Common entropy balance adaptive control method and system

Abstract

The invention provides a common entropy balance adaptive control method and system, and belongs to the technical field of control systems. The system comprises a multi-mode physiological sensing module, a control decision module, an execution module and a feedback module, the signal output end of the multi-mode physiological sensing module is connected with the signal input end of the control decision module, and the instruction output end of the control decision module is connected with the instruction input end of the execution module. The state feedback end of the execution module is connected with the signal input end of the feedback module. The multi-mode sensing layer collects respiratory related entropy change and energy signals, the control decision layer generates instructions through a feedforward pre-judgment and self-adaptive algorithm, the execution layer adjusts ventilation through a rotary proportional valve, and the feedback layer monitors and corrects parameters. The system realizes co-entropy balance of a respiratory system, reduces invalid entropy increase, promotes autonomous respiration, improves the offline success rate, and provides a new'assistance but not substitution 'normal form for respiratory support.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of control systems, in particular to a co-entropy balance adaptive control method and system. BACKGROUND

[0002] The core significance of co-entropy balance: According to the "five-dimensional coupling" theory of "future dynamics", the human respiratory system is an open dissipative system that needs to maintain a low-entropy steady state (co-entropy balance) through the exchange of matter and energy. The essence is energy flow balance and entropy generation minimization. In normal respiration, the entropy flow of gas exchange dominates, and the invalid entropy increase is controlled at a very low level.

[0003] Co-entropy imbalance problem of respiratory failure patients: Chronic obstructive pulmonary disease patients have increased expiratory resistance and accumulated residual volume due to airway obstruction and decreased lung elasticity, resulting in gas retention. Respiratory muscle fatigue patients have delayed inspiratory initiation due to decreased diaphragm strength, resulting in energy retention and increased invalid work. The core pathology of these patients is co-entropy imbalance of the respiratory system.

[0004] Defects of traditional ventilators: Traditional ventilators cannot achieve co-entropy balance. First, lag control exacerbates entropy increase, and feedback regulation responds only after physiological indicators change, resulting in delayed inspiratory initiation and energy retention. Second, fixed threshold disrupts energy flow, and the fixed assist threshold cannot match the patient's self-breathing work intensity, resulting in loss of low-entropy gas or gas retention. Third, replacement ventilation suppresses self-organization, and long-term replacement of spontaneous breathing can cause the respiratory center to "use and waste", and co-entropy imbalance is prone to recur after the patient is off the ventilator.

[0005] Necessity of the invention: Clinical studies have shown that restoring co-entropy balance is key to the rehabilitation of respiratory failure patients, and traditional ventilators have the above-mentioned defects, so an "assisted rather than replaced" adaptive assist ventilator is needed to solve the co-entropy imbalance problem. SUMMARY

[0006] The present application provides a co-entropy balance adaptive control method and system to solve the problems raised in the background art.

[0007] The specific technical solutions are as follows: A co-entropy balance adaptive control method and system, comprising: a multi-modal physiological perception module, a control decision module, an execution module, and a feedback module, wherein: The signal output end of the multi-modal physiological sensing module is connected with the signal input end of the control decision module, the instruction output end of the control decision module is connected with the instruction input end of the execution module, the state feedback end of the execution module is connected with the signal input end of the feedback module, the signal output end of the feedback module is connected with the signal input end of the control decision module, the multi-modal physiological sensing module collects the energy flow signal and the entropy change signal of the respiratory system and transmits them to the control decision module, the control decision module generates the assistance instruction based on the received signals through the feedforward prediction and the adaptive entropy balance algorithm and transmits the assistance instruction to the execution module, the execution module executes the ventilation adjustment action according to the assistance instruction, the feedback module collects the actual operation parameters of the execution module and the real-time entropy change index of the respiratory system and feeds them back to the control decision module, and the control decision module dynamically corrects the assistance instruction based on the feedback signal to maintain the total entropy generation of the respiratory system close to zero.

[0008] The adaptive control system for co-entropy balance, wherein the multi-modal physiological sensing module comprises a nasal pressure sensor, a tidal volume flow sensor and an expiratory flow sensor, the nasal pressure sensor is arranged at the nasal interface of the ventilator and is used to collect the pre-inspiration nasal pressure drop signal, the tidal volume flow sensor is arranged at the inspiration channel of the ventilator and is used to collect the tidal volume growth rate signal, and the expiratory flow sensor is arranged at the expiration channel of the ventilator and is used to collect the expiratory flow inflection point signal.

[0009] The adaptive control system for co-entropy balance, wherein the control decision module comprises a feedforward prediction unit and an adaptive entropy balance unit, the feedforward prediction unit is internally provided with a lag comparator and an SR state machine, the lag comparator receives the pre-inspiration nasal pressure drop signal transmitted by the nasal pressure sensor and compares it with a preset set threshold and a reset threshold, and outputs a state control signal to the SR state machine, and the SR state machine outputs an assistance opening or closing instruction according to the state control signal and the reset signal transmitted by the adaptive entropy balance unit.

[0010] The adaptive control system for co-entropy balance, wherein the preset set threshold of the lag comparator is negative 0.5 cmH2O, the preset reset threshold is negative 1 cmH2O, when the pre-inspiration nasal pressure drop signal is less than or equal to the set threshold, the lag comparator outputs an assistance opening trigger signal, when the pre-inspiration nasal pressure drop signal is greater than the reset threshold, the lag comparator outputs an assistance closing trigger signal, and the state switching delay of the SR state machine is not more than 20 ms.

[0011] The co-entropy balance adaptive control system, wherein the adaptive entropy balance unit receives a tidal volume growth rate signal transmitted by the tidal volume flow sensor, compares the tidal volume growth rate signal with a preset reference threshold, transmits a reset signal to the SR state machine when the tidal volume growth rate is less than the preset reference threshold, and maintains the current output state of the SR state machine when the tidal volume growth rate is greater than or equal to the preset reference threshold, and the preset reference threshold is 100 milliliters per second.

[0012] The co-entropy balance adaptive control system, wherein the adaptive entropy balance unit further comprises a hidden Markov model, the hidden Markov model predicts a real-time respiratory frequency by analyzing historical respiratory frequency data collected by the multi-modal physiological perception module, and the adaptive entropy balance unit reduces the assist threshold when the predicted respiratory frequency rises and increases the assist threshold when the predicted respiratory frequency falls.

[0013] The co-entropy balance adaptive control system, wherein the execution module comprises a rotary proportional valve driven by a stepper motor, the opening degree of the rotary proportional valve is adjustable within a range of 0-90 degrees, the response time is not more than 50 milliseconds, and the rotary proportional valve adjusts the valve core angle according to the assist instruction transmitted by the control decision module to change the auxiliary pressure and the ventilation flow.

[0014] The co-entropy balance adaptive control system, wherein the rotary proportional valve comprises an advance assist adjustment mode, a tidal adaptation adjustment mode, and an exhalation synchronization adjustment mode, in the advance assist adjustment mode, the rotary proportional valve adjusts the opening degree to 30 degrees and outputs an auxiliary pressure of 2 cmH2O after receiving the assist opening instruction, in the tidal adaptation adjustment mode, the rotary proportional valve dynamically adjusts the opening degree according to the adjustment instruction of the adaptive entropy balance unit, and in the exhalation synchronization adjustment mode, the rotary proportional valve is pre-opened to an opening degree of 5 degrees after receiving the exhalation flow inflection point signal transmitted by the exhalation flow sensor.

[0015] The present application also provides a co-entropy balance adaptive control method applied to the co-entropy balance adaptive control system, comprising the following steps: Firstly, the multi-modal physiological perception module collects the nasal pressure drop signal, the tidal volume growth rate signal, and the exhalation flow inflection point signal, and transmits the signals to the control decision module after signal conditioning; Secondly, the control decision module analyzes the nasal pressure drop signal through the feedforward prediction unit to trigger the assist opening instruction in advance and transmit it to the execution module; Thirdly, the control decision module analyzes the tidal volume growth rate signal through the adaptive entropy balance unit to dynamically adjust the assist threshold and transmit the adjustment instruction to the execution module; Fourthly, the execution module executes corresponding ventilation adjustment actions according to the assist opening instruction and the adjustment instruction. The fifth step is that the feedback module collects the actual operation parameters of the execution module and real-time total entropy generation and feeds back to the control decision module. The sixth step is that the control decision module generates a correction assistance instruction based on the feedback of the real-time total entropy generation to maintain the co-entropy balance state.

[0016] The co-entropy balance adaptive control method described above, wherein the real-time total entropy generation in the fifth step is calculated by the following method: ; wherein ΔS total is the real-time total entropy generation, k1 is the first entropy increase coefficient, k2 is the second entropy increase coefficient, k3 is the third entropy increase coefficient, ΔP nose is the nasal cavity pressure drop signal, is the tidal volume growth rate signal, dV th is a preset reference threshold, and ΔF exp is the expiratory flow inflection point signal; when the real-time total entropy generation is greater than zero, the control decision module adjusts the assistance threshold or the rotary proportional valve opening, and when the real-time total entropy generation approaches zero, the current parameters are maintained.

[0017] The present application has the following beneficial effects: 1. Achieving co-entropy balance: The multi-modal perception layer comprehensively captures the entropy change and energy flow signals at each stage of respiration, the feedforward prediction of the control decision layer reduces the start-up delay, the adaptive entropy balance matches the autonomous work intensity, and the closed-loop regulation of the execution feedback layer continuously corrects the parameters, ultimately making the total entropy generation of the respiratory system close to zero, restoring the low-entropy steady state, and solving the problem that traditional ventilators cannot achieve co-entropy balance.

[0018] 2. Reducing invalid entropy increase: The response delay of the inspiratory start-up is greatly shortened by the feedforward prediction, avoiding energy retention; the tidal volume adaptive regulation dynamically switches the assistance according to the entropy flow efficiency, avoiding the loss of low-entropy gas caused by excessive assistance or invalid work caused by insufficient assistance, and significantly reducing invalid entropy generation.

[0019] 3. Promoting the recovery of autonomous respiration: The design of "assistance rather than replacement" adjusts the assistance threshold and valve opening as needed, allowing the respiratory center to participate in the regulation of energy flow and entropy balance again, avoiding the degradation of respiratory center function caused by traditional replacement ventilation, and improving the success rate of patient weaning.

[0020] 4. Adapting to different pathological states: The multi-modal perception layer integrates pressure, flow, and frequency signals, and the control decision layer can dynamically adjust the control strategy according to the pathological characteristics of different patients (such as patients with chronic obstructive pulmonary disease and respiratory muscle fatigue), maintain the dynamic balance of entropy flow and invalid entropy increase, and have strong robustness. BRIEF DESCRIPTION OF DRAWINGS

[0021] Figure 1The schematic structural diagram of the co-entropy balance adaptive control system provided by the embodiment of the present application is shown in the figure. Figure 2 The flow chart of the co-entropy balance adaptive control method provided by the embodiment of the present application is shown in the figure. DETAILED DESCRIPTION

[0022] The technical solutions of the present application will be further described below in combination with the accompanying drawings and through specific embodiments.

[0023] The accompanying drawings are only used for exemplary illustration, and the representations are only schematic diagrams, not physical diagrams, and cannot be understood as limitations on the present application; in order to better illustrate the embodiments of the present application, some components of the accompanying drawings may be omitted, enlarged or reduced, and do not represent the size of the actual product; for those skilled in the art, it is understandable that some well-known structures and their descriptions in the accompanying drawings may be omitted.

[0024] The same or similar reference numerals in the accompanying drawings of the embodiments of the present application correspond to the same or similar components; in the description of the present application, it should be understood that if the terms "upper", "lower", "left", "right", "inner", "outer" and the like indicate the orientation or positional relationship shown in the drawings, they are only for the convenience of describing the present application and simplifying the description, and do not indicate or imply that the devices or elements referred to must have a particular orientation, be constructed and operated in a particular orientation, therefore the terms describing the positional relationship in the drawings are only used for exemplary illustration, and cannot be understood as limitations on the present application, for those skilled in the art, the specific meanings of the above terms can be understood according to the specific circumstances.

[0025] In the description of the present application, unless otherwise explicitly specified and limited, if the term "connection" and the like appear to indicate the connection relationship between components, the term should be broadly understood, for example, it can be fixedly connected, or it can be detachably connected, or it can be integrated; it can be mechanically connected, or it can be electrically connected; it can be directly connected, or it can be indirectly connected through an intermediate medium; it can be the communication inside two components or the interaction relationship between two components. For those skilled in the art, the specific meaning of the above terms in the present application can be understood according to the specific circumstances.

[0026] Embodiment 1 The co-entropy balance adaptive control system provided by the present embodiment, as shown in the figure, comprises a multi-modal physiological perception module, a control decision module, an execution module and a feedback module, wherein: Figure 1 The co-entropy balance adaptive control system provided by the present embodiment, as shown in the figure, comprises a multi-modal physiological perception module, a control decision module, an execution module and a feedback module, wherein: The signal output end of the multi-modal physiological sensing module is connected with the signal input end of the control decision module, the instruction output end of the control decision module is connected with the instruction input end of the execution module, the state feedback end of the execution module is connected with the signal input end of the feedback module, the signal output end of the feedback module is connected with the signal input end of the control decision module, the multi-modal physiological sensing module collects energy flow signals and entropy change signals of the respiratory system and transmits them to the control decision module, the control decision module generates a boosting instruction based on the received signals through a feedforward prediction and an adaptive entropy balance algorithm and transmits the boosting instruction to the execution module, the execution module executes a ventilation adjustment action according to the boosting instruction, the feedback module collects actual operating parameters of the execution module and real-time entropy change indicators of the respiratory system and feeds them back to the control decision module, and the control decision module dynamically corrects the boosting instruction based on the feedback signals to maintain a total entropy generation of the respiratory system close to zero.

[0027] The technical solution constructs a closed-loop system of “multi-modal physiological sensing-control decision-execution-feedback”, collects energy flow and entropy change signals related to respiration through a multi-modal module, generates an instruction by a control decision module in combination with a feedforward prediction and an adaptive entropy balance algorithm, implements adjustment by an execution module, returns actual parameters by a feedback module to correct the instruction, and finally maintains a total entropy generation close to zero. Compared with the traditional mode of lag control, fixed threshold or complete replacement of the ventilator, the closed-loop system can comprehensively capture key signals in the respiratory process, predict the respiratory intention in advance and dynamically adapt the respiratory state, avoid energy retention due to lag or entropy imbalance due to fixed adjustment, realize the co-entropy balance of the respiratory system, and take “assistance” as the core to maintain the self-respiratory function rather than inhibit it, thereby reducing the possibility of recurrence of entropy imbalance after offline.

[0028] Specifically, in the embodiment, the multi-modal physiological sensing module includes a nasal cavity pressure sensor, a tidal volume flow sensor and an exhalation flow sensor. The nasal cavity pressure sensor is arranged at a nasal cavity interface of the ventilator and is used to collect a pre-inspiration nasal cavity pressure drop signal. The tidal volume flow sensor is arranged at an inspiration channel of the ventilator and is used to collect a tidal volume growth rate signal. The exhalation flow sensor is arranged at an exhalation channel of the ventilator and is used to collect an exhalation flow inflection point signal.

[0029] The scheme deploys different sensors at key nodes of respiration, respectively collects signals related to inspiration start, entropy flow efficiency and exhalation entropy recovery, accurately acquires core entropy change and energy flow information at each stage of respiration, avoids control deviation caused by one-sided signal collection by a single sensor, provides comprehensive and accurate raw data support for the subsequent control decision module, and ensures the pertinence of subsequent algorithm adjustment.

[0030] Specifically, in the embodiment, the control decision module includes a feedforward prediction unit and an adaptive entropy balance unit. The feedforward prediction unit is internally provided with a hysteresis comparator and an SR state machine. The hysteresis comparator receives a nasal cavity pressure drop signal transmitted by a nasal cavity pressure sensor and compares the signal with a preset set threshold and a reset threshold, and outputs a state control signal to the SR state machine. The SR state machine outputs a boost opening or closing instruction according to the state control signal and a reset signal transmitted by the adaptive entropy balance unit.

[0031] The feedforward prediction unit in the scheme can identify the inhalation intention in advance, avoiding the hysteresis problem that the traditional feedback control needs to wait for the change of the physiological index before responding. The SR state machine cooperates with the signal of the adaptive unit to output the instruction, which can make the boost switch adjustment linked with the actual demand of breathing, reduce the delay in the inhalation start-up stage, avoid the invalid entropy increase caused by the disconnection between the boost and the autonomous breathing rhythm, and make the boost adjustment more in line with the natural rhythm of breathing.

[0032] Specifically, in the embodiment, the preset set threshold of the hysteresis comparator is negative 0.5 cmH2O, and the preset reset threshold is negative 1 cmH2O. When the nasal cavity pressure drop signal is less than or equal to the set threshold, the hysteresis comparator outputs a boost opening trigger signal. When the nasal cavity pressure drop signal is greater than the reset threshold, the hysteresis comparator outputs a boost closing trigger signal. The state switching delay of the SR state machine is not more than 20 ms.

[0033] The setting of the specific threshold in the scheme can accurately distinguish between the real inhalation start-up signal and the noise signal, avoid the false triggering or missed triggering of the boost due to the noise, and ensure the accuracy of the feedforward prediction. The short state switching delay can further shorten the interval between the inhalation start-up and the boost response, reduce the energy retention in the start-up stage, reduce the invalid entropy generation in the stage, and improve the reliability and timeliness of the feedforward control.

[0034] Specifically, in the embodiment, the adaptive entropy balance unit receives a tidal volume growth rate signal transmitted by a tidal volume flow sensor, compares the tidal volume growth rate signal with a preset reference threshold, and transmits a reset signal to the SR state machine when the tidal volume growth rate is less than the preset reference threshold. When the tidal volume growth rate is greater than or equal to the preset reference threshold, the adaptive entropy balance unit maintains the current output state of the SR state machine. The preset reference threshold is 100 ml / s.

[0035] The scheme dynamically adjusts the boost switch according to the entropy flow efficiency (reflected by the tidal volume growth rate), matches the work intensity of autonomous breathing, avoids excessive assistance to cause the "inertia" of the respiratory muscle when the autonomous breathing can maintain sufficient entropy flow, and promotes autonomous work by closing the boost to avoid the invalid work caused by insufficient assistance when the autonomous breathing entropy flow is insufficient, so as to maintain the balance between the entropy flow and the invalid entropy increase, and protect the autonomous breathing ability from being inhibited.

[0036] Specifically, in the present embodiment, the adaptive entropy balancing unit also internally incorporates a hidden Markov model, which predicts real-time respiratory frequency by analyzing historical respiratory frequency data collected by the multi-modal physiological perception module. When the predicted respiratory frequency rises, the adaptive entropy balancing unit lowers the assistance threshold. When the predicted respiratory frequency falls, the adaptive entropy balancing unit raises the assistance threshold.

[0037] This scheme can adjust the assistance threshold in advance by predicting changes in respiratory frequency (such as increased energy demand due to exercise or reduced energy supply due to fatigue), thereby avoiding entropy imbalance caused by maintaining the original parameters of assistance when the respiratory state has changed. Lowering the threshold when the frequency rises prevents excessive assistance from causing low-entropy gas loss, and raising the threshold when the frequency falls prevents insufficient assistance from causing ineffective work, making the adjustment of assistance more predictive and adapting to the respiratory demand under different physiological states.

[0038] Specifically, in the present embodiment, the execution module includes a rotary proportional valve driven by a stepper motor. The opening degree of the rotary proportional valve is adjustable within a range of zero to ninety degrees, and the response time is not more than fifty milliseconds. The rotary proportional valve adjusts the valve core angle according to the assistance instruction transmitted by the control decision module to change the auxiliary pressure and ventilation flow.

[0039] Compared with traditional fixed pressure output or low-precision adjustment components, the continuously adjustable opening degree of this scheme can achieve fine adjustment of auxiliary pressure and flow, avoiding the impact of sudden pressure changes on breathing; fast response can promptly implement the instructions of the control decision module, avoiding the delay between instructions and execution that causes adjustment lag, ensuring that the assistance parameters can match the changes in respiratory state in real time, and improving the precision and timeliness of the execution link.

[0040] Specifically, in the present embodiment, the rotary proportional valve includes an advance assistance adjustment mode, a tidal adaptation adjustment mode, and an exhalation synchronization adjustment mode. In the advance assistance adjustment mode, the rotary proportional valve adjusts the opening degree to thirty degrees and outputs an auxiliary pressure of two centimeters of water column after receiving the assistance opening instruction. In the tidal adaptation adjustment mode, the rotary proportional valve dynamically adjusts the opening degree according to the adjustment instruction of the adaptive entropy balancing unit. In the exhalation synchronization adjustment mode, the rotary proportional valve pre-opens to an opening degree of five degrees after receiving the exhalation flow inflection point signal transmitted by the exhalation flow sensor.

[0041] The three modes provided by this scheme cover the entire respiratory cycle - the advance assistance mode provides reserve pressure for inspiration start to reduce start-up delay; the tidal adaptation mode dynamically adjusts according to the entropy flow state to match real-time demand; and the exhalation synchronization mode prepares for the next round of inspiration to reduce energy retention. Through targeted adjustment in stages, the ventilation process is more coherent, and the production of invalid entropy in each stage is controlled, further promoting the total entropy production to approach zero and improving the stability of co-entropy balancing.

[0042] Embodiment 2 The embodiment provides a co-entropy balance adaptive control method, which is applied to the co-entropy balance adaptive control system in the embodiment 1, as shown in the figure, and comprises the following steps. Figure 2 In the first step, the multi-modal physiological sensing module collects the nasal pressure drop signal, the tidal volume growth rate signal and the expiratory flow inflection point signal, and transmits the signals to the control decision module after signal conditioning. In the second step, the control decision module analyzes the nasal pressure drop signal through the feedforward prediction unit, and transmits the power-on instruction to the execution module in advance. In the third step, the control decision module analyzes the tidal volume growth rate signal through the adaptive entropy balance unit, dynamically adjusts the power threshold and transmits the adjustment instruction to the execution module. In the fourth step, the execution module executes the corresponding ventilation adjustment action according to the power-on instruction and the adjustment instruction. In the fifth step, the feedback module collects the actual running parameters of the execution module and the real-time total entropy generation and feeds back to the control decision module. In the sixth step, the control decision module corrects the power instruction based on the feedback of the real-time total entropy generation, and maintains the co-entropy balance state.

[0043] The process realizes the closed-loop control from signal input to instruction output and then to feedback correction, and each step is closely connected - signal collection and conditioning ensure signal quality, feedforward and adaptive ensure control accuracy, execution implements adjustment, feedback and correction maintain continuous adaptation. Through process control, it avoids the adjustment confusion caused by the disconnection of each link, so that the system can continuously and stably maintain the co-entropy balance, and provides a coherent and reliable control logic for respiratory support.

[0044] Specifically, in the embodiment, the real-time total entropy generation in the fifth step is calculated by the following method: ; wherein ΔS total is the real-time total entropy generation, k1 is the first entropy increase coefficient, k2 is the second entropy increase coefficient, k3 is the third entropy increase coefficient, ΔP nose is the nasal pressure drop signal, is the tidal volume growth rate signal, dV th is the preset reference threshold, and ΔF exp is the expiratory flow inflection point signal; when the real-time total entropy generation is greater than zero, the control decision module adjusts the power threshold or the rotation proportion valve opening degree, and when the real-time total entropy generation tends to zero, the current parameters are maintained.

[0045] ​Quantify the total entropy production by formula, avoid the subjective judgment deviation of the entropy balance state, make the entropy balance or not have a clear quantitative basis; based on the quantitative results, adjust the parameters (threshold or opening degree when entropy increases, maintain when balanced), which can ensure the adjustment direction accurate, avoid the entropy imbalance caused by blind adjustment, so as to continuously control the total entropy production in the range close to zero, and provide quantitative support for the stable maintenance of the co-entropy balance.

[0046] In summary, the co-entropy balance adaptive control method and system provided by the embodiment has the following advantages: 1. Achieve co-entropy balance: Through the multi-modal perception layer, the entropy change and energy flow signals of each stage of respiration are comprehensively captured, the feedforward prediction of the control decision layer reduces the start delay, the adaptive entropy balance matches the autonomous work intensity, and the closed-loop adjustment of the execution feedback layer continuously corrects the parameters, finally makes the total entropy production of the respiratory system close to zero, restores the low-entropy steady state, and solves the problem that the traditional ventilator cannot achieve co-entropy balance.

[0047] 2. Reduce invalid entropy increase: The response delay of the inspiratory start is greatly shortened by the feedforward prediction, avoiding energy retention; the tidal adaptive adjustment dynamically switches the assistance according to the entropy flow efficiency, avoiding the loss of low-entropy gas caused by excessive assistance or invalid work caused by insufficient assistance, significantly reducing invalid entropy production.

[0048] 3. Promote the recovery of autonomous respiration: The design of "assistance rather than replacement" adjusts the assistance threshold and valve opening degree as needed, so that the respiratory center participates in the regulation of energy flow and entropy balance again, avoiding the respiratory center function degradation caused by traditional replacement ventilation, and improving the success rate of patient offline.

[0049] 4. Adapt to different pathological states: The multi-modal perception layer fuses pressure, flow, and frequency signals, and the control decision layer can dynamically adjust the control strategy according to the pathological characteristics of different patients (such as chronic obstructive pulmonary disease and respiratory muscle fatigue patients), maintain the dynamic balance of entropy flow and invalid entropy increase, and has strong robustness.

[0050] Working principle: The application constructs a closed-loop control system of "multi-modal perception-feedforward prediction-adaptive entropy balance-execution feedback", and each module cooperates to realize co-entropy balance, the specific principle is as follows: 1. Multi-modal physiological perception layer: three types of sensors are used to collect physiological signals reflecting "energy flow" and "entropy change" in real time - nasal valve + pressure sensor is installed at the entrance of the nasal cavity to detect the pressure drop before inspiration (reflecting the energy start demand); tidal volume flow sensor is installed in the inspiration channel to monitor the tidal volume growth rate (reflecting the entropy flow efficiency); exhalation flow sensor is installed in the exhalation channel to monitor the exhalation flow inflection point (reflecting the entropy recovery efficiency); the collected signals are filtered by a low-pass filter to remove noise, and then transmitted to the control decision layer.

[0051] 2. Control decision layer: "Advance assistance" and "dynamic entropy balance" are realized through two core algorithms - one is bistable cross feedforward, which presets "assist off" and "assist on" bistability, uses a lag comparator to predict inhalation intention (trigger assist on when nasal pressure drops to a set threshold), combines SR state machine to quickly switch assist state, reduces start-up delay and energy retention; two is tidal adaptive entropy balance, which compares tidal volume growth rate with reference threshold, closes assist to promote autonomous work when entropy flow is insufficient, and maintains assist when entropy flow is sufficient; at the same time, the historical respiratory frequency is analyzed through the hidden Markov model to predict the real-time respiratory frequency, and the assist threshold is reduced when the frequency is high, and the assist threshold is increased when the frequency is low, to match the autonomous work intensity.

[0052] 3. Execution feedback layer: The execution layer converts control commands into ventilation actions through a rotary proportional valve driven by a stepper motor (opening can be continuously adjusted), which is divided into three modes: advance assistance (when receiving assist on command, valve opening is adjusted to a certain value to output auxiliary pressure), tidal adaptation (valve opening is dynamically adjusted according to tidal volume growth rate), and exhalation synchronization (valve is pre-opened when receiving exhalation flow inflection point signal). The feedback layer monitors tidal volume growth rate, nasal pressure drop, and exhalation flow inflection point to calculate real-time total entropy production, adjusts assist threshold or valve opening if entropy increases, and maintains parameters if it approaches isentropic balance, forming a closed-loop feedback.

[0053] Method of use: 1. System deployment: First, connect the patient interface of the ventilator to the patient's airway, deploy a nasal cavity check valve + pressure sensor at the nasal inlet, deploy a tidal volume flow sensor in the inhalation channel, and deploy an exhalation flow sensor in the exhalation channel; at the same time, confirm that the main control chip of the control decision layer, the rotary proportional valve of the execution layer, and the touch screen of the user interaction layer are all connected normally.

[0054] 2. Parameter setting and start: Medical personnel set parameters such as target pressure and alarm threshold through the touch screen, and start the ventilator; the system automatically starts the signal acquisition function of the multi-modal physiological perception layer, and the sensors start to collect nasal pressure, tidal volume growth rate, and exhalation flow signals, which are transmitted to the control decision layer after being filtered by a low-pass filter to remove noise.

[0055] 3. Automatic control process: The control decision layer processes the collected signals in real time - through bistable cross feedforward to predict inhalation intention, trigger assist command in advance, and the rotary proportional valve enters advance assistance mode; then, according to the tidal volume growth rate, the tidal adaptive adjustment is carried out, and the rotary proportional valve switches to the tidal adaptation mode to dynamically adjust the opening; in the exhalation phase, after receiving the exhalation flow inflection point signal, the rotary proportional valve switches to the exhalation synchronization mode to pre-open the valve, preparing for the next round of inhalation.

[0056] 4. Monitoring and adjustment: medical staff can view the total entropy production, tidal volume growth rate, gas exchange related indicators and other co-entropy indicators in real time through the touch screen; the system continuously monitors the actual parameters through the feedback layer, automatically calculates the total entropy production, and automatically adjusts the assist threshold or rotary proportional valve opening if the entropy increases to maintain the co-entropy balance; if manual intervention is required, medical staff can modify the relevant parameters through the touch screen.

[0057] The above merely describes the preferred embodiments of the present application, and is not intended to limit the embodiments and protection scope of the present application. It should be noted by those skilled in the art that any equivalent replacement and obvious changes made according to the content of the present application should be included in the protection scope of the present application.

Claims

1. A co-entropy balance adaptive control system, characterized in that, It includes a multimodal physiological sensing module, a control decision module, an execution module, and a feedback module, among which: The signal output of the multimodal physiological sensing module is connected to the signal input of the control decision module. The command output of the control decision module is connected to the command input of the execution module. The state feedback of the execution module is connected to the signal input of the feedback module. The signal output of the feedback module is connected to the signal input of the control decision module. The multimodal physiological sensing module collects energy flow signals and entropy change signals of the respiratory system and transmits them to the control decision module. Based on the received signals, the control decision module generates assistance commands through feedforward prediction and adaptive entropy balance algorithms and transmits them to the execution module. The execution module performs ventilation regulation actions according to the assistance commands. The feedback module collects the actual operating parameters of the execution module and the real-time entropy change index of the respiratory system and feeds them back to the control decision module. Based on the feedback signals, the control decision module dynamically corrects the assistance commands to maintain a co-entropy balance state where the total entropy generation of the respiratory system approaches zero.

2. The co-entropy balance adaptive control system according to claim 1, characterized in that, The multimodal physiological sensing module includes a nasal pressure sensor, a tidal volume flow sensor, and an expiratory flow sensor. The nasal pressure sensor is located at the nasal interface of the ventilator to collect the nasal pressure drop signal before inspiration. The tidal volume flow sensor is located at the inspiratory channel of the ventilator to collect the tidal volume growth rate signal. The expiratory flow sensor is located at the expiratory channel of the ventilator to collect the expiratory flow inflection point signal.

3. The co-entropy balance adaptive control system according to claim 2, characterized in that, The control decision module includes a feedforward prediction unit and an adaptive entropy balance unit. The feedforward prediction unit has a built-in comparator with hysteresis and an SR state machine. The comparator with hysteresis receives the nasal pressure drop signal transmitted by the nasal pressure sensor and compares it with a preset threshold and a reset threshold. It outputs a state control signal to the SR state machine. The SR state machine outputs an assist-on or off command based on the state control signal and the reset signal transmitted by the adaptive entropy balance unit.

4. The co-entropy balance adaptive control system according to claim 3, characterized in that, The preset threshold with hysteresis comparator is -1.5 cm H2O, and the preset reset threshold is -1 cm H2O. When the nasal pressure drop signal is less than or equal to the set threshold, the hysteresis comparator outputs an assist-on trigger signal. When the nasal pressure drop signal is greater than the reset threshold, the hysteresis comparator outputs an assist-off trigger signal. The state switching delay of the SR state machine does not exceed 20 milliseconds.

5. The co-entropy balance adaptive control system according to claim 3, characterized in that, The adaptive entropy balancing unit receives the tidal volume growth rate signal transmitted by the tidal volume flow sensor and compares the tidal volume growth rate signal with a preset reference threshold. When the tidal volume growth rate is less than the preset reference threshold, the adaptive entropy balancing unit transmits a reset signal to the SR state machine. When the tidal volume growth rate is greater than or equal to the preset reference threshold, the adaptive entropy balancing unit maintains the current output state of the SR state machine. The preset reference threshold is 100 milliliters per second.

6. The co-entropy balance adaptive control system according to claim 5, characterized in that, The adaptive entropy balancing unit also incorporates a hidden Markov model. The hidden Markov model predicts the real-time respiratory rate by analyzing historical respiratory rate data collected by the multimodal physiological sensing module. When the predicted respiratory rate increases, the adaptive entropy balancing unit lowers the assist threshold; when the predicted respiratory rate decreases, the adaptive entropy balancing unit raises the assist threshold.

7. The co-entropy balance adaptive control system according to claim 1, characterized in that, The execution module includes a rotary proportional valve, which is driven by a stepper motor. The opening adjustment range of the rotary proportional valve is from zero to ninety degrees, and the response time is no more than fifty milliseconds. The rotary proportional valve adjusts the valve core angle according to the assist command transmitted by the control decision module to change the auxiliary pressure and ventilation flow.

8. The co-entropy balance adaptive control system according to claim 7, characterized in that, The rotary proportional valve includes three modes: advance assist adjustment mode, tide adaptation adjustment mode, and exhalation synchronization adjustment mode. In advance assist adjustment mode, after receiving the assist activation command, the rotary proportional valve adjusts the opening to 30 degrees and outputs an auxiliary pressure of 2 centimeters of water column. In tide adaptation adjustment mode, the rotary proportional valve dynamically adjusts the opening according to the adjustment command of the adaptive entropy balance unit. In exhalation synchronization adjustment mode, after receiving the exhalation flow inflection point signal transmitted by the exhalation flow sensor, the rotary proportional valve pre-opens to a 5-degree opening.

9. A co-entropy balance adaptive control method, characterized in that, Applied to the co-entropy balance adaptive control system according to any one of claims 1 to 8, Includes the following steps: The first step involves the multimodal physiological sensing module acquiring signals of nasal pressure decrease, tidal volume growth rate, and expiratory flow inflection point, which are then transmitted to the control decision module after signal conditioning. The second step is that the control decision module analyzes the nasal pressure drop signal through the feedforward prediction unit and triggers the assist activation command to be transmitted to the execution module in advance. Third, the control decision module analyzes the tidal volume growth rate signal through the adaptive entropy balance unit, dynamically adjusts the assist threshold, and transmits adjustment instructions to the execution module. Fourth, the execution module performs the corresponding ventilation adjustment actions according to the power assist activation command and adjustment command; The fifth step involves the feedback module collecting the actual operating parameters and real-time total entropy generated by the execution module and feeding them back to the control decision module. The sixth step involves the control decision module generating corrective assistance commands based on the real-time total entropy feedback to maintain a co-entropy balance.

10. The co-entropy balance adaptive control method according to claim 9, characterized in that, The real-time total entropy in step five is calculated as follows: ;where ΔS total For real-time total entropy generation, k1 is the first entropy increase coefficient, k2 is the second entropy increase coefficient, k3 is the third entropy increase coefficient, and ΔP nose This is a signal of decreased nasal pressure. The tidal volume growth rate signal, dV th ΔF is the preset baseline threshold. exp This is a signal of the inflection point in expiratory flow. When the real-time total entropy is greater than zero, the control decision module adjusts the assist threshold or the opening of the rotary proportional valve; when the real-time total entropy approaches zero, the current parameters are maintained.

Images