An adaptive audio adjustment sleep-aiding method based on sleep state recognition

Abstract

The application belongs to the technical field of smart home, wearable device, biological signal processing and artificial intelligence control, and specifically discloses a self-adaptive audio adjustment sleep-aiding method based on sleep state recognition, which comprises the following steps: a non-invasive touch interaction mode is introduced to simplify the starting process of the sleep-aiding process; key physiological indexes such as the heart rate and body movement of a user are continuously monitored, and subtle changes in the data are analyzed to determine the physiological state of the user and the reaction of the user to environmental changes, so that the change in the acoustic environment is ensured to be always within the comfortable range that can be accepted by the user; the amplitude of the volume adjustment is controlled to be below the perception threshold of the user, and a special disturbance observation window is set to evaluate the physiological influence of each fine adjustment operation, so that inappropriate adjustment can be found and cancelled in time before substantial interference is caused. The application significantly improves the reliability of the user experience and sleep-aiding effect.

Description

TECHNICAL FIELD

[0001] The present application belongs to the technical field of smart home, wearable devices, biological signal processing and artificial intelligence control, and relates to a self-adaptive audio adjustment sleep-aiding method based on sleep state recognition. BACKGROUND

[0002] Currently, in the technical practice of assisting sleep by creating an acoustic environment, the core challenge is how to achieve intelligent, personalized and non-disturbing management of sleep-aiding audio. From wakefulness to deep sleep, it is a fragile process of physiological state dynamic change, and the sensitivity to external acoustic stimulation also changes accordingly. A fixed audio environment cannot perfectly match this dynamic process. The key problem is the lack of a closed-loop control system that can perceive the user's sleep process in real time and adjust the acoustic environment accordingly, thus avoiding interference and even interruption of the user's sleep process due to excessive or inappropriate changes in audio.

[0003] The commonly used solutions or traditional methods in the industry mainly rely on user presets and manual operation. For example, users select a sleep-aiding audio through a mobile application or voice commands from a smart speaker, set a fixed playback volume and duration or a timed shutdown. Some more advanced methods monitor sleep state through wearable devices and, after determining that the user is asleep, directly turn off the audio or switch it to a very low volume according to the preset program. These methods require explicit user interaction during the start-up and adjustment process, and their adjustment logic is based on preset rules or simple sleep state determination, rather than continuous physiological feedback.

[0004] The drawbacks of traditional methods are obvious. First, the start-up process is cumbersome, and users still need to interact with electronic device screens or voice assistants before sleep, which can affect sleepiness. Second, the audio adjustment strategy is too rigid, and a fixed volume cannot adapt to the dynamic changes in user sensitivity during the sleep transition period, which may be too loud when the user is about to fall asleep. Abrupt shutdown or switching based on simple sleep determination may also be a new acoustic disturbance, disturbing users in light sleep due to delayed or inaccurate judgment. These methods are essentially open-loop or semi-open-loop systems, lacking the ability to immediately understand and feedback the subtle physiological disturbances caused by adjustment operations, and cannot truly create an acoustic comfort envelope that closely matches individual real-time needs. SUMMARY

[0005] In order to overcome the above-mentioned defects of the prior art, in order to achieve the above-mentioned purposes, the present application proposes the following technical scheme: A self-adaptive audio adjustment sleep-aiding method based on sleep state recognition, comprising: S1, obtaining a user sleep intention: when a wearable device worn on the user's body physically touches or directly connects with an audio device through Bluetooth, a composite instruction preset in the wearable device is read through near field communication technology to generate a handshake authorization instruction.

[0006] S2, establishing a physiological signal exclusive channel: based on the handshake authorization instruction, a low-latency wireless data connection is established between the wearable device and the audio device, and an initial acoustic environment is initialized.

[0007] S3, identifying physiological stable trend: the wearable device collects physiological signal stream and analyzes the heart rate signal fluctuation range and the comprehensive change amount of body motion signal in the physiological signal stream, identifies the sleep transition trend, and continuously transmits the real-time analyzed sleep transition trend to the audio device through the physiological signal exclusive channel.

[0008] S4, generating tentative adjustment parameters: in response to the sleep transition trend, locking the reference safe volume, and starting the fine-tuning cycle logic to generate acoustic fine-tuning parameters for slightly reducing the audio volume.

[0009] S5, performing acoustic environment fine-tuning: the audio device applies the acoustic fine-tuning parameters to perform a non-perceptible volume reduction operation on the playing sleep-aiding audio to generate a disturbance observation window.

[0010] S6, determining physiological disturbance feedback: in the disturbance observation window, the physiological signal stream data in the physiological signal exclusive channel is continuously analyzed to determine whether the physiological stress reaction caused by the volume reduction exists, and a regulation effectiveness report is generated.

[0011] S7, iteratively shaping acoustic comfort envelope: according to the result of the regulation effectiveness report, it is decided whether to set the current reduced volume as the new safe reference, and the steps S4 to S6 are cycled back until the audio volume is reduced to the preset mute threshold, forming an acoustic comfort envelope closely fitting the user's individual sensitivity curve.

[0012] Compared with the prior art, the present application has the following advantages: (1) The present application introduces a non-invasive touch interaction method, which greatly simplifies the starting process of the sleep-aiding process. The user only needs to physically touch the wearable device and the audio device once, which can complete the identity authorization, device pairing and function start at the same time. The whole process is intuitive and does not need to go through the steps of lighting the screen, unlocking the device, finding the application, etc., effectively reducing the cognitive interaction with electronic devices before sleep, avoiding the interference of blue light and complex operation on the user's sleep intention.

[0013] (2) The present application realizes unprecedented refinement and individualization control of the sleep-aiding acoustic environment by establishing a closed-loop feedback regulation system based on real-time physiological signals. Instead of relying on preset rigid rules, the system continuously monitors key physiological indicators such as heart rate and body movement of the user, analyzes the subtle changes in these data to determine the user's physiological state and response to environmental changes. This dynamic adjustment based on biological feedback ensures that each volume adjustment is a data-supported trial, ensuring that the changes in the acoustic environment are always within the user's physiological comfort zone, thus dynamically shaping an acoustic comfort envelope that perfectly fits the user's individual sleep curve for the night.

[0014] (3) The present application effectively solves the core problem of safely reducing environmental volume without disturbing the user by adopting an iterative adjustment strategy. By controlling the volume adjustment amplitude below the user's perception threshold and setting up a special disturbance observation window to evaluate the physiological impact of each fine-tuning operation, the system can discover and reverse inappropriate adjustments in time before causing substantial interference. This cautious and gradual adjustment logic ensures that the entire sleep-aiding audio volume reduction process is smooth, safe and imperceptible, maximizing the protection of the user's fragile transition from wakefulness to sleep from acoustic environment mutations.

[0015] (4) The present application significantly improves the reliability of user experience and sleep-aiding effect by seamlessly integrating user authorization, device connection, physiological monitoring and environmental control into an automated intelligent system. From the first touch when the user expresses the intention to fall asleep, the entire system takes over all subsequent complex judgments and operations without any further intervention from the user. The system not only intelligently identifies sleep trends, but also autonomously explores and finds the most suitable volume reduction path, ultimately reducing the environmental volume to silence after the user enters stable sleep, achieving full-process automation from start to finish and providing an efficient and personalized sleep-aiding solution for users. BRIEF DESCRIPTION OF DRAWINGS

[0016] In order to more clearly illustrate the technical solutions of the embodiments of the present application, the following will briefly introduce the drawings needed to be used in the embodiment description. Obviously, the drawings in the following description are only some embodiments of the present application, and other drawings can also be obtained by those skilled in the art without creative labor on the basis of these drawings.

[0017] Figure 1 The present application is a method for implementing step flowchart. DETAILED DESCRIPTION

[0018] With reference to the drawings of the embodiments of the present application, the technical solutions in the embodiments of the present application will be clearly and completely described, obviously, the described embodiments are only a part of the embodiments of the present application, rather than all the embodiments. Based on the embodiments of the present application, all other embodiments obtained by those skilled in the art without creative work belong to the scope of protection of the present application.

[0019] Please refer to Figure 1 As shown in the drawings, the adaptive audio adjustment sleep-aiding method based on sleep state recognition provided by the present application comprises: S1, obtaining the sleep intention of a user: when a wearable device worn on the body of the user physically touches an audio device or directly connects with the audio device through Bluetooth, a composite instruction preset in the wearable device is read through near field communication technology to generate a handshake authorization instruction.

[0020] In a preferred embodiment, the generation of the handshake authorization instruction comprises: S1.1, identifying a near field contact event between the wearable device and the audio device within a preset physical distance, and activating the near field communication function when the near field contact event is identified or the Bluetooth direct connection is identified.

[0021] S1.2, parsing the composite instruction containing the Bluetooth connection address and the start sleep-aiding mode identifier from the near field communication tag integrated in the wearable device.

[0022] S1.3, encapsulating the composite instruction as a handshake authorization instruction carrying a unique timestamp, which represents the authorization of the user to the system to take over the control.

[0023] In a further preferred embodiment, the identification of the near field contact event between the wearable device and the audio device within a preset physical distance specifically comprises: integrating a near field communication tag in the wearable device and integrating a near field communication reader in the audio device, when the body part of the user wearing the wearable device is close to the designated sensing area of the audio device, and the distance between the two enters the preset physical distance, the near field communication reader in the audio device senses the near field communication tag in the wearable device, and this sensing event is marked as a near field contact event.

[0024] Specifically, this step aims to safely and conveniently start the sleep-aiding process through a physical touch or a Bluetooth direct connection between the wearable device and the audio device. After the physical touch or the Bluetooth direct connection occurs, a passive near field communication tag is first integrated in a wearable device, and a near field communication reader is integrated in an audio device. The wearable device is an electronic device worn on the body of a user, such as a smart bracelet or a smart watch, which has a near field communication tag and a physiological sensor integrated therein. The audio device is an electronic device capable of playing audio, such as a smart speaker, which has a near field communication reader and a wireless communication module embedded therein.

[0025] The operation flow is as follows: the user will wear the wearable device on the wrist or other body parts, close to the designated induction area of the audio device, when the distance between the two enters the pre-set physical distance, for example, less than 5 cm, the near field communication reader in the audio device will sense the near field communication tag in the wearable device, and this sensing event is recognized by the system as a near field contact event, and the near field communication function of the two devices is immediately fully activated to prepare for data exchange. The pre-set physical distance is the effective sensing distance for triggering the near field communication function, which is set to 5 cm based on the near field communication technology standard, to ensure the explicitness of the user's operation intention and avoid accidental touch. Near field communication technology is a short-range, high-frequency radio communication technology that allows electronic devices to perform non-contact point-to-point data transmission within a few centimeters.

[0026] After successful activation, the audio device will read a pre-stored composite instruction from the near field communication tag of the wearable device. The composite instruction is parsed to extract two core data fields, namely the Bluetooth connection address for subsequent wireless connection and the start sleep aid mode identifier for starting specific functions. To ensure the effectiveness and security of the instruction, the system obtains the current accurate system time, generates a unique timestamp, and encapsulates the timestamp with the just parsed composite instruction into a new data packet. This finally generated data packet, i.e. a handshake authorization instruction, represents the user's explicit intention to authorize the audio device to take control of the sleep aid process at a specific time.

[0027] The composite instruction is a data record pre-stored in the near field communication tag of the wearable device, and its data structure contains two fixed fields, one field storing the Bluetooth connection address and the other field storing a specific Boolean or enumerated value representing the start of the sleep aid function.

[0028] The handshake authorization instruction is a dynamically generated data structure composed of data parsed from the composite instruction and an additional unique timestamp, used to establish a trust relationship between devices and deliver the initial instruction. The data format of this instruction is a JSON object containing Bluetooth address, mode identifier and timestamp.

[0029] For example, a smart bracelet worn by a user as a wearable device has a preset composite instruction in its internal near-field communication tag, which is "DA:A1:B2:C3:D4:E5, SleepMode=1". The user touches the bracelet as a smart speaker that is an audio device, and the distance between the two is less than the preset physical distance of 5 cm. The smart speaker recognizes the near-field contact event, activates the near-field communication function, reads and analyzes the composite instruction, obtains the Bluetooth connection address "DA:A1:B2:C3:D4:E5" and the sleep aid mode identifier "SleepMode=1". The system obtains the current time "2023-10-27T22:30:05.123Z" as a unique timestamp, and encapsulates the three data into a handshake authorization instruction.

[0030] S2, Establish a physiological signal dedicated channel: based on the handshake authorization instruction, a low-latency wireless data connection is established between the wearable device and the audio device, and an initial acoustic environment is initialized.

[0031] In a preferred embodiment, the establishment of the physiological signal dedicated channel comprises: S2.1, the audio device analyzes the handshake authorization instruction, extracts the Bluetooth connection address and automatically completes the pairing with the wearable device to establish a data connection.

[0032] S2.2, the audio device starts playing the preset sleep aid audio according to the sleep aid mode identifier in the handshake authorization instruction, forming an initial acoustic environment for shielding environmental noise.

[0033] S2.3, the wearable device starts continuously transmitting real-time data to the audio device at a predetermined frequency through the data connection, forming a physiological signal dedicated channel.

[0034] The predetermined frequency is the fixed rate at which the wearable device sends real-time data, which is set to 1 Hz. This setting is based on the analysis of hundreds of sets of sleep experiment data, balancing the ability to capture key physiological changes and saving the energy consumption of the wearable device.

[0035] In a further preferred embodiment, the physiological signal stream collected by the wearable device comprises: the wearable device starts its built-in optical heart rate sensor and accelerometer to collect real-time physiological data of the user, including heart rate signals and body motion signals, which are multi-axis acceleration value signals of the user's body movement.

[0036] The collected data is formatted into a unified user physiological signal stream.

[0037] Specifically, after receiving the handshake authorization instruction, the central processor in the audio device immediately parses the instruction and accurately extracts two key pieces of information, namely the Bluetooth connection address encapsulated by the previous link and the start sleep aid mode identifier. Using the extracted Bluetooth connection address, the audio device initiates a wireless pairing request to the designated wearable device. Since the handshake authorization instruction itself implies the user's authorization intention, the wearable device will automatically accept the pairing request without any additional confirmation operation by the user, thereby quickly establishing a low-latency wireless data connection between the two. The low-latency wireless data connection is a communication link that ensures a very low time delay from the sending end to the receiving end. In this scheme, low-power Bluetooth technology is used to control the delay within 50 milliseconds, ensuring the real-time nature of physiological signals.

[0038] At the same time, the audio device checks the parsed start sleep aid mode identifier. After confirming that the identifier is valid, in the case of physical touch, the audio device will call a pre-set sleep aid audio file stored in its internal storage, such as a segment of smooth white noise or rain sound recording, and start playing it externally through the speaker; in the case of Bluetooth direct connection, the audio device will call the online audio selected by the user. The pre-set sleep aid audio is one or more audio files pre-stored in the audio device, and the acoustic characteristics of which have been shown to help relax and fall asleep, such as continuous and no obvious change of pink noise. This continuous background sound constitutes an initial acoustic environment, which is mainly used to mask the sudden or irregular noise in the environment that may interfere with sleep.

[0039] After the low-latency wireless data connection is successfully established, the wearable device immediately starts its built-in physiological sensors, such as optical heart rate sensors and multi-axis accelerometers, to start real-time collection of user physiological data, and formats the collected physiological data into a unified user physiological signal stream. The user physiological signal stream is a series of data packets arranged in chronological order, each data packet containing multiple physiological indicators collected at a specific time, such as heart rate and body movement acceleration values in the standard coordinate system X-axis, Y-axis, Z-axis.

[0040] The stable data link from the wearable device to the audio device, which is specifically used to transmit real-time data, is the physiological signal dedicated channel. The physiological signal dedicated channel is a special data path temporarily established for sleep aid function, which uses low-latency wireless data connection to continuously and unidirectionally transmit the user physiological signal stream from the wearable device to the audio device after parsing.

[0041] For example, the audio device receives a handshake authorization instruction containing a Bluetooth connection address "DA:A1:B2:C3:D4:E5" and a sleep mode identifier "SleepMode=1". The audio device then automatically completes pairing with the wearable device corresponding to the Bluetooth connection address, establishing a low-latency wireless data connection. At the same time, according to the sleep mode identifier "SleepMode=1", the audio device starts playing a preset sleep-aiding audio, i.e. a gentle stream sound, to form an initial acoustic environment. After the connection is established, the wearable device starts collecting user physiological signal streams in real time at a predetermined frequency of 1 Hz, such as a data packet containing a heart rate of 75 beats per minute, body motion acceleration data of X-axis 0.1g, Y-axis -0.05g, and Z-axis 0.98g, and sends the data packet to the audio device through the connection after data analysis, thereby forming a physiological signal exclusive channel.

[0042] S3, identify the physiological stable trend: collect physiological signal streams through the wearable device and analyze the heart rate signal fluctuation range and the integrated change amount of body motion signals, identify the sleep transition trend, and continuously transmit the real-time analyzed sleep transition trend to the audio device through the physiological signal exclusive channel.

[0043] The physiological stable trend refers to the overall change process of the user's physiological indicators, particularly the heart rate and body activity, from active or irregular state to smooth and stable state. The system judges the sleep intention by identifying this trend.

[0044] In a preferred embodiment, the sleep transition trend is identified by S3.1, using a sliding time window to process continuous physiological signal stream data, calculating the difference between the maximum and minimum values of the user's physiological signal stream heart rate signal in each continuous sliding time window to obtain the heart rate fluctuation range.

[0045] By comparing the heart rate fluctuation range of the current sliding time window with that of the previous several sliding time windows, it is determined whether the heart rate fluctuation range presents a continuous decreasing trend and finally stabilizes in a preset narrow interval. If it is confirmed that the state presents, it is determined that the user's heart rate fluctuation converges.

[0046] S3.2, synchronously analyze whether the integrated change amount of the body motion signal in the continuous sliding time window presents a state of continuously being lower than a preset static threshold. If it is confirmed that the state presents, it is determined that the user's body tends to be static.

[0047] S3.3, when both the user's heart rate fluctuation convergence and the user's body tending to be static are met, it is determined that the user enters the transition stage from wakefulness to sleep, and a sleep transition trend identifying the stage is generated.

[0048] Specifically, the wearable device receives and caches the user's physiological signal stream continuously. To perform trend analysis, the system employs a sliding time window to process these continuous data. The sliding time window is a fixed length time period, but it slides constantly over time. After each slide, the latest data is included in the window while the old data beyond the window is discarded. This is used to dynamically capture and analyze the continuous data in the physiological signal stream. The window is set to 60 seconds. This setting is based on statistical analysis of 200 sets of sleep monitoring data, which found that the physiological changes before falling asleep can be most effectively observed at this time scale.

[0049] First, the system calculates the difference between the maximum and minimum values of the heart rate signal in each continuous sliding time window. This gives the fluctuation range of the heart rate. The system compares the heart rate fluctuation range of the current sliding time window with that of the previous several sliding time windows to determine whether the range shows a consistent decreasing trend, i.e., convergence, and eventually stabilizes in a pre-set narrow interval.

[0050] While analyzing the heart rate, the system also processes the body motion signal in the same sliding time window. This signal comes from the accelerometer in the wearable device. The system calculates the total change in all acceleration readings in this time period to quantify the user's physical activity level. Then, it compares this total change with a pre-set static threshold. If the total change consistently falls below the pre-set static threshold, it is determined that the user's body is in a static state. The static threshold is used to distinguish between conscious physical activity and unconscious micro-movements during sleep. Its value is set to 0.05g, which is obtained by calibrating the body motion signal data of a large number of users during sleep.

[0051] Only when both the heart rate fluctuation convergence and the body's tendency to be static are determined to be satisfied at the same time, and this satisfaction state lasts for a pre-set confirmation time, will the system finally determine that the user has entered the transition phase from wakefulness to sleep. Once the determination is successful, the system will generate a clear internal identifier, i.e., the sleep transition trend, to trigger the subsequent acoustic environment adjustment process. The sleep transition trend is a data identifier generated by the system when the above two conditions of heart rate and body motion are met at the same time. Its data structure is a Boolean value "True", which clearly indicates that the user has entered the pre-sleep state.

[0052] After recognizing the sleep transition trend, the wearable device continuously sends to the audio device through the established connection at a predetermined frequency, e.g., once per second.

[0053] The heart rate fluctuation convergence refers to the difference between the maximum value and the minimum value of the heart rate gradually decreases and eventually stabilizes in a smaller range within a continuous analysis period. The body tends to be stationary refers to a state where the body activity index calculated from the data collected by the accelerometer continuously falls below a certain threshold, indicating that the user has no major body movements.

[0054] The wearable device continuously receives the user's physiological signal stream transmitted through the physiological signal dedicated channel. The system sets the sliding time window to 60 seconds. Within the sliding time window from 22:35 to 22:36, the system calculates that the user's heart rate fluctuation range is 15 beats per minute, and the integrated change of the body motion signal is 0.1g, which is higher than the stationary threshold of 0.05g, and does not meet the condition. Within the sliding time window from 22:36 to 22:37, the heart rate fluctuation range is reduced to 10 beats per minute, but the integrated change of the body motion signal is still 0.08g, which does not meet the condition. Within the sliding time window from 22:37 to 22:38, the system calculates that the heart rate fluctuation range further converges to 6 beats per minute, and the integrated change of the body motion signal decreases to 0.03g, which is lower than the stationary threshold. Since both the heart rate fluctuation convergence and the body tends to be stationary conditions are met at this moment, the system determines that the user enters the transition stage from wakefulness to sleep, and generates a sleep transition trend identifier.

[0055] S4, generating a tentative adjustment parameter: in response to the sleep transition trend, locking the reference safe volume, and starting the fine-tuning cycle logic to generate an acoustic fine-tuning parameter for slightly reducing the audio volume.

[0056] The fine-tuning cycle logic is a program control structure that runs in a loop in the mode of "generating parameters - executing adjustments - observing feedback - decision making" until the preset target is reached.

[0057] The decrement algorithm is a core algorithm used to calculate the amplitude of each volume adjustment in the fine-tuning cycle logic, which is implemented by a non-linear decrement function in this scheme.

[0058] In a preferred embodiment, the step of generating an acoustic fine-tuning parameter for slightly reducing the audio volume in response to the sleep transition trend, locking the reference safe volume, and starting the fine-tuning cycle logic includes: S4.1, after identifying the sleep transition trend, locking the current audio playback volume as the reference safe volume.

[0059] S4.2, first call the preset non-linear decrement function to calculate a small adjustment value that is less than a fixed percentage of the current volume.

[0060] S4.3, encapsulate the small adjustment value as an acoustic fine-tuning parameter containing the adjustment direction and the adjustment amplitude to make the first tentative change to the initial acoustic environment.

[0061] Specifically, the core of this step is to generate an instruction for gently adjusting the acoustic environment to avoid startling the user who has just entered the light sleep state. The operation process is as follows: once the system internally generates a sleep transition trend identifier representing that the user has entered the sleep transition stage, the audio device will immediately perform a locking action, record the volume value of the sleep-aiding audio being played at this moment, and define it as the baseline safe volume. The baseline safe volume is the current audio playback volume recorded by the system when the sleep transition trend is identified, which serves as the starting point and safe upper limit for all subsequent fine-tuning operations. This baseline safe volume is acceptable to the user in a wakeful state, so it serves as the highest upper limit for subsequent adjustments, ensuring that any adjustment does not exceed this comfortable range.

[0062] After locking the baseline safe volume, the system first calls an internally preset non-linear decreasing function, which takes the current baseline safe volume as input and outputs a small adjustment value. This function is designed to be non-linear, meaning that the higher the volume, the smaller the proportion of the calculated adjustment value to the total volume, and vice versa, to achieve more delicate and humanized adjustment. The calculation result of the function is strictly limited to ensure that its absolute value is less than two fixed percentages of the current volume, to ensure that the volume change is weak enough not to be easily detected by the user.

[0063] The preset non-linear decreasing function is a mathematical function, for example where is the current volume, is a coefficient, and its characteristic is that the larger the input value , the slower the growth rate of the output value . The basis for setting this function is based on the Weber's Law in auditory psychology, which states that a person's perception of change is related to the absolute value of the change and the original stimulus intensity. This function aims to keep the perception of volume adjustment at a similar low level.

[0064] After the calculation is completed, the system encapsulates the small adjustment value obtained to create a new data structure, namely the acoustic fine-tuning parameter. This parameter explicitly contains two key pieces of information: one is the adjustment direction, which is fixed as "decrease" at this stage, and the other is the adjustment amplitude, which is the specific value calculated by the non-linear decreasing function. This encapsulated acoustic fine-tuning parameter is then passed to the audio playback control module, preparing for the first exploratory change to the initial acoustic environment in the next step.

[0065] For example, after the system generates the sleep transition trend identification, it immediately locks the current audio playback volume, which is 50 units, and sets this value as the baseline safe volume. The system first calls the preset nonlinear decreasing function, inputting the current volume 50. The function outputs a small adjustment value, for example, 0.8 units, which is less than two percent of the current volume 50 (i.e., 1 unit). Subsequently, the system encapsulates this result into an acoustic fine-tuning parameter, in which the adjustment direction is "decrease" and the adjustment amplitude is 0.8. After the acoustic fine-tuning parameter is generated, the system is ready to proceed to the next step, using it to actually adjust the volume.

[0066] S5, perform acoustic environment fine-tuning: the audio device applies the acoustic fine-tuning parameter to perform a barely perceptible volume reduction operation on the playing sleep-aiding audio, generating a perturbation observation window.

[0067] The acoustic environment fine-tuning refers to making a very small, barely perceptible adjustment to the key parameters (such as volume, frequency components) of the playing audio environment.

[0068] The barely perceptible volume reduction operation is a special volume adjustment method, whose adjustment amplitude is carefully designed to be lower than the minimum audible difference of human ears to sound intensity changes, thereby avoiding shocking the user in a light sleep state.

[0069] In a preferred embodiment, the audio device applies the acoustic fine-tuning parameter to perform a barely perceptible volume reduction operation on the playing sleep-aiding audio, generating a perturbation observation window, including: S5.1, based on the acoustic fine-tuning parameter, slightly reducing the audio playback volume from the baseline safe volume.

[0070] S5.2, immediately start a pre-designed time period at the moment when the volume reduction operation is completed.

[0071] S5.3, define the time period as a perturbation observation window specially used for monitoring physiological feedback.

[0072] Specifically, the step aims to make a slight change to the acoustic environment that is almost imperceptible to the user's subjective consciousness, and to create conditions for subsequent physiological response evaluation. The operation process is as follows: after the audio playback control module of the audio device receives the acoustic fine-tuning parameter, the system accurately analyzes the parameter to obtain the adjustment direction of "lower" and the specific adjustment amplitude. Subsequently, the audio playback control module immediately executes the volume adjustment instruction, and reduces the volume of the sleep-aiding audio currently being played by the adjustment amplitude defined in the acoustic fine-tuning parameter from the recorded reference safe volume level. For example, if the reference safe volume is 50 units and the adjustment amplitude is 0.8 units, the new playback volume will be accurately set to 49.2 units. The execution process of this volume reduction operation is very fast, usually completed within tens of milliseconds.

[0073] At the precise moment when the volume reduction operation is completed, the system triggers an internal timer to immediately start a timing period lasting tens of seconds. The setting of the timing period is based on the statistical results of sleep physiology research, that is, the subconscious physiological stress response caused by weak external stimuli usually appears within a few seconds to tens of seconds after the stimulus occurs. The start of this timing period marks the beginning of a special monitoring phase, and the entire timing period is defined by the system as a disturbance observation window. The disturbance observation window is a specific monitoring period started immediately after a slight acoustic environment adjustment, and its function is to observe whether the adjustment causes physiological disturbance to the user.

[0074] During this window, the main task of the system will change to passive monitoring, focusing on collecting and analyzing the user's physiological signals to determine whether this slight acoustic environment change has caused any subconscious interference to the user. The establishment of this disturbance observation window is a key bridge connecting active acoustic environment adjustment and passive user physiological feedback.

[0075] For example, the audio device receives an acoustic fine-tuning parameter, in which the adjustment direction is "lower" and the adjustment amplitude is 0.8. The audio device immediately executes the volume reduction operation to accurately reduce the playback volume from the reference safe volume of 50 units to 49.2 units. At the moment when the volume is adjusted to 49.2 units, the system immediately starts a timer to begin a 30-second timing period. This period of time, which starts from the moment when the volume adjustment is completed and lasts for 30 seconds, is defined as a disturbance observation window, during which the system will closely monitor the changes in the data transmitted by the physiological signal exclusive channel.

[0076] S6, determining physiological disturbance feedback: in the disturbance observation window, the physiological signal stream data in the physiological signal exclusive channel is continuously analyzed to determine whether the physiological stress response caused by the volume reduction exists, and a regulation effectiveness report is generated.

[0077] The determining physiological disturbance feedback is the whole process of analyzing the user's physiological signals to assess the intensity of their response to external stimuli.

[0078] In a preferred embodiment, the determining physiological disturbance feedback comprises: during the whole disturbance observation window, processing the user's physiological signals stream received from the physiological signal dedicated channel in real time by the audio device to calculate the heart rate variability index in real time.

[0079] The short-term heart rate variability index value calculated within the disturbance observation window is obtained and dynamically compared with the stable heart rate variability index baseline value calculated within a set time period before the start of the disturbance observation window. If the short-term heart rate variability index value shows a fluctuation trend beyond the pre-set difference range compared to the stable heart rate variability index baseline value, it is determined that the physiological stress response to the volume micro-decrease exists, otherwise it is determined that it does not exist.

[0080] When the physiological stress response to the volume micro-decrease exists, a regulation invalidity marker is generated, otherwise a regulation validity marker is generated, thereby integrating a regulation validity report containing the regulation validity marker or the regulation invalidity marker.

[0081] Specifically, the core task of this step is to analyze the user's physiological feedback in depth to determine whether the fine tuning of the acoustic environment is successful. The operation process is as follows: during the whole duration of the disturbance observation window, the analysis module in the audio device will process the user's physiological signals stream received from the physiological signal dedicated channel in real time and continuously, and focus the analysis on the heart rate variability index.

[0082] The heart rate variability index is an index that measures the slight changes in the interval time between successive heartbeats, which is usually obtained by calculating the standard deviation or coefficient of variation of the heart rate signal within a set time period. It can reflect the activity and balance of the autonomic nervous system. In a stable state, the heart rate variability index value is relatively high and stable. When subjected to external pressure or interference, the sympathetic nervous activity increases, which will cause the heart rate variability index value to decrease instantaneously or produce sharp fluctuations, so it is an extremely sensitive biological index for monitoring whether the subconscious level is disturbed.

[0083] The system will calculate the short-term heart rate variability index value within the disturbance observation window and dynamically compare it with the stable heart rate variability index baseline value at a time period before the start of the disturbance observation window. The system looks for the physiological stress response that may be triggered by the external environment change, i.e. the volume micro-decrease event. This response usually appears as an instantaneous and abnormal fluctuation in the heart rate variability index data, such as a sudden and short-term, large amplitude decrease in the heart rate variability index value. The physiological stress response is an unconscious and automatic protective response of the organism to external stimuli (such as sound changes) in the autonomic nervous system, usually accompanied by rapid changes in heart rate, respiration, etc.

[0084] Based on the monitoring results, the system makes a binary decision. In the first case, if no abnormal fluctuation is detected by the system throughout the entire perturbation observation window, it is determined that the current volume micro-decrease operation has not caused any measurable disturbance to the user's sleep transition process. Based on this determination, the system generates an adjustment effectiveness report and writes a "adjustment effective" label in the report. In the second case, if abnormal fluctuation is detected at any time within the window, the system determines that the current volume micro-decrease has caused subconscious disturbance to the user, even if the user is not awake. At this time, the system generates an adjustment effectiveness report labeled "adjustment ineffective" and, as an immediate response measure, sends an instruction to the audio playback control module to restore the volume of the sleep-aiding audio to the level before the micro-decrease operation. The generated adjustment effectiveness report is used to record whether the current acoustic fine-tuning is successful, i.e., whether it has caused disturbance to the user, and whether it is effective or ineffective, will be used as a key input for the next decision.

[0085] For example, within a 30-second perturbation observation window starting from the moment when the volume is set to 49.2 units, the system continuously analyzes the user's heart rate variability (HRV) data. One possible case is that the user's HRV data remains stable throughout the 30 seconds without any significant abnormal fluctuation. Therefore, the system determines that the current volume micro-decrease has not caused disturbance to the user and generates an adjustment effectiveness report labeled "adjustment effective". Another possible case is that at the 12th second after the start of the perturbation observation window, the system detects a short but sharp drop in the user's HRV value. The system identifies this as a physiological stress response triggered by the volume change and determines that the current micro-tuning has caused subconscious disturbance to the user. Therefore, the system generates an adjustment effectiveness report labeled "adjustment ineffective" and immediately performs a restoration operation to restore the audio volume from 49.2 units to 50 units before the micro-decrease.

[0086] S7, iteratively shaping the acoustic comfort envelope: based on the results of the adjustment effectiveness report, decide whether to set the current reduced volume as the new safe baseline, and loop back to steps S4-S6 until the audio volume is reduced to the preset mute threshold, forming an acoustic comfort envelope that closely fits the user's individual sensitivity curve.

[0087] The iterative shaping of the acoustic comfort envelope is a process of gradually constructing a dynamically changing set of audio environment parameters (such as volume) through a series of continuous, feedback-based micro-tuning operations, and the boundary of this set just fits the user's auditory comfort zone in a specific physiological state (such as sleep).

[0088] The cycle returns to S4-S6 steps means that the program flow jumps back to S4 step to start a new round of parameter generation, fine-tuning execution, and observation of feedback after completing the decision of S7.

[0089] In a preferred embodiment, the iterative shaping of the acoustic comfort envelope comprises: S7.1, setting the current volume after the micro-decrease operation as the new baseline safe volume when the adjustment effectiveness report contains the adjustment valid flag.

[0090] After confirming the new baseline safe volume, a preset delay period is started, and the fine-tuning cycle logic is started again to generate the next smaller acoustic fine-tuning parameter.

[0091] S7.2, when the adjustment effectiveness report contains the adjustment invalid flag, the length of the preset delay period is extended, and the cycle returns to the step of generating the tentative adjustment parameter.

[0092] Specifically, this step is the decision and cycle control center of the entire sleep-aiding audio adjustment process, and the purpose is to dynamically and gradually shape an acoustic environment most suitable for the current state of the user based on physiological disturbance feedback. The decision means that the system decides whether to consolidate the current adjustment results or to cancel and adjust the strategy according to the content of the adjustment effectiveness report.

[0093] The operation flow has two paths according to the result of the adjustment effectiveness report. The first path, if the system receives an adjustment valid flag in the adjustment effectiveness report, it indicates that the last volume micro-decrease did not disturb the user. The system adopts a confirmation strategy to set the current lower volume value after the micro-decrease as the new baseline safe volume. This means that the next fine-tuning will take this new baseline as the starting point, so that the volume will be further decreased. After confirming the new baseline, the system starts a delay period to wait for a period of time, which aims to ensure that the user's state is stable in the new acoustic environment, and then returns to the step of generating the tentative adjustment parameter, based on the new lower baseline safe volume, to start the fine-tuning cycle logic again to generate the next smaller acoustic fine-tuning parameter.

[0094] The second path, if the adjustment effectiveness report flag is "adjustment invalid", the system adopts a cancellation and avoidance strategy. Since the volume has been restored immediately in the physiological disturbance feedback step, no volume operation is needed in this step. But the system will record this failed attempt and significantly extend the length of the delay period, and then return to the step of generating the tentative adjustment parameter. The purpose of extending the delay period is to give the user more time to adapt and stabilize, avoiding cumulative disturbance caused by continuous adjustment attempts.

[0095] The delay period is a waiting time set by the system after a complete "adjust-feedback" cycle before starting the next cycle. When the adjustment is valid, the delay period is set to a standard short value, for example 60 seconds; when the adjustment is invalid, the delay period is doubled, for example set to 120 seconds, to provide more sufficient recovery time.

[0096] The whole method through this "micro-decrease-observation-confirmation / cancellation" closed-loop iterative process, constantly exploring and approaching the user's sensitivity to sound in the sleep transition period, the volume of the sleep aid audio is gradually and safely reduced from the initial level, until the volume value reaches a preset mute threshold, for example, complete silence or lower than the ambient noise level. The preset mute threshold is a lower limit of the volume level, when the volume of the sleep aid audio is reduced to this value, the system considers that the sleep aid purpose has been achieved and should stop reducing the volume, the threshold is usually set to 0 or a very low value, for example 5 units, to avoid the sudden disappearance of the audio itself in an absolutely quiet environment. When the volume is reduced to this mute threshold, the cycle is terminated.

[0097] The smooth descending curve from the initial volume to the mute threshold formed by the whole dynamic adjustment process, constitutes an acoustic comfort envelope closely fitting the user's individual sensitivity curve. The acoustic comfort envelope is a trajectory or range that describes the audio volume that can keep the user comfortable and not disturbed over time during the whole sleep aid process. This envelope is dynamically generated and highly individualized, reflecting the user's changing sensitivity to sound from falling asleep to deepening sleep.

[0098] For example, if the received adjustment validity report is marked as "adjustment valid", the system confirms the current volume of 49.2 units as the new baseline safe volume. After waiting for a 60-second delay period, the program flow returns to step S4 to calculate the next acoustic fine-tuning parameter based on the 49.2-unit volume. If the received adjustment validity report is marked as "adjustment invalid" (at this time the volume has been restored to 50 units in S6), the system does not confirm the volume, but starts a 120-second extended delay period, and then returns to step S4, still based on the 50-unit volume to try to generate adjustment parameters. This "micro-decrease-observation-confirmation / cancellation" cycle is repeated, the volume may be gradually reduced from 50 to 49.2, then to 48.5, and so on, until it is finally safely reduced to the preset mute threshold of 5 units, at which point the cycle ends. The whole descending trajectory from volume 50 to 5 successfully shapes an acoustic comfort envelope for the user's sleep-in process that night.

[0099] It should be noted that: the above formula, through the principle of dimensional consistency and mathematical standardization means (for example, normalization processing, dimensionless parameter conversion or unit system unification), different properties of physical quantities can be translated into unitless standard values or the same dimension superimposable parameters, so as to eliminate the interference of different dimensions on the operation logic, so that the formula retains the original data distribution characteristics while having mathematical operation rationality and objective law adaptability. The above merely illustrates the exemplary embodiments of the present application, and cannot limit the scope of the present application.

Claims

1. An adaptive audio adjustment sleep-aiding method based on sleep state recognition, characterized in that, The method comprises the following steps: S1, obtaining a user sleep intention: when a wearable device worn on the user's body physically touches or directly connects with an audio device through Bluetooth, a composite instruction pre-stored in the wearable device is read through near field communication technology to generate a handshake authorization instruction; S2, establishing a physiological signal exclusive channel: based on the handshake authorization instruction, a low-latency wireless data connection is established between the wearable device and the audio device, and an initial acoustic environment is initialized; S3, identifying a physiological stable trend: physiological signal streams are collected by the wearable device, and the heart rate signal fluctuation range and the comprehensive change amount of body motion signals in the physiological signal streams are analyzed to identify a sleep transition trend, and the real-time analyzed sleep transition trend is continuously transmitted to the audio device through the physiological signal exclusive channel; The sleep transition trend is identified, comprising: S3.1, using a sliding time window to process continuous physiological signal stream data, calculating the difference between the maximum and minimum values of the user's physiological signal stream heart rate signal in each continuous sliding time window to obtain the heart rate fluctuation range; By comparing the heart rate fluctuation range of the current sliding time window with that of the previous several sliding time windows, it is determined whether the heart rate fluctuation range presents a continuous decreasing trend and finally stabilizes in a pre-set narrow interval, and if it is confirmed that the state presents, it is determined that the user's heart rate fluctuation converges; S3.2, synchronously analyzing whether the comprehensive change amount of the body motion signal in the continuous sliding time window presents a state of continuously being lower than a pre-set static threshold, and if it is confirmed that the state presents, it is determined that the user's body tends to be static; S3.3, when the two conditions of user heart rate fluctuation convergence and user body tending to be static are met at the same time, it is determined that the user enters a transition stage from wakefulness to sleep, and a sleep transition trend identifying the stage is generated; S4, generating a tentative adjustment parameter: in response to the sleep transition trend, locking a reference safe volume, and starting a fine-tuning cycle logic to generate an acoustic fine-tuning parameter for slightly reducing the audio volume; The response to the sleep transition trend, locking the reference safe volume, and starting the fine-tuning cycle logic to generate the acoustic fine-tuning parameter for slightly reducing the audio volume, comprising: S4.1, after identifying the sleep transition trend, locking the current audio playback volume as the reference safe volume; S4.2, first calling a pre-set non-linear decreasing function to calculate a small adjustment value which is less than a fixed percentage of the current volume; S4.3, encapsulating the small adjustment value as an acoustic fine-tuning parameter containing an adjustment direction and an adjustment amplitude to make a first tentative change to the initial acoustic environment; S5, performing acoustic environment fine-tuning: the audio device applies the acoustic fine-tuning parameter to perform an imperceptible volume reduction operation on the playing sleep-aiding audio to generate a perturbation observation window; The audio device applies the acoustic fine-tuning parameter to perform an imperceptible volume reduction operation on the playing sleep-aiding audio to generate a perturbation observation window, comprising: S5.1, based on the acoustic fine-tuning parameter, slightly reducing the audio playback volume from the reference safe volume; S5.2, immediately starting a pre-designed time period at the moment when the volume reduction operation is completed. S5.3, defining the timing period as a perturbation observation window dedicated to monitoring physiological feedback; S6, determining physiological perturbation feedback: in the perturbation observation window, continuously analyzing physiological signal stream data in the physiological signal dedicated channel to determine whether physiological stress response caused by the slight volume reduction exists, and generating a regulation effectiveness report; The determining physiological perturbation feedback comprises: During the entire perturbation observation window, the user's physiological signal stream received from the physiological signal dedicated channel is processed in real time by the audio device to calculate the heart rate variability index in real time; The short-term heart rate variability index value calculated in the perturbation observation window is obtained, and it is dynamically compared with the stable heart rate variability index baseline value calculated in the set time period before the start of the perturbation observation window. If the short-term heart rate variability index value compared with the stable heart rate variability index baseline value presents a fluctuation trend exceeding the preset difference range, it is determined that the physiological stress response caused by the slight volume reduction exists, otherwise it is determined that it does not exist; When the physiological stress response caused by the slight volume reduction exists, a regulation invalidation mark is generated, otherwise a regulation effectiveness mark is generated, so as to integrate a regulation effectiveness report containing the regulation effectiveness mark or the regulation invalidation mark; S7, iterative shaping of acoustic comfort envelope: according to the result of the regulation effectiveness report, it is decided whether the current reduced volume is set as a new safety benchmark, and the steps S4 to S6 are returned cyclically until the audio volume is reduced to the preset mute threshold, forming an acoustic comfort envelope closely fitting the user's individual sensitivity curve.

2. The self-adaptive audio adjusting sleep-aiding method based on sleep state recognition according to claim 1, characterized in that, The generating handshake authorization instruction comprises: S1.1, identifying a near-field contact event between the wearable device and the audio device within a preset physical distance, activating the near-field communication function when the near-field contact event is identified or the Bluetooth direct connection is identified; S1.2, parsing the composite instruction containing the Bluetooth connection address and the sleep aid mode identification from the near-field communication tag integrated in the wearable device; S1.3, encapsulating the composite instruction as a handshake authorization instruction carrying a unique timestamp, representing the user's authorization of the system to take over control.

3. The self-adaptive audio adjusting sleep-aiding method based on sleep state recognition according to claim 2, characterized in that, The near-field contact event between the wearable device and the audio device within a preset physical distance is specifically: integrating a near-field communication tag in the wearable device and a near-field communication reader in the audio device, when the user places the body part wearing the wearable device close to the designated sensing area of the audio device, and the distance between them enters the preset physical distance, the near-field communication reader in the audio device senses the near-field communication tag in the wearable device, and the sensing action is marked as a near-field contact event.

4. The self-adaptive audio adjusting sleep-aiding method based on sleep state recognition according to claim 2, characterized in that, The establishing a physiological signal dedicated channel comprises: S2.1, the audio device parses the handshake authorization instruction, extracts the Bluetooth connection address and automatically completes the pairing with the wearable device to establish data connection; S2.2, the audio device starts to play the preset sleep aid audio according to the sleep aid mode identification in the handshake authorization instruction, forming an initial acoustic environment for shielding environmental noise; S2.3, the wearable device starts to continuously transmit real-time data to the audio device at a predetermined frequency through the data connection, forming a physiological signal dedicated channel.

5. The self-adaptive audio adjusting sleep-aiding method based on sleep state recognition according to claim 1, characterized in that, The physiological signal stream is collected by a wearable device, comprising: The wearable device starts its built-in optical heart rate sensor and accelerometer to collect the user's physiological data in real time, including heart rate signals and body movement signals, which are multi-axis acceleration value signals of the user's body movement; The collected data is formatted into a unified user physiological signal stream.

6. The self-adaptive audio adjusting sleep-aiding method based on sleep state recognition of claim 1, wherein, The acoustic comfort envelope is iteratively shaped, comprising: S7.1, in the adjustment effectiveness report containing the adjustment effective mark, set the current volume value after the volume micro-decrease operation as the new baseline safe volume; After confirming the new baseline safe volume, start a preset delay period, and then start the micro-adjustment cycle logic again to generate the next smaller acoustic micro-adjustment parameter; S7.2, in the adjustment effectiveness report containing the adjustment invalid mark, extend the length of the preset delay period, and then loop back to the trial adjustment parameter generation step.

Images