Low-power-consumption newborn sleep management system

Abstract

The invention discloses a low-power-consumption newborn sleep management system, which relates to the technical field of newborn sleep management and comprises a multi-mode sensing module, a data processing and analyzing module, an environment regulation and control module and a power management module. The system collects environment and physiological data through a low-power-consumption sensor array, analyzes the sleep state in combination with layered feature extraction and dynamic threshold judgment technologies, and optimizes the sleep environment by using a closed-loop feedback mechanism. The power management module adopts an energy recovery and dynamic power supply strategy to reduce energy consumption. According to the application, comprehensive monitoring of the sleep state of the newborn and intelligent environment optimization are realized, the problems of high power consumption and poor real-time performance of an existing system are solved, and the requirements of low power consumption and high reliability are met.

Description

Technical Field

[0001] This invention belongs to the field of medical health and intelligent device technology, specifically a low-power neonatal sleep management system. Background Technology

[0002] Neonatal sleep management systems can utilize IoT and smart health monitoring technologies to monitor and optimize an infant's sleep status and environment in real time. However, existing sleep management systems still have certain limitations in terms of power consumption control, data acquisition accuracy, and comprehensive monitoring of an infant's sleep status, making it difficult to meet the practical requirements of low power consumption and high reliability.

[0003] A search revealed a management support system with publication number CN111684507B, published on January 12, 2024. This patent monitors the driver's driving and physical states using a first and second monitoring device, respectively, and integrates a server and a manager's terminal for focused monitoring. However, this technical solution is primarily geared towards driving scenarios, and its monitoring targets and application scenarios differ significantly from those of newborn sleep management. Furthermore, the system has high energy consumption, posing a challenge for application in low-power newborn sleep management scenarios requiring long-term operation. Additionally, this solution lacks a refined analysis and feedback mechanism for infant sleep states, making it difficult to meet the specific needs of newborn sleep management.

[0004] A search revealed a management assistance system with publication number CN112041907B, published on October 15, 2021. This patent records dynamic images of the driver using a camera and combines this with monitoring information to determine whether the driver's status requires attention. However, the core of this technical solution lies in the acquisition and processing of dynamic images, which consumes a lot of energy and is not suitable for newborn sleep management scenarios. Furthermore, the system does not include monitoring and regulating the sleep environment (such as temperature, humidity, and sound), nor does it possess the ability to acquire and analyze infant physiological signals (such as heart rate and respiratory rate) with low power consumption. Therefore, it has certain limitations in supporting the actual needs of newborn sleep management.

[0005] The aforementioned problems indicate that existing management systems still require further improvement in terms of power consumption control, targeted data acquisition, and comprehensive monitoring and optimization of newborn sleep states. Therefore, this invention provides a low-power newborn sleep management system, aiming to better meet the needs of the newborn sleep management field for efficient, low-power, and intelligent systems by optimizing sensor design, reducing system power consumption, improving data acquisition accuracy, and achieving intelligent regulation of the sleep environment and infant physiological state. Summary of the Invention

[0006] This invention provides a low-power neonatal sleep management system and method. Addressing the shortcomings of existing neonatal sleep management systems in power consumption control, data acquisition accuracy, and comprehensive monitoring and optimization of infant sleep states, this solution employs a multimodal sensor array to achieve efficient acquisition of sleep environment and infant physiological signals, combined with an adaptive sampling frequency adjustment mechanism to reduce system energy consumption. To address the issues of high latency and insufficient real-time performance in existing systems during data processing, this solution proposes a sleep state analysis module based on a lightweight algorithm, improving data processing efficiency through hierarchical feature extraction and dynamic threshold determination techniques. Furthermore, to address the lack of intelligent regulation capabilities for the sleep environment in existing systems, this solution introduces a closed-loop feedback regulation mechanism based on environmental perception. This mechanism monitors environmental parameters in real time and generates adjustment commands based on a preset model, driving the execution unit to complete environmental optimization.

[0007] This invention provides a low-power neonatal sleep management system, comprising a multimodal sensing module, a data processing and analysis module, an environmental control module, and a power management module. The multimodal sensing module includes a temperature sensor, a humidity sensor, a sound sensor, a heart rate monitoring unit, and a respiratory rate monitoring unit. These sensors are integrated into a wearable device via a flexible circuit board and transmit data to the central processing unit using Bluetooth Low Energy protocol. The data processing and analysis module performs preliminary processing on the collected data using a hierarchical feature extraction algorithm, identifies the infant's sleep state using dynamic threshold determination technology, and transmits the results to the environmental control module. The environmental control module includes a temperature control unit, a humidification unit, and a noise suppression unit, which adjusts the sleep environment by receiving instructions from the data processing and analysis module. The power management module employs energy recovery technology combined with lithium battery power supply, dynamically adjusting the power supply strategy by monitoring the real-time power consumption of each module to extend the system's battery life.

[0008] This invention provides a low-power neonatal sleep management method, which includes the following steps: Step S1: Multimodal data acquisition; Step S2: Data processing and sleep state analysis; Step S3: Generation of environmental control instructions; Step S4: Adjustment of environmental parameters; Step S5: Power management and energy consumption optimization. In step S1, the multimodal data acquisition involves collecting ambient temperature via a temperature sensor, ambient humidity via a humidity sensor, ambient noise intensity via a sound sensor, the infant's heart rate signal via a heart rate monitoring unit, and the infant's respiratory signal via a respiratory rate monitoring unit. All data is uploaded to the central processing unit at fixed time intervals. In step S2, the data processing and sleep state analysis utilize a hierarchical feature extraction algorithm to process the collected data. First, basic features such as temperature change trends, humidity fluctuation ranges, heart rate variability, and respiratory cycle patterns are extracted. Then, dynamic threshold determination technology is used to identify the infant's sleep state, including light sleep, deep sleep, and wakefulness. In step S3, the environmental control command generation is based on the current sleep state and environmental parameters. The required environmental adjustment amount is calculated using a preset model to generate specific control commands. In step S4, the environmental parameter adjustment is based on the generated control commands, driving the temperature control unit to adjust the ambient temperature, driving the humidification unit to adjust the ambient humidity, and driving the noise suppression unit to reduce ambient noise. In step S5, the power management and energy consumption optimization involves monitoring the real-time power consumption of each module and dynamically adjusting the sensor sampling frequency and data transmission rate to reduce system energy consumption while ensuring data acquisition accuracy.

[0009] In the design of the multimodal sensing module, the temperature sensor, humidity sensor, and sound sensor are all manufactured using low-power MEMS technology. They are installed using an embedded design and connected to the central processing unit via a flexible circuit board. The thickness of the flexible circuit board does not exceed 0.5 mm to ensure wearing comfort. The heart rate monitoring unit and respiratory rate monitoring unit use photoplethysmography (PPG) and impedance measurement to collect heart rate and respiratory signals, respectively. The electrode pads of both units are in contact with the skin through conductive silicone, with a contact area of ​​1 square centimeter. The contact pressure is adjusted by a spring structure to ensure the stability of signal acquisition.

[0010] In the data processing and analysis module, the first layer of the hierarchical feature extraction algorithm extracts the time-domain features of the raw data, including the mean, variance, and peak value; the second layer extracts the frequency-domain features, including the dominant frequency component and spectral energy distribution; the third layer combines the time-domain and frequency-domain features to generate a comprehensive feature vector for subsequent dynamic threshold determination; the dynamic threshold determination technology obtains a set of adaptive thresholds through training on historical data, and combines them with the feature vector of the current data to determine the infant's sleep state, with an accuracy rate of no less than 95%.

[0011] In the environmental control module, the temperature control unit uses a semiconductor cooling chip as the core component, which achieves heating or cooling functions by changing the direction of current. Its maximum power is 5 watts and the response time is 3 seconds. The humidification unit uses ultrasonic atomization technology, which atomizes water molecules through high-frequency vibration and releases them into the air. Its maximum atomization volume is 300 ml / hour and the noise is less than 30 decibels. The noise suppression unit uses active noise reduction technology, which collects ambient noise through a microphone and generates reverse sound waves to cancel it out. The noise reduction depth can reach 20 decibels.

[0012] In the power management module, energy recovery technology uses piezoelectric materials to convert the baby's slight movements into electrical energy, which is stored in the lithium battery. A single movement can generate about 0.1 milliwatts of energy. The lithium battery has a capacity of 2000 mAh and supports continuous system operation for 72 hours. The dynamic power supply strategy adjusts the power supply priority according to the real-time power consumption of each module, prioritizing the normal operation of the data acquisition and environmental control modules, while reducing the power consumption of non-critical modules.

[0013] The beneficial effects of this invention are:

[0014] This invention achieves comprehensive monitoring of newborn sleep status and intelligent optimization of the sleep environment through the above-mentioned technical means, solving the shortcomings of existing systems in terms of power consumption control, data acquisition accuracy and real-time performance, and meeting the actual needs of low power consumption and high reliability. Attached Figure Description

[0015] Figure 1 This is a schematic diagram of the module structure of the low-power neonatal sleep management system of the present invention.

[0016] The attached figures are labeled as follows:

[0017] 1. Multimodal sensing module; 2. Data processing and analysis module; 3. Environmental control module; 4. Power management module; 5. Central processing unit; 6. Temperature control unit; 7. Humidification unit; 8. Noise suppression unit. Detailed Implementation

[0018] The present invention will be further described below with reference to the accompanying drawings and specific embodiments. The illustrative embodiments and descriptions herein are used to explain the present invention, but are not intended to limit the present invention.

[0019] This invention provides a low-power neonatal sleep management system, the specific implementation of which is combined with Figure 1 Please provide a detailed explanation. Figure 1 The system's modular structure diagram is shown, including a multimodal sensing module 1, a data processing and analysis module 2, an environmental control module 3, and a power management module 4. The modules are connected by circuits to form a complete system, and the central processing unit 5 serves as the core control unit responsible for coordinating the work of each module.

[0020] The multimodal sensing module 1 consists of a temperature sensor, a humidity sensor, a sound sensor, a heart rate monitoring unit, and a respiratory rate monitoring unit. These sensors are all manufactured using low-power MEMS technology and integrated into a wearable device via a flexible circuit board. The thickness of the flexible circuit board is no more than 0.5 mm to ensure wearing comfort. The temperature, humidity, and sound sensors are embedded on the surface of the flexible circuit board, located at different positions within the device to collect environmental parameters. The heart rate and respiratory rate monitoring units contact the infant's skin via conductive silicone electrodes, with a contact area of ​​1 square centimeter. The contact pressure is adjusted by a spring structure to ensure stable signal acquisition. All sensors are connected to the central processing unit 5 via the flexible circuit board, using the Bluetooth Low Energy protocol for data transmission. During data transmission, the sensors upload the collected data to the central processing unit 5 at fixed time intervals; the sampling frequency can be dynamically adjusted according to actual needs.

[0021] The data processing and analysis module 2 uses a hierarchical feature extraction algorithm to perform preliminary processing on the collected data. First, time-domain features, including mean, variance, and peak value, are extracted from the raw data. Then, frequency-domain features, such as the dominant frequency component and spectral energy distribution, are extracted. Finally, a comprehensive feature vector is generated by combining the time-domain and frequency-domain features. The dynamic threshold determination technology is trained based on historical data to obtain a set of adaptive thresholds and combines them with the feature vector of the current data to determine the infant's sleep state. This process is completed by the central processing unit 5, and the determination results include three states: light sleep, deep sleep, and wakefulness, which are then transmitted to the environmental control module 3. The operation of the data processing and analysis module 2 relies on the computing power of the central processing unit 5, while simultaneously reducing energy consumption through a low-power design.

[0022] The environmental control module 3 includes a temperature control unit 6, a humidification unit 7, and a noise suppression unit 8, used to adjust the sleep environment according to instructions from the data processing and analysis module 2. The temperature control unit 6 uses a semiconductor cooling chip as its core component, achieving heating or cooling functions by changing the direction of current, with a maximum power of 5 watts and a response time of 3 seconds. The humidification unit 7 uses ultrasonic atomization technology, releasing water molecules into the air through high-frequency vibration, with a maximum atomization volume of 300 ml / hour and noise levels below 30 decibels. The noise suppression unit 8 uses active noise cancellation technology, collecting ambient noise through a microphone and generating reverse sound waves to cancel it out, achieving a noise reduction depth of up to 20 decibels. The temperature control unit 6, humidification unit 7, and noise suppression unit 8 are all connected to the central processing unit 5 via circuitry, receiving control instructions and executing corresponding operations. For example, when the ambient temperature is detected to be too high, the central processing unit 5 sends a cooling instruction to the temperature control unit 6, which then activates the cooling function to lower the ambient temperature.

[0023] The power management module 4 employs energy recovery technology combined with lithium battery power supply. It uses piezoelectric materials to convert the baby's slight movements into electrical energy, which is then stored in the lithium battery. A single movement can generate approximately 0.1 milliwatts of energy. The lithium battery has a capacity of 2000 mAh, supporting continuous system operation for 72 hours. The power management module 4 monitors the power consumption of each module in real time through circuitry and dynamically adjusts the power supply strategy based on the real-time power consumption. For example, when the power consumption of the multimodal sensing module 1 is high, the power management module 4 will prioritize powering it while reducing the power consumption of non-critical modules to extend the system's battery life. Furthermore, the power management module 4 also optimizes energy consumption by dynamically adjusting the sensor's sampling frequency and data transmission rate, reducing the overall system energy consumption while ensuring data acquisition accuracy.

[0024] In practical applications, the system operates as follows: First, the temperature sensor, humidity sensor, and sound sensor in the multimodal sensing module 1 collect ambient temperature, humidity, and noise intensity, respectively, while the heart rate monitoring unit and respiratory rate monitoring unit collect the infant's heart rate and respiratory signals. This data is uploaded to the central processing unit 5 at fixed time intervals via Bluetooth Low Energy. Upon receiving the data, the central processing unit 5 processes it using a hierarchical feature extraction algorithm, extracting time-domain and frequency-domain features and generating a comprehensive feature vector. Subsequently, it identifies the infant's sleep state using dynamic thresholding technology. When the determination result indicates a light sleep state, the central processing unit 5 generates control commands based on current environmental parameters, such as increasing ambient humidity or reducing ambient noise. These control commands are transmitted to the environmental control module 3 via circuitry. The temperature control unit 6, humidification unit 7, and noise suppression unit 8 execute corresponding operations to optimize the sleep environment. Simultaneously, the power management module 4 monitors the power consumption of each module in real time and dynamically adjusts the power supply strategy to extend the system's battery life. For example, when the system is under low load, the power management module 4 reduces the sensor sampling frequency and data transmission rate to save energy.

[0025] The multimodal sensing module 1 is designed with both wearing comfort and signal acquisition stability in mind. Temperature, humidity, and sound sensors are embedded in different locations on the flexible circuit board for comprehensive environmental parameter acquisition. The heart rate and respiratory rate monitoring units contact the skin via conductive silicone electrodes; the contact pressure is adjusted by a spring structure to prevent signal distortion due to poor contact. The thickness of the flexible circuit board is controlled to within 0.5 mm, ensuring the device is lightweight and fits snugly against a baby's skin. Furthermore, the use of Bluetooth Low Energy further reduces power consumption during data transmission.

[0026] The hierarchical feature extraction algorithm and dynamic threshold determination technology are among the core technologies of the data processing and analysis module 2. The first layer of the hierarchical feature extraction algorithm extracts time-domain features such as mean, variance, and peak value from the raw data; the second layer extracts frequency-domain features such as the dominant frequency component and spectral energy distribution; and the third layer combines the time-domain and frequency-domain features to generate a comprehensive feature vector. The dynamic threshold determination technology obtains a set of adaptive thresholds through training on historical data and combines them with the feature vector of the current data to determine the infant's sleep state. This process requires the central processing unit 5 to have high computing power and low power consumption to meet real-time requirements.

[0027] The design of the environmental control module 3 prioritizes both adjustment effectiveness and user experience. The temperature control unit 6 utilizes a semiconductor cooling chip for heating and cooling, offering a short response time and moderate power, enabling rapid adjustment of the ambient temperature. The humidification unit 7 employs ultrasonic atomization technology, providing a large atomization volume with low noise, minimizing disturbance to the infant. The noise suppression unit 8 utilizes active noise cancellation technology, collecting ambient noise through a microphone and generating a reverse sound wave to cancel it out, achieving a noise reduction depth of up to 20 decibels. These three units are connected to the central processing unit 5 via circuitry, receiving control commands and quickly executing corresponding operations to achieve closed-loop feedback regulation of the sleep environment.

[0028] The energy recovery technology and dynamic power supply strategy of Power Management Module 4 are key to the system's low-power design. The energy recovery technology uses piezoelectric materials to convert the baby's slight movements into electrical energy, which is stored in the lithium battery to supplement system power consumption. The dynamic power supply strategy adjusts power supply priority based on the real-time power consumption of each module, prioritizing the normal operation of data acquisition and environmental control modules while reducing the power consumption of non-critical modules. Furthermore, Power Management Module 4 also optimizes energy consumption by dynamically adjusting the sensor sampling frequency and data transmission rate, minimizing overall system power consumption while ensuring data acquisition accuracy.

[0029] The various modules of this system are connected by circuits to form a complete closed-loop control system. Multimodal sensing module 1 is responsible for data acquisition, data processing and analysis module 2 is responsible for data processing and sleep state analysis, environmental control module 3 is responsible for environmental parameter adjustment, and power management module 4 is responsible for energy consumption optimization. The modules achieve efficient data transmission and collaborative operation through circuits and communication protocols, thereby realizing comprehensive monitoring of newborn sleep states and intelligent optimization of the sleep environment.

[0030] To enable those skilled in the art to fully understand and implement this invention, the specific implementation principle of this invention will be further explained below in conjunction with a specific application scenario.

[0031] In actual operation, the ambient temperature, humidity, and noise intensity data are first collected by the temperature sensor, humidity sensor, and sound sensor in the multimodal sensing module 1, respectively. Simultaneously, the heart rate monitoring unit and respiratory rate monitoring unit collect the infant's heart rate and respiratory signals. These sensors are all integrated onto a flexible circuit board with a thickness controlled within 0.5 mm, and the data is uploaded to the central processing unit 5 via Bluetooth Low Energy protocol. Upon receiving the data, the central processing unit 5 processes the raw data according to a hierarchical feature extraction algorithm. In the first layer, time-domain features, such as mean, variance, and peak value, are extracted from the temperature, humidity, noise intensity, heart rate, and respiratory rate data. In the second layer, frequency-domain features, such as the dominant frequency component and spectral energy distribution, are extracted. In the third layer, the time-domain and frequency-domain features are combined to generate a comprehensive feature vector. Subsequently, based on dynamic threshold determination technology, the central processing unit 5 uses a set of adaptive thresholds trained from historical data, combined with the feature vector of the current data, to determine the infant's sleep state, including light sleep, deep sleep, and wakefulness.

[0032] When the baby is determined to be in a light sleep state, the central processing unit 5 generates control commands based on the current environmental parameters. For example, if the ambient temperature is too high, the central processing unit 5 sends a cooling command to the temperature control unit 6. The core component of the temperature control unit 6 is a semiconductor cooling chip, which achieves heating or cooling functions by changing the direction of current, with a response time of 3 seconds and a maximum power of 5 watts. Upon receiving the command, the temperature control unit 6 immediately activates the cooling function to quickly lower the ambient temperature. If the ambient humidity is too low, the central processing unit 5 sends a humidification command to the humidification unit 7. The humidification unit 7 uses ultrasonic atomization technology, which atomizes water molecules through high-frequency vibration and releases them into the air. The maximum atomization volume is 300 ml / hour, and the operating noise is below 30 decibels, ensuring that it will not disturb the baby's sleep. If the ambient noise is too loud, the central processing unit 5 sends a noise reduction command to the noise suppression unit 8. The noise suppression unit 8 uses active noise reduction technology, which collects ambient noise through a microphone and generates reverse sound waves to cancel it out, achieving a noise reduction depth of up to 20 decibels, thereby effectively reducing ambient noise.

[0033] Meanwhile, the power management module 4 monitors the power consumption of each module in real time and dynamically adjusts the power supply strategy based on the real-time power consumption. For example, when the power consumption of the multimodal sensing module 1 is high, the power management module 4 prioritizes powering it while reducing the power consumption of non-critical modules to extend the system's battery life. Furthermore, the power management module 4 converts the baby's slight movements into electrical energy using piezoelectric materials. A single movement can generate approximately 0.1 milliwatts of energy, which is stored in a 2000 mAh lithium battery, supporting continuous system operation for 72 hours. The power management module 4 also optimizes energy consumption by dynamically adjusting the sensor's sampling frequency and data transmission rate, minimizing overall system energy consumption while ensuring data acquisition accuracy.

[0034] In the above process, the design of the multimodal sensing module 1 is particularly crucial. Temperature, humidity, and sound sensors are embedded in different locations on the flexible circuit board to comprehensively collect environmental parameters. The heart rate and respiratory rate monitoring units contact the infant's skin via conductive silicone electrodes, with a contact area of ​​1 square centimeter. The contact pressure is adjusted by a spring structure to ensure the stability of signal acquisition. The lightweight and conformable design of the flexible circuit board makes the device comfortable to wear, avoiding discomfort to the infant due to prolonged wear.

[0035] The hierarchical feature extraction algorithm and dynamic threshold determination technology are among the core technologies of the data processing and analysis module 2. The first layer of the hierarchical feature extraction algorithm extracts time-domain features such as mean, variance, and peak value from the raw data; the second layer extracts frequency-domain features such as the dominant frequency component and spectral energy distribution; and the third layer combines the time-domain and frequency-domain features to generate a comprehensive feature vector. The dynamic threshold determination technology obtains a set of adaptive thresholds through training on historical data and combines them with the feature vector of the current data to determine the infant's sleep state. This process requires the central processing unit 5 to have high computing power and low power consumption to meet real-time requirements.

[0036] The design of the environmental control module 3 prioritizes both adjustment effectiveness and user experience. The temperature control unit 6 utilizes a semiconductor cooling chip for heating and cooling, offering a short response time and moderate power, enabling rapid adjustment of the ambient temperature. The humidification unit 7 employs ultrasonic atomization technology, providing a large atomization volume with low noise, minimizing disturbance to the infant. The noise suppression unit 8 utilizes active noise cancellation technology, collecting ambient noise through a microphone and generating a reverse sound wave to cancel it out, achieving a noise reduction depth of up to 20 decibels. These three units are connected to the central processing unit 5 via circuitry, receiving control commands and quickly executing corresponding operations to achieve closed-loop feedback regulation of the sleep environment.

[0037] The energy recovery technology and dynamic power supply strategy of Power Management Module 4 are key to the system's low-power design. The energy recovery technology uses piezoelectric materials to convert the baby's slight movements into electrical energy, which is stored in the lithium battery to supplement system power consumption. The dynamic power supply strategy adjusts power supply priority based on the real-time power consumption of each module, prioritizing the normal operation of data acquisition and environmental control modules while reducing the power consumption of non-critical modules. Furthermore, Power Management Module 4 also optimizes energy consumption by dynamically adjusting the sensor sampling frequency and data transmission rate, minimizing overall system power consumption while ensuring data acquisition accuracy.

[0038] In summary, the various modules of this system are interconnected via circuitry to form a complete closed-loop control system. Multimodal sensing module 1 is responsible for data acquisition, data processing and analysis module 2 is responsible for data processing and sleep state analysis, environmental control module 3 is responsible for adjusting environmental parameters, and power management module 4 is responsible for energy consumption optimization. The modules communicate efficiently and collaboratively via circuitry and communication protocols, thereby achieving comprehensive monitoring of newborn sleep states and intelligent optimization of the sleep environment.

[0039] The technical solutions of the present invention are not limited to the specific embodiments described above. Any technical modifications made in accordance with the technical solutions of the present invention fall within the protection scope of the present invention.

Claims

1. A low-power neonatal sleep management system, characterized in that: It includes a multimodal sensing module (1), a data processing and analysis module (2), an environmental control module (3), and a power management module (4); The multimodal sensing module (1) includes a temperature sensor, a humidity sensor, a sound sensor, a heart rate monitoring unit, and a respiratory rate monitoring unit. Each sensor is integrated into a wearable device via a flexible circuit board and transmits data to the central processing unit (5) via Bluetooth Low Energy protocol. The data processing and analysis module (2) uses a hierarchical feature extraction algorithm to perform preliminary processing on the collected data, identifies the infant's sleep state through dynamic threshold determination technology, and transmits the results to the environmental control module (3). The environmental control module (3) includes a temperature control unit (6), a humidification unit (7), and a noise suppression unit (8), which adjusts the sleep environment by receiving instructions from the data processing and analysis module (2). The power management module (4) uses energy recovery technology combined with lithium battery power supply and dynamically adjusts the power supply strategy by monitoring the real-time power consumption of each module.

2. The low-power neonatal sleep management system according to claim 1, characterized in that: The temperature sensor, humidity sensor and sound sensor in the multimodal sensing module (1) are all manufactured using low-power MEMS technology. The installation method is an embedded design, and the thickness of the flexible circuit board is 0.5 mm.

3. The low-power neonatal sleep management system according to claim 1, characterized in that: The heart rate monitoring unit and respiratory rate monitoring unit acquire signals using photoplethysmography and impedance measurement, respectively. The electrode pads are in contact with the skin via conductive silicone, with a contact area of ​​1 square centimeter, and the contact pressure is adjusted by a spring structure.

4. The low-power neonatal sleep management system according to claim 1, characterized in that: The hierarchical feature extraction algorithm of the data processing and analysis module (2) includes a first layer for extracting time-domain features, a second layer for extracting frequency-domain features, and a third layer for combining time-domain and frequency-domain features to generate a comprehensive feature vector. The dynamic threshold determination technology is based on historical data training to obtain a set of adaptive thresholds.

5. A low-power neonatal sleep management system according to claim 1, characterized in that: The temperature control unit (6) uses a semiconductor cooling chip as the core component, with a maximum power of 5 watts and a response time of 3 seconds; the humidification unit (7) uses ultrasonic atomization technology, with a maximum atomization volume of 300 ml per hour and an operating noise of less than 30 decibels; the noise suppression unit (8) uses active noise reduction technology, with a noise reduction depth of 20 decibels.

6. A low-power neonatal sleep management system according to claim 1, characterized in that: The power management module (4) converts the baby's movements into electrical energy through piezoelectric materials, generating 0.1 milliwatts of energy per movement. The lithium battery has a capacity of 2000 mAh and supports continuous operation of the system for 72 hours.

7. A low-power neonatal sleep management method, characterized in that: The method includes the following steps: Step S1: Multimodal data acquisition; Step S2: Data processing and sleep state analysis; Step S3: Generation of environmental control instructions; Step S4: Adjust environmental parameters; Step S5: Power Management and Energy Optimization.

8. A low-power neonatal sleep management method according to claim 7, characterized in that: In step S1, the multimodal data acquisition is performed by acquiring ambient temperature through a temperature sensor, ambient humidity through a humidity sensor, ambient noise intensity through a sound sensor, heart rate signal of the infant through a heart rate monitoring unit, and respiratory signal of the infant through a respiratory rate monitoring unit. All data are uploaded to the central processing unit (5) at fixed time intervals.

9. A low-power neonatal sleep management method according to claim 7, characterized in that: In step S5, the power management and energy consumption optimization reduce system energy consumption by monitoring the real-time power consumption of each module and dynamically adjusting the sampling frequency and data transmission rate of the sensors, while ensuring data acquisition accuracy.

Images