processed through a linear interpolation algorithm to compensate for the inconsistency of the sampling frequency to ensure data accuracy and real-time performance.

[0030] In this embodiment, the temperature and humidity sensor is mainly used to monitor the temperature and humidity parameters of oxygen in the pipeline.

[0031] Specifically, relative humidity is calculated using the following formula: ; in, is the relative humidity, expressed as a percentage; is the actual water vapor partial pressure in oxygen; is the saturated water vapor pressure, which is related to the oxygen temperature.

[0032] In order to adapt to different environmental conditions, the temperature and humidity sensor has a self-calibration function. The sensor can be calibrated according to a known standard humidity environment. The calibration result corrects the humidity measurement value through an interpolation algorithm. The calibrated humidity data is input into the subsequent multi-physics field coupling modeling module to compensate for the impact of temperature and humidity changes on oxygen density and viscosity.

[0033] The pressure sensor in this embodiment is used to measure the static pressure of oxygen in the pipeline and the ambient pressure. Generally, the pressure sensor adopts micro-pressure sensing technology, which can accurately capture the tiny pressure fluctuations in the oxygen supply system.

[0034] Specifically, the measurement range of the pressure sensor includes: Static pressure in the pipeline, used to evaluate the pressure distribution during the oxygen flow; The ambient pressure outside the pipeline is used to compensate for the interference of external pressure changes on flow monitoring.

[0035] The output signal of the pressure sensor will undergo multi-stage filtering to eliminate high-frequency interference and ensure the stability and consistency of the measured data.

[0036] In order to achieve effective fusion of multi-physical field data, in this embodiment, the multi-physical field sensing module also includes a data preprocessing unit for performing unified preliminary processing on the flow, temperature, humidity and pressure data collected by the sensor.

[0037] Specifically, the data preprocessing unit performs the following operations: De-noising and linear interpolation of vortex signals collected by flow sensors; Calibrate and linearize the data of temperature and humidity sensors; Perform multi-stage filtering on the data of the pressure sensor; All sensor output data are unified to the same time step and sorted by timestamp.

[0038] The data preprocessing unit also includes a temperature compensation module, which is used to correct the influence of temperature change on flow and viscosity. The compensation formula is as follows: ; in, is the current temperature Oxygen viscosity under The base temperature Viscosity under is the temperature sensitivity coefficient.

[0039] Central processing module, including: Multi-physics coupling modeling unit, used to establish dynamic coupling models between flow field, thermal field and pressure field; A dynamic flow prediction unit, used to predict the dynamic change of oxygen flow based on the state space model and time series data; An optimal control unit, used to dynamically adjust monitoring parameters according to the prediction results to optimize monitoring accuracy; In order to achieve accurate modeling and dynamic monitoring of multi-physics field data, the present invention designs a central processing module. As the core computing unit of the system, this module is responsible for processing data from the multi-physics field sensing module, including flow, temperature, humidity and pressure information. The central processing module constructs an intelligent analysis framework of the system through functional units such as multi-physics field coupling modeling, dynamic flow prediction and optimal control, providing high-precision and high-reliability support for real-time monitoring of oxygen flow.

[0040] The central processing module is closely integrated with the multi-physics field sensing module. Its input is pre-processed multi-physics field data, and its output includes real-time oxygen flow prediction value and control signal, which directly affects the performance of the output and alarm module.

[0041] In this embodiment, the multi-physics field coupling modeling unit is used to establish a dynamic coupling model between the flow field, thermal field and pressure field to describe the multi-physics field characteristics of the oxygen flow in the pipeline. By introducing fluid dynamics, heat conduction and pressure distribution models, the unit can simulate the complexity of multi-field interactions.

[0042] Specifically, the flow field modeling is based on the following Navier–Stokes equations: ; in, Indicates oxygen density; represents the velocity field vector; represents the pressure field; Indicates oxygen viscosity; Represents external forces, such as flow disturbances caused by spoilers.

[0043] In flow field modeling, the pipe boundary condition adopts no-slip boundary condition (i.e., the velocity at the pipe wall is zero). As an option, the inlet flow velocity can be set to be constant or dynamically changed, which is determined by the flow data collected by the sensor.

[0044] In this embodiment, the thermal field modeling is based on the Fourier heat conduction equation: ; in, represents the temperature field; is the thermal diffusion coefficient, defined as ,in is the thermal conductivity, is the specific heat capacity; is the heat source term, which is generated by the friction heat between the sensor and the oxygen flow in the pipeline.

[0045] Pressure field modeling is coupled with flow field modeling to drive oxygen flow through pressure gradient while considering the dynamic modulation of flow velocity by static pressure and ambient pressure.

[0046] Multi-physics coupling modeling discretizes and solves the above equations through the finite element method (FEM). The solution process includes the following steps: Setting of initial conditions, including initial temperature field , velocity field and pressure field ; Use numerical iteration method to update multiple field variables until convergence condition is reached; Output dynamic distribution of flow field, thermal field and pressure field.

[0047] In some embodiments, in order to improve computing efficiency, part of the computing tasks can be migrated to the cloud, and the coupling model can be accelerated through distributed computing.

[0048] In this embodiment, the dynamic flow prediction unit is used to predict the dynamic change of the oxygen flow rate according to the time series data, so as to solve the real-time problem caused by the sensor response lag.

[0049] Specifically, the dynamic prediction describes the dynamic characteristics of oxygen flow based on the state space model: ; ; in, is the system state vector, including the flow rate ,temperature and pressure ; is the input signal, such as environmental disturbance; is the output flow value; and represents the nonlinear disturbance term.

[0050] In this embodiment, a recursive neural network (RNN) is used to learn the characteristics of the nonlinear disturbance term. Generally, the input of the RNN is the time series data collected by the multi-physics field sensor module, and the output is the predicted value of the oxygen flow at the next moment.

[0051] The RNN model is trained through the following steps: Data preprocessing: normalize historical data to eliminate dimensional differences; Model initialization: setting initial weights and biases; Loss function optimization: The model parameters are optimized through the error square sum objective function. The loss function is: ; in, It is the loss function, which is used to measure the error between the model prediction value and the true value; is the total number of samples, that is, the number of samples used for training or testing; It is The true value of the sample (GroundTruth); It is The predicted value of a sample (PredictedValue); is the squared error between the true value and the predicted value.

[0052] In this embodiment, the optimal control unit is used to optimize the operating parameters of the system according to the dynamic prediction results, including the sampling frequency of the sensor, filter parameters, etc.

[0053] Specifically, the control algorithm is based on the following cost function: ; in, represents the state error term; Indicates the control energy consumption item; and are the start and end time of the control respectively.

[0054] The feedback gain matrix is ​​obtained by solving the Riccati equation : ; The optimal control law is: ; Among them, the control signal Used to adjust the operating parameters of the sensor in real time.

[0055] Output and alarm module, used to display oxygen flow data in real time, trigger abnormal alarms, and send data to external terminals; The output and alarm module of the present invention is designed to realize the real-time presentation of the calculation results of the central processing module and the alarm triggering of abnormal conditions. Through the visual output of real-time monitoring data such as oxygen flow, temperature, humidity and pressure, it is convenient for medical staff and management systems to grasp the patient's oxygen supply in a timely manner. In addition, the output and alarm module also supports wireless data transmission, and the monitoring results can be sent to a remote terminal to realize remote management functions. The operation of this module depends on the real-time data provided by the multi-physical field sensing module and the dynamic prediction and optimization results output by the central processing module.

[0056] In this embodiment, the real-time data display subunit of the output and alarm module displays the oxygen flow rate and related parameters through a multi-channel display screen.

[0057] In general, the display screen outputs data in the following order: Current real-time oxygen flow value (unit: L / min); Temperature and humidity parameters, including temperature (unit: °C) and relative humidity (unit: %); The pressure value in the pipeline (unit: kPa), including static pressure and dynamic pressure change values.

[0058] The display screen can simultaneously present the trend of data changes, and intuitively reflect the historical changes of flow, temperature and humidity in the form of line graphs or bar graphs. For example, the display period can be set to 1 hour, 6 hours or 24 hours, and the historical data is dynamically generated by the central processing module.

[0059] The real-time data output function also supports interface customization, and medical staff can select the priority of data display according to their needs, such as displaying only oxygen flow or additionally displaying temperature and humidity.

[0060] Specifically, the data output is processed using the following mathematical model: The flow data is output by the dynamic flow prediction unit in the central processing module; The temperature and humidity data are calculated using the following formula: ; in, is the relative humidity; is the actual water vapor pressure; is the saturated water vapor pressure, and is related to temperature related.

[0061] In this embodiment, the abnormal alarm subunit is used to send out an alarm signal when the monitoring data deviates from the set threshold value, prompting medical staff to take intervention measures.

[0062] In general, alarm conditions include but are not limited to: The oxygen flow rate is lower than the preset threshold, such as 0.5 L / min; The pressure in the pipeline exceeds the safety threshold, such as 120 kPa; The temperature or humidity value exceeds the normal range, for example, the temperature is higher than 40 ℃.

[0063] The alarm signal may include an audible and visual alarm, such as a combination of a buzzer and a flashing light. In a possible implementation, the intensity of the alarm signal is related to the degree of deviation from the threshold, for example, when the flow rate drops to 0.2 L / min, a double alarm signal is triggered.

[0064] Specifically, the alarm triggering mechanism is based on the following mathematical logic: ; in, Indicates real-time monitoring value; and They are the upper and lower alarm thresholds respectively; Indicates that an alarm is triggered.

[0065] In this embodiment, the output and alarm module also includes a data transmission subunit for uploading the monitoring data to a remote terminal or a cloud platform.

[0066] Specifically, data transmission uses wireless communication technologies, such as Wi-Fi or Bluetooth, to support the following functions: Real-time transmission of current oxygen flow, temperature, humidity and pressure data; Upload historical data regularly, for example once an hour; Receive update instructions from the cloud, such as adjustments to alarm thresholds.

[0067] To ensure the security of data transmission, the module introduces encryption algorithms, such as AES-256, to encrypt data packets. During transmission, the data integrity is verified by timestamps to prevent data loss or tampering.

[0068] The data transmission subunit also supports offline mode, that is, when the network is interrupted, the system stores the data in the local storage unit and automatically uploads it after the network is restored.

[0069] Please see attached Figure 2 , intelligent multi-modal medical oxygen flow monitoring device, including: Sensor module for collecting oxygen flow, temperature, humidity and pressure data; The sensor module is the basic data input unit of the system. It uses multiple types of sensors to capture multi-physical field information of oxygen flow and environmental conditions in real time, providing accurate and reliable raw data support for subsequent modeling and prediction. At the same time, high-sensitivity sensor technology ensures data acquisition accuracy under low flow rate or dynamic flow rate conditions, thereby improving the overall performance of the system.

[0070] A coupled modeling unit is used to establish a coupled model of oxygen flow field, thermal field and pressure field; The coupling modeling unit is responsible for comprehensively processing the multi-physics field data provided by the sensor module, combining theories such as fluid dynamics, heat conduction and pressure distribution to establish a mathematical model to accurately describe the coupling relationship of oxygen flow. Through the coupling model, the influence of temperature, humidity and pressure on oxygen flow can be dynamically reflected, providing multi-dimensional input parameter support for subsequent prediction units.

[0071] A prediction unit, used for dynamically predicting oxygen flow based on a time series; The prediction unit predicts the future oxygen flow rate change trend in real time by learning the historical time series data provided by the sensor module and using deep learning algorithms such as recursive neural networks. This function effectively solves the problem of sensor response lag and provides accurate prediction data support for the control unit, thereby improving the system's real-time performance and dynamic adaptability.

[0072] a control unit for optimizing sensor sampling and signal processing parameters through feedback control; The control unit dynamically adjusts the sampling frequency, filtering parameters and signal processing strategy of the sensor based on the output of the prediction unit by designing a feedback control algorithm. By optimizing the control process, the monitoring error and calculation delay are reduced, and the robustness of the system to external interference is enhanced, providing high-precision monitoring data for the output module.

[0073] Output module, used to display oxygen flow and alarm information; The output module is responsible for presenting the analysis results of the central processing unit in a visual form, including real-time display of monitoring data such as oxygen flow, temperature, humidity and pressure, and triggering alarm signals under abnormal conditions. Through sound and light alarms and remote communication functions, it ensures that medical staff can obtain information in a timely manner and take intervention measures, thereby improving the safety and convenience of patient oxygen supply management.

[0074] Please see attached Figure 3 , an intelligent multi-modal medical oxygen flow monitoring method, comprising the following steps: S1, collect oxygen flow, temperature, humidity and pressure data in real time through the sensor module; Step S1 is the basis of the entire method. It uses multi-physics field sensors to obtain key parameters in the oxygen flow process in real time, including oxygen flow rate, temperature, humidity and pressure information in the pipeline, to provide accurate raw data support for subsequent dynamic modeling and prediction. At the same time, the collected multi-dimensional data can reflect the complex dynamic conditions of environmental changes and oxygen use, laying the foundation for the accuracy of the monitoring system.

[0075] S2. Establish a coupled dynamic model of flow field, thermal field and pressure field based on the collected data; Step S2 constructs a multi-physics field dynamic coupling model of oxygen in the pipeline by analyzing the oxygen flow rate, temperature, humidity and pressure data collected by the sensor, combining fluid dynamics, heat conduction and pressure distribution theory. Through the coupling model, the influence of temperature change on oxygen density and viscosity and the modulation effect of pressure gradient on flow rate can be accurately described, thereby providing data support with clear physical meaning for dynamic prediction.

[0076] S3, based on time series data and state space model, using recurrent neural network to predict the dynamic changes of oxygen flow; Step S3 uses the time series data provided by the sensor and the output of the coupling model to dynamically predict the flow trend using a recurrent neural network (RNN). By learning and modeling the historical data of oxygen flow, RNN can effectively capture the characteristics of nonlinear and dynamic changes, predict the flow change at the next moment, and provide accurate input reference for feedback control, thereby improving the real-time and adaptability of the system.

[0077] S4, optimize monitoring parameters through feedback control algorithm, adjust the sampling frequency and signal filtering strategy of the sensor; Step S4 uses the feedback control algorithm to optimize the key parameters in the monitoring process, including the sampling frequency and signal processing strategy of the sensor, based on the dynamic flow data provided by the prediction unit. By adjusting the monitoring parameters in real time, the response speed and measurement accuracy of the system in a dynamic flow rate environment can be effectively improved, while reducing the impact of external interference on the data, providing high-quality monitoring results for the output stage.

[0078] S5, output real-time oxygen flow and trigger an alarm when an abnormality is detected; Step S5 presents the prediction and control results of the central processing unit in the form of real-time data, including the current values ​​of oxygen flow, temperature, humidity and pressure. At the same time, when the flow or other parameters exceed the safe range, the system automatically triggers an alarm signal, and reminds medical staff to take timely intervention measures through sound and light prompts or remote communication functions, thereby ensuring the safety of oxygen inhalation for patients and the reliability of medical services.

[0079] Although embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that various changes, modifications, substitutions and variations may be made to the embodiments without departing from the principles and spirit of the present invention, and that the scope of the present invention is defined by the appended claims and their equivalents.

Claims

1. Intelligent multi-modal medical oxygen flow monitoring system, characterized in that: The following steps are involved: Multi-physics field sensing module for real-time collection of oxygen flow, temperature, humidity and pressure data; Central processing module, including: Multi-physics coupling modeling unit, used to establish dynamic coupling models between flow field, thermal field and pressure field; A dynamic flow prediction unit, used to predict the dynamic change of oxygen flow based on the state space model and time series data; An optimal control unit, used to dynamically adjust monitoring parameters according to the prediction results to optimize monitoring accuracy; The output and alarm module is used to display oxygen flow data in real time, trigger abnormal alarms, and send data to external terminals.

2. The intelligent multi-modal medical oxygen flow monitoring system according to claim 1, characterized in that: The multi-physics field sensing module comprises: The flow sensor uses micro-electromechanical system technology, has the ability to detect trace flow rates, and obtains the oxygen flow rate based on the vortex frequency signal generated by the baffle. The relationship between the vortex frequency signal and the oxygen flow rate is: ； in, is the vortex frequency, is the Strouhal number, is the oxygen flow rate, is the characteristic size of the spoiler; The temperature and humidity sensor is used to collect temperature and humidity data in the pipeline and calculate the collected humidity signal through the following formula to obtain relative humidity: ； in, is the relative humidity, is the actual water vapor pressure, is the saturated water vapor pressure; Pressure sensor, used to measure static pressure and ambient pressure data in the pipeline in real time; A data preprocessing unit is used to preprocess the data sources from the flow sensor, the temperature and humidity sensor, and the pressure sensor, wherein the data preprocessing includes: The vortex frequency signal of the flow sensor is denoised, and the Kalman filter algorithm is used to optimize the collected signal. The flow velocity data after filtering is: ； in, is the estimated value of the state at the current moment, is the current measured value, is the Kalman gain, is the observation matrix; Perform linear interpolation on the sampling of temperature, humidity and pressure data to unify the time step; The flow and viscosity parameters are compensated based on temperature, and the compensation model is: ； in, For temperature The viscosity of oxygen under The base temperature The viscosity of is the coefficient of the effect of temperature change on viscosity.

3. The intelligent multi-modal medical oxygen flow monitoring system according to claim 1, characterized in that: The multi-physics field coupling modeling unit establishes the dynamic coupling relationship between the flow field, the thermal field and the pressure field through the following modeling method: The flow field is established based on the fluid dynamics model, and the flow field characteristics are determined by the oxygen flow rate and pressure gradient; The thermal field is established based on the heat conduction model, and the thermal field characteristics are determined by the temperature distribution and the heat diffusion characteristics of the pipeline; The influence of thermal field on flow field is introduced into coupling modeling through the dynamic changes of oxygen density and viscosity.

4. The intelligent multi-modal medical oxygen flow monitoring system according to claim 1, characterized in that: The dynamic traffic prediction unit uses a recursive neural network model to train time series data, and the recursive neural network is trained in the following manner: The input is the historical time series data of flow, temperature and pressure collected by the multi-physics field sensing module; The output is the predicted value of oxygen flow at the next moment; The network parameters are optimized using the loss function with the sum of squared errors as the target.

5. The intelligent multi-modal medical oxygen flow monitoring system according to claim 1, characterized in that: The optimal control unit dynamically adjusts the following parameters through the optimal control algorithm: Design a cost function based on flow monitoring error and control cost; Adjust the sensor sampling frequency and signal filtering parameters in real time according to the feedback gain matrix; The control law uses input state error to feedback and correct the output signal.

6. The intelligent multi-modal medical oxygen flow monitoring system according to claim 1, characterized in that: The output and alarm module comprises: Real-time display subunit, used to display current oxygen flow, temperature, humidity and pressure; An alarm subunit triggers an alarm signal when the oxygen flow rate is lower than a preset threshold; The data transmission subunit is used to send the monitoring data to the remote terminal through wireless communication technology.

7. The intelligent multi-modal medical oxygen flow monitoring system according to claim 1, characterized in that: The central processing module further includes a calibration module, which is used to: Perform static zero flow detection regularly; Automatically adjust sensor sensitivity based on calibration data; When the monitoring error exceeds the preset threshold, the recalibration process is triggered.

8. The intelligent multi-modal medical oxygen flow monitoring system according to claim 1, characterized in that: The flow sensor of the multi-physics field sensing module guides the flow of oxygen to form a vortex through a baffle, and calculates the oxygen flow rate by detecting the frequency and amplitude of the vortex, wherein the frequency of the vortex is determined by the flow rate.

9. An intelligent multi-modal medical oxygen flow monitoring device, applied to the intelligent multi-modal medical oxygen flow monitoring system according to any one of claims 1 to 8, characterized in that: include: Sensor module for collecting oxygen flow, temperature, humidity and pressure data; A coupled modeling unit is used to establish a coupled model of oxygen flow field, thermal field and pressure field; A prediction unit, used for dynamically predicting oxygen flow based on a time series; a control unit for optimizing sensor sampling and signal processing parameters through feedback control; Output module, used to display oxygen flow and alarm information.

10. An intelligent multi-modal medical oxygen flow monitoring method, applied to an intelligent multi-modal medical oxygen flow monitoring system as claimed in any one of claims 1 to 8, characterized in that: The following steps are involved: Collect oxygen flow, temperature, humidity and pressure data in real time through sensor modules; According to the collected data, a coupled dynamic model of flow field, thermal field and pressure field is established; Based on time series data and state space model, recurrent neural network is used to predict the dynamic changes of oxygen flow rate; Optimize monitoring parameters through feedback control algorithms and adjust the sampling frequency and signal filtering strategy of sensors; Outputs real-time oxygen flow and triggers an alarm when an abnormality is detected.

Images