An event-driven closed-loop control method and system for a respiratory care robot

Abstract

The application discloses an event-driven closed-loop control method and system for a respiratory care robot, and relates to the technical field of medical rehabilitation robot control. The method collects and pre-processes respiratory-related physiological signals of a patient in real time, identifies preset respiratory event types from the physiological signals, and includes an inhale start event, an exhale start event and a cough event. In response to the identified respiratory event, a respiratory care control mode corresponding to the respiratory event is triggered, and an initial control instruction is generated to drive an actuator to perform a care action. During the execution of the care action, the physiological signals are used as feedback to modify the control instruction, thereby realizing the linkage of event triggering and closed-loop adjustment. The respiratory care action is driven by the patient's own respiratory event, and is continuously adjusted during the action, thereby improving the synchronization, comfort and sputum removal effectiveness of the human-machine system, and reducing the invalid energy consumption and the risk of patient resistance.

Description

Technical Field

[0001] This invention relates to the field of medical rehabilitation robot control technology, specifically to an event-driven closed-loop control method and system for respiratory care robots. Background Technology

[0002] Respiratory care robots are a type of automated medical device used to assist patients with respiratory insufficiency in airway clearance, respiratory muscle training, and secretion removal. Products include high-frequency chest wall vibration expectoration machines and mechanical inhalation and exhalation assistive devices. These devices apply positive pressure, negative pressure, or high-frequency vibration to the patient's chest wall or airway to simulate physiological coughing or assisted breathing, thereby helping patients clear airway secretions and improve ventilation. They have important application value in the fields of neuromuscular diseases, chronic obstructive pulmonary disease, and intensive care.

[0003] Current control methods for respiratory care robots generally employ time-triggered or fixed-sequence open-loop control strategies. For example, a sputum-clearing vest inflates and deflates periodically at a preset frequency and intensity, failing to sense the patient's spontaneous breathing rhythm and cough impulse. Similarly, a cough suppressant machine executes actions according to a fixed positive-negative pressure sequence, lacking dynamic response to the patient's real-time airway status. These control methods are completely disconnected from the patient's own breathing intentions, failing to achieve on-demand triggering, easily leading to human-machine asynchrony, causing patient discomfort or even resistance, and resulting in treatment interruption. Furthermore, the lack of a closed-loop adjustment mechanism based on physiological signal feedback during nursing actions means that fixed control parameters cannot adaptively adjust to the patient's real-time changes in airway resistance and compliance, making it difficult to guarantee sputum clearance efficiency and ventilation effects, and failing to provide customized assistance strategies for different respiratory stages.

[0004] In summary, existing control methods for respiratory care robots have significant shortcomings in terms of human-machine synchronization, on-demand response capability, and dynamic closed-loop adjustment. There is an urgent need for a control method and system that can trigger nursing actions based on the patient's own respiratory events and continuously make closed-loop corrections based on physiological feedback during the execution of the actions, so as to meet the urgent clinical needs for safe, comfortable, and efficient respiratory care. Summary of the Invention

[0005] To address the common technical problems in existing respiratory care robot control methods, such as the disconnect between control actions and patient breathing intentions leading to human-machine asynchrony, and the lack of closed-loop adjustment based on physiological signal feedback during nursing action execution, resulting in inefficient sputum clearance, this invention designs an event-driven closed-loop control method and system for respiratory care robots. This method constructs a three-layer progressive control architecture at the macro level: event recognition, pattern triggering, and closed-loop correction. Through feature extraction of multimodal physiological signals and state machine logic determination, it identifies inspiratory initiation events, expiratory initiation events, and cough events. Furthermore, it establishes a one-to-one mapping relationship between each respiratory event and inspiratory pressure support mode, expiratory oscillation sputum clearance mode, and cough-assisted sputum clearance mode. During the execution of each control mode, based on the closed-loop correction mechanism, the respiratory care robot achieves a comprehensive technical effect: nursing actions are driven by the patient's own respiratory events; continuous closed-loop correction is performed based on real-time physiological feedback during action execution; and automatic switching to a safety assurance mode is implemented under abnormal conditions.

[0006] To solve the above-mentioned technical problems, the present invention provides the following technical solution: Firstly, an event-driven closed-loop control method for respiratory care robots, the specific steps of which are as follows: S100. Real-time acquisition of the patient's respiratory-related physiological signals and preprocessing of the respiratory-related physiological signals; S200. Based on the preprocessed respiratory-related physiological signals, identify at least one preset respiratory event and output an event type identifier corresponding to the identified respiratory event. The respiratory events include inhalation start event, exhalation start event, and cough event. S300. In response to the event type identifier, a respiratory care control mode corresponding to the event type identifier is triggered, and an initial control command matching the respiratory care control mode is generated. The respiratory care control mode includes an inspiratory pressure support mode corresponding to the inspiratory start event, an expiratory oscillation expectoration mode corresponding to the expiratory start event, and a cough-assisted expectoration mode corresponding to the cough event. S400. Based on the initial control command, drive the actuator of the respiratory care robot to perform the corresponding respiratory care action; S500 During the execution of the respiratory care action, the respiratory-related physiological signals collected in real time are used as feedback signals, and the initial control command is modified based on the feedback signals to adjust the action state of the actuator.

[0007] Furthermore, in S100, the patient's respiratory-related physiological signals are acquired in real time and preprocessed, including: The patient's airway pressure signal is collected by an airway pressure sensor, the patient's inspiratory flow signal is collected by a respiratory flow sensor, and the patient's respiratory electromyography signal is collected by a respiratory electromyography electrode. The airway pressure signal, the inspiratory flow rate signal, and the respiratory electromyography signal are amplified, bandpass filtered, and converted from analog to digital to obtain preprocessed digital physiological signals, which include digital airway pressure signal, digital inspiratory flow rate signal, and digital respiratory electromyography signal.

[0008] Furthermore, in S200, the process of identifying a preset respiratory event is as follows: Feature parameters are extracted from digital airway pressure signals, digital inspiratory flow signals, and digital respiratory electromyography signals. The feature parameters include the root mean square envelope of inspiratory flow, airway pressure, and respiratory electromyography signals. The feature parameters are compared with preset event determination thresholds and timing logic; When the inhalation flow rate exceeds a preset inhalation trigger flow rate threshold and the duration reaches a first preset duration, the inhalation start event is determined to have occurred, and an event type identifier corresponding to the inhalation start event is output. When the inspiratory start event has been identified, and the monitored inspiratory flow rate drops from the peak inspiratory flow of a single inspiratory process corresponding to the inspiratory start event to a preset percentage of the peak inspiratory flow, the expiration start event is determined to have occurred, and an event type identifier corresponding to the expiration start event is output. When the root mean square envelope of the respiratory electromyography signal exceeds a preset electromyography burst threshold, and the derivative of the airway pressure with respect to time exceeds a preset pressure change rate threshold, the cough event is determined to have occurred, and an event type identifier corresponding to the cough event is output.

[0009] Furthermore, in S300, the matching process for the initial control command is as follows: When the event type identifier corresponds to the inhalation start event, the inhalation pressure support mode is triggered to generate an initial control command for controlling the actuator to output a positive pressure airflow with a preset pressure waveform and a preset inhalation duration. When the event type identifier corresponds to the exhalation start event, the exhalation oscillation expectoration mode is triggered, generating an initial control command for controlling the actuator to superimpose a high-frequency oscillating airflow in the patient's airway and drive the expectoration vest worn on the patient's chest and abdomen to perform synchronous high-frequency oscillation. When the event type identifier corresponds to the cough event, the cough-assisted expectoration mode is triggered, generating an initial control command to control the actuator to first apply a first positive pressure airflow, switch to negative pressure suction after a second preset time, and simultaneously start high-frequency oscillation.

[0010] Furthermore, in S500, the initial control command is corrected based on the feedback signal, including during the execution of the inspiratory pressure support mode: The digital airway pressure signal is acquired in real time as a pressure feedback signal. The actual airway pressure represented by the pressure feedback signal is compared with the target pressure value of the preset pressure waveform at the current moment to obtain the pressure deviation. A proportional-integral-derivative control algorithm is used to calculate the adjustment amount based on the pressure deviation, so as to adjust the output pressure of the positive pressure airflow in real time, so that the actual airway pressure tracks the preset pressure waveform.

[0011] Furthermore, the proportional-integral-derivative control algorithm is based on The control quantity output to the actuator is calculated in real time, wherein... For a moment The control quantity output to the proportional regulating valve in the actuator is used to adjust the output pressure of the positive pressure airflow. For a moment The pressure deviation mentioned above, i.e. The target pressure value - the actual airway pressure, This is a preset proportional gain coefficient, used to generate a control component proportionally to the current value of the pressure deviation. The preset integral gain coefficient, The pressure deviation is defined as the time from the start of the respiratory care action to the end of the time interval. Integrating over time is used to eliminate steady-state error. The preset differential gain coefficient, For a moment The rate of change of the pressure deviation with respect to time is used to predict the error trend and provide damping, and the calculated... It is superimposed in real time onto the initial control command to achieve closed-loop correction of the positive pressure airflow output pressure.

[0012] Furthermore, in S500, the current control command is corrected based on the feedback signal, including during the execution of the expiratory oscillation expectoration mode: The digital airway pressure signal and digital inspiratory flow signal are acquired in real time as feedback signals, and the current average airway pressure and the current oscillation amplitude of the high-frequency oscillating airflow are extracted from the feedback signals. The average airway pressure is compared with a preset safe average pressure threshold, and the basal positive pressure level applied to the patient's airway is adjusted according to the comparison result. Based on the comparison between the current oscillation amplitude and the preset safe oscillation range, the oscillation amplitude and oscillation frequency of the high-frequency oscillating airflow are adjusted so that the pressure alternation in the airway is maintained within the preset safe oscillation range. Among them, the adjusted oscillation amplitude is based on Update To calculate the oscillation amplitude of the adjusted high-frequency oscillating airflow, The current oscillation amplitude of the high-frequency oscillating airflow. The average airway pressure is extracted from the digital airway pressure signal. The preset safe average pressure threshold is... The update is set to a preset amplitude adjustment coefficient, such that when the mean airway pressure... Higher than the preset safe average pressure threshold When the high-frequency oscillating airflow is in a certain state, the oscillation amplitude automatically decreases, and vice versa, thereby achieving closed-loop adaptive adjustment of the oscillation amplitude.

[0013] Furthermore, in S500, the current control command is corrected based on the feedback signal, including the following during the execution of the cough-assisted expectoration mode: Real-time monitoring of digital airway pressure signal and digital inspiratory flow signal, and calculation of the rate of decrease of digital airway pressure signal with respect to time and the rate of increase of digital inspiratory flow signal; When the absolute value of the rate of decrease of the digital airway pressure signal exceeds the preset pressure drop threshold and the rate of increase of the digital inspiratory flow signal exceeds the preset flow surge threshold, it is determined that the sputum discharge channel has been established, and the intensity of negative pressure suction is maintained while the intensity of high-frequency oscillation is reduced. When the rate of increase of the digital airway pressure signal within the preset observation window exceeds the preset pressure rise threshold, and the digital inhalation flow signal remains below the preset flow lower limit threshold, it is determined that there is a risk of airway obstruction. In this case, the intensity of the negative pressure suction is reduced, and the intensity of the high-frequency oscillation is increased.

[0014] On the other hand, an event-driven closed-loop control system for respiratory care robots includes: The signal acquisition module is used to acquire the patient's respiratory-related physiological signals in real time and preprocess the respiratory-related physiological signals; An event recognition module, connected to the signal acquisition module, is used to receive preprocessed respiratory-related physiological signals, recognize at least one preset respiratory event, and output an event type identifier corresponding to the recognized respiratory event. An event-driven controller, connected to the event identification module and the signal acquisition module, is used to receive the event type identifier, trigger the corresponding respiratory care control mode and generate an initial control command that matches the respiratory care control mode. During the operation of the respiratory care control mode, it receives respiratory-related physiological signals from the signal acquisition module as feedback, corrects the initial control command and outputs the final control command. An execution module, connected to the event-driven controller, includes at least one actuator for performing respiratory care actions according to the final control command.

[0015] Compared with existing technologies, this event-driven closed-loop control method for respiratory care robots has the following advantages: This invention constructs a two-layer mechanism of event recognition and pattern triggering, using the initiation of inspiration, the initiation of expiration, and coughing events as direct trigger signals for respiratory care actions. Multimodal physiological signals are acquired in real time through airway pressure sensors, respiratory flow sensors, and respiratory electromyography electrodes. Feature parameters of the root mean square envelope of inspiratory flow, airway pressure, and respiratory electromyography signals are extracted from these signals. Based on preset event judgment thresholds and temporal logic, the patient's respiratory intention is accurately identified. This ensures that the timing of each care action is entirely determined by the patient's own physiological intention, rather than mechanically relying on a fixed sequence, thereby improving patient tolerance and treatment compliance.

[0016] In inspiratory pressure support mode, this invention uses a proportional-integral-derivative control algorithm to adjust the control quantity of the proportional control valve in real time based on the pressure deviation between the actual airway pressure and the target pressure value, ensuring that the actual airway pressure accurately tracks the preset pressure waveform. In expiratory oscillation expectoration mode, based on the comparison between the average airway pressure and the preset safe average pressure threshold, the oscillation amplitude of the high-frequency oscillating airflow is adaptively adjusted to keep the airway pressure fluctuations within the preset safe oscillation range. In cough-assisted expectoration mode, the negative pressure suction intensity and high-frequency oscillation intensity are dynamically adjusted to achieve an intelligent balance between expectoration and safety. Based on real-time physiological signals, this invention can adapt to the constantly changing airway mechanical characteristics of the patient, improving expectoration efficiency and the effectiveness of ventilation support.

[0017] Other advantages, objectives and features of the invention will be set forth in part in the description which follows, and in part will be apparent to those skilled in the art from the following examination or study, or may be learned from the practice of the invention. Attached Figure Description

[0018] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the accompanying drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are merely some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without any creative effort.

[0019] Figure 1 A flowchart illustrating the steps of an event-driven closed-loop control method for respiratory care robots; Figure 2 This is a schematic diagram illustrating the mapping and switching between event types and control modes in an embodiment of the present invention; Figure 3 This is a data flow diagram of an event-driven closed-loop control system for a respiratory care robot according to an embodiment of the present invention. Detailed Implementation

[0020] To better understand the above technical solutions, a detailed description of the solutions will be provided below in conjunction with the accompanying drawings and specific embodiments. Obviously, the described embodiments are merely some, not all, of the embodiments of the present invention. All other embodiments obtained by those skilled in the art based on the embodiments of the present invention without creative effort are within the scope of protection of the present invention.

[0021] In the description of this invention, the term "preset" refers to a value or condition set and stored in the controller before system initialization or use based on clinical experience or individualized patient parameters. Its specific value can be adjusted by medical personnel within a safe range. The term "real-time" means that signal acquisition, processing, and control command updates are completed within the same control cycle. The control cycle of this invention is 10ms, ensuring that the system's response delay to changes in the patient's physiological state is less than one cycle. The term "feedback signal" refers to a digital physiological signal acquired by sensors and preprocessed, used to compare with a target value to generate a deviation signal, thereby driving closed-loop correction. The term "actuator" includes a proportional control valve, a positive / negative pressure switching valve, a high-frequency oscillation generator, and a sputum clearance vest actuator, which are controlled by the electrical control quantity (voltage or current) output by the controller. The term "connection" in this invention refers to the ability to transfer signals or energy between two modules or components, including direct electrical connections, indirect coupling through intermediate circuits, and wireless communication connections.

[0022] This embodiment provides an event-driven closed-loop control method for respiratory care robots, such as... Figure 1 As shown, the method specifically includes: S100. Real-time acquisition of the patient's respiratory-related physiological signals and preprocessing of the respiratory-related physiological signals; S200. Based on the preprocessed respiratory-related physiological signals, identify at least one preset respiratory event and output an event type identifier corresponding to the identified respiratory event. The respiratory events include inhalation start event, exhalation start event, and cough event. S300. In response to the event type identifier, a respiratory care control mode corresponding to the event type identifier is triggered, and an initial control command matching the respiratory care control mode is generated. The respiratory care control mode includes an inspiratory pressure support mode corresponding to the inspiratory start event, an expiratory oscillation expectoration mode corresponding to the expiratory start event, and a cough-assisted expectoration mode corresponding to the cough event. S400. Based on the initial control command, drive the actuator of the respiratory care robot to perform the corresponding respiratory care action; S500 During the execution of the respiratory care action, the respiratory-related physiological signals collected in real time are used as feedback signals, and the initial control command is modified based on the feedback signals to adjust the action state of the actuator.

[0023] In step S100 above, the acquisition of respiratory-related physiological signals relies on multiple types of sensors deployed at the patient end. These sensors include, but are not limited to, airway pressure sensors, respiratory flow sensors, and respiratory electromyography electrodes. The airway pressure sensor is located at the airway interface of the respiratory care robot and is used to acquire airway pressure signals at the patient end. These signals reflect the real-time pressure changes within the patient's airway during the respiratory process and nursing actions. The respiratory flow sensor is located in the gas flow path of the respiratory care robot and is used to acquire the patient's inspiratory flow signals. These signals characterize the volume of gas entering the patient's airway per unit time. The respiratory electromyography electrodes are attached to the patient's skin in the diaphragm region or the external intercostal muscle region and are used to acquire the patient's respiratory electromyography signals. These signals reflect the intensity of electrical activity of the patient's respiratory-related muscles.

[0024] In this embodiment, the preprocessing procedure for airway pressure signal, inspiratory flow rate signal, and respiratory electromyography signal is as follows: Signal amplification: The original electrical signals output by the airway pressure sensor, respiratory flow sensor, and respiratory electromyography electrode are amplified by a preamplifier. Bandpass filtering is used to set different bandpass filter parameters for different characteristic frequency ranges of signals. For airway pressure signals and inspiratory flow signals, bandpass filters with cutoff frequencies of 0.05Hz and 10Hz are used to filter out high-frequency noise and DC drift. For respiratory electromyography signals, bandpass filters with cutoff frequencies of 20Hz and 500Hz are used to filter out effective frequency components of electromyography signals while filtering out electrocardiogram interference and power frequency interference. Analog-to-digital conversion (ADC) transmits the amplified and filtered analog signal to the ADC for conversion at a preset sampling frequency. In this embodiment, the sampling frequency is set to 1000Hz and the sampling bit depth is 16 bits. The converted signal is a digital physiological signal, which includes a digital airway pressure signal, a digital inspiratory flow rate signal, and a digital respiratory electromyography signal. The converted digital physiological signal is stored in the controller's buffer in the form of data frames in chronological order for subsequent steps.

[0025] In step S200 above, based on the preprocessed respiratory-related physiological signals, at least one preset respiratory event is identified. The respiratory events include the onset of inhalation, the onset of exhalation, and a cough event. In this embodiment, the identification process includes: Feature parameters are extracted from digital airway pressure signals, digital inspiratory flow signals, and digital respiratory electromyography (EMG) signals. These feature parameters include the root mean square (RMS) envelope of the inspiratory flow, airway pressure, and respiratory EMG signals. The inspiratory flow is obtained directly from the current sampled value of the digital inspiratory flow signal, and the airway pressure is obtained directly from the current sampled value of the digital airway pressure signal. The RMS envelope of the respiratory EMG signals is calculated using a sliding window algorithm. Specifically, with the current sampling point as the center, the digital respiratory EMG signal sampling values ​​within each 50ms time window before and after the current sampling point are taken. The sum of the squares of all sampling values ​​within the window is calculated, divided by the number of sampling points, and then the square root is taken to obtain the RMS envelope value at that moment.

[0026] The extracted feature parameters are compared with preset event determination thresholds and temporal logic to determine the type of the currently occurring respiratory event. The determination process is executed according to the following temporal logic: The process for determining the inspiratory start event is as follows: The digital inspiratory flow rate signal is monitored in real time. When the inspiratory flow rate value represented by the digital inspiratory flow rate signal exceeds the preset inspiratory trigger flow rate threshold, and this exceeding state continues for a first preset duration, the event recognition module determines that an inspiratory start event has occurred. In this embodiment, the inspiratory trigger flow rate threshold is set to 5L / min, and the first preset duration is set to 50ms. This duration condition is used to avoid false triggering caused by noise or slight airflow fluctuations in the patient. When an inspiratory start event is determined to occur, an event type identifier corresponding to the inspiratory start event is generated. This event type identifier is in the form of a digital code.

[0027] The process of determining the start of exhalation event relies on the identified start of inhalation event: after the start of inhalation event has been identified, the change of the digital inhalation flow signal during that inhalation is continuously monitored; during normal inhalation, the inhalation flow rises from zero to the peak inhalation and gradually decreases back to zero; when the monitored inhalation flow drops from the peak inhalation of the inhalation process corresponding to the start of inhalation event to a preset percentage of the peak inhalation, the event recognition module determines that an start of exhalation event has occurred; in this embodiment, the preset percentage is set to 25%, that is, when the inhalation flow drops to 25% of the peak inhalation, the inhalation phase is determined to have ended and the exhalation phase to have begun, and when an start of exhalation event is determined to have occurred, an event type identifier corresponding to the start of exhalation event is generated.

[0028] The process for determining a cough event is as follows: The root mean square envelope of the respiratory electromyography (EMG) signal and the derivative of the airway pressure signal with respect to time are monitored. When the root mean square envelope of the respiratory EMG signal exceeds a preset EMG burst threshold, and the derivative of the airway pressure with respect to time exceeds a preset pressure change rate threshold, the event recognition module determines that a cough event has occurred. In this embodiment, the EMG burst threshold is set to three times the baseline value of the root mean square envelope of the respiratory EMG signal when the patient is breathing calmly. This baseline value is updated and calculated in real time when the patient is not coughing. The pressure change rate threshold is set to 5 cmH2O / s. A cough event is characterized by a sudden and strong contraction of the respiratory muscles, accompanied by a rapid increase in airway pressure. When both conditions are met simultaneously, a cough event can be accurately identified. When a cough event is determined to have occurred, an event type identifier corresponding to the cough event is generated.

[0029] The generated event type identifier is output to the event-driven controller as an interrupt signal or message queue. When multiple breathing events are identified simultaneously within a preset time window, arbitration is performed according to a preset priority rule. In this embodiment, the priority from high to low is as follows: cough event, inhalation start event, and exhalation start event.

[0030] In step S300 above, such as Figure 2 As shown, the event-driven controller receives an event type identifier. Internally, the event-driven controller stores a mapping table between event type identifiers and respiratory care control modes. These respiratory care control modes include an inspiratory pressure support mode corresponding to the start of inhalation event, an expiratory oscillation expectoration mode corresponding to the start of exhalation event, and a cough-assisted expectoration mode corresponding to the cough event. The controller also generates initial control commands matching the respiratory care control modes, including: When the event type identifier received by the event-driven controller corresponds to the inspiratory start event, the event-driven controller triggers the inspiratory pressure support mode. In the inspiratory pressure support mode, the event-driven controller generates an initial control command for controlling the actuator to output a positive pressure airflow with a preset pressure waveform and a preset inspiratory duration. The preset pressure waveform is in the form of a square wave, a ramp wave, or a sine wave. In this embodiment, the preset pressure waveform is a square wave, the preset pressure amplitude is 15 cmH2O, and the preset inspiratory duration is 1.2 s. The initial control command includes a target pressure value, a target inspiratory duration, and a pressure rise slope parameter. According to the initial control command, the actuator immediately outputs a positive pressure airflow when the patient begins inspiratory, assisting the patient in completing the inspiratory process.

[0031] When the event type identifier received by the event-driven controller corresponds to the start of expiration event, the event-driven controller triggers the expiratory oscillation clearance mode. In the expiratory oscillation clearance mode, the event-driven controller generates an initial control command to control the actuator to superimpose a high-frequency oscillating airflow in the patient's airway, and simultaneously drives the clearance vest worn on the patient's chest and abdomen to perform synchronous high-frequency oscillation. The initial control command includes the basal positive pressure level parameter, the high-frequency oscillation frequency parameter, the high-frequency oscillation amplitude parameter, and the clearance vest vibration amplitude parameter. In this embodiment, the basal positive pressure level is set to 5 cmH2O, the high-frequency oscillation frequency is set to 12 Hz, the high-frequency oscillation amplitude is set to 5 cmH2O (peak-to-peak value), and the clearance vest vibration amplitude is set to be linked with the high-frequency oscillation amplitude.

[0032] When the event type identifier received by the event-driven controller corresponds to a cough event, the event-driven controller triggers the cough-assisted expectoration mode. In the cough-assisted expectoration mode, the event-driven controller generates an initial control command to control the actuator to first apply a first positive pressure airflow. This initial control command includes first positive pressure phase parameters, negative pressure suction phase parameters, and high-frequency oscillation parameters. The first positive pressure phase parameters include a first positive pressure value, a second preset duration, and a pressure rise rate. In this embodiment, the first positive pressure value is set to 25 cmH2O, the second preset duration is set to 1.5 s, and the pressure rise rate is set to 50 cmH2O / s. The actuator applies the first positive pressure airflow to the patient's airway within the second preset duration, allowing the patient's lungs to fully inflate and expand. After the second preset duration, the initial control command controls the actuator to switch to negative pressure suction. The negative pressure value for negative pressure suction is set to -40 cmH2O. Throughout the entire process of positive pressure application and negative pressure suction, the initial control command synchronously starts high-frequency oscillation. The high-frequency oscillation frequency is set to 10 Hz, and the high-frequency oscillation amplitude is set to 3 cmH2O.

[0033] In step S400 above, the actuator of the respiratory care robot includes a proportional regulating valve, a positive and negative pressure switching valve, a high-frequency oscillation generator, and a sputum suction vest driver. The actuator receives the initial control command output by the event-driven controller, and after parsing and amplifying the initial control command, drives the corresponding physical actuator. The proportional regulating valve is set in the gas supply pipeline of the respiratory care robot and is used to regulate the pressure of the positive pressure airflow output to the patient's airway. The proportional regulating valve adjusts the valve opening according to the target pressure value in the initial control command so that the output pressure approaches the target value. The positive and negative pressure switching valve is set at the airway interface of the respiratory care robot and is used to switch between positive pressure output and negative pressure suction. The positive and negative pressure switching valve switches the airway from a positive pressure air source to a negative pressure suction source when the second preset time is reached, based on the mode switching signal in the initial control command. The high-frequency oscillation generator is set in the gas flow path of the respiratory care robot to superimpose a high-frequency alternating airflow on the basic airflow. Based on the high-frequency oscillation frequency and amplitude parameters in the initial control command, the high-frequency oscillation generator drives the high-frequency oscillation diaphragm to reciprocate, generating high-frequency pressure fluctuations. The sputum clearance vest actuator is connected to the sputum clearance vest worn on the patient's chest and abdomen to drive the airbag of the sputum clearance vest to periodically inflate and deflate. Based on the vibration amplitude parameters of the sputum clearance vest in the initial control command, the sputum clearance vest actuator controls the opening duration and frequency of the inflation and deflation valves.

[0034] During the execution of respiratory care actions by the actuator, the actual operating parameters of the actuator are monitored in real time by built-in sensors, and the monitoring results are fed back to the event-driven controller.

[0035] In step S500 above, the closed-loop correction process executes different correction strategies according to the current respiratory care control mode, specifically: Closed-loop correction process in inspiratory pressure support mode During the execution of the inspiratory pressure support mode, the event-driven controller acquires the digital airway pressure signal in real time as a pressure feedback signal. The event-driven controller compares the actual airway pressure represented by the pressure feedback signal with the target pressure value of the preset pressure waveform at the current moment to obtain the pressure deviation.

[0036] The proportional-integral-derivative control algorithm is adopted to calculate the adjustment amount based on the pressure deviation, so as to adjust the output pressure of the positive pressure airflow in real time, so that the actual airway pressure tracks the preset pressure waveform.

[0037] Proportional-integral-derivative control algorithm based on The control quantity output to the actuator is calculated in real time, whereby... For a moment The control signal output to the proportional control valve in the actuator is a voltage signal (V), which is used to drive the valve core displacement of the proportional control valve to regulate the output pressure of the positive pressure airflow. For a moment The pressure deviation, i.e. The value is the difference between the target pressure and the actual airway pressure. In this embodiment, time t is measured in seconds (s) and is counted from the start of the current inspiratory pressure support mode. The preset proportional gain coefficient, proportional term The control component is generated proportionally to the current value of the pressure deviation; the larger the deviation, the greater the adjustment force, and the smaller the deviation, the smaller the adjustment force. The preset integral gain coefficient, integral term Pressure deviation from the start time of respiratory care actions to time [missing information]. The integral over time is used to eliminate steady-state error. When a persistent static deviation exists in the system, the integral term accumulates until the deviation is eliminated. The preset differential gain coefficient, differential term For a moment The rate of change of pressure deviation with respect to time is used to predict the development trend of error and provide damping. When the pressure deviation increases rapidly, the derivative term generates a reverse control component to suppress overshoot. The event-driven controller will calculate the... The corrected final control command is superimposed on the initial control command in real time and output to the proportional regulating valve to achieve closed-loop correction of the positive pressure airflow output pressure.

[0038] In actual control, the parameters of the proportional-integral-derivative (PID) control algorithm are adaptively adjusted based on the patient's airway mechanics. When the patient's airway resistance increases, the system increases the flow rate by monitoring the digital inspiratory flow signal in real time. and The coefficient is used to enhance control response and automatically increases when patient airway compliance decreases. The coefficient is used to suppress pressure overshoot.

[0039] Closed-loop correction process in expiratory oscillation expectoration mode During the execution of the expiratory oscillation expectoration mode, the event-driven controller acquires digital airway pressure and digital inspiratory flow signals in real time as feedback signals. The event-driven controller extracts the current average airway pressure and the current oscillation amplitude of the high-frequency oscillating airflow from the feedback signals. The method for extracting the average airway pressure is as follows: the digital airway pressure signal is processed by moving average with a time window of 200ms to obtain the average airway pressure within this time window. After filtering out high-frequency oscillation components, the baseline pressure level within the airway is obtained. The current oscillation amplitude is extracted as follows: for the digital airway pressure signal, the pressure peak and trough values ​​are extracted in each oscillation cycle. Half of the difference between the peak and trough values ​​is calculated as the oscillation amplitude for that cycle. The average of the oscillation amplitudes over five consecutive cycles is then taken to obtain the current oscillation amplitude. The event-driven controller will average airway pressure. With the preset safe average pressure threshold Compare the results and adjust the baseline positive pressure level applied to the patient's airway accordingly. Exceed When, reduce the baseline positive pressure level; when Below At that time, the base positive pressure level is increased. The event-driven controller adjusts the current oscillation amplitude accordingly. Based on the comparison results with the preset safe oscillation range, the oscillation amplitude and frequency of the high-frequency oscillating airflow are adjusted to maintain the pressure alternation within the preset safe oscillation range. In this embodiment, the preset safe oscillation range is... - .

[0040] Adjusted oscillation amplitude based on To be updated, among which, To calculate the oscillation amplitude of the adjusted high-frequency oscillating airflow, This represents the current oscillation amplitude of the high-frequency oscillating airflow. The average airway pressure extracted from the digital airway pressure signal. To preset a safe average pressure threshold, In this embodiment, the preset amplitude adjustment coefficient is used. Set to 0.2, dimensionless, based on So that when the mean airway pressure Higher than the preset safe average pressure threshold At that time, the oscillation amplitude of the high-frequency oscillating airflow Compared to Automatically decreases; conversely, when the mean airway pressure Below the preset safe average pressure threshold hour, Compared to Automatically increases, this adjustment mechanism achieves closed-loop adaptive adjustment of the high-frequency oscillation amplitude, ensuring that the airway pressure changes are always within a safe range during expiratory oscillation expectoration.

[0041] While adjusting the oscillation amplitude, the event-driven controller also adjusts the oscillation frequency based on a comparison between the current oscillation amplitude and a preset safe oscillation range. When the current oscillation amplitude falls below the preset safe oscillation range lower limit, the oscillation frequency is reduced to 8Hz to enhance the oscillation amplitude response, while maintaining the oscillation amplitude adjustment. When the oscillation frequency exceeds the upper limit of the preset safe oscillation range, while maintaining the oscillation amplitude adjustment, the oscillation frequency is increased to 15Hz to improve sputum expectoration efficiency and reduce the pressure peak of a single oscillation.

[0042] The synchronous high-frequency oscillation intensity of the sputum clearance vest is linked to the oscillation amplitude of the high-frequency oscillation airflow within the airway, and the event-driven controller is based on the calculated... The drive control quantity for the sputum-clearing vest is generated according to a preset linear mapping relationship. In this embodiment, when for At that time, the drive control amount of the sputum clearance vest corresponds to an airbag inflation pressure of 50 kPa. When the volume is increased or decreased, the drive control quantity of the sputum-clearing vest increases or decreases proportionally.

[0043] In the cough-assisted expectoration mode, the closed-loop correction process involves the event-driven controller monitoring the digital airway pressure signal and the digital inspiratory flow signal in real time. The event-driven controller calculates the rate of decrease of the digital airway pressure signal with respect to time and the rate of increase of the digital inspiratory flow signal. The rise rate of the digital inspiratory flow rate signal is obtained by differentiating the digital airway pressure signal. The event-driven controller obtains the data by differentiating the digital inhalation flow signal and then executes judgment and adjustment logic based on the calculated parameter values. When the absolute value of the rate of decrease of the digital airway pressure signal exceeds a preset pressure drop threshold, and the rate of increase of the digital inspiratory flow signal exceeds a preset flow surge threshold, the event-driven controller determines that a sputum discharge channel has been established. In this embodiment, the pressure drop threshold is set to 20 cmH2O / s, and the flow surge threshold is set to 15 L / min / s. The establishment of a sputum discharge channel indicates that large sputum clumps in the patient's airway have been loosened and have begun to move outward, resulting in a sharp decrease in airway resistance. At this time, the event-driven controller maintains the intensity of negative pressure suction unchanged, while reducing the intensity of high-frequency oscillation to 50% of its current value. Maintaining the intensity of negative pressure suction ensures that sputum continues to be discharged under negative pressure. Reducing the intensity of high-frequency oscillation avoids damage to the airway mucosa due to excessive oscillation when airway resistance has already decreased.

[0044] When the rate of increase of the digital airway pressure signal within the preset observation window exceeds the preset pressure rise threshold, and the digital inspiratory flow signal remains below the preset lower flow threshold, the event-driven controller determines that there is a risk of airway obstruction. In this embodiment, the preset observation window is set to 500ms, the pressure rise threshold is set to 10cmH2O / s, and the lower flow threshold is set to 3L / min. If the flow rate remains below the lower flow threshold for more than 300ms, the risk of airway obstruction is indicated by sputum forming a blockage at the narrow part of the airway, leading to an abnormal increase in airway pressure and a decrease in flow. At this time, the event-driven controller reduces the intensity of negative pressure suction to 60% of the current value, while increasing the intensity of high-frequency oscillation to 150% of the current value. Reducing the intensity of negative pressure suction can prevent airway collapse due to excessive negative pressure suction in the case of airway obstruction, while increasing the intensity of high-frequency oscillation can enhance the breaking and loosening effect on the blocking sputum, helping to relieve airway obstruction.

[0045] Throughout the entire cough-assisted expectoration mode, the event-driven controller continuously executes monitoring, judgment, and adjustment logic with a control cycle of 100ms until the cough-assisted expectoration mode ends.

[0046] Through the steps S100 to S500 described above, the event-driven closed-loop control method for respiratory care robots provided in this embodiment achieves a comprehensive control effect: nursing actions are triggered by the patient's own respiratory events, continuous closed-loop correction is performed based on real-time physiological feedback during the execution of the actions, and the adjustment strategy is automatically switched under abnormal conditions.

[0047] In a preferred embodiment, this embodiment provides an event-driven closed-loop control system for respiratory care robots, such as... Figure 3 As shown, the system includes a signal acquisition module, an event recognition module, an event-driven controller, and an execution module, wherein: The signal acquisition module is used to acquire the patient's respiratory-related physiological signals in real time and preprocess the respiratory-related physiological signals; An event recognition module, connected to the signal acquisition module, is used to receive preprocessed respiratory-related physiological signals, recognize at least one preset respiratory event, and output an event type identifier corresponding to the recognized respiratory event. An event-driven controller, connected to the event identification module and the signal acquisition module, is used to receive the event type identifier, trigger the corresponding respiratory care control mode and generate an initial control command that matches the respiratory care control mode. During the operation of the respiratory care control mode, it receives respiratory-related physiological signals from the signal acquisition module as feedback, corrects the initial control command and outputs the final control command. An execution module, connected to the event-driven controller, includes at least one actuator for performing respiratory care actions according to the final control command.

[0048] In this embodiment, the signal acquisition module continuously acquires the patient's respiratory-related physiological signals at a sampling frequency of 1000Hz and simultaneously outputs the digital physiological signals to the event recognition module and the event-driven controller. The event recognition module performs feature extraction and event determination on the received digital physiological signals every 50ms. The event-driven controller operates in an event-driven manner when no control mode is triggered, and only triggers the corresponding control mode upon receiving an event type identifier. During the execution of the control mode, the event-driven controller operates continuously in a closed-loop control manner, with each control cycle lasting 10ms. Within each control cycle, it reads the latest feedback signal and updates the final control command output to the execution module.

[0049] The above description is merely a preferred embodiment of the present invention and is not intended to limit the present invention in any way. Although the present invention has been disclosed above with reference to preferred embodiments, it is not intended to limit the present invention. Any person skilled in the art can make some modifications or alterations to the above-disclosed technical content to create equivalent embodiments without departing from the scope of the present invention. Any simple modifications, equivalent changes and alterations made to the above embodiments based on the technical essence of the present invention without departing from the scope of the present invention shall still fall within the scope of the present invention.

Claims

1. An event-driven closed-loop control method for respiratory care robots, characterized in that, The specific steps of this method are as follows: S100. Real-time acquisition of the patient's respiratory-related physiological signals and preprocessing of the respiratory-related physiological signals; S200. Based on the preprocessed respiratory-related physiological signals, identify at least one preset respiratory event and output an event type identifier corresponding to the identified respiratory event. The respiratory events include inhalation start event, exhalation start event, and cough event. S300. In response to the event type identifier, a respiratory care control mode corresponding to the event type identifier is triggered, and an initial control command matching the respiratory care control mode is generated. The respiratory care control mode includes an inspiratory pressure support mode corresponding to the inspiratory start event, an expiratory oscillation expectoration mode corresponding to the expiratory start event, and a cough-assisted expectoration mode corresponding to the cough event. S400. Based on the initial control command, drive the actuator of the respiratory care robot to perform the corresponding respiratory care action; S500 During the execution of the respiratory care action, the respiratory-related physiological signals collected in real time are used as feedback signals, and the initial control command is modified based on the feedback signals to adjust the action state of the actuator.

2. The event-driven closed-loop control method for a respiratory care robot according to claim 1, characterized in that... In step S100, the patient's respiratory-related physiological signals are acquired in real time and preprocessed, including: The patient's airway pressure signal is collected by an airway pressure sensor, the patient's inspiratory flow signal is collected by a respiratory flow sensor, and the patient's respiratory electromyography signal is collected by a respiratory electromyography electrode. The airway pressure signal, the inspiratory flow rate signal, and the respiratory electromyography signal are amplified, bandpass filtered, and converted from analog to digital to obtain preprocessed digital physiological signals, which include digital airway pressure signal, digital inspiratory flow rate signal, and digital respiratory electromyography signal.

3. The event-driven closed-loop control method for a respiratory care robot according to claim 1, characterized in that, In step S200, the process of identifying a preset respiratory event is as follows: Feature parameters are extracted from digital airway pressure signals, digital inspiratory flow signals, and digital respiratory electromyography signals. The feature parameters include the root mean square envelope of inspiratory flow, airway pressure, and respiratory electromyography signals. The feature parameters are compared with preset event determination thresholds and timing logic; When the inhalation flow rate exceeds a preset inhalation trigger flow rate threshold and the duration reaches a first preset duration, the inhalation start event is determined to have occurred, and an event type identifier corresponding to the inhalation start event is output. When the inspiratory start event has been identified, and the monitored inspiratory flow rate drops from the peak inspiratory flow of a single inspiratory process corresponding to the inspiratory start event to a preset percentage of the peak inspiratory flow, the expiration start event is determined to have occurred, and an event type identifier corresponding to the expiration start event is output. When the root mean square envelope of the respiratory electromyography signal exceeds a preset electromyography burst threshold, and the derivative of the airway pressure with respect to time exceeds a preset pressure change rate threshold, the cough event is determined to have occurred, and an event type identifier corresponding to the cough event is output.

4. The event-driven closed-loop control method for a respiratory care robot according to claim 1, characterized in that, In S300, the matching process for the initial control command is as follows: When the event type identifier corresponds to the inhalation start event, the inhalation pressure support mode is triggered to generate an initial control command for controlling the actuator to output a positive pressure airflow with a preset pressure waveform and a preset inhalation duration. When the event type identifier corresponds to the exhalation start event, the exhalation oscillation expectoration mode is triggered, generating an initial control command for controlling the actuator to superimpose a high-frequency oscillating airflow in the patient's airway and drive the expectoration vest worn on the patient's chest and abdomen to perform synchronous high-frequency oscillation. When the event type identifier corresponds to the cough event, the cough-assisted expectoration mode is triggered, generating an initial control command to control the actuator to first apply a first positive pressure airflow, switch to negative pressure suction after a second preset time, and simultaneously start high-frequency oscillation.

5. The event-driven closed-loop control method for a respiratory care robot according to claim 1, characterized in that, In step S500, the initial control command is corrected based on the feedback signal, and during the execution of the inspiratory pressure support mode, the following is included: The digital airway pressure signal is acquired in real time as a pressure feedback signal. The actual airway pressure represented by the pressure feedback signal is compared with the target pressure value of the preset pressure waveform at the current moment to obtain the pressure deviation. A proportional-integral-derivative control algorithm is used to calculate the adjustment amount based on the pressure deviation, so as to adjust the output pressure of the positive pressure airflow in real time, so that the actual airway pressure tracks the preset pressure waveform.

6. The event-driven closed-loop control method for a respiratory care robot according to claim 5, characterized in that, The proportional-integral-derivative control algorithm is based on The control quantity output to the actuator is calculated in real time, wherein... For a moment The control quantity output to the proportional regulating valve in the actuator is used to adjust the output pressure of the positive pressure airflow. For a moment The pressure deviation mentioned above, i.e. The target pressure value - the actual airway pressure, This is a preset proportional gain coefficient, used to generate a control component proportionally to the current value of the pressure deviation. The preset integral gain coefficient, The pressure deviation is defined as the time from the start of the respiratory care action to the end of the time interval. Integral over time, The preset differential gain coefficient, For a moment The rate of change of the pressure deviation with respect to time will be calculated. It is superimposed on the initial control command in real time.

7. The event-driven closed-loop control method for a respiratory care robot according to claim 1, characterized in that, In S500, the current control command is corrected based on the feedback signal, including during the execution of the expiratory oscillation expectoration mode: The digital airway pressure signal and digital inspiratory flow signal are acquired in real time as feedback signals, and the current average airway pressure and the current oscillation amplitude of the high-frequency oscillating airflow are extracted from the feedback signals. The average airway pressure is compared with a preset safe average pressure threshold, and the basal positive pressure level applied to the patient's airway is adjusted according to the comparison result. Based on the comparison between the current oscillation amplitude and the preset safe oscillation range, the oscillation amplitude and oscillation frequency of the high-frequency oscillating airflow are adjusted so that the pressure alternation in the airway is maintained within the preset safe oscillation range. Among them, the adjusted oscillation amplitude is based on Update To calculate the oscillation amplitude of the adjusted high-frequency oscillating airflow, The current oscillation amplitude of the high-frequency oscillating airflow. The average airway pressure is extracted from the digital airway pressure signal. The preset safe average pressure threshold is... This is the preset amplitude adjustment coefficient.

8. The event-driven closed-loop control method for a respiratory care robot according to claim 1, characterized in that, In step S500, the current control command is corrected based on the feedback signal, including the following during the execution of the cough-assisted expectoration mode: Real-time monitoring of digital airway pressure signal and digital inspiratory flow signal, and calculation of the rate of decrease of digital airway pressure signal with respect to time and the rate of increase of digital inspiratory flow signal; When the absolute value of the rate of decrease of the digital airway pressure signal exceeds the preset pressure drop threshold, and the rate of increase of the digital inspiratory flow signal exceeds the preset flow surge threshold, it is determined that the sputum discharge channel has been established, and the intensity of negative pressure suction is maintained. When the rate of increase of the digital airway pressure signal within the preset observation window exceeds the preset pressure rise threshold, and the digital inspiratory flow signal is lower than the preset flow lower limit threshold, it is determined that there is a risk of airway obstruction.

9. An event-driven closed-loop control system for a respiratory care robot, applicable to the event-driven closed-loop control method for a respiratory care robot as described in any one of claims 1-8, characterized in that, The system includes: The signal acquisition module is used to acquire the patient's respiratory-related physiological signals in real time and preprocess the respiratory-related physiological signals; An event recognition module, connected to the signal acquisition module, is used to receive preprocessed respiratory-related physiological signals, recognize at least one preset respiratory event, and output an event type identifier corresponding to the recognized respiratory event. An event-driven controller, connected to the event identification module and the signal acquisition module, is used to receive the event type identifier, trigger the corresponding respiratory care control mode and generate an initial control command that matches the respiratory care control mode. During the operation of the respiratory care control mode, it receives respiratory-related physiological signals from the signal acquisition module as feedback, corrects the initial control command and outputs the final control command. An execution module, connected to the event-driven controller, includes at least one actuator for performing respiratory care actions according to the final control command.

Images