Intelligent sleep regulation and control system based on infrared ray and sound monitoring

Abstract

The invention belongs to the technical field of smart home, and particularly relates to an intelligent sleep regulation and control system based on infrared ray and sound monitoring, which integrates infrared ray monitoring, sound recognition, dryness and humidity measurement, luminosity sensing, sound recording, temperature sensing and artificial intelligence analysis technologies. According to the method, the sleep stage, the sleep quality and the sleep posture of the user are accurately recognized, and the indoor environment parameters are automatically adjusted, so that the purposes of optimizing the sleep experience, improving the getting-up comfort level and achieving health early warning are achieved.

Description

Technical Field

[0001] The present invention belongs to the technical field of smart home, and specifically relates to an intelligent sleep regulation system based on infrared and sound monitoring. Background Art

[0002] Traditional alarm clocks and sleep monitoring devices play important roles in modern life, but their functions and technical means still have obvious limitations. Traditional alarm clocks mainly rely on a single audio stimulus (such as a ringtone) to wake up users. This sudden and high-intensity stimulation method often causes users to be suddenly pulled back from deep sleep to the waking state, resulting in discomfort and even affecting the mental state of the whole day. At the same time, although existing sleep monitoring devices can record users' sleep data (such as heart rate, breathing rate, etc.) to a certain extent, their data analysis dimensions are limited, and they lack the ability to comprehensively regulate sleep cycles and environmental factors.

[0003] Obviously, traditional alarm clocks and sleep monitoring devices can only provide basic waking functions and limited sleep data analysis, lacking the ability to dynamically adjust the environment according to the actual sleep state of users to adapt to the best waking time. Therefore, it is particularly important to develop a system that can real-time monitor and analyze sleep data, and accordingly automatically adjust the indoor environment, enabling users to naturally enter the light sleep state at a preset time and thus be easily awakened. Summary of the Invention

[0004] To make up for the deficiencies of the prior art, the present invention proposes an intelligent sleep regulation system based on infrared and sound monitoring. The present invention is mainly used to solve the problem that existing smart home products cannot dynamically adjust the environment according to the actual sleep state of users so that users can wake up at the best time.

[0005] The technical solution adopted by the present invention to solve its technical problems is: The present invention provides an intelligent sleep regulation system based on infrared and sound monitoring, including a data acquisition module, an intelligent analysis and processing module, and an environmental regulation system;

[0006] The data acquisition module includes a body data acquisition unit and an environmental data acquisition unit; the body data acquisition unit is used to real-time monitor and record users' physiological indicators; the environmental data acquisition unit is used to collect data of environmental factors related to sleep; the data collected by the body data acquisition unit and the environmental data acquisition unit are both transmitted to the intelligent analysis and processing module;

[0007] The intelligent analysis and processing module determines the sleep quality of users according to the data collected by the body data acquisition unit, and then gives environmental adjustment parameters according to the environmental data collected by the environmental data acquisition unit, and transmits the parameters to the environmental regulation system;

[0008] The environmental regulation system adjusts the influencing factors in the user's sleep environment according to the parameters transmitted by the intelligent analysis and processing module.

[0009] The intelligent analysis and processing module learns the relationship between body posture and sleep quality by using a self-developed machine learning model; the intelligent analysis and processing module judges the user's sleep quality according to the data collected by the body data acquisition unit, identifies the user's sleep pattern, and then determines how to adjust the environmental parameters so that the user can obtain better sleep; the intelligent analysis and processing module then determines the parameters that need to be further adjusted in the environment according to the environmental data collected by the environmental data acquisition unit and in combination with the above-mentioned determined high-quality sleep parameters, and then transmits the parameters to the environmental regulation system, thereby ensuring that the user can obtain better sleep; moreover, by indirectly extending or reducing the duration of a sleep cycle, it is possible to fine-tune the environmental parameters through the environmental regulation system to guide the user to gradually enter the light sleep state for waking up under the condition of ensuring sleep quality, so that the user can be comfortably woken up while obtaining good sleep.

[0010] Preferably, the body data acquisition unit includes an infrared sensor; the infrared sensor is used to monitor the user's sleep posture, body movement frequency and sleep depth; the detection data of the infrared sensor is transmitted to the intelligent analysis and processing module; the intelligent analysis and processing module analyzes and processes the detection data of the infrared sensor through infrared data analysis technology.

[0011] Preferably, the infrared data analysis technology includes data preprocessing, peak recognition and spectrum interpretation;

[0012] The data preprocessing includes baseline correction, noise removal and spectral smoothing;

[0013] The peak recognition includes the first derivative method and the second derivative method.

[0014] By analyzing and processing the detection data of the infrared sensor through infrared data analysis technology, after accurately identifying the changes in the user's sleep posture and body movement frequency, the user's sleep depth can be evaluated. Moreover, after analysis and processing, it is possible to generate a curve of the user's sleep posture changes, a statistical chart of body movement frequency and a sleep depth evaluation report, which helps the user to fully understand their own sleep situation. Long-term recording of the user's body posture and environment can also help predict sleep trends and evaluate the impact of the current sleep posture on sleep quality and possible health problems.

[0015] Preferably, the body data collection unit further includes a sound monitor; the sound monitor is used to collect the breathing sound, snoring sound and external noise during the user's sleep; the detection data of the sound monitor is transmitted to the intelligent analysis and processing module; the intelligent analysis and processing module analyzes and processes the detection data of the sound monitor through speech recognition technology and sound analysis technology.

[0016] Preferably, the speech recognition technology includes the following steps:

[0017] S1: Construct a multi-modal acoustic feature extraction framework, and preprocess the original audio signal using an improved adaptive noise cancellation algorithm; separate the user's breathing sound from the environmental noise through dual-microphone beamforming technology;

[0018] S2: Adopt time-frequency domain hybrid coding technology, and then combine the composite features of Mel frequency cepstral coefficients and Teager energy operator to effectively capture the non-linear features of snoring;

[0019] S3: Construct an acoustic event classification model using a deep residual convolutional network.

[0020] Using speech recognition technology, specifically through steps such as preprocessing (noise reduction, normalization, etc.), feature extraction (MFCC, etc.) and pattern matching (acoustic models such as HMM, DNN, etc.) of the collected sound signals, key indicators such as the user's breathing frequency and snoring intensity can be identified.

[0021] In addition, using sound analysis technology, the type and intensity of the noise are identified, and then its impact on the user's sleep is evaluated. Specifically, in the processing of sound recording and recognition data, this solution constructs a multi-modal acoustic feature extraction framework and preprocesses the original audio signal using an improved adaptive noise cancellation algorithm (ANC 3.0); separates the user's breathing sound from the environmental noise through dual-microphone beamforming technology. In the feature extraction stage, time-frequency domain hybrid coding technology is introduced, and the composite features of Mel frequency cepstral coefficients (MFCC) and Teager energy operator (TEO) are combined to effectively capture the non-linear features of snoring. A deep residual convolutional network (ResNet-50 improved type) is used to construct an acoustic event classification model, which is pre-trained on a 1 million-hour sleep audio dataset through transfer learning, achieving a breathing event detection accuracy of 98.7% and a snoring classification F1-score of 0.94. In particular, this solution proposes a voiceprint-breathing coupling analysis technology, which establishes a user-specific vocal cord vibration model through a long short-term memory network (LSTM), and can distinguish physiological snoring from pathological sleep apnea events. This technology has been clinically verified to have a consistency of 89.3% compared with traditional PSG monitoring.

[0022] Preferably, the environmental data acquisition unit includes a temperature and humidity sensor and a photometric sensor; the temperature and humidity sensor is used to monitor the humidity and temperature of the indoor environment in real time; the photometric sensor is used to monitor the indoor light intensity; the data collected by the temperature and humidity sensor and the photometric sensor are transmitted to the intelligent analysis and processing module.

[0023] The intelligent analysis and processing module compares the collected data with a preset comfortable range to determine whether the current environment is suitable for sleeping. According to the change of light intensity, the intelligent analysis and processing module judges whether it is necessary to adjust the curtains or lights to create a suitable sleeping environment.

[0024] Preferably, the intelligent analysis and processing module uses machine learning algorithms to establish a user sleep model, and then accurately identifies the user's sleep stage and sleep quality.

[0025] The intelligent analysis and processing module is used to comprehensively analyze various types of received data, and uses machine learning algorithms (such as support vector machines, decision trees, etc.) to establish a user sleep model to accurately identify the user's sleep stage (deep sleep, light sleep, REM, etc.) and sleep quality.

[0026] The intelligent analysis and processing module integrates the time series processing ability of Transformer and the spatial feature extraction advantage of the graph convolutional network (GCN) to construct a spatio-temporal attention fusion network (STAF-Net); moreover, the intelligent analysis and processing module maps heterogeneous data such as infrared thermal imaging data (spatial resolution 0.5 cm 2 )、acoustic feature vectors (128 dimensions), environmental sensor readings (10 Hz sampling rate) to a unified latent space, and synchronously completes sleep stage classification (using an improved AASM standard), sleep quality assessment (constructing an SQI index containing 23 indicators) and environmental parameter optimization strategy generation through a multi-task learning framework. Among them, the breakthrough dynamic environment response algorithm (DERA) uses a method combining reinforcement learning and fuzzy control to achieve millisecond-level decision response on the NVIDIA Jetson edge computing platform, and can dynamically adjust the coupling relationship of four-dimensional environmental parameters according to real-time physiological data. For example, when detecting REM-period eye movements, it automatically enhances the 8-12 Hz sound field component to extend the dream stage.

[0027] Wake up the user during the light sleep stage or create a light sleep window at the user-specified time point through environmental changes to make getting up more comfortable. At the same time, generate a user sleep stage distribution map, a sleep quality assessment report and personalized sleep improvement suggestions.

[0028] Preferably, the intelligent sleep regulation system further includes an external control system and a user interface;

[0029] The external control system enables users to adjust the influencing factors in the environment at any time through voice interaction or terminal software according to their own needs; in addition to its own automatic control function, users of this intelligent sleep regulation system can also perform voice control at any time according to their own needs to adjust parameters such as the light, temperature, humidity, and sound in the bedroom. In addition, this system is also equipped with software that supports Apple and Android systems, facilitating consumers to receive sleep reports through their mobile phones and adjust factors such as the environment and time.

[0030] The user interface is used for direct interaction between the user and the intelligent sleep regulation system. This intelligent sleep regulation system provides an intuitive operation interface and mobile application, allowing users to set personal preferences, view sleep reports, adjust wake-up times, and set other environmental parameters, enhancing the personalization of the user experience.

[0031] The beneficial effects of the present invention are as follows:

[0032] 1. In the present invention, the intelligent analysis and processing module learns the relationship between body posture and sleep quality by using a self-developed machine learning model; the intelligent analysis and processing module determines the user's sleep quality based on the data collected by the body data acquisition unit, identifies the user's sleep pattern, and then determines how to adjust the environmental parameters to enable the user to obtain better sleep; the intelligent analysis and processing module then determines the parameters that need to be further adjusted in the environment based on the environmental data collected by the environmental data acquisition unit and in combination with the above-determined high-quality sleep parameters, and then transmits the parameters to the environmental regulation system, thus ensuring that the user can obtain better sleep; moreover, by indirectly extending or reducing the duration of a sleep cycle, the environmental regulation system can finely adjust the environmental parameters to guide the user to gradually enter the light sleep state for waking up, so that the user can be comfortably woken up while obtaining high-quality sleep.

[0033] 2. The intelligent sleep regulation system described in the present invention can comprehensively and accurately evaluate the user's sleep state and environmental adaptability, and automatically adjust indoor environmental parameters (such as temperature and humidity, light intensity, sound environment, etc.) accordingly to achieve the purposes of optimizing the sleep experience, improving the comfort of waking up, and providing health warnings.

[0034] 3. The present invention analyzes and processes the detection data of the infrared sensor through infrared data analysis technology. After accurately identifying the changes in the user's sleep posture and body movement frequency, the sleep depth of the user is evaluated. Moreover, after analysis and processing, it can generate a curve of the user's sleep posture changes, a statistical chart of body movement frequency, and a sleep depth evaluation report, which helps the user fully understand their own sleep situation. Long-term recording of the user's body posture and environment can also help predict sleep trends and evaluate the impact of the current sleep posture on sleep quality and possible health problems. Brief Description of the Drawings

[0035] The present invention will be further described below in conjunction with the accompanying drawings.

[0036] Figure 1 It is a functional architecture diagram of the intelligent sleep regulation system of the present invention.

[0037] Specific embodiments

[0038] In order to make the technical means, creative features, achieved purposes and effects realized by the present invention easy to understand, the present invention will be further described below in conjunction with specific embodiments.

[0039] As Figure 1 shown, an intelligent sleep regulation system based on infrared and sound monitoring includes a data acquisition module, an intelligent analysis and processing module, and an environmental regulation system;

[0040] The data acquisition module includes a body data acquisition unit and an environmental data acquisition unit; the body data acquisition unit is used to monitor and record the physiological indicators of the user in real time; the environmental data acquisition unit is used to collect data on the influencing factors related to sleep in the environment; the data collected by the body data acquisition unit and the environmental data acquisition unit are both transmitted to the intelligent analysis and processing module;

[0041] The intelligent analysis and processing module judges the sleep quality of the user according to the data collected by the body data acquisition unit, and then gives the environmental adjustment parameters according to the environmental data collected by the environmental data acquisition unit, and transmits the parameters to the environmental regulation system;

[0042] The environmental regulation system adjusts the influencing factors in the sleep environment of the user according to the parameters transmitted by the intelligent analysis and processing module.

[0043] During operation, the intelligent analysis and processing module learns the relationship between body posture and sleep quality by using a self-developed machine learning model; the intelligent analysis and processing module judges the sleep quality of the user according to the data collected by the body data acquisition unit, identifies the sleep pattern of the user, and then determines how to adjust the environmental parameters so that the user can obtain better sleep; the intelligent analysis and processing module then determines the parameters that need to be further adjusted in the environment according to the environmental data collected by the environmental data acquisition unit and in combination with the above-mentioned high-quality sleep parameters judged, and then transmits the parameters to the environmental regulation system, so as to ensure that the user can obtain better sleep; and by indirectly extending or reducing the duration of a sleep cycle, it is possible to fine-tune the environmental parameters through the environmental regulation system to guide the user to gradually enter the light sleep state for waking up under the condition of ensuring sleep quality, so that the user can be comfortably woken up while obtaining high-quality sleep.

[0044] AsFigure 1 As shown, the body data acquisition unit includes an infrared sensor; the infrared sensor is used to monitor the user's sleep posture, body movement frequency, and sleep depth; the detection data of the infrared sensor is transmitted to the intelligent analysis and processing module; the intelligent analysis and processing module analyzes and processes the detection data of the infrared sensor through infrared data analysis technology.

[0045] The infrared data analysis technology includes data preprocessing, peak recognition, and spectrum interpretation;

[0046] The data preprocessing includes baseline correction, noise removal, and spectral smoothing;

[0047] The peak recognition includes the first derivative method and the second derivative method.

[0048] By analyzing and processing the detection data of the infrared sensor through infrared data analysis technology, after accurately identifying the changes in the user's sleep posture and body movement frequency, the user's sleep depth can be further evaluated. Moreover, after analysis and processing, it can generate a curve of the user's sleep posture changes, a statistical chart of body movement frequency, and a sleep depth evaluation report, which helps the user fully understand their own sleep situation. Long-term recording of the user's body posture and environment can also help predict sleep trends and evaluate the impact of the current sleep posture on sleep quality and possible health problems.

[0049] As Figure 1 shown, the body data acquisition unit further includes a sound monitor; the sound monitor is used to collect the user's breathing sounds, snoring sounds, and external noises during sleep; the detection data of the sound monitor is transmitted to the intelligent analysis and processing module; the intelligent analysis and processing module analyzes and processes the detection data of the sound monitor through speech recognition technology and sound analysis technology.

[0050] The speech recognition technology includes the following steps:

[0051] S1: Construct a multi-modal acoustic feature extraction framework, and preprocess the original audio signal using an improved adaptive noise cancellation algorithm; separate the user's breathing sound from the environmental noise through a dual-microphone beamforming technology;

[0052] S2: Adopt a time-frequency domain hybrid coding technology, and combine the composite features of Mel frequency cepstral coefficients and Teager energy operator to effectively capture the non-linear features of snoring sounds;

[0053] S3: Construct an acoustic event classification model using a deep residual convolutional network.

[0054] By using speech recognition technology, specifically through steps such as preprocessing (noise reduction, normalization, etc.), feature extraction (MFCC, etc.), and pattern matching (acoustic models such as HMM and DNN) on the collected sound signals, key indicators such as the user's breathing frequency and snoring intensity can be identified.

[0055] In addition, by using sound analysis technology, the type and intensity of noise are identified, and then its impact on the user's sleep is evaluated. Specifically, in the processing of sound recording and recognition data, this solution preprocesses the original audio signal by constructing a multi-modal acoustic feature extraction framework and using an improved adaptive noise cancellation algorithm (ANC 3.0); separates the user's breathing sound and environmental noise through a dual-microphone beamforming technology. In the feature extraction stage, a time-frequency domain hybrid coding technology is introduced, and a composite feature combining Mel Frequency Cepstral Coefficients (MFCC) and Teager Energy Operator (TEO) is used to effectively capture the non-linear features of snoring. An acoustic event classification model is constructed using a deep residual convolutional network (improved ResNet-50), and through transfer learning, it is pre-trained on a 100,000-hour sleep audio dataset, achieving a breathing event detection accuracy of 98.7% and a snoring classification F1-score of 0.94. In particular, this solution proposes a voiceprint-breathing coupling analysis technology, which establishes a personalized vocal cord vibration model for the user through a Long Short-Term Memory network (LSTM), and can distinguish physiological snoring from pathological sleep apnea events. This technology has been clinically verified to have a consistency of 89.3% compared with traditional PSG monitoring.

[0056] The environmental data acquisition unit includes a temperature and humidity sensor and a light intensity sensor; the temperature and humidity sensor is used to monitor the humidity and temperature of the indoor environment in real time; the light intensity sensor is used to monitor the indoor light intensity; the data collected by the temperature and humidity sensor and the light intensity sensor are transmitted to the intelligent analysis and processing module.

[0057] The intelligent analysis and processing module compares the collected data with a preset comfortable range to determine whether the current environment is suitable for sleeping. According to the change in light intensity, the intelligent analysis and processing module determines whether it is necessary to adjust the curtains or lights to create a suitable sleep environment.

[0058] The intelligent analysis and processing module uses machine learning algorithms to establish a user sleep model, and then accurately identifies the user's sleep stage and sleep quality.

[0059] The intelligent analysis and processing module is used to comprehensively analyze various types of received data, and uses machine learning algorithms (such as support vector machines, decision trees, etc.) to establish a user sleep model, and accurately identify the user's sleep stage (deep sleep, light sleep, REM, etc.) and sleep quality.

[0060] The intelligent analysis and processing module integrates the time series processing ability of Transformer and the spatial feature extraction advantage of the Graph Convolutional Network (GCN) to construct a Spatio-Temporal Attention Fusion Network (STAF-Net); moreover, the intelligent analysis and processing module maps heterogeneous data such as infrared thermal imaging data (spatial resolution 0.5 cm 2 ), acoustic feature vectors (128 dimensions), and environmental sensor readings (10 Hz sampling rate) to a unified latent space, and synchronously completes sleep stage classification (using an improved AASM standard), sleep quality assessment (constructing an SQI index containing 23 indicators), and generation of environmental parameter optimization strategies through a multi-task learning framework. Among them, the breakthrough Dynamic Environment Response Algorithm (DERA) uses a method combining reinforcement learning and fuzzy control to achieve millisecond-level decision-making response on the NVIDIA Jetson edge computing platform, and can dynamically adjust the coupling relationship of four-dimensional environmental parameters according to real-time physiological data. For example, when detecting REM-stage eye movements, it automatically enhances the 8-12 Hz sound field component to extend the dream stage.

[0061] Wake the user during the light sleep stage or create a light sleep window by changing the environment at the user-specified time point to make getting up more comfortable. At the same time, generate a user sleep stage distribution map, a sleep quality assessment report, and personalized sleep improvement suggestions.

[0062] As Figure 1 shown, the intelligent sleep regulation system further includes an external control system and a user interface;

[0063] The external control system is used for the user to adjust the influencing factors in the environment at any time through voice interaction or terminal software according to their own needs; in addition to its own automatic control function, the user can also adjust the parameters such as the light, temperature, humidity, and sound in the bedroom through voice control at any time according to their own needs. In addition, the system is also equipped with software supporting Apple and Android systems, which is convenient for consumers to receive sleep reports through mobile phones and adjust factors such as the environment and time.

[0064] The user interface is used for direct interaction between the user and the intelligent sleep regulation system. The intelligent sleep regulation system provides an intuitive operation interface and a mobile application, allowing users to set personal preferences, view sleep reports, adjust wake-up time, and other environmental parameter settings, enhancing the personalization of the user experience.

[0065] Implementation case:

[0066] Taking the user setting the wake-up time at 6:30 as an example, the intelligent sleep control system captures the user's body surface temperature distribution (gradient change from 36.2℃→36.5℃ in the core area) and body movement frequency (0.8 times / minute) in real time through the infrared thermal imaging array deployed around the bed. Simultaneously, the high-sensitivity sound monitor array collects the respiratory sound spectrum characteristics (base frequency 0.25Hz±0.03Hz) and snoring energy distribution (peak frequency band 120-300Hz).

[0067] The above data is processed by the MSF-Transformer multi-source fusion algorithm, combined with the micro-motion signal (amplitude <5μm) monitored by the flexible piezoelectric mattress, and it is determined that the user enters the N2 sleep stage at 5:58 (confidence 89.3%). At this time, the environmental control system triggers the light sleep transition protocol: the quantum dot smart window film of the environmental control system starts the electrochromic response, and the color temperature migrates from 1800K to 4500K at a rate of 0.5K / s, and cooperates with the full-spectrum LED array to increase the illumination at a gradient of 0.08lux / s. This process ensures that the brightness change rate is always lower than 0.12cd / m through the light environment control algorithm based on Lyapunov stability. 2 / s visual perception threshold; the temperature control subsystem of the environmental control system links the distributed air conditioner and the microfluidic temperature control mattress, and adopts a respiratory phase synchronization strategy to apply a sub-perceptible temperature rise of 0.02℃ / min during the user's exhalation phase, so that the ambient temperature is gradually adjusted from 20℃ to 22.5℃ awakening comfort zone; the sound field module of the environmental control system plays 4-8Hz base white noise through a beamforming array, and mixes in a dynamically migrating forest soundscape (the intensity of the bird's singing source is attenuated at -1.2dB / min, and the azimuth angle is continuously shifted from 270° to 90°). Its frequency band distribution is dynamically adjusted through real-time respiratory rate feedback (the center frequency migrates from the δ wave zone to the α wave zone at a rate of 0.08Hz / s). At 6:25, the infrared sensor detected that the user's body movement frequency rose to 1.2 times / minute and the sound monitor detected that the user's respiratory rhythm variation coefficient (RMSSD) reached 42ms, and it was determined that the light sleep window had been formed (91.7% probability). The piezoelectric actuator of the environmental control system was immediately activated to generate a traveling wave tactile stimulation with an amplitude of 28μm and a frequency of 1.8Hz, and simultaneously released β-caryophyllene aromatic molecules (concentration gradient 0.8ppm→2.2ppm). Finally, at the awakening time of 6:30, the user's core body temperature rose to a physiological awakening state of 36.8℃. The changes in environmental parameters throughout the process were controlled by four-dimensional quantization (temperature step 0.01℃, sound pressure step 0.05dB) to ensure no perceptual interference, achieving a natural awakening comfort score of 4.6 / 5.

[0068] The embodiments of the present invention have been described above in conjunction with the accompanying drawings. However, the present invention is not limited to the above specific embodiments. The above specific embodiments are merely illustrative and not restrictive. Under the inspiration of the present invention, those of ordinary skill in the art can also make many forms without departing from the spirit of the present invention and the scope protected by the claims. All of these are within the protection scope of the present invention.

Claims

1. Intelligent sleep control system based on infrared and sound monitoring, characterized by: It includes data acquisition module, intelligent analysis and processing module and environmental control system; The data acquisition module includes a body data acquisition unit and an environment data acquisition unit; the body data acquisition unit is used to monitor and record the user's physiological indicators in real time; the environment data acquisition unit is used to collect data on factors affecting sleep in the environment; the data collected by the body data acquisition unit and the environment data acquisition unit are transmitted to the intelligent analysis and processing module; The intelligent analysis and processing module determines the user's sleep quality based on the data collected by the body data collection unit, and then provides environmental adjustment parameters based on the environmental data collected by the environmental data collection unit, and transmits the parameters to the environmental control system; The environmental control system adjusts the influencing factors in the user's sleeping environment according to the parameters transmitted by the intelligent analysis and processing module.

2. The intelligent sleep control system based on infrared and sound monitoring according to claim 1 is characterized in that: The body data acquisition unit includes an infrared sensor; the infrared sensor is used to monitor the user's sleeping posture, body movement frequency and sleep depth; the detection data of the infrared sensor is transmitted to the intelligent analysis and processing module; the intelligent analysis and processing module analyzes and processes the detection data of the infrared sensor through infrared data analysis technology.

3. The intelligent sleep control system based on infrared and sound monitoring according to claim 2 is characterized in that: The infrared data analysis technology includes data preprocessing, peak recognition and spectrum interpretation; The data preprocessing includes baseline correction, noise removal and spectrum smoothing; The peak value identification includes a first-order derivative method and a second-order derivative method.

4. The intelligent sleep control system based on infrared and sound monitoring according to claim 1 is characterized in that: The body data collection unit also includes a sound monitor; the sound monitor is used to collect the user's breathing sound, snoring sound and external noise during sleep; the detection data of the sound monitor is transmitted to the intelligent analysis and processing module; The intelligent analysis and processing module analyzes and processes the detection data of the sound monitor through voice recognition technology and sound analysis technology.

5. The intelligent sleep control system based on infrared and sound monitoring according to claim 4 is characterized in that: The speech recognition technology comprises the following steps: S1: Build a multimodal acoustic feature extraction framework and use an improved adaptive noise cancellation algorithm to preprocess the original audio signal; separate the user's breathing sound from the ambient noise through dual-microphone beamforming technology; S2: It uses time-frequency domain hybrid coding technology, combined with the composite features of Mel-frequency cepstral coefficients and Teager energy operator, to effectively capture the nonlinear characteristics of snoring. S3: Use deep residual convolutional network to build an acoustic event classification model.

6. The intelligent sleep control system based on infrared and sound monitoring according to claim 1 is characterized in that: The environmental data acquisition unit includes a temperature and humidity sensor and a photometry sensor; the temperature and humidity sensor is used to monitor the humidity and temperature of the indoor environment in real time; the photometry sensor is used to monitor the indoor light intensity; the data collected by the temperature and humidity sensor and the photometry sensor are transmitted to the intelligent analysis and processing module.

7. The intelligent sleep control system based on infrared and sound monitoring according to claim 1 is characterized in that: The intelligent analysis and processing module uses a machine learning algorithm to establish a user sleep model, thereby accurately identifying the user's sleep stage and sleep quality.

8. The intelligent sleep control system based on infrared and sound monitoring according to claim 1 is characterized in that: It also includes external control systems and user interaction interfaces; The external control system is used for users to adjust the influencing factors in the environment at any time according to their own needs through voice interaction or terminal software; The user interaction interface is used for direct interaction between the user and the intelligent sleep regulation system.

Images