LED display screen control module of intelligent wearable device

Abstract

The invention particularly relates to an LED display screen control module of intelligent wearable equipment, which comprises an action sensing feedback unit for capturing and identifying user actions, an accelerometer, a gyroscope, a temperature compensation circuit, a low-pass filter, an action sensing microcontroller and a vibration motor interface, and further comprises an analysis unit for analyzing the actions and states of a user, comprising a data processing chip and further comprises an anomaly detection module, and the anomaly detection module is integrated on the data processing chip and used for recognizing actions in an abnormal state based on historical data and real-time data. The data processing chip firstly executes waveform analysis, then selectively carries out standard deviation calculation, frequency domain analysis, time domain feature extraction and wavelet transform based on the waveform analysis, not only can improve efficiency, but also can reduce unnecessary resource consumption and reduce misoperation, and adopts a self-adaptive learning algorithm and an intelligent neglect mechanism, so that the reliability of the system is improved. And the equipment can automatically adjust the acceleration threshold according to the use habit of the user.

Description

Technical Field

[0001] The present invention belongs to the field of display screen control, and in particular relates to an LED display screen control module for an intelligent wearable device. Background Art

[0002] With the rapid development of the smart wearable device market, users have higher and higher requirements for device experience. Especially in the control of LED display screens, traditional control methods have deficiencies in many aspects. These problems not only affect the user experience, but also limit the battery life and overall performance of the device. Specifically, they are manifested in the following aspects:

[0003] Frequent misoperations: Traditional control methods rely on single sensor data and are easily affected by environmental noise, resulting in frequent misoperations. For example, when a user is doing daily activities, the device may mistakenly think that the user is performing a specific operation and trigger the screen wake-up or other functions.

[0004] Existing devices have significant delays in recognizing user actions and responding, affecting the user's instant experience.

[0005] Lack of intelligence: Many current smart wearable devices lack sufficient intelligent design and are unable to dynamically adjust screen control strategies based on actual user usage, resulting in a poor user experience.

[0006] Insufficient personalized services: Traditional LED display control solutions fail to fully utilize machine learning technology to predict user usage habits, making the services provided by the equipment not personalized enough and unable to meet the needs of different users. Summary of the invention

[0007] The purpose of the present invention is to provide an LED display screen control module for a smart wearable device to solve the problems raised in the above background technology.

[0008] In order to solve the above technical problems, the present invention provides the following technical solutions:

[0009] An LED display screen control module for a smart wearable device, comprising:

[0010] A motion sensing feedback unit for capturing and identifying user motions, the motion sensing feedback unit comprising:

[0011] An accelerometer, used to capture the user's acceleration information in three-dimensional space;

[0012] Gyroscope, used to assist the accelerometer in motion recognition and provide angular velocity information;

[0013] Temperature compensation circuit, used to ensure that the sensor can still work accurately in different temperature environments;

[0014] A low-pass filter to filter out high-frequency noise in the accelerometer and gyroscope data;

[0015] A motion sensing microcontroller that processes sensor data and performs multi-level threshold setting, adaptive learning algorithms, and dual-sensor data fusion;

[0016] Vibration motor interface, used to trigger vibration feedback to confirm that the action has been recognized when a specific action is recognized;

[0017] Also included is an analysis unit for analyzing the user's actions and states, the analysis unit comprising:

[0018] A data processing chip for first performing waveform analysis, and then selectively performing standard deviation calculation, frequency domain analysis, time domain feature extraction, and wavelet transformation based on the waveform analysis;

[0019] Deep learning interactive interface, used to interact with online deep learning servers and run deep learning models to recognize complex patterns;

[0020] Memory, used to store machine learning models and historical data;

[0021] Anomaly detection module,The anomaly detection module is integrated in the data processing chip, and is used to identify actions under abnormal conditions based on historical data and real-time data.

[0022] Furthermore, the temperature compensation circuit includes a temperature sensor, and a temperature compensation algorithm is implemented in a microcontroller of the temperature compensation circuit. The algorithm dynamically adjusts the output value of the sensor based on a pre-calibrated temperature-output relationship curve or table.

[0023] Furthermore, the motion sensing microcontroller processes the sensor data, performs multi-level threshold setting, adaptive learning algorithm and dual sensor data fusion, specifically:

[0024] The motion sensing microcontroller collects raw data from the accelerometer and gyroscope via an I2C or SPI interface;

[0025] The received data is first temperature corrected by a temperature compensation circuit and then removed from high-frequency noise by a low-pass filter;

[0026] Implement multi-level threshold setting: The motion sensing microcontroller dynamically adjusts the thresholds of acceleration and angular velocity based on the user's usage habits and current activity intensity; the thresholds are used to distinguish different motion modes;

[0027] Adaptive learning algorithm: Through the built-in machine learning algorithm, the motion-sensing microcontroller learns the user's typical motion patterns and adjusts the threshold accordingly;

[0028] Dual sensor data fusion: Use complementary filter or extended Kalman filter algorithm to fuse the data of accelerometer and gyroscope.

[0029] Furthermore, the data processing chip is used to first perform waveform analysis, and then selectively perform standard deviation calculation, frequency domain analysis, time domain feature extraction, and wavelet transformation based on the waveform analysis. Specifically:

[0030] The data processing chip first performs waveform analysis on the raw data collected from the accelerometer and gyroscope; by analyzing the shape, amplitude and duration of this data, it can preliminarily determine whether the user is stationary;

[0031] When waveform analysis indicates that the user is not stationary, the data processing chip calculates the standard deviation of the acceleration and angular velocity data; the standard deviation can be used to quantify the intensity of the movement and thus distinguish different types of movement;

[0032] If the standard deviation exceeds a predetermined threshold, further analysis is performed on the specific action type:

[0033] When the waveform analysis shows a periodic motion, the frequency domain analysis of the periodic motion is performed, and the data processing chip will perform a fast Fourier transform to convert the time domain signal into a frequency domain signal;

[0034] By extracting the time domain features of the waveform, the data processing chip can identify the peak, valley, and zero-crossing features in the action;

[0035] When waveform analysis finds the presence of non-periodic transient motion, and the motion is jumping or swinging the arms, wavelet transform is applied to analyze the transient features in the motion.

[0036] Furthermore, the interactive online deep learning server runs deep learning models to recognize complex patterns, specifically:

[0037] The deep learning server deploys data-trained deep learning models, including convolutional neural networks or recurrent neural networks, which can recognize complex patterns in user actions.

[0038] Real-time data transmission between smart wearable devices and deep learning servers; when smart wearable devices capture user action data, this data will be sent to the server for further processing; the deep learning model on the server side will infer the transmitted data to recognize complex patterns; complex patterns include gesture recognition; the server completes the data processing, recognizes the complex patterns in the user's actions, and feeds the results back to the smart wearable device.

[0039] Furthermore, the anomaly detection module uses an isolation forest or local anomaly factor algorithm to identify whether the current action deviates from the user's normal action pattern.

[0040] Furthermore, it also includes a wake-up unit for quickly responding to user actions and activating the screen. The wake-up unit includes: a backlight control circuit for controlling the turning on and off of the screen backlight according to the instructions of the screen wake-up unit.

[0041] Furthermore, it also includes a power management unit for intelligently switching the device to a sleep mode, the power management unit including:

[0042] Power management chip, used to monitor and manage the power supply of each part;

[0043] Voltage regulator, used to dynamically adjust the output voltage according to load conditions;

[0044] The sleep controller is used to implement a multi-level sleep mechanism and control the sleep duration.

[0045] Beneficial Effects

[0046] The data processing chip first performs waveform analysis, and then selectively performs standard deviation calculation, frequency domain analysis, time domain feature extraction, and wavelet transform based on the waveform analysis, which can not only improve efficiency but also reduce unnecessary resource consumption, improve the accuracy and robustness of motion recognition, and improve the accuracy and robustness of motion recognition by combining the data of the accelerometer and gyroscope and using data fusion algorithms (such as complementary filters or extended Kalman filters).

[0047] Even in complex environments, it can accurately judge user actions and reduce misidentification.

[0048] To reduce misoperation, the adaptive learning algorithm and intelligent ignoring mechanism are adopted to enable the device to automatically adjust the acceleration threshold according to the user's usage habits, thereby reducing misoperation.

[0049] The device can learn the user's typical movement patterns and adjust the threshold accordingly, further reducing the probability of misoperation.

[0050] Accelerating signal processing enables the device to respond to user actions in the first place, improving the user experience.

[0051] Through intelligent sleep, dynamic voltage regulation, energy consumption prediction and optimization, energy conservation can be maximized and the battery life of the equipment can be extended.

[0052] Dynamically adjusting the power supply strategy and sleep mechanism enables the device to enter a low-power state when not in use, thereby saving power.

[0053] Provide personalized services, predict user status in real time through machine learning technology, and manage device behavior based on personal habits.

[0054] The device can dynamically adjust the screen control strategy according to the user's actual usage and provide more personalized services.

[0055] Anomaly detection algorithms (such as isolation forest or local anomaly factor algorithms) are introduced to identify actions under abnormal conditions and take corresponding measures in a timely manner. BRIEF DESCRIPTION OF THE DRAWINGS

[0056] Figure 1 It is a system block diagram of the present invention. DETAILED DESCRIPTION

[0057] The following will be combined with the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative work are within the scope of protection of the present invention.

[0058] The purpose of the present invention is to provide a LED display screen control module for a smart wearable device disclosed in the present application, such as Figure 1 As shown, including:

[0059] A motion sensing feedback unit for capturing and identifying user motions, the motion sensing feedback unit further comprising:

[0060] An accelerometer, used to capture the user's acceleration information in three-dimensional space;

[0061] Gyroscope, used to assist the accelerometer in motion recognition and provide angular velocity information;

[0062] Temperature compensation circuit, used to ensure that the sensor can still work accurately in different temperature environments;

[0063] A low-pass filter to filter out high-frequency noise in the accelerometer and gyroscope data;

[0064] A motion sensing microcontroller that processes sensor data and performs multi-level threshold setting, adaptive learning algorithms, and dual-sensor data fusion;

[0065] Vibration motor interface, used to trigger vibration feedback to confirm that the action has been recognized when a specific action is recognized;

[0066] Also included is an analysis unit for analyzing the user's actions and states, the analysis unit further comprising:

[0067] A data processing chip for first performing waveform analysis, and then selectively performing standard deviation calculation, frequency domain analysis, time domain feature extraction, and wavelet transformation based on the waveform analysis;

[0068] Deep learning interactive interface, used to interact with online deep learning servers and run deep learning models to recognize complex patterns;

[0069] Memory, used to store machine learning models and historical data;

[0070] Anomaly detection module,The anomaly detection module is integrated in the data processing chip, and is used to identify actions under abnormal conditions based on historical data and real-time data.

[0071] The following is a detailed description: A LED display screen control module for a smart wearable device, comprising:

[0072] A motion sensing feedback unit for capturing and identifying user motions, the motion sensing feedback unit further comprising:

[0073] An accelerometer, used to capture the user's acceleration information in three-dimensional space;

[0074] Gyroscope, used to assist the accelerometer in motion recognition and provide angular velocity information;

[0075] Temperature compensation circuit, used to ensure that the sensor can still work accurately in different temperature environments;

[0076] The temperature compensation circuit includes:

[0077] Temperature sensor: The integrated temperature sensor is used to monitor the ambient temperature around the sensor module in real time. This helps to obtain the current operating temperature, thus providing a basis for temperature compensation.

[0078] A temperature compensation algorithm is implemented in the temperature compensation circuit microcontroller, which is based on a pre-calibrated temperature-output relationship curve or table and can dynamically adjust the output value of the sensor.

[0079] A low-pass filter to filter out high-frequency noise in the accelerometer and gyroscope data;

[0080] Key points for selecting low-pass filter:

[0081] Selection of the cutoff frequency of the low-pass filter: According to the characteristics of the accelerometer and gyroscope, select a suitable cutoff frequency. Usually this frequency should be lower than the upper limit of the human body movement frequency to ensure that useful signal components are not filtered out.

[0082] Filter order: The order of a filter determines its roll-off rate, which is how quickly the signal decays after the cutoff frequency. For wearable devices, choosing a second-order or higher filter can more effectively remove high-frequency noise while maintaining low phase distortion.

[0083] Filter type: Butterworth, Chebyshev or Bessel filters can be used, among which Butterworth filter is often selected because of its flat passband response. Chebyshev filter allows a certain fluctuation in the passband, but has a faster roll-off rate in the stopband. Bessel filter performs well in signal delay and is suitable for occasions that require minimum phase distortion.

[0084] The low-pass filter can be implemented by analog circuits or by digital signal processing on a microcontroller.

[0085] Dynamic adjustment: Considering that different users may have different exercise patterns, the parameters of the low-pass filter should support dynamic adjustment to optimize the filtering effect according to the actual application scenario. For example, when it is detected that the user is doing strenuous exercise, the cutoff frequency can be appropriately lowered to better suppress high-frequency noise.

[0086] The low-pass filter can not only effectively filter out high-frequency noise in the accelerometer and gyroscope data, but also ensure the complete retention of useful signals, thereby providing higher quality input for subsequent data processing, ultimately achieving more accurate motion recognition and a better user experience.

[0087] Motion sensing microcontroller, used to process sensor data, perform multi-level threshold setting, adaptive learning algorithm and dual sensor data fusion; as follows:

[0088] The motion sensing microcontroller collects raw data from the accelerometer and gyroscope via an I2C or SPI interface;

[0089] The received data is firstly temperature-corrected by the temperature compensation circuit, and then the high-frequency noise is removed by the low-pass filter to ensure the quality of the input data;

[0090] The motion sensing microcontroller dynamically adjusts the thresholds of acceleration and angular velocity based on the user's usage habits and current activity intensity; these thresholds are used to distinguish different motion modes, such as stillness, walking, and running;

[0091] Environmental adaptability: Threshold setting should not only consider user behavior, but also the environmental factors of the device, such as whether the device is worn on the arm or leg, and whether the user is in an indoor or outdoor environment;

[0092] User behavior modeling: Through built-in machine learning algorithms, the motion-sensing microcontroller can learn the user's typical motion patterns and adjust the thresholds accordingly;

[0093] Adopt complementary filter or extended Kalman filter algorithm to fuse the data of accelerometer and gyroscope to improve the accuracy and robustness of action recognition;

[0094] Data fusion is performed in real time inside the motion sensing microcontroller, ensuring that the user's movements can be quickly and accurately recognized even in complex environments;

[0095] The motion sensing microcontroller can not only accurately capture and identify the user's movements, but also continuously optimize itself according to the user's usage habits, thus providing a more intelligent and personalized user experience;

[0096] The vibration motor interface is used to trigger vibration feedback to confirm that the action has been recognized when a specific action is recognized.

[0097] An LED display screen control module of a smart wearable device also includes an analysis unit for analyzing the user's actions and states, the analysis unit including:

[0098] The data processing chip is used to first perform waveform analysis, and then selectively perform standard deviation calculation, frequency domain analysis, time domain feature extraction, and wavelet transformation based on the waveform analysis; the details are as follows:

[0099] Initial waveform analysis is used to identify basic motion patterns. The data processing chip first performs waveform analysis on the raw data collected from the accelerometer and gyroscope. By analyzing the shape, amplitude, and duration of this data, it can preliminarily determine whether the user is stationary.

[0100] When waveform analysis indicates that the user is not stationary, the data processing chip calculates the standard deviation of the acceleration and angular velocity data; the standard deviation can be used to quantify the intensity of the movement and thus distinguish different types of movement;

[0101] If the standard deviation exceeds a predetermined threshold, further analysis is performed on the specific action type:

[0102] When waveform analysis shows periodic motion, frequency domain analysis is performed. For periodic motion (such as walking or running), the data processing chip will perform a fast Fourier transform (FFT) to convert the time domain signal into a frequency domain signal; this makes it easier to identify the frequency characteristics of the motion, such as cadence;

[0103] By extracting the time domain features of the waveform, the data processing chip can identify the peak, valley, and zero-crossing features in the action; these features are very useful for distinguishing different actions;

[0104] When waveform analysis finds the presence of non-periodic transient motion, and the motion is jumping or swinging the arms, wavelet transform is applied.

[0105] Wavelet transform is used to analyze transient features in actions. For those actions with sudden nature, such as jumping or swinging arms quickly, wavelet transform can help capture these features;

[0106] Build a basic framework, which includes reading sensor data, performing waveform analysis, calculating standard deviation, performing frequency domain analysis (Fast Fourier Transform FFT), time domain feature extraction, and wavelet transform. Here we use a sample code snippet to demonstrate this process, sample code:

[0107]

[0108]

[0109]

[0110]

[0111]

[0112] This example code shows the basic logic of how to process sensor data.

[0113] Deep learning interactive interface, used to interact with online deep learning servers and run deep learning models to recognize complex patterns; specifically:

[0114] Data-trained deep learning models, such as convolutional neural networks (CNNs) or recurrent neural networks (RNNs), are deployed on deep learning servers; these CNNs or RNNs are able to recognize complex patterns in user actions, even those that are difficult to identify through simple statistical analysis;

[0115] Through wireless communication technology, real-time data transmission can be achieved between smart wearable devices and deep learning servers; when the device captures user action data, the data will be sent to the server for further processing;

[0116] The server-side deep learning model will infer the transmitted data to identify more complex action patterns; these patterns include but are not limited to gesture recognition, complex motion pattern recognition, and other high-level action analysis;

[0117] The models on the server can be updated regularly or on demand to reflect the latest user behavior patterns or improved algorithms; these updated models can be pushed to smart wearable devices via OTA to ensure that the devices always have the latest recognition capabilities;

[0118] Once the server has processed the data and identified complex patterns in the user's movements, it feeds the results back to the smart wearable device;

[0119] The interactive online deep learning server can not only provide powerful pattern recognition capabilities, but also dynamically adjust the model according to the user's behavior to improve the intelligence level of the device. This enables smart wearable devices to more accurately identify user actions, reduce misoperations, and dynamically adjust screen control strategies according to the user's actual activities, thereby significantly improving the user experience.

[0120] Memory, used to store machine learning models and historical data;

[0121] Anomaly detection module, which is integrated in the data processing chip, is used to identify abnormal actions based on historical data and real-time data; the anomaly detection module uses the Isolation Forest or Local Outlier Factor (LOF) algorithm to identify whether the current action deviates from the user's normal action pattern. These algorithms can detect data points that are significantly different from historical data, thereby identifying potential abnormal behaviors. Specifically:

[0122] Isolation Forest: This is a decision tree-based algorithm that segments data by randomly selecting features and feature values ​​to isolate outliers from other data points. The advantage of the Isolation Forest algorithm is its high efficiency and high sensitivity to outliers. By training the Isolation Forest model, the anomaly detection module can quickly identify action data that is significantly different from the user's normal action pattern.

[0123] Local Outlier Factor (LOF): This is a density-based algorithm that detects outliers by comparing the local density of a data point with the local density of its neighbors. The LOF algorithm can identify data points that appear sparse in a local area, i.e., outliers. By calculating the LOF value of a data point and comparing it with a threshold, the anomaly detection module can determine whether a certain action data is abnormal.

[0124] An LED display screen control module of a smart wearable device also includes a wake-up unit for quickly responding to user actions and activating the screen, and the wake-up unit includes:

[0125] The backlight control circuit is used to control the screen backlight to turn on and off according to the instructions of the screen wake-up unit;

[0126] An LED display screen control module of an intelligent wearable device also includes a power management unit for intelligently switching the device to a sleep mode, and the power management unit includes:

[0127] Power management chip, used to monitor and manage the power supply of each part;

[0128] Voltage regulator, used to dynamically adjust the output voltage according to load conditions;

[0129] The sleep controller is used to implement a multi-level sleep mechanism and control the sleep duration.

[0130] Specific circuit selection of each circuit or module mentioned in this application:

[0131] 1. Accelerometer and gyroscope

[0132] Accelerometer: ADXL345 (triaxial accelerometer, suitable for consumer electronics), Bosch BMA250E, etc.

[0133] Gyroscope: MPU-6050 (integrates three-axis gyroscope and three-axis accelerometer, widely used in various embedded systems), L3GD20H, etc.

[0134] 2. Temperature compensation circuit

[0135] Temperature sensor: TMP36 / TMP35 / TMP37 (linear voltage output temperature sensor, simple and easy to use), LM35, etc.

[0136] Temperature compensation algorithm: can be implemented in microcontroller using any language that supports C / C++ programming.

[0137] 3. Low-pass filter

[0138] Filter: You can choose Butterworth, Chebyshev or Bessel filters. The specific implementation can be done on a microcontroller through digital signal processing (DSP), or use a dedicated analog filter chip, such as TI's OPA series operational amplifier to implement an analog filter.

[0139] 4. Motion sensing microcontroller

[0140] Microcontroller: You can choose ARM Cortex-M series MCU, such as STM32F103 (Bluepill), STM32F4 series, etc. These MCUs have good performance and rich peripheral interfaces.

[0141] 5. Vibration motor interface

[0142] Driver IC: You can choose DRV8871 (single-channel full-bridge motor driver) or other driver ICs suitable for driving small vibration motors.

[0143] 6. Data processing chip

[0144] Data processing: You can use a high-performance microcontroller or DSP, such as TI's C2000 series or STM32F7 series. These processors have sufficient computing power to perform waveform analysis, standard deviation calculation, FFT and other processing.

[0145] 7. Deep Learning Interactive Interface

[0146] Wireless communication: You can use a Bluetooth BLE module (such as nRF52 series) or a Wi-Fi module (such as ESP32) for wireless communication with the server.

[0147] 8. Memory

[0148] Memory: You can choose NOR Flash (such as AT25DF161D) or NAND Flash (such as K9F5608U0D) to store machine learning models and historical data.

[0149] 9. Anomaly Detection Module

[0150] Integrated in data processing chip: can be implemented through software without additional hardware.

[0151] 10. Backlight control circuit

[0152] Backlight control: PWM signal can be used to control the LED brightness. Common driver ICs include TPS65186 (used for backlight control of LCD panels).

[0153] 11. Power Management Unit

[0154] Power management chip: You can choose Texas Instruments' (TI) TPS61088 (boost converter) or other applicable power management IC.

[0155] Voltage Regulator: You can choose a low dropout regulator (LDO) such as the LM2937, or a switch mode power supply (SMPS) such as the TPS62160.

[0156] Sleep Controller: The sleep mechanism can be implemented in software or by using a microcontroller with low power modes.

[0157] The core technical effects of this application are mainly reflected in the following aspects:

[0158] The data processing chip first performs waveform analysis, and then selectively performs standard deviation calculation, frequency domain analysis, time domain feature extraction, and wavelet transform based on the waveform analysis, which can not only improve efficiency but also reduce unnecessary resource consumption, improve the accuracy and robustness of motion recognition, and improve the accuracy and robustness of motion recognition by combining the data of the accelerometer and gyroscope and using data fusion algorithms (such as complementary filters or extended Kalman filters).

[0159] Even in complex environments, it can accurately judge user actions and reduce misidentification.

[0160] To reduce misoperation, the adaptive learning algorithm and intelligent ignoring mechanism are adopted to enable the device to automatically adjust the acceleration threshold according to the user's usage habits, thereby reducing misoperation.

[0161] The device can learn the user's typical movement patterns and adjust the threshold accordingly, further reducing the probability of misoperation.

[0162] Accelerating signal processing enables the device to respond to user actions in the first place, improving the user experience.

[0163] Through intelligent sleep, dynamic voltage regulation, energy consumption prediction and optimization, energy conservation can be maximized and the battery life of the equipment can be extended.

[0164] Dynamically adjusting the power supply strategy and sleep mechanism enables the device to enter a low-power state when not in use, thereby saving power.

[0165] Provide personalized services, predict user status in real time through machine learning technology, and manage device behavior based on personal habits.

[0166] The device can dynamically adjust the screen control strategy according to the user's actual usage and provide more personalized services.

[0167] Anomaly detection algorithms (such as isolation forest or local anomaly factor algorithms) are introduced to identify actions under abnormal conditions and take corresponding measures in a timely manner.

[0168] In summary, this application aims to achieve more accurate recognition of user actions through a series of innovative designs and technical improvements, improve the response speed and energy efficiency of the device, and provide more personalized and secure services, thereby significantly improving the overall user experience of smart wearable devices. Through the realization of these technical effects, smart wearable devices can not only more accurately identify user intentions and reduce misoperations, but also dynamically adjust screen control strategies according to the user's actual activities to achieve intelligent management.

Claims

1. An LED display screen control module for a smart wearable device, characterized in that: include: A motion sensing feedback unit for capturing and identifying user motions, the motion sensing feedback unit comprising: An accelerometer, used to capture the user's acceleration information in three-dimensional space; Gyroscope, used to assist the accelerometer in motion recognition and provide angular velocity information; Temperature compensation circuit, used to ensure that the sensor can still work accurately in different temperature environments; A low-pass filter to filter out high-frequency noise in the accelerometer and gyroscope data; A motion sensing microcontroller that processes sensor data and performs multi-level threshold setting, adaptive learning algorithms, and dual-sensor data fusion; Vibration motor interface, used to trigger vibration feedback to confirm that the action has been recognized when a specific action is recognized; Also included is an analysis unit for analyzing the user's actions and states, the analysis unit comprising: A data processing chip for first performing waveform analysis, and then selectively performing standard deviation calculation, frequency domain analysis, time domain feature extraction, and wavelet transformation based on the waveform analysis; Deep learning interactive interface, used to interact with online deep learning servers and run deep learning models to recognize complex patterns; Memory, used to store machine learning models and historical data; Anomaly detection module,The anomaly detection module is integrated in the data processing chip, and is used to identify actions under abnormal conditions based on historical data and real-time data.

2. The LED display screen control module of a smart wearable device according to claim 1, characterized in that: The temperature compensation circuit includes a temperature sensor, and a temperature compensation algorithm is implemented in the temperature compensation circuit microcontroller. The algorithm dynamically adjusts the output value of the sensor based on a pre-calibrated temperature-output relationship curve or table.

3. The LED display screen control module of a smart wearable device according to claim 1, characterized in that: The motion sensing microcontroller processes sensor data, performs multi-level threshold setting, adaptive learning algorithms, and dual sensor data fusion. Specifically: The motion sensing microcontroller collects raw data from the accelerometer and gyroscope via an I2C or SPI interface; The received data is first temperature corrected by a temperature compensation circuit and then removed from high-frequency noise by a low-pass filter; Implement multi-level threshold setting: The motion sensing microcontroller dynamically adjusts the thresholds of acceleration and angular velocity based on the user's usage habits and current activity intensity; the thresholds are used to distinguish different motion modes; Adaptive learning algorithm: Through the built-in machine learning algorithm, the motion-sensing microcontroller learns the user's typical motion patterns and adjusts the threshold accordingly; Dual sensor data fusion: Use complementary filter or extended Kalman filter algorithm to fuse the data of accelerometer and gyroscope.

4. The LED display screen control module of a smart wearable device according to claim 1, characterized in that: The data processing chip is used to perform waveform analysis first, and then selectively perform standard deviation calculation, frequency domain analysis, time domain feature extraction, and wavelet transformation based on the waveform analysis. Specifically: The data processing chip first performs waveform analysis on the raw data collected from the accelerometer and gyroscope; by analyzing the shape, amplitude and duration of this data, it can preliminarily determine whether the user is stationary; When waveform analysis indicates that the user is not stationary, the data processing chip calculates the standard deviation of the acceleration and angular velocity data; the standard deviation can be used to quantify the intensity of the movement and thus distinguish different types of movement; If the standard deviation exceeds a predetermined threshold, further analysis is performed on the specific action type: When the waveform analysis shows a periodic motion, the frequency domain analysis of the periodic motion is performed, and the data processing chip will perform a fast Fourier transform to convert the time domain signal into a frequency domain signal; By extracting the time domain features of the waveform, the data processing chip can identify the peak, valley, and zero-crossing features in the action; When waveform analysis finds the presence of non-periodic transient motion, and the motion is jumping or swinging the arms, wavelet transform is applied. Wavelet transform is used to analyze transient characteristics in motion.

5. The LED display screen control module of a smart wearable device according to claim 1, characterized in that: An interactive online deep learning server that runs deep learning models to recognize complex patterns. Specifically: The deep learning server deploys data-trained deep learning models, including convolutional neural networks or recurrent neural networks, which can recognize complex patterns in user actions. Real-time data transmission between smart wearable devices and deep learning servers; when smart wearable devices capture user action data, this data will be sent to the server for further processing; the deep learning model on the server side will infer the transmitted data to recognize complex patterns; complex patterns include gesture recognition; the server completes the data processing, recognizes the complex patterns in the user's actions, and feeds the results back to the smart wearable device.

6. The LED display screen control module of a smart wearable device according to claim 1, characterized in that: The anomaly detection module uses the isolation forest or local anomaly factor algorithm to identify whether the current action deviates from the user's normal action pattern.

7. The LED display screen control module of a smart wearable device according to claim 1, characterized in that: It also includes a wake-up unit for quickly responding to user actions and activating the screen. The wake-up unit includes: a backlight control circuit for controlling the turning on and off of the screen backlight according to the instructions of the screen wake-up unit.

8. The LED display screen control module of a smart wearable device according to claim 1, characterized in that: Also included is a power management unit for intelligently switching the device to sleep mode, the power management unit including: Power management chip, used to monitor and manage the power supply of each part; Voltage regulator, used to dynamically adjust the output voltage according to load conditions; The sleep controller is used to implement a multi-level sleep mechanism and control the sleep duration.

Images