Intelligent breathing machine humidifying device with self-adaptive adjusting function and method

Abstract

The invention relates to the technical field of medical equipment, in particular to an intelligent breathing machine humidifying device with a self-adaptive adjusting function and a method. The intelligent breathing machine humidifying device comprises a breathing machine main body and a humidifying module, and the humidifying module is arranged on the breathing machine main body; the humidifying module comprises a multi-source sensing module, a preprocessing and calculation rehearsal module, an instant feedback and AI control module, an energy consumption calculation and regulation early warning module and a humidifying and water supplementing module. The multi-source sensing module is used for collecting the temperature, humidity and breathing waveform of inhaled gas at the near end of the airway of a patient in real time. The multi-dimensional data of the patient is collected in real time, the digital twin model is used for predicting the humidification trend, output is dynamically adjusted on the basis, a traditional fixed preset mode is changed through the design, closed-loop self-adaptive adjustment is achieved, strategies such as pulse compensation are combined, the humidification efficiency is optimized, it is ensured that the optimal humidification state continues, and the humidification effect is improved. Therefore, the complication risk is effectively reduced, and the treatment safety and patient comfort are improved.

Description

Technical Field

[0001] This invention relates to the field of medical device technology, specifically to an intelligent ventilator humidification device and method with adaptive adjustment function. Background Technology

[0002] Humidification is a crucial component of mechanical ventilation support. Its core objective is to provide patients with gas at a suitable temperature and humidity to maintain the normal function of airway mucosal cilia, prevent excessively thick or thin sputum, and thus reduce the risk of complications such as airway damage, lung infection, and airway obstruction. Traditional humidification techniques mainly rely on preset fixed temperatures or simple settings for constant output, and their humidification process is an open-loop system. This method ignores the influence of individual patient differences, dynamic evolution of the disease, and fluctuations in environmental parameters, making it difficult to achieve precise and personalized humidification treatment. It can easily lead to insufficient or excessive humidification, affecting treatment efficacy and patient comfort.

[0003] For example, a humidification system and control method for a ventilator, with application number CN202111586166.2 and authorization announcement date of 20240329, includes: a ventilator that actively outputs compressed airflow; a humidifier unit equipped with a sterile sealed water tank and a microporous leakage tube sealed to the output end of the sealed water tank; a breathing tube that directionally delivers humidified airflow, with at least a portion of the microporous leakage tube disposed within the breathing tube; the humidifier unit, the ventilator, and the breathing tube are connected by a sealed connector; and a breathing mask that receives the humidified airflow and is connected to the output end of the breathing tube. This invention provides uniform humidification, can precisely control the humidification degree of the airflow within the pipeline, and effectively improves humidification efficiency.

[0004] Traditional humidification control of ventilators is mostly based on presets or simple feedback, lacking synchronous perception and deep integration of multi-source information such as proximal airway temperature and humidity, respiratory waveform, and environmental parameters. It cannot achieve true closed-loop adaptive adjustment. Therefore, it is urgent to design an intelligent ventilator humidification device and method with adaptive adjustment function to solve the above problems. Summary of the Invention

[0005] The purpose of this invention is to provide an intelligent ventilator humidification device and method with adaptive adjustment function to overcome the above-mentioned shortcomings in the prior art.

[0006] To achieve the above objectives, the present invention provides the following technical solution: A smart ventilator humidification device with adaptive adjustment function includes a ventilator body and a humidification module. The humidification module is mounted on the ventilator body and includes a multi-source sensing module, a preprocessing and calculation pre-simulation module, a real-time feedback and AI control module, an energy consumption calculation and regulation early warning module, and a humidification and water replenishment module, wherein: The multi-source sensing module is used to collect the temperature and humidity of the inhaled gas and the respiratory waveform at the proximal end of the patient's airway in real time; The multi-source sensing module is also used to detect the humidifier tank water level, heating plate temperature, pipeline condensate status, and ward environment temperature and humidity. The multi-source sensing module includes a patient-side sensing unit, a device-side sensing unit, and an environment-side sensing unit, wherein: The patient-side sensing unit includes a miniature temperature and humidity sensor, a respiratory waveform sensor, and an end-tidal gas detection sensor, wherein: The miniature temperature and humidity sensor is integrated into the proximal end of the ventilator tubing to monitor the temperature and humidity of the inhaled gas. The respiratory waveform sensor is used to collect respiratory flow and pressure waveforms to identify airflow fluctuations and pressure rises caused by thick sputum. The respiratory waveform sensor identifies waveform jitter characteristics caused by dryness or coughing; The end-tidal gas detection sensor is used to assess airway moisture loss. The device-side sensing unit includes a humidification tank water level sensor, a heating plate temperature sensor, and an optical or weight sensor for monitoring the amount of condensate in the pipeline. The environmental sensing unit includes an environmental temperature and humidity sensor and an atmospheric pressure sensor.

[0007] The preprocessing and computational simulation module is connected to the multi-source sensing module; The preprocessing and calculation simulation module is used to fuse and preprocess the collected data; The preprocessing and calculation simulation module performs high-fidelity simulation based on the constructed digital twin model of the humidification process to predict the changes in airway temperature and humidity, water consumption and energy consumption trends at different time scales in the future. The preprocessing and computational simulation module is configured as follows: The preprocessing and calculation simulation module preprocesses the data collected by the multi-source sensing module, specifically including: Set the differentiated sampling frequency, filtering and noise reduction, outlier removal and data normalization according to the dynamic characteristics of the parameters; The preprocessing and calculation simulation module constructs and runs a digital twin physical model of the humidification process that integrates thermodynamic and mass transfer equations, and is used to calculate the quantitative relationship between heating power, water temperature, evaporation rate, pipeline heat loss and condensate generation. The preprocessing and computational simulation module is also coupled with an LSTM time-series prediction model trained on historical data to perform multi-time-scale predictions, specifically including: Short-term (1-30 minutes) prediction of airway temperature and humidity changes and condensation risk; Mid-term (30-60 minutes) forecast of humidification tank water consumption and energy consumption trends; Long-term (hourly) prediction of the impact of patient activity or changes in condition on humidification needs.

[0008] The real-time feedback and AI control module is connected to the preprocessing and calculation simulation module; The real-time feedback and AI control module is used to dynamically set the target humidification range based on individual patient information and prediction results; The real-time feedback and AI control module also adjusts the heating and humidification output in real time based on the model predictive control algorithm, optimizes efficiency through intake phase pulse compensation, and performs anomaly detection and safety intervention. The real-time feedback and AI control module is configured as follows: The real-time feedback and AI control module has a built-in mode library containing personalized humidification parameters for different disease types, and can dynamically set and load target humidification strategies according to the patient's disease type, current ventilation mode and activity status. The real-time feedback and AI control module adopts a model predictive control algorithm and generates and sends control commands to the humidification and water replenishment module based on the prediction results of the preprocessing and calculation pre-simulation module. The control commands include an inhalation phase pulse power compensation strategy to optimize the instantaneous humidification efficiency under high flow support. The real-time feedback and AI control module also implements a closed-loop control of prediction-execution-correction, compares the sensor feedback with the predicted value in real time, and triggers strategy readjustment when the deviation exceeds the limit; The real-time feedback and AI control module also presets multiple controllable modes such as comfort priority, balanced energy saving, and energy saving priority, and has abnormal detection and early warning functions for insufficient humidification, excessive humidification, excessive condensate, and equipment failure.

[0009] The energy consumption calculation and control early warning module is connected to the real-time feedback and AI control module and the preprocessing and calculation pre-simulation module. The energy consumption calculation and control early warning module is used to establish an energy consumption model. Under the constraint of humidification demand, it performs multi-objective optimization, generates and executes energy-saving strategies including predictive preheating and peak-valley electricity optimization, and issues early warnings for energy efficiency anomalies or maintenance needs. The energy consumption calculation and control early warning module is configured as follows: The energy consumption calculation and control early warning module performs multi-objective optimization. The objective function is to minimize the total energy consumption while satisfying the airway temperature and humidity safety constraints, and to accurately calculate the energy consumption distribution of each component such as heating, sensing, and control. The energy consumption calculation and control early warning module generates and executes energy-saving strategies, which include: Based on the hospital's energy management strategy of "valley leveling" or "peak shaving and valley filling" during peak and valley periods; Predictive preheating or power adaptive looping strategies are executed based on predicted changes in humidification demand. A strategy of integrating with the hospital's energy management system to participate in demand response.

[0010] The humidification and water replenishment module is connected to the real-time feedback and AI control module and the energy consumption calculation and regulation early warning module. The humidification and water replenishment module receives control commands and precisely executes humidification through an adjustable power heating plate and atomizing components, thereby achieving automatic maintenance of the water level in the humidification tank and management of sterile water.

[0011] The humidification and water replenishment module includes an intelligent water level control unit, wherein: The intelligent water level control unit uses metal wire electrodes for water level sampling. Through a control circuit based on the AT89C2051 microcontroller and the K117 high input impedance field-effect transistor, it connects to an external sterile water storage device to realize automatic water replenishment, automatic shutdown when full, low water level alarm, and anti-dry burning protection functions for the humidification tank.

[0012] The humidification and water replenishment module also includes a condensate water management unit, wherein; The condensate management unit monitors the amount of condensate in the breathing tubing using optical or weight sensors, and triggers an automatic drainage mechanism when the amount of condensate reaches a preset threshold.

[0013] The humidification and water replenishment module is configured to implement a humidification fluid replacement cycle management strategy, which extends the safe replacement cycle of the humidification fluid to days based on the monitoring and calculation of humidification water pollution factors.

[0014] In another embodiment of the present invention, the real-time feedback and AI control module and the energy consumption calculation and regulation early warning module are communicatively connected to the remote monitoring system, which can transmit humidification status, early warning information, energy consumption data and equipment status to the nurse station or cloud platform in real time, and can receive instructions from the hospital energy management system.

[0015] A humidification method for an intelligent ventilator with adaptive adjustment function includes the following steps: Step 1. Device initialization and multi-source data synchronous acquisition: Start the ventilator, load the personalized humidification parameter mode library corresponding to the patient's disease type and ventilation mode, initialize the target humidification range, safety threshold and operating mode, and synchronously acquire patient-side data, device-side data and environmental-side data at differentiated sampling frequencies through the multi-source sensing module; Step 2. Data Fusion and Multi-Time-Scale Digital Twin Computation: The collected multi-source data is preprocessed by filtering, denoising, outlier removal and normalization. The preprocessed data is then input into a digital twin model of the humidification process that integrates thermodynamic and mass transfer physical equations. This model is then coupled with an LSTM time-series prediction model trained on historical data for high-fidelity simulation and multi-time-scale prediction. Step 3. Personalized target setting and model predictive control-based strategy generation: Based on the patient's real-time status and the prediction results of Step 2, the target humidification parameter range is dynamically set or adjusted. Based on the model predictive control algorithm, the system minimizes total energy consumption while meeting the target humidification parameter range and equipment safety constraints. Multi-objective optimization is performed to generate an instantaneous control strategy that includes the heating plate power adjustment curve, atomization output command and inspiratory phase pulse power compensation scheme for a period of time in the future. At the same time, an energy-saving scheduling strategy that is linked to the peak and valley periods of hospital electricity consumption is generated. Step 4. Control Strategy Verification and Energy-Saving Optimization Execution: The generated real-time control strategy and energy-saving scheduling strategy are virtually simulated and verified in the digital twin model. After verification, the control command is sent to the execution unit. The execution unit precisely controls the humidification process by adjusting the power of the heating plate and the output of the atomizing component, and performs pulsed power compensation in the inspiratory phase to optimize the instantaneous humidification efficiency under high flow rate ventilation. At the same time, the energy-saving scheduling strategy and the demand response strategy of communication and coordination with the hospital energy management system are executed. Step 5. Intelligent water level maintenance, condensate management and closed-loop correction: The intelligent water level control unit, based on the AT89C2051 microcontroller and the K117 high input impedance field-effect transistor, automatically replenishes water from the sterile water storage device to the humidification tank according to the calculated water consumption trend, realizing automatic shutdown when the water is full and low water level warning. At the same time, the condensate volume in the pipeline is monitored by sensors. When the threshold is reached, the automatic discharge mechanism is triggered. The sensor feedback data and the predicted value of the digital twin model are compared in real time. If the deviation exceeds the preset threshold, the control strategy is recalculated and adjusted to achieve closed-loop control. Step Six. Status Monitoring, Remote Interaction, and Model Iteration: Continuously monitor the humidification effect, system energy consumption, and equipment operating status. Real-time detection and early warning are provided for insufficient humidification, excessive humidification, excessive condensate, abnormal energy efficiency, and equipment failure. Humidification status data, early warning information, energy consumption reports, and equipment logs are transmitted to the remote monitoring system in real time, and remote intervention commands are received. Based on the calculation and assessment of the risk of humidified water pollution, the system dynamically manages and prompts the safe replacement cycle of the humidification liquid. At the same time, the device records operating data and updates the digital twin model and prediction model through online learning to optimize the personalized strategy library.

[0016] In the above technical solution, the present invention provides an intelligent ventilator humidification device and method with adaptive adjustment function, the beneficial effects of which are: (1) This invention collects multi-dimensional patient data in real time and uses a digital twin model to predict humidification trends. Based on this, the output is dynamically adjusted. This design changes the traditional fixed preset mode and realizes closed-loop adaptive adjustment. Combined with strategies such as pulse compensation, it optimizes humidification efficiency and ensures the continuous best humidification state, thereby effectively reducing the risk of complications and improving treatment safety and patient comfort.

[0017] (2) By establishing an energy consumption model, this invention can intelligently optimize the system energy consumption while ensuring the humidification effect. It can automatically adjust the power according to the predicted demand and formulate energy-saving strategies in combination with the peak and valley periods of hospital electricity consumption. It can even be linked with the energy management system, which significantly reduces the energy consumption and electricity cost of equipment operation and realizes green and economical operation.

[0018] (3) The present invention realizes functions such as automatic water level maintenance and automatic condensate management, and can intelligently assess the humidification fluid replacement cycle. The system has comprehensive abnormal detection and early warning capabilities, and supports remote data transmission and monitoring, which greatly reduces the frequency of manual operation and intervention by medical staff, reduces the risk of human error, and improves the safety and efficiency of clinical management.

[0019] (4) The present invention has the ability to iteratively optimize through online learning, enabling the system to continuously adapt to different patient groups and clinical scenarios, becoming more and more "intelligent" with use. At the same time, the device, as a network node, achieves bidirectional communication with the remote monitoring system and the hospital energy management system. This not only facilitates centralized monitoring and remote intervention by medical staff, but also provides valuable real-time data streams for equipment performance analysis, refined energy management and clinical research at the hospital level, which helps to promote the digital and intelligent transformation of respiratory therapy management. Attached Figure Description

[0020] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments recorded in this invention. For those skilled in the art, other drawings can be obtained based on these drawings.

[0021] Figure 1 This is a three-dimensional structural schematic diagram of an embodiment of an intelligent ventilator humidification device and method with adaptive adjustment function according to the present invention; Figure 2 This is a schematic diagram of the humidification component structure provided in an embodiment of the intelligent ventilator humidification device and method with adaptive adjustment function of the present invention; Figure 3 This is a flowchart illustrating an embodiment of the intelligent ventilator humidification device and method with adaptive adjustment function according to the present invention. Detailed Implementation

[0022] To enable those skilled in the art to better understand the technical solution of the present invention, the present invention will be further described in detail below with reference to the accompanying drawings.

[0023] like Figure 1-2 As shown in the figure, an intelligent ventilator humidification device with adaptive adjustment function provided by an embodiment of the present invention includes a ventilator body 1 and a humidification module 2. The humidification module 2 is no longer a simple heating and humidification unit, but an intelligent system integrating sensing, calculation, decision-making and control. The humidification module 2 is disposed on the ventilator body 1 and includes a multi-source sensing module 210, a preprocessing and calculation pre-simulation module 220, an instant feedback and AI control module 230, an energy consumption calculation and regulation early warning module 240, and a humidification and water replenishment module 250, wherein: The multi-source sensing module 210 is used to collect the temperature and humidity of the inhaled gas and the respiratory waveform in the proximal airway of the patient in real time; The multi-source sensing module 210 is also used to detect the humidifier tank water level, heating plate temperature, pipeline condensate status, and ward environment temperature and humidity. The multi-source sensing module 210 includes a patient-side sensing unit, a device-side sensing unit, and an environment-side sensing unit, wherein: The patient-side sensing unit includes a miniature temperature and humidity sensor, a respiratory waveform sensor, and an end-expiratory gas detection sensor, among which: Miniature temperature and humidity sensors are integrated into the patient's end of the ventilator tubing to monitor the temperature and humidity of the inhaled gas; The respiratory waveform sensor is used to collect respiratory flow and pressure waveforms to identify airflow fluctuations and pressure rises caused by thick sputum. The respiratory waveform sensor identifies waveform jitter characteristics caused by dryness or coughing; End-tidal gas detection sensors are used to assess airway moisture loss. The equipment-side sensing unit includes a humidification tank water level sensor, a heating plate temperature sensor, and an optical or weight sensor for monitoring the amount of condensate in the pipeline. The environmental sensing unit includes an environmental temperature and humidity sensor and an atmospheric pressure sensor.

[0024] The preprocessing and computational simulation module 220 is connected to the multi-source sensing module 210; The preprocessing and calculation simulation module 220 is used to fuse and preprocess the collected data; The preprocessing and calculation simulation module 220 performs high-fidelity simulation based on the constructed digital twin model of the humidification process to predict the changes in airway temperature and humidity, water consumption and energy consumption trends at different time scales in the future. The preprocessing and computational simulation module 220 is configured as follows: The preprocessing and computational simulation module 220 preprocesses the data collected by the multi-source sensing module 210, specifically including: Set the differentiated sampling frequency, filtering and noise reduction, outlier removal and data normalization according to the dynamic characteristics of the parameters; The preprocessing and calculation simulation module 220 constructs and runs a digital twin physical model of the humidification process that integrates thermodynamic and mass transfer equations, and is used to calculate the quantitative relationship between heating power, water temperature, evaporation rate, pipeline heat loss and condensate generation. The preprocessing and computational simulation module 220 is also coupled with an LSTM time series prediction model trained on historical data to perform multi-time-scale predictions, specifically including: Short-term (1-30 minutes) prediction of airway temperature and humidity changes and condensation risk; Mid-term (30-60 minutes) forecast of humidification tank water consumption and energy consumption trends; Long-term (hourly) prediction of the impact of patient activity or changes in condition on humidification needs.

[0025] The real-time feedback and AI control module 230 is connected to the preprocessing and calculation simulation module 220; The real-time feedback and AI control module 230 is used to dynamically set the target humidification range based on individual patient information and prediction results; The real-time feedback and AI control module 230 also adjusts the heating and humidification output in real time based on the model predictive control algorithm, optimizes efficiency through intake phase pulse compensation, and performs anomaly detection and safety intervention. The real-time feedback and AI control module 230 is configured as follows: The real-time feedback and AI control module 230 has a built-in mode library containing personalized humidification parameters for different disease types, and can dynamically set and load target humidification strategies according to the patient's disease type, current ventilation mode and activity status. The real-time feedback and AI control module 230 adopts a model predictive control algorithm and generates and sends control commands to the humidification and water replenishment module 250 based on the prediction results of the preprocessing and calculation pre-simulation module 220. The control commands include an inhalation phase pulse power compensation strategy to optimize the instantaneous humidification efficiency under high flow support. The real-time feedback and AI control module 230 also implements a prediction-execution-correction closed-loop control, compares sensor feedback with predicted values ​​in real time, and triggers strategy readjustment when the deviation exceeds the limit; The real-time feedback and AI control module 230 also presets multiple controllable modes such as comfort priority, balanced energy saving, and energy saving priority, and has abnormal detection and early warning functions for insufficient humidification, excessive humidification, excessive condensate, and equipment failure.

[0026] The energy consumption calculation and control early warning module 240 is connected to the real-time feedback and AI control module 230 and the preprocessing and calculation simulation module 220; The energy consumption calculation and control early warning module 240 is used to establish an energy consumption model. Under the constraint of meeting humidification demand, it performs multi-objective optimization, generates and executes energy-saving strategies including predictive preheating and peak-valley electricity optimization, and issues early warnings for energy efficiency anomalies or maintenance needs. The energy consumption calculation and control early warning module 240 is configured as follows: The energy consumption calculation and control early warning module 240 performs multi-objective optimization. The objective function is to minimize the total energy consumption while meeting the airway temperature and humidity safety constraints, and to accurately calculate the energy consumption distribution of each component such as heating, sensing, and control. The energy consumption calculation and control early warning module 240 generates and executes energy-saving strategies, which include: Based on the hospital's energy management strategy of "valley leveling" or "peak shaving and valley filling" during peak and valley periods; Predictive preheating or power adaptive looping strategies are executed based on predicted changes in humidification demand. A strategy of integrating with the hospital's energy management system to participate in demand response.

[0027] The humidification and water replenishment module 250 is connected to the real-time feedback and AI control module 230 and the energy consumption calculation and regulation early warning module 240; The humidification and water replenishment module 250 receives control commands and precisely performs humidification through an adjustable power heating plate and atomizing components, and realizes automatic maintenance of the water level in the humidification tank and sterile water management.

[0028] The humidification and water replenishment module 250 includes an intelligent water level control unit, wherein: The intelligent water level control unit uses metal wire electrodes to sample the water level. Through a control circuit based on the AT89C2051 microcontroller and the K117 high input impedance field-effect transistor, it connects to an external sterile water storage device to realize automatic water replenishment, automatic shutdown when full, low water level alarm, and anti-dry burning protection functions for the humidification tank.

[0029] The humidification and water replenishment module 250 also includes a condensate management unit, wherein; The condensate management unit monitors the amount of condensate in the breathing tubing using optical or weight sensors and triggers an automatic drainage mechanism when the amount of condensate reaches a preset threshold.

[0030] The humidification and water replenishment module 250 is configured to implement a humidification fluid replacement cycle management strategy, which extends the safe replacement cycle of the humidification fluid to 7 days based on the monitoring and calculation of humidification water pollution factors.

[0031] In another embodiment of the present invention, the real-time feedback and AI control module 230 and the energy consumption calculation and regulation early warning module 240 are communicatively connected to the remote monitoring system, which can transmit the humidification status, early warning information, energy consumption data and equipment status to the nurse station or cloud platform in real time, and can receive instructions from the hospital energy management system.

[0032] In one embodiment of the present invention, the refined design of the multi-source sensing module 210 is the foundation for its accurate sensing. It includes three sensing units: a patient-side sensing unit, a device-side sensing unit, and an environmental sensing unit. The patient-side sensing unit is the key innovation, integrating a miniature temperature and humidity sensor near the patient to directly monitor the core parameters of the patient's inhaled gas. The respiratory waveform sensor (usually integrated into the ventilator's flow / pressure sensor) is not only used for routine monitoring but is also configured to identify changes in airway resistance caused by thick sputum (manifested as specific waveform distortion or pressure rise characteristics) and abnormal waveform jitter caused by dryness or coughing. The end-expiratory gas detection sensor (optional, such as a miniature humidity or condensation sensor) is used to indirectly assess airway moisture loss. The device-side sensing unit includes a continuous water level sensor in the humidifier, a high-precision temperature sensor on the heating plate, and an optical sensor or miniature weighing sensor located at the low point of the tubing to monitor condensation accumulation. The environmental sensing unit is deployed near the device or integrated into the housing to monitor ambient temperature, humidity, and atmospheric pressure, which significantly affect tubing heat dissipation and evaporation rate.

[0033] In another embodiment of the present invention, the configuration of the preprocessing and computational simulation module 220 embodies the idea of ​​integrating data-driven and physical models. First, it performs specialized preprocessing on multi-source asynchronous and heterogeneous data: setting differentiated sampling frequencies based on the dynamic characteristics of the parameters (e.g., high-frequency sampling of patient temperature and humidity, and low-frequency sampling of water level), using digital filtering (e.g., Kalman filtering) for noise reduction, eliminating outliers through statistical methods (e.g., the 3σ criterion), and performing normalization to provide high-quality data for subsequent calculations. Second, it constructs and runs a digital twin physical model of the humidification process. This model, based on the energy and mass conservation equations, accurately describes the process from heat generation by the heating plate, water evaporation, gas heating and humidification, ... The complete physical process from pipeline heat and moisture exchange to condensate generation can be quantitatively analyzed to understand the interactions between various variables. More importantly, this module couples a long short-term memory network time-series prediction model trained on a large amount of historical operating data. The physical model provides mechanistic constraints and real-time status, while the LSTM model learns complex nonlinear time-series relationships. The combination of the two enables high-fidelity predictions across multiple time scales: short-term (1-30 minutes) predictions are used for immediate control, such as preventing condensation or humidity overshoot; medium-term (30-60 minutes) predictions are used for water consumption and energy consumption forecasting to guide energy-saving scheduling; and long-term (several hours) predictions are used to assess the macroscopic impact of patient activities (such as suctioning and turning over) or potential changes in the patient's condition on humidification needs.

[0034] In another embodiment of the present invention, the real-time feedback and AI control module 230 is the core of realizing personalized and adaptive control. It has a built-in, continuously expanding library of personalized humidification parameter modes, storing optimized humidification strategy templates for different diseases (such as ARDS and COPD) and different ventilation modes (such as VCV, PCV, and high-flow). The system can automatically load the initial template based on patient information and dynamically fine-tune the target range according to the patient's current activity status (resting, active, coughing). This module employs a model predictive control algorithm: in each control cycle, it uses a digital twin model to predict the system's output trajectory under various potential control inputs over a future period, and then optimizes it by tracking the target humidification range, minimizing energy consumption, and avoiding condensation. A series of optimal future control variables (heating plate power, atomization amount) are solved, and the first control variable is applied to the actual system. In particular, to address the problem of short gas residence time during high-flow ventilation, a "pulse power compensation strategy for the intake phase" is specifically designed. This involves briefly increasing the heating power during the intake phase to achieve higher instantaneous evaporation efficiency during peak airflow. The entire control process is a closed loop of "prediction-execution-correction": the sensor feedback value is compared with the model prediction value in real time. If the deviation exceeds the allowable range, the model parameter correction or control strategy readjustment is triggered. In addition, the module presets multiple controllable modes such as "comfort priority", "balanced energy saving", and "energy saving priority", and integrates real-time detection and graded early warning logic for insufficient humidification, excessive humidification, excessive condensate, and equipment failure.

[0035] In another embodiment of the present invention, the energy consumption calculation and control early warning module 240 is specifically responsible for energy-side management and optimization. It establishes a detailed energy consumption decomposition model and can accurately calculate the energy consumption of each sub-component such as the heating plate, sensor, control circuit, and water replenishment device. Its core task is to perform multi-objective optimization. The objective function is usually set to minimize the total energy consumption of the system under the premise of strictly meeting the airway temperature and humidity safety constraints. Based on this, it can generate and execute a variety of advanced energy-saving strategies: 1) Link with the hospital's time-of-use electricity pricing or energy management system to implement energy management strategies such as "valley leveling" (appropriately increasing the base temperature to store heat during periods of low electricity prices) or "peak shaving and valley filling" (avoiding high-power operation during peak electricity consumption periods); 2) Based on the prediction of humidification demand by the preprocessing module, a "predictive preheating" strategy is implemented. For example, the power is increased gradually and in advance before the expected increase in demand due to patient activity, so as to avoid emergency high-power heating; or a "power adaptive loop" strategy is implemented to make the power output smoother while ensuring the humidification effect. 3) When connected to the hospital's energy management system, it can be used as a controllable load to participate in demand response (DR) events.

[0036] In another embodiment of the present invention, the intelligent water level control unit in the humidification and water replenishment module 250 is one of the key hardware components for achieving unattended operation. This unit uses a reliable and low-cost metal wire electrode as a water level detection sensor. Its control circuit uses the AT89C2051 microcontroller as the core controller, which is responsible for logic judgment. A switch control circuit composed of a high input impedance field-effect transistor K117 is used to drive the solenoid valve or micro pump connected to the external sterile water storage bottle. The working principle is as follows: the microcontroller determines the water level by detecting the conduction state between the electrodes. When the water level is lower than the low water level electrode, a water replenishment command is triggered to open the solenoid valve / water pump. When the water level reaches the high water level electrode, water replenishment stops. This design realizes the functions of automatic stop when the water is full, low water level audible and visual alarm, and linkage heating plate shutdown to prevent dry burning, ensuring water level safety.

[0037] In another embodiment of the present invention, the humidification and water replenishment module 250 further includes a condensate management unit. This unit monitors the amount of condensate in real time through an optical sensor (monitoring the liquid level) or a miniature weight sensor (monitoring the weight of the condensate tank) located at the lowest point of the breathing tubing. When the monitored amount of condensate reaches a preset safety threshold (such as close to 70% of the tubing capacity), the control unit will trigger an automatic drainage mechanism. This mechanism may automatically open a drainage solenoid valve to lead the condensate to the collection bag, or control a miniature pump to remove the condensate, thereby effectively reducing the risk of aspiration, bacterial growth, and the impact of "water hammer" on the breathing waveform caused by tubing vibration due to excessive condensate, further reducing the nursing burden.

[0038] In another embodiment of the present invention, the humidification and water replenishment module 250 is also configured to implement an innovative humidification fluid replacement cycle management strategy. Traditionally, humidification fluid needs to be replaced every 24 hours to prevent contamination, but this consumes a lot of human resources and consumables. This device models and monitors the main factors of humidification water contamination (such as environmental colony sedimentation, operational contamination, back contamination, etc.), and combines real-time monitoring data such as humidification tank opening time, water replenishment frequency, and ambient air quality to dynamically assess the contamination risk. Under controllable risk conditions, the system can intelligently extend the safe replacement cycle of humidification fluid (e.g., to 7 days) and issue a replacement reminder when the actual risk threshold is reached. This can significantly reduce consumable consumption and nursing workload while ensuring patient safety.

[0039] like Figure 3 The method for humidifying a smart ventilator with adaptive adjustment, as shown, includes the following steps: Step 1. Device initialization and multi-source data synchronous acquisition: Start the ventilator, load the personalized humidification parameter mode library corresponding to the patient's disease type and ventilation mode, initialize the target humidification range, safety threshold and operating mode, and synchronously acquire patient-side data, device-side data and environmental-side data at differentiated sampling frequencies through the multi-source sensing module; Step 2. Data Fusion and Multi-Time-Scale Digital Twin Computation: The collected multi-source data is preprocessed by filtering, denoising, outlier removal and normalization. The preprocessed data is then input into a digital twin model of the humidification process that integrates thermodynamic and mass transfer physical equations. This model is then coupled with an LSTM time-series prediction model trained on historical data for high-fidelity simulation and multi-time-scale prediction. Step 3. Personalized target setting and model predictive control-based strategy generation: Based on the patient's real-time status and the prediction results of Step 2, the target humidification parameter range is dynamically set or adjusted. Based on the model predictive control algorithm, the system minimizes total energy consumption while meeting the target humidification parameter range and equipment safety constraints. Multi-objective optimization is performed to generate an instantaneous control strategy that includes the heating plate power adjustment curve, atomization output command and inspiratory phase pulse power compensation scheme for a period of time in the future. At the same time, an energy-saving scheduling strategy that is linked to the peak and valley periods of hospital electricity consumption is generated. Step 4. Control Strategy Verification and Energy-Saving Optimization Execution: The generated real-time control strategy and energy-saving scheduling strategy are virtually simulated and verified in the digital twin model. After verification, the control command is sent to the execution unit. The execution unit precisely controls the humidification process by adjusting the power of the heating plate and the output of the atomizing component, and performs pulsed power compensation in the inspiratory phase to optimize the instantaneous humidification efficiency under high flow rate ventilation. At the same time, the energy-saving scheduling strategy and the demand response strategy of communication and coordination with the hospital energy management system are executed. Step 5. Intelligent water level maintenance, condensate management and closed-loop correction: The intelligent water level control unit, based on the AT89C2051 microcontroller and the K117 high input impedance field-effect transistor, automatically replenishes water from the sterile water storage device to the humidification tank according to the calculated water consumption trend, realizing automatic shutdown when the water is full and low water level warning. At the same time, the condensate volume in the pipeline is monitored by sensors. When the threshold is reached, the automatic discharge mechanism is triggered. The sensor feedback data and the predicted value of the digital twin model are compared in real time. If the deviation exceeds the preset threshold, the control strategy is recalculated and adjusted to achieve closed-loop control. Step Six. Status Monitoring, Remote Interaction, and Model Iteration: Continuously monitor the humidification effect, system energy consumption, and equipment operating status. Real-time detection and early warning are provided for insufficient humidification, excessive humidification, excessive condensate, abnormal energy efficiency, and equipment failure. Humidification status data, early warning information, energy consumption reports, and equipment logs are transmitted to the remote monitoring system in real time, and remote intervention commands are received. Based on the calculation and assessment of the risk of humidified water pollution, the system dynamically manages and prompts the safe replacement cycle of the humidification liquid. At the same time, the device records operating data and updates the digital twin model and prediction model through online learning to optimize the personalized strategy library.

[0040] The foregoing has only described certain exemplary embodiments of the present invention by way of illustration. Undoubtedly, those skilled in the art can modify the described embodiments in various ways without departing from the spirit and scope of the present invention. Therefore, the foregoing drawings and descriptions are illustrative in nature and should not be construed as limiting the scope of protection of the claims of the present invention.

Claims

1. A smart ventilator humidification device with adaptive adjustment function, comprising a ventilator body (1) and a humidification module (2), characterized in that: The humidification module (2) is mounted on the main body (1) of the ventilator. The humidification module (2) includes a multi-source sensing module (210), a preprocessing and calculation pre-simulation module (220), an instant feedback and AI control module (230), an energy consumption calculation and regulation early warning module (240), and a humidification and water replenishment module (250), wherein: The multi-source sensing module (210) is used to collect the temperature and humidity of the inhaled gas and the respiratory waveform at the proximal end of the patient's airway in real time; The multi-source sensing module (210) is also used to detect the humidifier tank water level, heating plate temperature, pipeline condensate status, and ward environment temperature and humidity. The preprocessing and calculation simulation module (220) is connected to the multi-source sensing module (210). The preprocessing and calculation simulation module (220) is used to fuse and preprocess the collected data; The preprocessing and calculation simulation module (220) performs high-fidelity simulation based on the constructed digital twin model of the humidification process to predict the changes in airway temperature and humidity, water consumption and energy consumption trends at different time scales in the future. The real-time feedback and AI control module (230) is connected to the preprocessing and calculation simulation module (220). The real-time feedback and AI control module (230) is used to dynamically set the target humidification range based on individual patient information and prediction results; The real-time feedback and AI control module (230) also adjusts the heating and humidification output in real time based on the model predictive control algorithm, optimizes efficiency through intake phase pulse compensation, and performs anomaly detection and safety intervention. The energy consumption calculation and control early warning module (240) is connected to the real-time feedback and AI control module (230) and the preprocessing and calculation simulation module (220). The energy consumption calculation and control early warning module (240) is used to establish an energy consumption model, which performs multi-objective optimization under the constraint of humidification demand, generates and executes energy-saving strategies including predictive preheating and peak-valley power consumption optimization, and issues early warnings for energy efficiency anomalies or maintenance needs. The humidification and water replenishment module (250) is connected to the real-time feedback and AI control module (230) and the energy consumption calculation and regulation early warning module (240). The humidification and water replenishment module (250) receives control commands and precisely performs humidification through an adjustable power heating plate and atomizing components, and realizes automatic maintenance of the water level in the humidification tank and sterile water management.

2. The intelligent humidification device for a ventilator with adaptive adjustment function according to claim 1, characterized in that, The multi-source sensing module (210) includes a patient-side sensing unit, a device-side sensing unit, and an environment-side sensing unit, wherein: The patient-side sensing unit includes a miniature temperature and humidity sensor, a respiratory waveform sensor, and an end-tidal gas detection sensor, wherein: The miniature temperature and humidity sensor is integrated into the proximal end of the ventilator tubing to monitor the temperature and humidity of the inhaled gas. The respiratory waveform sensor is used to collect respiratory flow and pressure waveforms to identify airflow fluctuations and pressure rises caused by thick sputum. The respiratory waveform sensor identifies waveform jitter characteristics caused by dryness or coughing; The end-tidal gas detection sensor is used to assess airway moisture loss. The device-side sensing unit includes a humidification tank water level sensor, a heating plate temperature sensor, and an optical or weight sensor for monitoring the amount of condensate in the pipeline. The environmental sensing unit includes an environmental temperature and humidity sensor and an atmospheric pressure sensor.

3. The intelligent humidification device for a ventilator with adaptive adjustment function according to claim 2, characterized in that, The preprocessing and computational simulation module (220) is configured as follows: The preprocessing and calculation simulation module (220) preprocesses the data collected by the multi-source sensing module (210), specifically including: Set the differentiated sampling frequency, filtering and noise reduction, outlier removal and data normalization according to the dynamic characteristics of the parameters; The preprocessing and calculation simulation module (220) constructs and runs a digital twin physical model of the humidification process that integrates thermodynamic and mass transfer equations, and is used to calculate the quantitative relationship between heating power, water temperature, evaporation rate, pipeline heat loss and condensate generation. The preprocessing and computational simulation module (220) is also coupled with an LSTM time series prediction model trained on historical data to perform multi-time scale prediction, specifically including: Short-term (1-30 minutes) prediction of airway temperature and humidity changes and condensation risk; Mid-term (30-60 minutes) forecast of humidification tank water consumption and energy consumption trends; Long-term (hourly) prediction of the impact of patient activity or changes in condition on humidification needs.

4. The intelligent humidification device for a ventilator with adaptive adjustment function according to claim 3, characterized in that, The real-time feedback and AI control module (230) is configured as follows: The real-time feedback and AI control module (230) has a built-in mode library containing personalized humidification parameters for different disease types, and can dynamically set and load target humidification strategies according to the patient's disease type, current ventilation mode and activity status. The real-time feedback and AI control module (230) adopts a model predictive control algorithm and generates and sends control commands to the humidification and water replenishment module (250) based on the prediction results of the preprocessing and calculation pre-simulation module (220). The control commands include an inhalation phase pulse power compensation strategy to optimize the instantaneous humidification efficiency under high flow support. The real-time feedback and AI control module (230) also implements a prediction-execution-correction closed-loop control, compares the sensor feedback with the predicted value in real time, and triggers strategy readjustment when the deviation exceeds the limit; The real-time feedback and AI control module (230) also presets multiple controllable modes such as comfort priority, balanced energy saving and energy saving priority, and has abnormal detection and early warning functions for insufficient humidification, excessive humidification, excessive condensate water and equipment failure.

5. The intelligent humidification device for a ventilator with adaptive adjustment function according to claim 4, characterized in that, The energy consumption calculation and control early warning module (240) is configured as follows: The energy consumption calculation and control early warning module (240) performs multi-objective optimization. The objective function is to minimize the total energy consumption while satisfying the airway temperature and humidity safety constraints, and to accurately calculate the energy consumption distribution of each component such as heating, sensing, and control. The energy consumption calculation and control early warning module (240) generates and executes energy-saving strategies, which include: Based on the hospital's energy management strategy of "valley leveling" or "peak shaving and valley filling" during peak and valley periods; Predictive preheating or power adaptive looping strategies are executed based on predicted changes in humidification demand. A strategy of integrating with the hospital's energy management system to participate in demand response.

6. The intelligent humidification device for a ventilator with adaptive adjustment function according to claim 5, characterized in that, The humidification and water replenishment module (250) includes an intelligent water level control unit, wherein: The intelligent water level control unit uses metal wire electrodes for water level sampling. Through a control circuit based on the AT89C2051 microcontroller and the K117 high input impedance field-effect transistor, it connects to an external sterile water storage device to realize automatic water replenishment, automatic shutdown when full, low water level alarm, and anti-dry burning protection functions for the humidification tank.

7. The intelligent humidification device for a ventilator with adaptive adjustment function according to claim 6, characterized in that, The humidification and water replenishment module (250) also includes a condensate management unit, wherein; The condensate management unit monitors the amount of condensate in the breathing tubing using optical or weight sensors, and triggers an automatic drainage mechanism when the amount of condensate reaches a preset threshold.

8. The intelligent humidification device for a ventilator with adaptive adjustment function according to claim 7, characterized in that, The humidification and water replenishment module (250) is configured to implement a humidification liquid replacement cycle management strategy, which extends the safe replacement cycle of the humidification liquid to 7 days based on the monitoring and calculation of humidification water pollution factors.

9. A smart ventilator humidification device with adaptive adjustment function according to claim 8, characterized in that, The real-time feedback and AI control module (230) and the energy consumption calculation and regulation early warning module (240) are connected to the remote monitoring system and can transmit the humidification status, early warning information, energy consumption data and equipment status to the nurse station or cloud platform in real time, and can receive instructions from the hospital energy management system.

10. A humidification method for an intelligent ventilator with adaptive adjustment function, characterized in that, Includes the following steps: Step 1. Device initialization and multi-source data synchronous acquisition: Start the ventilator, load the personalized humidification parameter mode library corresponding to the patient's disease type and ventilation mode, initialize the target humidification range, safety threshold and operating mode, and synchronously acquire patient-side data, device-side data and environmental-side data at differentiated sampling frequencies through the multi-source sensing module; Step 2. Data Fusion and Multi-Time-Scale Digital Twin Computation: The collected multi-source data is preprocessed by filtering, denoising, outlier removal and normalization. The preprocessed data is then input into a digital twin model of the humidification process that integrates thermodynamic and mass transfer physical equations. This model is then coupled with an LSTM time-series prediction model trained on historical data for high-fidelity simulation and multi-time-scale prediction. Step 3. Personalized target setting and model predictive control-based strategy generation: Based on the patient's real-time status and the prediction results of Step 2, the target humidification parameter range is dynamically set or adjusted. Based on the model predictive control algorithm, the system minimizes total energy consumption while meeting the target humidification parameter range and equipment safety constraints. Multi-objective optimization is performed to generate an instantaneous control strategy that includes the heating plate power adjustment curve, atomization output command and inspiratory phase pulse power compensation scheme for a period of time in the future. At the same time, an energy-saving scheduling strategy that is linked to the peak and valley periods of hospital electricity consumption is generated. Step 4. Control Strategy Verification and Energy-Saving Optimization Execution: The generated real-time control strategy and energy-saving scheduling strategy are virtually simulated and verified in the digital twin model. After verification, the control command is sent to the execution unit. The execution unit precisely controls the humidification process by adjusting the power of the heating plate and the output of the atomizing component, and performs pulsed power compensation in the inspiratory phase to optimize the instantaneous humidification efficiency under high flow rate ventilation. At the same time, the energy-saving scheduling strategy and the demand response strategy of communication and coordination with the hospital energy management system are executed. Step 5. Intelligent water level maintenance, condensate management and closed-loop correction: The intelligent water level control unit, based on the AT89C2051 microcontroller and the K117 high input impedance field-effect transistor, automatically replenishes water from the sterile water storage device to the humidification tank according to the calculated water consumption trend, realizing automatic shutdown when the water is full and low water level warning. At the same time, the condensate volume in the pipeline is monitored by sensors. When the threshold is reached, the automatic discharge mechanism is triggered. The sensor feedback data and the predicted value of the digital twin model are compared in real time. If the deviation exceeds the preset threshold, the control strategy is recalculated and adjusted to achieve closed-loop control. Step Six. Status Monitoring, Remote Interaction, and Model Iteration: Continuously monitor the humidification effect, system energy consumption, and equipment operating status. Real-time detection and early warning are provided for insufficient humidification, excessive humidification, excessive condensate, abnormal energy efficiency, and equipment failure. Humidification status data, early warning information, energy consumption reports, and equipment logs are transmitted to the remote monitoring system in real time, and remote intervention commands are received. Based on the calculation and assessment of the risk of humidified water pollution, the system dynamically manages and prompts the safe replacement cycle of the humidification liquid. At the same time, the device records operating data and updates the digital twin model and prediction model through online learning to optimize the personalized strategy library.

Images