A method and system for regulating airflow in a respirator

Abstract

The application discloses a respirator airflow regulation method and system, applied to the technical field of respirators, target risk prediction information is acquired, wherein the target risk information comprises posture risk prediction information and voice intention prediction information; target gas supply parameters are determined according to the target risk prediction information; and the gas supply parameters of the respirator are adjusted according to the target gas supply parameters. The respirator gas supply safety is improved at least to a certain extent, the voice intelligibility during posture change is effectively improved under the premise of preventing gas invasion, and the effective utilization of the gas source is also considered.

Description

Technical Field

[0001] This application relates to the field of respirator technology, and more specifically, to a respirator airflow regulation method and system. Background Technology

[0002] Existing respirator airflow control methods and systems primarily rely on user physiological and environmental parameters to regulate air supply and meet basic breathing needs. However, in complex and dynamically changing environments, such as scenarios where users are in confined spaces and undergo significant posture changes (e.g., bending, twisting, crawling) while simultaneously needing to communicate verbally, existing technologies face significant challenges. Firstly, changes in user posture can reduce the seal between the respirator mask and face or impair airway patency. Existing systems fail to effectively detect or predict these risks associated with posture changes and cannot adjust air supply strategies in a timely manner to prevent the intrusion of harmful external gases or overcome additional breathing resistance. Secondly, airflow noise generated during respirator supply can interfere with user voice communication, reducing the clarity of information transmission. Existing systems, while ensuring air supply, fail to optimize for voice communication needs, such as reducing noise interference when the user attempts to speak. Therefore, existing technologies struggle to address the complex relationship between changes in mask sealing or patency caused by posture changes and airflow noise during voice communication. They cannot effectively improve voice clarity during posture changes while ensuring gas supply safety and preventing gas intrusion, and they also cannot ensure the effective utilization of the gas source.

[0003] To address the aforementioned issues, existing technologies urgently need improvement. Summary of the Invention

[0004] To address the shortcomings of existing technologies, this application provides a method and system for regulating the airflow of a respirator, which effectively improves speech clarity during posture changes while ensuring safe air supply and preventing gas intrusion, and also takes into account the effective utilization of the air source.

[0005] This application provides a method for regulating airflow in a ventilator, applied to a ventilator, comprising:

[0006] Obtain target risk prediction information, which includes posture risk prediction information and voice intent prediction information;

[0007] Determine the target gas supply parameters based on the target risk prediction information;

[0008] Adjust the respirator's air supply parameters according to the target air supply parameters;

[0009] In the case where the target risk prediction information is attitude risk prediction information, the respirator's air supply parameters are adjusted according to the first target air supply parameters.

[0010] When the target risk prediction information is voice intent prediction information, the respirator's air supply parameters are adjusted according to the second target air supply parameters.

[0011] When the target risk prediction information consists of posture risk prediction information and voice intent prediction information, the respirator's air supply parameters are adjusted according to the third target air supply parameters.

[0012] By predicting posture and voice intent risks and adjusting the gas supply accordingly, the above solution can more intelligently cope with complex scenarios, improving security and communication performance.

[0013] To improve the solution, this application also proposes a method for regulating the airflow of a respirator, the respirator including a mask and a detection device, the detection device being used to acquire the mask's sealing information, breathing patency information, and acoustic information; the step of obtaining target air supply parameters based on the target risk prediction information includes at least one of the following:

[0014] When the target risk information is attitude risk prediction information, the first target air supply parameter is determined based on the sealing information and the breathing patency information;

[0015] When the target risk information is voice intent prediction information, the second target gas supply parameters are determined based on the acoustic information;

[0016] When the target risk information is posture risk prediction information and voice intent prediction information, the third target air supply parameters are determined based on the sealing information, the breathing patency information and the acoustic information.

[0017] The above scheme provides a specific way to determine the target gas supply parameters under different risk modes, making the method more operable.

[0018] To improve the solution, this application also proposes a respirator airflow regulation method. When the target risk information is speech intent prediction information, a second target air supply parameter is determined based on acoustic information, including:

[0019] Acquire information about continuous changes in body posture;

[0020] Based on acoustic information and continuous changes in body posture, the predicted trend of change is obtained;

[0021] The second target gas supply parameters are determined based on acoustic information and predicted trends.

[0022] The above scheme takes into account the continuous changes in body posture when adjusting the relevant air supply parameters after determining the voice intent, thus improving the accuracy of prediction.

[0023] To improve the solution, this application also proposes a method for regulating respirator airflow, which obtains a predicted trend based on acoustic information and continuous changes in body posture, including:

[0024] Acquire event characteristic information of external acoustic events in the space where the respirator is located;

[0025] Based on event characteristic information, determine the first level of impact of external acoustic events on acoustic information;

[0026] Based on the first level of impact, acoustic information, and continuous changes in body posture, the predicted trend of change is obtained.

[0027] By incorporating the impact of external acoustic events into the prediction model, the analysis of acoustic information becomes more accurate.

[0028] To improve the solution, this application also proposes a method for regulating respirator airflow, which, based on the first level of influence and continuous changes in acoustic information and body posture, yields a predicted trend, including:

[0029] When the first influence level is greater than or equal to the preset first influence level, the acoustic information is adjusted according to the first influence level to obtain the adjusted acoustic information;

[0030] Based on the adjusted acoustic information and the continuous changes in body posture, the predicted trend of change is obtained.

[0031] The above solution provides an acoustic information adjustment mechanism to address strong interference from external acoustic events, thereby reducing misjudgments.

[0032] To improve the solution, this application also proposes a method for regulating respirator airflow, which, based on event characteristic information, determines the first level of influence of external acoustic events on acoustic information, including:

[0033] The event feature information is analyzed to obtain the feature components of the event feature information;

[0034] Based on the feature components and the preset external acoustic event type library, the type of external acoustic event and the first parameter relationship are determined. The first parameter relationship represents the acoustic feature representation and the first influence level relationship of the acoustic information of different external acoustic event types included in the external acoustic event type library.

[0035] Based on the type of external acoustic event, the characteristic components of the event's characteristic information, and the relationship of the first parameter, the first level of influence of the external acoustic event on the acoustic information is determined.

[0036] The above approach provides a method for quantifying the impact of external acoustic events based on event characteristics and a type library, thereby improving the objectivity of the assessment.

[0037] To improve the solution, this application also proposes a method for regulating the airflow of a respirator. Based on the type of the external acoustic event, the characteristic components of the event's feature information, and the relationship of the first parameter, the method determines the first level of influence of the external acoustic event on the acoustic information, including:

[0038] The first propagation characteristic characterization of the current actual acoustic propagation characteristics in the space where the respirator is located and the second propagation characteristic characterization of the preset reference acoustic environment associated with the first parameter relationship are obtained.

[0039] The first difference is obtained based on the first propagation characteristic characterization and the second propagation characteristic characterization, and the propagation characteristic correction factor is determined based on the first difference.

[0040] The relationship between the type of external acoustic event, the feature components of the event feature information, and the first parameter is calculated to obtain the initial influence level of the external acoustic event on the acoustic information;

[0041] The initial impact level is adjusted based on the propagation characteristic correction factor to determine the first impact level of external acoustic events on acoustic information.

[0042] The above scheme takes into account the acoustic propagation characteristics of the actual environment, corrects the impact assessment of external acoustic events, and improves accuracy.

[0043] To improve the solution, this application also proposes a method for regulating ventilator airflow, which determines a propagation characteristic correction factor based on a first difference, including:

[0044] The first difference is decomposed to obtain multiple difference components;

[0045] Determine the component weight corresponding to each difference component based on each difference component.

[0046] For each difference component and its weight, the propagation characteristic correction factor is determined.

[0047] The above scheme provides a method for calculating the propagation characteristic correction factor based on the difference components and their weights, making the correction more precise.

[0048] To improve the solution, this application also proposes a method for regulating ventilator airflow, which determines the component weights corresponding to the differential components based on the differential components, including:

[0049] Acquire historical data of the difference components within a preset time period;

[0050] Analyze the changing trends of the differential components based on historical data;

[0051] The component weights corresponding to the differences are determined based on the changing trends.

[0052] The above approach utilizes historical data analysis to determine the weights of the differential components by analyzing their changing trends, making the correction factor more adaptable.

[0053] To improve the solution, this application also provides a respirator airflow regulation system, including:

[0054] The acquisition module is used to acquire target risk prediction information, which includes posture risk prediction information and voice intent prediction information.

[0055] The target gas supply parameter determination module is used to determine the target gas supply parameters based on the target risk prediction information;

[0056] The air supply parameter adjustment module is used to adjust the air supply parameters of the respirator according to the target air supply parameters;

[0057] In the case where the target risk prediction information is attitude risk prediction information, the respirator's air supply parameters are adjusted according to the first target air supply parameters.

[0058] When the target risk prediction information is voice intent prediction information, the respirator's air supply parameters are adjusted according to the second preset target air supply parameters.

[0059] When the target risk prediction information consists of posture risk prediction information and voice intent prediction information, the respirator's air supply parameters are adjusted according to the third target air supply parameters.

[0060] The above scheme provides a system platform for implementing the above methods.

[0061] In summary, the respirator airflow control method and system provided in this application effectively solve the problem that existing technologies struggle to coordinate the complex relationship between changes in mask sealing or patency caused by posture changes and airflow noise during voice communication. This is achieved by acquiring posture risk prediction information and voice intent prediction information, and determining and adjusting the respirator's air supply parameters based on this information. As a result, it effectively improves voice clarity during posture changes while ensuring air supply safety and preventing gas intrusion, and also takes into account the effective utilization of the air source. Attached Figure Description

[0062] To more clearly illustrate the technical solutions in the embodiments of the present invention, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0063] Figure 1 This is a flowchart of a ventilator airflow regulation method provided in one embodiment of this application;

[0064] Figure 2 This is a flowchart of determining a second target gas supply parameter provided in one embodiment of this application;

[0065] Figure 3 This is a flowchart illustrating the predicted trend of change provided in one embodiment of this application;

[0066] Figure 4 This is a flowchart of obtaining the predicted trend of change provided in another embodiment of this application;

[0067] Figure 5 This is a flowchart of determining the first level of influence of an external acoustic event on acoustic information, provided in one embodiment of this application;

[0068] Figure 6 This is a flowchart of another embodiment of the present application for determining the first level of influence of an external acoustic event on acoustic information;

[0069] Figure 7 This is a flowchart of determining a propagation characteristic correction factor based on a first difference, provided in one embodiment of this application;

[0070] Figure 8 This is a flowchart of determining the component weights corresponding to the difference components based on the difference components, provided in one embodiment of this application. Detailed Implementation

[0071] The technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, and not all embodiments. Based on the embodiments of this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.

[0072] Furthermore, the described features, structures, or characteristics may be combined in one or more embodiments in any suitable manner.

[0073] In the following description, numerous specific details are provided to give a thorough understanding of embodiments of the present application. However, those skilled in the art will recognize that the technical solutions of the present application can be practiced without one or more of the specific details, or other methods, components, apparatuses, steps, etc. may be employed. In other instances, well-known methods, apparatuses, implementations, or operations are not shown or described in detail to avoid obscuring various aspects of the present application.

[0074] The block diagrams shown in the accompanying drawings are merely functional entities and do not necessarily correspond to physically independent entities. That is, these functional entities can be implemented in software, in one or more hardware modules or integrated circuits, or in different network and / or processor devices and / or microcontroller devices.

[0075] The flowcharts shown in the accompanying drawings are merely illustrative and do not necessarily include all content and operations / steps, nor do they necessarily have to be performed in the described order. For example, some operations / steps can be broken down, while others can be combined or partially combined; therefore, the actual execution order may change depending on the specific circumstances.

[0076] In the description of this application, it should be understood that the terms "first" and "second" are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of indicated technical features. Therefore, a feature defined as "a" or "second" may explicitly or implicitly include one or more of that feature. In the description of this application, unless otherwise stated, "a plurality of" means two or more.

[0077] like Figure 1 As shown, Figure 1 This is a flowchart of a ventilator airflow regulation method provided in one embodiment of this application; the ventilator airflow regulation method may include, but is not limited to, steps S110, S120 and S130.

[0078] Step S110: Obtain target risk prediction information, wherein the target risk information includes posture risk prediction information and voice intent prediction information;

[0079] Step S120: Determine the target gas supply parameters based on the target risk prediction information;

[0080] Step S130: Adjust the respirator's air supply parameters according to the target air supply parameters;

[0081] In the case where the target risk prediction information is attitude risk prediction information, the respirator's air supply parameters are adjusted according to the first target air supply parameters.

[0082] When the target risk prediction information is voice intent prediction information, the respirator's air supply parameters are adjusted according to the second target air supply parameters.

[0083] When the target risk prediction information consists of posture risk prediction information and voice intent prediction information, the respirator's air supply parameters are adjusted according to the third target air supply parameters.

[0084] Among them, target risk prediction information refers to the set of information used to predict the risks or special needs that a user may face during breathing. It can include posture risk prediction information and voice intent prediction information. Posture risk prediction information refers to information that predicts the breathing-related risks that may be caused by changes in the user's body posture. It can be achieved by analyzing the user's body posture sensor data, motion trajectory data, etc., such as through inertial measurement unit data and visual recognition data. It is mainly used to assess the potential impact of posture changes on mask sealing or airway patency. Voice intent prediction information refers to information that predicts the user's intention to make a voice communication. It can be achieved by analyzing the user's acoustic signal characteristics, facial muscle activity signals, breathing pattern changes, etc., such as through microphone sound signals and electromyography sensor signals. It is mainly used to determine whether the user is about to speak in order to optimize the acoustic environment during communication. Target air supply parameters refer to the set of parameters determined based on the predicted risks or needs and used to adjust the air supply status of the respirator. It can include air supply pressure, air supply flow rate, gas mixing ratio, etc. The purpose is to adapt the air supply status of the respirator to the user's current or upcoming state.

[0085] The first target gas supply parameter refers to the target gas supply parameter determined when attitude risks are predicted. Its purpose is to address the risks caused by attitude changes, such as compensating for insufficient oxygen supply caused by leakage by increasing pressure or increasing oxygen concentration. The second target gas supply parameter refers to the target gas supply parameter determined when voice intent is predicted. Its purpose is to optimize the environment during voice communication, such as reducing aerodynamic noise by decreasing flow rate. The third target gas supply parameter refers to the target gas supply parameter determined when both attitude risks and voice intent are predicted. Its purpose is to determine the gas supply parameter while taking into account both attitude risk management and voice communication optimization. For example, if attitude changes lead to a decrease in sealing and the user is about to communicate with the outside world, the oxygen supply ratio can be adjusted to reduce flow rate to reduce noise while ensuring the normal oxygen content required for the user's breathing.

[0086] By coordinating and comprehensively considering posture risk prediction information and voice intent prediction information, the target gas supply parameters adapted to different risk combinations are dynamically determined, achieving the effect of improving voice communication clarity and saving gas resources while ensuring gas supply safety and smooth operation.

[0087] In some embodiments, a method for regulating airflow in a respirator includes the step of obtaining target air supply parameters based on target risk prediction information, which includes:

[0088] When the target risk information is attitude risk prediction information, the first target air supply parameters are determined based on sealing information and breathing patency information.

[0089] When the target risk information is voice intent prediction information, the second target gas supply parameters are determined based on the acoustic information;

[0090] When the target risk information is posture risk prediction information and voice intent prediction information, the third target air supply parameters are determined based on sealing information, breathing patency information and acoustic information.

[0091] The sealing information refers to the degree of fit between the mask and the user's face, which can be obtained using pressure sensors, gas flow sensors, or image recognition devices. Its purpose is to assess whether there is a leak in the mask. The breathing patency information refers to the smoothness of the airflow through the user's airway, which can be obtained using flow sensors, pressure sensors, or bioelectrical impedance sensors. Its purpose is to assess whether there is abnormal resistance or pattern in the user's breathing. The acoustic information of the mask refers to the acoustic environment information inside or near the respirator mask, which can include the user's voice, breathing airflow noise, external sounds, etc. It can be obtained using microphones or acoustic sensors. Its purpose is to assess the sound conditions inside the mask, especially noise related to voice communication.

[0092] This application's solution achieves refined control of respirator airflow by combining information collected by the respirator's built-in detection device with target risk prediction information. Specifically, the respirator includes a mask and a detection device. The detection device acquires information about the mask's seal, breathing patency, and acoustic properties. This information directly reflects the respirator's current operating status and the user's breathing status. When the target risk information is posture risk prediction information, the system focuses on the mask's seal and breathing patency, and determines a first target air supply parameter based on the seal and breathing patency information acquired by the detection device. For example, if the seal is poor, the system can determine to increase the air supply pressure; if breathing patency is low, the system can determine to increase the air supply flow rate. When the target risk information is voice intent prediction information, the system focuses on the acoustic environment inside the mask and determines a second target air supply parameter based on the acoustic information acquired by the detection device. For example, by analyzing the acoustic information, the system determines whether there is airflow noise interference and adjusts the air supply parameters accordingly to reduce noise. When the target risk information consists of posture risk prediction and voice intent prediction, the system comprehensively considers the mask's seal, breathing patency, and the acoustic environment within the mask. Based on the seal information, breathing patency information, and acoustic information obtained from the detection device, it determines the third target air supply parameters. This ensures air supply safety, prevents harmful gas intrusion, maintains sufficient ventilation, improves voice communication clarity, and maximizes the conservation of gas cylinder resources. This approach, combining the respirator's own detection information with risk prediction information, makes the determination of air supply parameters more accurate, fully utilizes the respirator's hardware capabilities, and thus improves the overall performance and effectiveness of the respirator.

[0093] like Figure 2 As shown, Figure 2 This is a flowchart of determining the second target gas supply parameter provided in one embodiment of this application; when the target risk information is voice intent prediction information, the second target gas supply parameter is determined based on the acoustic information, which may include, but is not limited to, steps S210, S220 and S230.

[0094] Step S210: Obtain information on continuous changes in body posture;

[0095] Step S220: Based on the acoustic information and the continuous change information of body posture, obtain the predicted change trend;

[0096] Step S230: Determine the second target gas supply parameters based on acoustic information and predicted change trends.

[0097] Among them, the information on continuous changes in body posture refers to the data on changes in the user's body position and posture over time. Specifically, it can be obtained by an inertial measurement unit, posture sensor, or posture recognition module based on visual analysis. Its purpose is to capture changes in the user's breathing patency or mask seal that may be caused by changes in body position. The prediction of change trend refers to the prediction of the direction and degree of changes in the user's breathing needs or state over a future period of time based on current and historical acoustic information and information on continuous changes in body posture. Specifically, it can be analyzed and inferred through machine learning models or based on expert systems. Its purpose is to achieve early perception and response to changes in the user's state.

[0098] This application's solution, when the target risk information is speech intent prediction information, no longer relies solely on acoustic information to determine the second target ventilation parameter. Instead, it first acquires continuous changes in the user's body posture. This posture change information, along with the acoustic information, is input into an analysis module. This module comprehensively analyzes both types of information; for example, acoustic information may indicate that the user is attempting to speak, while posture information may show that the user is engaged in strenuous exercise or is in a restricted posture. Through joint analysis of this information, the system can obtain a predictive trend, such as predicting that the user is about to enter a hyperventilation demand state or that breathing may be restricted. Finally, based on the original acoustic information and the obtained predictive trend, the system determines a more accurate second target ventilation parameter. This approach, combining posture information and predictive trends, allows the system to more comprehensively assess the user's actual state and potential needs, and avoids discontinuous changes in system parameters that could lead to lag in ventilation parameter adjustments.

[0099] In the above embodiments, information on continuous changes in body posture can be acquired by an inertial measurement unit (IMU) worn on the user's body or respirator. The IMU outputs data including acceleration and angular velocity, which are processed by algorithms to obtain posture angles, motion states, etc. Acoustic information can be acquired by microphones inside or outside the mask. The IMU data and the acoustic signals collected by the microphones are input into a trained neural network model. This model, trained with a large amount of real data on different postures and voice states, can output a predicted value representing the trend of the user's breathing demand changes in the next few seconds. Then, the system can determine specific second target air supply parameters based on the current acoustic information and the predicted trend, using a lookup table method or another decision model. For example, if an increase in breathing demand is predicted in the next few seconds, the air supply pressure or flow rate is increased, or the air supply pressure or flow rate is continuously adjusted with an acceleration, rather than adjusting the air supply pressure or flow rate only when the actual increase in demand is detected. By combining continuous changes in body posture with acoustic information and obtaining predicted trends, the system can more comprehensively perceive changes in the user's state and demand, improving the accuracy and timeliness of air supply parameter adjustments.

[0100] like Figure 3 As shown, Figure 3 This is a flowchart of obtaining a predicted change trend provided in one embodiment of this application; the predicted change trend is obtained based on acoustic information and continuous change information of body posture, which may include, but is not limited to, steps S310, S320 and S330.

[0101] Step S310: Obtain event characteristic information of external acoustic events in the space where the respirator is located;

[0102] Step S320: Based on event feature information, determine the first level of influence of external acoustic events on acoustic information;

[0103] Step S330: Based on the first level of influence, and based on the acoustic information and the continuous change information of body posture, obtain the predicted change trend.

[0104] Among them, the event characteristic information of external acoustic events refers to the attribute data carried by non-user breathing-related sound signals generated in the external environment of the respirator. It can be characterized by parameters such as the spectral characteristics, time-domain characteristics, energy distribution, duration, and instantaneous peak value of the audio signal. The first impact level refers to the quantitative representation of the degree to which the external acoustic event interferes with the acoustic information received by the respirator or the sensors used to acquire acoustic information. It can be represented by numerical values, level classifications, or probability values.

[0105] In the process of predicting trends based on acoustic information and continuous changes in body posture, the proposed solution first acquires event characteristic information of external acoustic events in the space where the respirator is located. This is because sounds in the external environment, such as alarms, ambient noise, and shouts from other people, can mix in or interfere with the acoustic information collected by the respirator or sensors, especially the acoustic information used to identify the user's voice intent. By analyzing the characteristics of these external acoustic events, potential sources of interference can be identified. Next, based on these event characteristic information, the first level of influence of the external acoustic events on the acoustic information is determined. This level quantifies the degree to which the intensity or nature of the external interference affects the usability of the acoustic information. For example, a high-intensity sudden noise may render the acoustic information completely unusable, while continuous low-intensity background noise may only reduce the signal-to-noise ratio. Finally, when obtaining the predicted trend, the solution no longer relies solely on the original acoustic information and continuous changes in body posture, but incorporates the first level of influence. This means that the prediction model or algorithm will adjust its dependence on acoustic information or make corresponding corrections based on the degree of external interference. For example, when the first level of influence is high, the weight of acoustic information in the prediction can be reduced, or body posture information can be prioritized for initial prediction, or the acoustic information can be filtered or denoised before use. It is precisely because of the introduction of the identification and quantification of external acoustic interference, and its use to correct the prediction process, that the predicted trend obtained from the continuous changes in acoustic information and body posture is more accurate. This provides a more reliable input for subsequently determining the second target air supply parameters based on acoustic information and predicted trends, improving the accuracy of the second target air supply parameters, and thus optimizing the airflow control effect of the respirator, especially in scenarios where voice intent is required for control.

[0106] Acquiring event feature information of external acoustic events in the space where the respirator is located can be achieved by collecting ambient sound signals through a microphone configured outside the respirator and performing time-frequency analysis on the signals to extract parameters including, but not limited to, energy distribution, dominant frequency components, and duration as event feature information. Based on the event feature information, determining the first level of influence of external acoustic events on the acoustic information can be done by comparing the extracted event feature information with a pre-defined external acoustic event type library to identify the event type. Then, based on the event type and its feature intensity, a pre-defined mapping table or a trained classifier is consulted to obtain a quantified influence level, such as an integer value from 1 to 5, with higher values ​​indicating greater influence. When predicting the trend based on the first influence level, the acoustic information, and the continuous changes in body posture, if the determined first influence level exceeds a certain threshold, the collected acoustic information can be digitally processed, for example, using spectral subtraction or deep learning noise reduction algorithms to reduce external noise interference, resulting in adjusted acoustic information. Then, the adjusted acoustic information, along with information on the continuous changes in body posture, is input into a predictive model. This model can be a recurrent neural network or a long short-term memory network, used to analyze this time-series data and output a predicted trend of changes in the user's breathing pattern or voice intent.

[0107] like Figure 4 As shown, Figure 4 This is a flowchart of obtaining a predicted change trend provided in another embodiment of this application; the predicted change trend is obtained based on a first influence level, acoustic information, and continuous change information of body posture, and may include, but is not limited to, steps S410 and S420.

[0108] Step S410: When the first influence level is greater than or equal to the preset first influence level, the acoustic information is adjusted according to the first influence level to obtain the adjusted acoustic information.

[0109] Step S420: Based on the adjusted acoustic information and the continuous change information of body posture, the predicted change trend is obtained.

[0110] The preset first impact level refers to a pre-defined threshold used to determine whether the impact of external acoustic events on acoustic information reaches a level requiring correction. This threshold can be implemented using a specific numerical value, a range, or a level indicator, aiming to distinguish between situations requiring acoustic information adjustment and those not requiring adjustment. Adjusting acoustic information based on the first impact level involves processing the original acoustic information according to the specific degree of impact of external acoustic events to reduce or eliminate external interference components. This can be achieved through filtering, noise reduction, spectral correction, or model-based methods, with the goal of obtaining acoustic information that better reflects the user's own state. The adjusted acoustic information... Acoustic information refers to the acoustic information obtained after adjustment processing according to the first level of influence. It can be a processed audio signal, a corrected acoustic feature vector, or acoustic data with some noise components removed. Its purpose is to provide a cleaner and more accurate input for subsequent predictions. Predicting the trend of change refers to the prediction of possible future changes in the ventilator user based on the processed acoustic information and the continuous change information of body posture. Specifically, it can be a predicted value representing the direction and magnitude of changes in physiological parameters such as respiratory rate, tidal volume, and ventilation, or a predicted result representing the user's intention (such as trying to speak or cough). Its purpose is to provide a basis for adjusting the ventilator's air supply parameters.

[0111] Through the above technical solution, when the impact of external acoustic events on acoustic information reaches a certain level, the acoustic information can be effectively corrected, thereby obtaining a more accurate prediction of the changing trend, and thus achieving a more precise adjustment of the respirator's air supply parameters, improving the performance of respiratory support and voice communication.

[0112] like Figure 5 As shown, Figure 5 This is a flowchart of determining the first level of influence of an external acoustic event on acoustic information according to an embodiment of this application; the determination of the first level of influence of an external acoustic event on acoustic information based on event feature information may include, but is not limited to, steps S510, S520 and S530.

[0113] Step S510: Analyze the event feature information to obtain the feature components of the event feature information;

[0114] Step S520: Based on the feature components and the preset external acoustic event type library, determine the type of external acoustic event and the first parameter relationship, wherein the first parameter relationship characterizes the acoustic feature representation and the first influence level relationship of the acoustic information of different external acoustic event types included in the external acoustic event type library;

[0115] Step S530: Determine the first level of influence of the external acoustic event on the acoustic information based on the type of the external acoustic event, the characteristic components of the event characteristic information, and the relationship of the first parameter.

[0116] Among them, event feature information refers to the raw or preliminary processed data extracted from external acoustic events in the space where the respirator is located, used to describe the acoustic characteristics of the event. It can be implemented using time-domain waveform data, frequency-domain spectrum data, time-frequency analysis results, or a combination thereof, with the purpose of providing basic data for subsequent event analysis. Feature components refer to numerical values ​​or vectors that can quantify specific attributes of the event, obtained by decomposing or extracting event feature information. They can be implemented using acoustic parameters such as energy, frequency, amplitude, duration, and instantaneous rate of change, with the purpose of transforming complex acoustic event information into calculable and comparable quantitative indicators. The external acoustic event type library refers to a pre-established database or model containing multiple known external acoustic event types and their corresponding acoustic feature representations and the first influence level relationship of acoustic information. It can be implemented using lookup tables, rule sets, or machine learning models, with the purpose of identifying external acoustic event types and assessing their potential. The impact provides a reference basis; the first parameter relationship refers to the correlation mapping between the typical acoustic feature representation of a specific external acoustic event type recorded in the external acoustic event type database and the preset or average first impact level of that type of event on acoustic information. It can be implemented using numerical, hierarchical, or functional relationships, and its purpose is to provide a preliminary impact assessment standard based on event type; the acoustic feature representation refers to the standardized or modeled representation used to describe the typical acoustic characteristics of various events in the external acoustic event type database. It can be implemented using feature vectors, templates, or statistical models, and its purpose is to serve as a basis for matching and comparing with the feature components of actual events; the first impact level refers to the index that quantifies the degree of interference or impact of external acoustic events on acoustic information inside the respirator mask (especially acoustic information used for voice communication). It can be implemented using numerical scoring, hierarchical classification (such as low, medium, high), or probability values, and its purpose is to provide a basis for subsequent adjustments to acoustic information.

[0117] The solution proposed in this application analyzes the event characteristic information of external acoustic events to obtain quantified feature components, which provides a data foundation for accurate identification and evaluation of external sounds. Based on these feature components and combined with a pre-defined library of external acoustic event types, the system can identify the specific type of external acoustic event and obtain a pre-defined first parameter relationship between that type of event and the impact level of acoustic information. This process is equivalent to classifying and performing a preliminary impact assessment of external sounds.

[0118] Furthermore, the scheme comprehensively considers the identified event type, the specific feature components of the parsed event feature information, and the preset first parameter relationship to determine the final first influence level of the external acoustic event on the acoustic information. This means that the evaluation result does not solely depend on the event type, but rather combines the actual acoustic intensity and characteristics of the event, refining and modifying the preset influence relationship. For example, events of the same "shouting" type may have different feature components (such as amplitude) and thus different degrees of influence on the acoustic information. In this way, the scheme can more accurately quantify the actual interference level of external acoustic events. This more precise first influence level provides a basis for subsequent acoustic information adjustments, making the processing of acoustic information more targeted, effectively reducing the interference of external noise on voice communication, and thus improving the accuracy of voice intent prediction.

[0119] The specific process is as follows: First, external ambient sound is collected through a microphone to obtain raw event characteristic information, such as a segment of audio waveform data. Next, this audio waveform data is analyzed, for example, by performing Fourier transform or wavelet analysis, to extract its energy distribution, peak amplitude, duration, etc., within different frequency ranges as feature components of the event characteristic information. Then, these feature components are matched against a pre-defined external acoustic event type library. This library may contain event types such as "fire alarm," "person shouting," "falling object," and "radio communication," each type associated with typical acoustic characteristics (such as the specific frequency range and pulse pattern of a fire alarm) and a preliminary first impact level (such as a high impact level for a fire alarm and a low impact level for radio communication). By comparing the extracted feature components with acoustic feature representations in the type library, the system can determine the most likely type of the current external acoustic event and obtain the preliminary impact level corresponding to that type. Finally, based on the identified event type, the actual measured feature components, and the first parameter relationship corresponding to that type, the system adjusts the preliminary impact level using a preset calculation model or rule, thereby determining the final first impact level of the external acoustic event on the acoustic information. For example, if the event is identified as "a person shouting," the preliminary impact level is "medium," but the actual amplitude is much higher than the typical value, the final determined first impact level may be adjusted to "medium-high" or "high."

[0120] Through the above technical solution, this application can determine the accurate impact level of external acoustic events on acoustic information based on a refined analysis of the types of external acoustic events and their specific acoustic characteristics. This solves the problem in existing technologies where simple assessment of the impact of external sounds leads to inaccurate subsequent acoustic information adjustments, providing a more reliable input for subsequent speech intent prediction based on acoustic information, thereby improving the accuracy and reliability of respirator airflow control in complex noise environments.

[0121] like Figure 6 As shown, Figure 6 This is a flowchart of another embodiment of the present application for determining the first level of influence of an external acoustic event on acoustic information; the first level of influence of an external acoustic event on acoustic information is determined according to the type of the external acoustic event, the feature components of the event feature information and the first parameter relationship, which may include, but is not limited to, steps S610, S620, S630 and S640.

[0122] Step S610: Obtain the first propagation characteristic characterization of the current actual acoustic propagation characteristics in the space where the respirator is located and the second propagation characteristic characterization of the preset reference acoustic environment associated with the first parameter relationship;

[0123] Step S620: Obtain the first difference based on the first propagation characteristic characterization and the second propagation characteristic characterization, and determine the propagation characteristic correction factor based on the first difference;

[0124] Step S630: Calculate the relationship between the type of external acoustic event, the feature components of the event feature information, and the first parameter to obtain the initial influence level of the external acoustic event on the acoustic information;

[0125] Step S640: Adjust the initial impact level based on the propagation characteristic correction factor to determine the first impact level of external acoustic events on acoustic information.

[0126] The first propagation characteristic characterization refers to the quantitative description of the sound propagation characteristics within the space where the respirator is currently located. It can be characterized using parameters such as reverberation time, frequency response curve, and sound pressure level attenuation rate. The second propagation characteristic characterization refers to the quantitative description of the sound propagation characteristics of a preset reference acoustic environment related to the first parameter. It can be characterized using the same set of parameters as the first propagation characteristic characterization. These parameters can be derived from standard acoustic environment data or pre-measured reference environment data. The first difference refers to the degree of difference between the first and second propagation characteristic characterizations. It can be calculated using parameter difference, vector distance, similarity index, etc. The propagation characteristic correction factor is a correction coefficient used to adjust the initial influence level to reflect the influence of the actual acoustic propagation characteristics. It can be a multiplier, an addend, or a more complex function. This factor is determined based on the first difference.

[0127] The specific implementation of determining the first level of influence of external acoustic events on acoustic information is as follows: First, a preset test signal (e.g., a frequency sweep signal) is played through the microphone built into the respirator, and the response signal received by the same or another microphone as the signal propagates in space is recorded. This response signal is analyzed to calculate the reverberation time of the current space at different frequencies, which serves as the first propagation characteristic characterization. Simultaneously, the reverberation time data of a standard anechoic chamber, associated with the currently used first parameter relationship (e.g., a lookup table based on standard anechoic chamber data), is consulted, serving as the second propagation characteristic characterization. Next, the average or maximum difference between the reverberation time of the current space and the reverberation time of the standard anechoic chamber within the key frequency range is calculated to obtain the first difference. Based on this first difference, a preset correction factor lookup table is consulted, which maps different reverberation time difference values ​​to corresponding propagation characteristic correction factors. For example, if the reverberation time of the current space is much greater than that of the standard anechoic chamber, the correction factor may be greater than 1, indicating slow sound attenuation and the potential amplification of the impact of external events; if it is much smaller, the correction factor may be less than 1. Then, based on the type of the external acoustic event (e.g., identified as an alarm sound), the characteristic components of the event's feature information (e.g., loudness of 80dB and dominant frequency of 1kHz), and a preset first parameter relationship (e.g., a lookup table showing that the initial impact level of this type of event in the reference environment is moderate), the initial impact level of the external acoustic event on the acoustic information is calculated (e.g., a value of 0.6). Finally, this initial impact level is multiplied by the previously determined propagation characteristic correction factor (or other forms of combined calculation) to obtain the final first impact level of the external acoustic event on the acoustic information. Through the above technical solution, when determining the first impact level of the external acoustic event on the acoustic information based on the type of the external acoustic event, the characteristic components of the event's feature information, and the first parameter relationship, the current actual acoustic propagation characteristics of the space where the respirator is located and the propagation characteristics of the preset reference acoustic environment can be obtained. Based on the difference between the two, a propagation characteristic correction factor is determined, and this correction factor is used to adjust the initial impact level. This allows the determined first impact level to reflect the impact of the actual acoustic environment on sound propagation, overcoming the problem of potential bias caused by relying solely on preset relationships for evaluation, and improving the accuracy of the impact level assessment. Accurate impact level assessment provides a reliable basis for subsequent acoustic information adjustments based on this level. For example, when a high-impact level event is detected, the airflow pattern can be adjusted more precisely to reduce noise interference, thereby improving speech clarity, or the airflow pattern can be maintained at a low impact level to conserve air supply, ultimately optimizing the airflow control effect of the respirator under different acoustic environments.

[0128] like Figure 7 As shown, Figure 7This is a flowchart of determining a propagation characteristic correction factor based on a first difference, provided in one embodiment of this application; the first difference is obtained based on the first propagation characteristic characterization and the second propagation characteristic characterization, and the propagation characteristic correction factor level is determined based on the first difference, which may include, but is not limited to, steps S710, S720 and S730.

[0129] Step S710: Decompose the first difference to obtain multiple difference components;

[0130] Step S720: Determine the component weight corresponding to each difference component based on each difference component.

[0131] Step S730: Calculate the propagation characteristic correction factor for each difference component and component weight.

[0132] The first difference refers to the overall difference between the actual acoustic propagation characteristics and the propagation characteristics of the preset reference acoustic environment. It can be represented by the difference, ratio, or a more complex functional relationship between the two acoustic propagation characteristics. The difference component refers to a finer-grained difference term obtained after decomposing the overall first difference. It can be represented by frequency domain difference, time domain difference, energy difference, or differences in specific acoustic parameters (such as reverberation time and signal-to-noise ratio). The component weight is a numerical value used to measure the contribution of each difference component to the final propagation characteristic correction factor. It can be determined using a preset constant, an empirically set proportional coefficient, or a dynamic value obtained through data analysis and model training. The propagation characteristic correction factor is a correction coefficient used to adjust the initial influence level of external acoustic events on acoustic information. It can be applied using a multiplicative factor, an additive factor, or a more complex nonlinear function.

[0133] The method for determining the propagation characteristic correction factor can be implemented as follows: First, obtain the frequency domain response curve of the actual acoustic propagation characteristics and the frequency domain response curve of a preset reference acoustic environment. The difference between these two curves is taken as the first difference. Then, the frequency domain difference is decomposed into multiple frequency band difference components. For example, the entire frequency band can be divided into low-frequency, mid-frequency, and high-frequency bands, and the average or maximum difference within each band is calculated as the corresponding difference component. Next, the component weight is determined based on the degree of influence of each frequency band on speech intelligibility. For example, the weight of the mid-frequency band can be set higher than that of the low-frequency and high-frequency bands. The weight can also be determined by analyzing historical data, such as obtaining weights based on acoustic information collected under different acoustic environments and corresponding speech recognition results. Finally, the difference component of each frequency band is multiplied by its corresponding component weight, and all products are summed to obtain the final propagation characteristic correction factor.

[0134] The above technical solution decomposes the overall primary difference into multiple difference components, determines the corresponding component weight based on the importance of each component, and then obtains the propagation characteristic correction factor through weighted calculation. This processing method can more precisely analyze and utilize the differences between the actual acoustic environment and the reference environment, enabling the determined propagation characteristic correction factor to more accurately reflect the complex relationship between different propagation characteristics and the impact of acoustic events. This improves the accuracy of the propagation characteristic correction factor, thereby enhancing the accuracy of determining the impact level of external acoustic events based on acoustic information.

[0135] like Figure 8 As shown, Figure 8 This is a flowchart of determining the component weights corresponding to the difference components based on the difference components according to an embodiment of this application; determining the component weights corresponding to each difference component based on each difference component may include, but is not limited to, steps S810, S820 and S830.

[0136] Step S810: Obtain historical data of the difference components within a preset time period;

[0137] Step S820: Analyze the changing trends of the difference components based on historical data;

[0138] Step S830: Determine the component weights corresponding to the difference components based on the changing trend.

[0139] Among them, the difference component refers to the various components obtained after decomposing the first difference between the actual acoustic propagation characteristics and the propagation characteristics of the reference acoustic environment. It can represent different dimensions or manifestations of acoustic propagation characteristics, such as attenuation differences, phase delay differences, and reverberation time differences within a specific frequency range. Its purpose is to refine the complex overall difference into units that can be independently analyzed and processed. Historical data refers to a series of numerical sets obtained by continuously or periodically measuring, recording, or calculating specific difference components within a preset time period. It can be stored in the form of time series data, aiming to provide dynamic information on the changes of the difference component over time. The trend of change refers to... The analysis of historical data of the differential component reveals its evolution pattern or stability characteristics over a predetermined time period. This can be characterized using statistical methods (such as mean, variance, standard deviation, and rate of change calculation) or signal processing methods (such as trend line analysis and volatility analysis). The purpose is to assess the reliability and importance of the differential component. Component weight refers to a numerical value assigned to the differential component based on its changing trend. This value reflects the relative importance or contribution proportion of the differential component in calculating the propagation characteristic correction factor. It can be represented by a normalization coefficient or a scaling factor. The purpose is to dynamically adjust the influence of different differential components on the final correction result.

[0140] Specifically, acquiring historical data of the differential components within a preset time period can be achieved by continuously collecting acoustic signals through sensors and periodically calculating the values ​​of the differential components. These values ​​are then stored in a circular buffer, the size of which corresponds to a preset time length, such as storing data from the most recent 60 seconds. Analyzing the changing trends of the differential components based on historical data can be done by calculating the standard deviation of the differential component values ​​within the circular buffer, or by calculating the difference between its maximum and minimum values ​​within the preset time period, thereby measuring its volatility. Determining the component weights corresponding to the differential components based on the changing trends can be achieved by establishing a mapping relationship. For example, if the calculated standard deviation is less than a certain threshold, a higher weight is assigned (e.g., 0.8-1.0); if the standard deviation is greater than the threshold, the weight is reduced linearly or non-linearly according to the magnitude of the standard deviation (e.g., the weight is inversely proportional to the standard deviation), thus ensuring that the more stable the differential components, the higher their weights.

[0141] The above technical solution can dynamically adjust the importance of each difference component in the calculation of the propagation characteristic correction factor according to the historical trend of each component, thereby calculating the propagation characteristic correction factor more accurately, improving the accuracy of the assessment of the impact level of external acoustic events, and ultimately improving the performance of the respirator airflow regulation.

[0142] The proposed solution acquires historical data of the differential component over a preset time period, enabling the observation and recording of its changes over time. Based on this historical data, the changing trends of the differential component can be analyzed. Since the changing trends of different differential components reflect the stability and reliability of their impact on acoustic propagation, it becomes possible to determine the component weights corresponding to the differential components based on these trends. By assigning higher weights to differential components with stable changing trends and lower weights to those with drastic changing trends, it ensures that more reliable and representative differential components play a dominant role in calculating the propagation characteristic correction factor, while reducing interference from unstable or transient factors. This dynamic weight allocation mechanism based on historical data and changing trends allows the propagation characteristic correction factor to more accurately reflect the complexity of the current actual acoustic propagation environment and to make a more precise comparison with the reference environment, thereby improving the accuracy of assessing the impact level of external acoustic events on the internal acoustic information of the respirator. This more accurate impact level assessment further enhances the reliability of determining target air supply parameters based on voice intent prediction information, ultimately optimizing the performance of the respirator's airflow control, enabling it to more accurately respond to the user's voice communication needs while balancing safety and air supply efficiency.

[0143] In some embodiments, this application further proposes a respirator airflow regulation system, comprising:

[0144] The acquisition module is used to acquire target risk prediction information, which includes posture risk prediction information and voice intent prediction information.

[0145] The target gas supply parameter determination module is used to determine the target gas supply parameters based on the target risk prediction information;

[0146] The air supply parameter adjustment module is used to adjust the air supply parameters of the respirator according to the target air supply parameters;

[0147] In the case where the target risk prediction information is attitude risk prediction information, the respirator's air supply parameters are adjusted according to the first target air supply parameters.

[0148] When the target risk prediction information is voice intent prediction information, the respirator's air supply parameters are adjusted according to the second target air supply parameters.

[0149] When the target risk prediction information consists of posture risk prediction information and voice intent prediction information, the respirator's air supply parameters are adjusted according to the third target air supply parameters.

[0150] The acquisition module is a unit used to collect or receive external information and convert it into signals that the system can process. It can be implemented using sensor interface circuits, data acquisition units, or communication interfaces. The target air supply parameter determination module is a unit that performs logical judgments and calculations based on input information and outputs control parameters. It can be implemented using microcontrollers, digital signal processors, or application-specific integrated circuits. The air supply parameter adjustment module is a unit that drives the actuator according to the control parameters to change the respirator's air supply state. It can be implemented using motor drive circuits, valve control circuits, or actuator drive units.

[0151] The proposed solution implements the key steps of the respirator airflow control method by dividing them into an acquisition module, a target air supply parameter determination module, and an air supply parameter adjustment module. The acquisition module is responsible for acquiring target risk prediction information in real time, reflecting user posture changes and voice intent, providing input for subsequent control. The target air supply parameter determination module receives information from the acquisition module, analyzes different risk scenarios (postural risk, voice intent risk, or both) according to preset logic or algorithms, and calculates the corresponding target air supply parameters. The air supply parameter adjustment module controls the respirator's air supply actuators based on the target air supply parameters output by the target air supply parameter determination module, such as adjusting fan speed or valve opening, thereby changing parameters such as air flow and pressure. This modular system structure allows the method steps to be physically implemented and executed efficiently. The information flow between modules is clear, facilitating coordination and data processing, thus ensuring the method can be effectively applied to real-world scenarios and solving practical application problems that are difficult to overcome by relying solely on method descriptions, such as inter-module coordination and rapid data processing and transmission. This system implementation approach enables faster and more accurate responses to user status changes and environmental risks, improving the real-time performance and accuracy of respirator airflow control.

[0152] The acquisition module may include a sensor interface circuit connecting the attitude sensor and microphone, and a data acquisition chip that digitizes the sensor signals. The target air supply parameter determination module may employ a microcontroller, which internally runs an algorithm program to determine the target air supply parameters based on attitude risk prediction information and voice intent prediction information. The air supply parameter adjustment module may include a motor drive circuit connected to the respirator fan, or a proportional valve control circuit connected to the respirator control valve. The sensor interface circuit acquires signals from the attitude sensor and microphone, and the data acquisition chip converts the analog signals into digital signals before sending them to the microcontroller. Based on the received digital signals, the microcontroller executes the algorithm program to calculate the target air supply parameters for the current state. Then, the microcontroller sends control signals to the motor drive circuit or proportional valve control circuit through an output interface to drive the fan or control valve, thereby adjusting the respirator's air supply parameters. For example, when the microcontroller determines that there is an attitude risk, it outputs a control signal to cause the motor drive circuit to increase the fan speed and increase the air supply flow to cope with possible mask leakage or increased breathing resistance. When it determines that there is a voice intent, it outputs a control signal to cause the proportional valve control circuit to adjust the valve opening to optimize the acoustic environment inside the mask.

[0153] The above technical solution provides a ventilator airflow control system. This system, through a modular design, concretizes each step of the ventilator airflow control method, thereby achieving precise control of the ventilator's air supply parameters. The system can acquire user posture risk prediction information and voice intent prediction information in a timely and accurate manner, and flexibly determine and adjust target air supply parameters according to different risk situations. This system implementation method enables the method to be executed effectively, solving problems that may arise in practical applications such as module coordination, rapid data processing and transmission, improving the real-time performance, accuracy, and adaptability of ventilator airflow control, and ensuring the user's respiratory safety and comfort.

[0154] The above description is merely an embodiment of this application and is not intended to limit the scope of protection of this application. Various modifications and variations can be made to this application by those skilled in the art. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of this application should be included within the scope of protection of this application.

Claims

1. A method for regulating airflow in a respirator, applied to a respirator, characterized in that, include: Obtain target risk prediction information, wherein the target risk information includes posture risk prediction information and voice intent prediction information; Determine the target gas supply parameters based on the target risk prediction information; Adjust the respirator's air supply parameters according to the target air supply parameters; Wherein, when the target risk prediction information is attitude risk prediction information, the air supply parameters of the respirator are adjusted according to the first target air supply parameters; When the target risk prediction information is voice intent prediction information, the respirator's air supply parameters are adjusted according to the second target air supply parameters. When the target risk prediction information is posture risk prediction information and voice intent prediction information, the air supply parameters of the respirator are adjusted according to the third target air supply parameters. The respirator includes a mask and a detection device, the detection device being used to acquire the mask's sealing information, breathing patency information, and acoustic information; obtaining the target air supply parameters based on the target risk prediction information includes at least one of the following: When the target risk information is attitude risk prediction information, the first target air supply parameter is determined based on the sealing information and the breathing patency information; When the target risk information is voice intent prediction information, the second target gas supply parameters are determined based on the acoustic information; When the target risk information is posture risk prediction information and voice intent prediction information, the third target air supply parameters are determined based on the sealing information, the breathing patency information and the acoustic information.

2. The method according to claim 1, characterized in that, When the target risk information is voice intent prediction information, determining the second target gas supply parameters based on the acoustic information includes: Acquire information on continuous changes in body posture; Based on the acoustic information and the continuous change information of the body posture, the predicted change trend is obtained; The second target gas supply parameters are determined based on the acoustic information and the predicted change trend.

3. The method according to claim 2, characterized in that, The step of predicting the trend of change based on the acoustic information and the continuous change information of the body posture includes: Acquire event characteristic information of external acoustic events in the space where the respirator is located; Based on the event characteristic information, a first level of influence of the external acoustic event on the acoustic information is determined; Based on the first influence level, the acoustic information, and the continuous change information of the body posture, a predicted change trend is obtained.

4. The method according to claim 3, characterized in that, The step of obtaining the predicted trend based on the first influence level, the acoustic information, and the continuous change information of the body posture includes: When the first influence level is greater than or equal to the preset first influence level, the acoustic information is adjusted according to the first influence level to obtain the adjusted acoustic information; Based on the adjusted acoustic information and the continuous change information of the body posture, the predicted change trend is obtained.

5. The method according to claim 3, characterized in that, The step of determining the first level of influence of the external acoustic event on the acoustic information based on the event feature information includes: The event feature information is parsed to obtain the feature components of the event feature information; Based on the feature components and a preset external acoustic event type library, the type of the external acoustic event and a first parameter relationship are determined, wherein the first parameter relationship characterizes the acoustic feature representation of different external acoustic event types included in the external acoustic event type library and the first influence level relationship of the acoustic information; Based on the type of the external acoustic event, the feature components of the event feature information, and the relationship of the first parameter, the first level of influence of the external acoustic event on the acoustic information is determined.

6. The method according to claim 5, characterized in that, Based on the type of the external acoustic event, the feature components of the event feature information, and the first parameter relationship, a first influence level of the external acoustic event on the acoustic information is determined, including: The first propagation characteristic characterization of the current actual acoustic propagation characteristics in the space where the respirator is located and the second propagation characteristic characterization of the preset reference acoustic environment associated with the first parameter relationship are obtained. Based on the first propagation characteristic characterization and the second propagation characteristic characterization, a first difference is obtained, and a propagation characteristic correction factor is determined according to the first difference; The relationship between the type of the external acoustic event, the feature components of the event feature information, and the first parameter is calculated to obtain the initial influence level of the external acoustic event on the acoustic information; The initial impact level is adjusted based on the propagation characteristic correction factor to determine the first impact level of the external acoustic event on the acoustic information.

7. The method according to claim 6, characterized in that, The step of determining the propagation characteristic correction factor based on the first difference includes: The first difference is decomposed to obtain multiple difference components; Determine the component weight corresponding to each of the aforementioned difference components based on each of the difference components; The propagation characteristic correction factor is determined by calculating each difference component and its weight.

8. The method according to claim 7, characterized in that, The step of determining the component weight corresponding to the difference component based on the difference component includes: Obtain historical data of the difference components within a preset time period; Based on the historical data, analyze the changing trends of the difference components; The component weights corresponding to the difference components are determined based on the changing trend.

9. A respirator airflow regulation system for performing the respirator airflow regulation method as described in any one of claims 1-8, characterized in that, include: The acquisition module is used to acquire target risk prediction information, wherein the target risk information includes posture risk prediction information and voice intent prediction information; The target gas supply parameter determination module is used to determine the target gas supply parameters based on the target risk prediction information; An air supply parameter adjustment module is used to adjust the air supply parameters of the respirator according to the target air supply parameters. Wherein, when the target risk prediction information is attitude risk prediction information, the air supply parameters of the respirator are adjusted according to the first target air supply parameters; When the target risk prediction information is voice intent prediction information, the air supply parameters of the respirator are adjusted according to the second preset target air supply parameters. When the target risk prediction information is posture risk prediction information and voice intent prediction information, the respirator's air supply parameters are adjusted according to the third target air supply parameters.

Images