A low-power intelligent terminal device fast wake-up starting method
By analyzing historical wake-up data of smart terminal devices, calculating wake-up frequency and complexity, and dynamically adjusting the sleep level, the problem of inconsistency between sleep level settings and user needs is solved, achieving a balance between low power consumption and fast wake-up.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- BEIJING LEISHEN BOFENG INFORMATION TECH CO LTD
- Filing Date
- 2026-06-11
- Publication Date
- 2026-07-14
Smart Images
Figure CN122394980A_ABST
Abstract
Description
Technical Field
[0001] This invention relates to the field of electronic digital data processing technology, and specifically to a fast wake-up and startup method for low-power smart terminal devices. Background Technology
[0002] With the development of IoT technology, smart terminal devices are widely used in smart homes, health monitoring, industrial sensing, and outdoor monitoring. To meet the low power consumption requirements of smart terminal devices, existing technologies involve putting these devices into a sleep state when not in use, shutting down unnecessary modules to reduce power consumption and extend battery life. The devices are then powered on and the corresponding modules are activated when needed.
[0003] In existing technologies, smart terminal devices often have multiple sleep levels, with different wake-up speeds and power consumptions. The specific sleep level used is usually set by the user and remains unchanged after setting. However, if the sleep level setting is inconsistent with the user's current usage needs, such as using a fast wake-up sleep level when user needs are low, it will lead to increased sleep power consumption, affecting the device's usage time and lifespan. Conversely, using a low-power sleep level when user needs are high will result in a slower wake-up speed, impacting the user experience. Summary of the Invention
[0004] To address the aforementioned technical problems, the present invention aims to provide a fast wake-up and startup method for low-power smart terminal devices, the specific technical solution of which is as follows: In a first aspect, the present invention provides a fast wake-up and startup method for a low-power smart terminal device, comprising the following steps: 1) Based on the wake-up data of a set number of historical periods of smart terminal devices, calculate the wake-up frequency ratio of each time period of each period and the complexity of each effective wake-up of each time period of each period, and combine the wake-up frequency ratio and the complexity to obtain the wake-up intensity of each effective wake-up of each time period of each period of the smart terminal devices. 2) Based on the wake-up intensity of each effective wake-up in each time period of each cycle of the smart terminal device, calculate the effective wake-up distribution characteristics in each time period within a cycle, obtain the fast wake-up start-up demand score of the smart terminal device in each time period within a cycle, and then determine the sleep level of the smart terminal device in each time period within a cycle.
[0005] In conjunction with the first aspect mentioned above, in some possible implementations, the effective wake-up distribution characteristics of each time period within a cycle include the density characteristics and peak characteristics of the effective wake-up distribution of each time period within a cycle. Methods for obtaining the fast wake-up and startup demand scores of smart terminal devices at various time periods within a cycle include: By integrating the density and peak characteristics of the effective wake-up distribution in each time period within a cycle, a rapid wake-up start-up demand score is obtained for each time period within a cycle.
[0006] In conjunction with the first aspect mentioned above, among some possible implementation methods, the methods for calculating the complexity of each effective wake-up in each time period of each cycle include: Based on a set number of historical wake-up data from smart terminal devices, filter out the valid wake-up data for the set number of historical periods. By integrating the percentage of wake-up modules, the execution time after wake-up, and the amount of data processed after wake-up in each effective wake-up period of each cycle, the complexity of each effective wake-up in each effective wake-up period of each cycle can be obtained.
[0007] In conjunction with the first aspect mentioned above, among some possible implementations, methods for determining the sleep level of a smart terminal device at different times within a cycle include: Cluster analysis is performed on the fast wake-up start-up demand scores for each time period within a cycle to obtain each cluster. Based on the fast wake-up start-up demand scores in each cluster, the hibernation level for each time period is determined. The higher the fast wake-up start-up demand score in a cluster, the lower the determined hibernation level.
[0008] In conjunction with the first aspect mentioned above, among some possible implementations, methods for calculating the density characteristics of the effective wake-up distribution at any given time interval within a cycle include: For a given period, the average wake-up intensity of all valid wake-ups during that period and its adjacent periods is calculated. Then, the average wake-up intensity of a set number of historical periods is averaged to obtain the density characteristics of the valid wake-up distribution during that period within a given period.
[0009] In conjunction with the first aspect mentioned above, among some possible implementations, methods for calculating the peak characteristics of the effective wake-up distribution for any time period within a cycle include: Find the maximum value of the wake-up intensity of all valid wake-ups in a given historical period, and obtain the peak characteristics of the valid wake-up distribution in that period within a period.
[0010] In conjunction with the first aspect mentioned above, among some possible implementation methods, the method of obtaining the fast wake-up start-up demand score of a smart terminal device for each time period within a cycle by integrating the density characteristics and peak characteristics of the effective wake-up distribution in each time period within a cycle includes: The density and peak characteristics of each time period within a cycle are weighted and summed to obtain the fast wake-up and startup demand score of the smart terminal device in each time period within a cycle. The weights for the weighted summation are derived from the correlation between the density and peak characteristics of each time period within a cycle and the proportion of the overall wake-up frequency of each time period within a cycle. The percentage of total wake-up frequency in any given period within a cycle is the average percentage of wake-up frequency in any given period across a set number of historical cycles.
[0011] In conjunction with the first aspect mentioned above, in some possible implementations, the wake-up data includes at least the amplitude of the wake-up signal, the rising edge slope, the falling edge slope, and the duration. Methods for filtering effective wake-up data include: Wake-up data with an amplitude greater than or equal to the noise threshold, a rising edge slope within the set rising edge slope range, a falling edge slope within the set falling edge slope range, and a duration greater than the set duration are selected as valid wake-up data.
[0012] In conjunction with the first aspect mentioned above, in some possible implementations, if the number of historical cycles of the smart terminal device is less than a set number, the sleep level of the smart terminal device will be configured to a preset sleep level.
[0013] In conjunction with the first aspect mentioned above, among some possible implementation methods, the clustering analysis method is K-means clustering analysis.
[0014] The present invention has the following beneficial effects: Based on the wake-up data of a set number of historical periods of smart terminal devices, the present invention calculates the wake-up frequency ratio of each time period in each period and the complexity of each effective wake-up in each time period of each period, and then integrates them to obtain the wake-up intensity of each time period of each period of the smart terminal devices. This realizes the quantification of the user's usage of smart terminal devices in a set number of historical periods. Furthermore, based on the wake-up intensity of each time period in each period, the effective wake-up distribution characteristics of each time period in a period are calculated. This realizes the regular statistical analysis of a set number of historical periods and obtains the fast wake-up start-up demand score of each time period in a period to reflect the user's general usage of smart terminal devices. This enables the setting of different sleep levels in a period according to the user's usage habits, so that the sleep level corresponds to the user's usage needs in each time period, and minimizes sleep power consumption while meeting wake-up needs. Attached Figure Description
[0015] Figure 1 This is a flowchart of the method steps in an embodiment of the present invention. Detailed Implementation
[0016] To clearly illustrate the technical features of this solution, the invention will be described in detail below through specific implementation methods and in conjunction with the accompanying drawings.
[0017] This invention uses a certain smart terminal device as an example to illustrate a fast wake-up and startup method for a low-power smart terminal device. The smart terminal device includes a main CPU (Central Processing Unit) and various peripheral modules, including a memory module, a display module, a communication module, a sensing module, and a storage module. The memory module is RAM, in which data is lost after power-off, while the storage module is flash memory, in which data is not lost after power-off. The storage module is used to store wake-up data from historical cycles, the communication address of the communication module, and the configuration information of the sensing module.
[0018] Since the use of smart terminal devices typically follows a certain periodicity—for example, the wake-up time of smart door locks is usually consistent every day—this invention uses a set number of historical wake-up data periods to determine the wake-up status of smart terminal devices at different times within a period, and then formulates a sleep strategy for different times within a period. Specific method steps are as follows: Figure 1 As shown, the process includes the following steps: Step 1: Obtain a set number of historical wake-up data from smart terminal devices, and then filter out the valid wake-up data from the set number of historical periods.
[0019] In this embodiment, a 24-hour day is used as a cycle, and the set quantity is 7 as an example. Therefore, the wake-up data for the set quantity of historical cycles includes wake-up data from the past week. Wake-up data includes the type, amplitude, trigger time, rising edge slope, duration, execution duration after wake-up, data processing volume after wake-up, and falling edge slope of the wake-up signal. The execution duration after wake-up is the time from the first module being woken up and powered on to the last module being powered off and going into sleep mode. Based on the type of wake-up signal, the number of peripheral modules to be woken up by the wake-up signal is determined. Wake-up signal types include GPIO level, touch, motion, wireless command, sensor trigger, etc. Each type requires different peripheral modules to be woken up. Statistical analysis can determine the number of peripheral modules woken up by each type of wake-up signal (also called the number of wake-up modules).
[0020] The time period to which the wake-up signal belongs is determined based on the trigger time of the wake-up signal. The number of wake-up modules, the execution duration after wake-up, and the amount of data processed after wake-up are used to calculate the complexity of the wake-up signal.
[0021] In this embodiment, each time period is 10 minutes, thus dividing a cycle (one day) into 144 time periods.
[0022] Based on the rising slope, amplitude, duration, and falling slope of the wake-up signal, valid wake-up data is selected from all wake-up data to avoid invalid wake-ups such as signal interference fluctuations and accidental touches affecting the formulation of sleep strategies.
[0023] Methods for filtering effective wake-up data include: If the amplitude of the wake-up signal is greater than or equal to the preset noise threshold, the rising edge slope is within the preset rising edge slope range, the falling edge slope is within the preset falling edge slope range, and the duration of the wake-up signal is greater than the set duration, then the wake-up signal is determined to be a valid wake-up signal, and the data of the wake-up signal is valid wake-up data.
[0024] The preset noise threshold is the maximum amplitude of the wake-up signal caused by common noise in smart terminal devices. When the amplitude of the wake-up signal is less than the noise threshold, it is determined that the wake-up signal is caused by noise and is not a valid wake-up signal.
[0025] The rising edge slope range and falling edge slope range are determined based on the rising edge slope range and falling edge slope range of the effective wake-up signal of the smart terminal device measured under experimental conditions. The upper limit of the rising edge slope range is determined by adding a certain margin (e.g., 5%) to the upper limit of the rising edge slope range based on the measured effective wake-up signal, and the lower limit of the rising edge slope range is determined by subtracting a certain margin (e.g., 5%) from the lower limit of the rising edge slope range based on the measured effective wake-up signal. The upper limit of the falling edge slope range is determined by adding a certain margin (e.g., 5%) to the upper limit of the falling edge slope range based on the measured effective wake-up signal, and the lower limit of the falling edge slope range is determined by subtracting a certain margin (e.g., 5%) from the lower limit of the falling edge slope range based on the measured effective wake-up signal.
[0026] The duration is set based on the minimum duration of the measured effective wake-up signal minus a certain margin (e.g., 5%).
[0027] If the rising edge slope of the wake-up signal is not within the preset rising edge slope range or the falling edge slope is not within the preset falling edge slope range, it indicates that the slope is too slow (such as slow environmental drift) or too steep (such as electrostatic transient glitches), and the wake-up signal is marked as invalid.
[0028] If the duration of the wake-up signal is less than or equal to the set duration, it is determined to be a wake-up signal caused by accidental touch by the user, and not an actual valid wake-up signal.
[0029] If the smart terminal device has no wake-up data for a set number of historical periods, or if the number of historical periods is less than the set number, it will enter sleep mode according to the preset sleep level and continue to collect wake-up data until the number of historical periods equals the set number, thus obtaining the set number of wake-up data for historical periods. In this embodiment, the preset sleep level is level two sleep mode, which provides a good balance between sleep power consumption and wake-up speed.
[0030] Step 2: Based on the valid wake-up data selected in Step 1, determine the wake-up intensity of each valid wake-up in each time period of each cycle.
[0031] 2.1) Calculate the wake-up frequency percentage of each period in each cycle of the smart terminal device.
[0032] For a single cycle, the percentage of wake-up frequency in each time period is equal to the ratio of the number of effective wake-ups in each time period to the total number of effective wake-ups in that cycle. The formula is expressed as: In the formula, For the first The first cycle The percentage of wake-up frequency in each time period For the first The first cycle Effective wake-up count in a given time period For the first The total number of effective wake-ups in each cycle. If the first cycle... Total number of effective wake-ups per cycle If it is 0, then set the first The wake-up frequency percentage for each time period within each cycle is 0.
[0033] 2.2) Calculate the complexity of each effective wake-up in each time period of each cycle of the smart terminal device.
[0034] For a single effective wake-up, the more wake-up modules there are, the longer the execution time after wake-up, the larger the amount of data processed after wake-up, and the greater the complexity. Therefore, in this embodiment, the method for calculating the complexity of a single wake-up includes: In the formula, For the first The first cycle The first time period The complexity of each effective wake-up, For the first The first cycle The first time period The number of wake-up modules that are successfully woken up. The total number of peripheral modules in the smart terminal device (i.e., the total number of peripheral modules, in this embodiment) ), For the first The first cycle The first time period The percentage of wake-up modules that are effectively woken up. For the first The first cycle The first time period The amount of data processed after a single effective wake-up. This is the maximum amount of data processed after wake-up, preset at the factory by the smart terminal device. For the first The first cycle The first time period The execution duration after a valid wake-up. This is the maximum execution time after wake-up that is preset at the factory for smart terminal devices. If, after a certain wake-up, the number of wake-up modules is 0, the execution time after wake-up is 0, or the amount of data processed after wake-up is 0, it means that the smart terminal device has not actually worked after wake-up, and therefore the complexity of that wake-up is also 0.
[0035] 2.3) By integrating the wake-up frequency percentage of each time period within each cycle and the complexity of each effective wake-up within that time period, the wake-up intensity of each effective wake-up in each time period of each cycle is obtained. Specific methods include: In the formula, For the first The first cycle The first time period The wake-up strength of a single effective wake-up.
[0036] Step 3: Based on the wake-up intensity of each effective wake-up in each time period of each cycle calculated in Step 2, calculate the effective wake-up distribution characteristics in each time period within a cycle, and then obtain the fast wake-up start-up demand score in each time period within a cycle.
[0037] In this embodiment, the effective wake-up distribution characteristics of each time period within a cycle include the density characteristics and peak characteristics of the effective wake-up distribution of each time period within a cycle.
[0038] 3.1) Based on the wake-up intensity of each effective wake-up in each time period of each cycle calculated in step 2, calculate the density characteristics of the effective wake-up distribution in each time period within a cycle.
[0039] The density characteristics of the effective wake-up distribution in each time period within a cycle are used to characterize the continuous usage of smart terminal devices by users in each time period. The calculation method includes: First, for a given period, calculate the average wake-up intensity of all valid wake-ups for each time period and a predetermined number of adjacent time periods. In this implementation, the predetermined number is 1. The formula is as follows: In the formula, For the first The first cycle The average wake-up intensity of all valid wake-ups within a given time period and the number of adjacent time periods before and after it. For the first The first cycle The first time period before the previous time period The wake-up strength of a single effective wake-up. For the first The first cycle The first time period after the first time period The wake-up strength of a single effective wake-up. , , The first The first cycle The time period, the first The previous time period, the first time period The total number of effective wake-ups in the next time period. If no smart terminal devices are woken up during a given time period and a set number of adjacent time periods before and after it (e.g., users do not use smart terminal devices late at night), the average wake-up strength is set to 0.
[0040] For the first period of each cycle, the preceding period is the last period of the previous cycle. For the first cycle that does not have a preceding cycle, , For the last period of each cycle, the next period becomes the first period of the next cycle. For the last cycle where there is no next cycle, , .
[0041] Since historical data for a single period may contain significant data biases and cannot represent the general usage habits of users for smart terminal devices, this invention calculates the mean of wake-up intensity for all effective wake-ups within a set number of historical periods and a set number of adjacent periods before and after each period. This yields the density characteristics of the effective wake-up distribution. The specific calculation method includes: In the formula, For the first time in a period Density characteristics of the effective wake-up distribution over a given time period. In this embodiment, in order to set the quantity, .
[0042] 3.2) Based on the wake-up intensity of each effective wake-up in each time period of each cycle calculated in step two, calculate the peak characteristics of the effective wake-up distribution in each time period within a cycle.
[0043] For the first in a period The methods for calculating the peak characteristics of the effective wake-up distribution over a given time period include: Find the first historical period of a given number of times The maximum value of the awakening intensity in the time period is used as the _th The peak characteristics of the effective wake-up distribution over a given time period are expressed by the formula: In the formula, For the first time in a period Peak characteristics of the effective wake-up distribution over a time period. This indicates the operation to find the maximum value. When the set quantity is in the historical period of the [number]th [period], [the operation is performed]. When there is no valid wake-up data in any time period, set It is 0.
[0044] 3.3) By combining the density characteristics of the effective wake-up distribution in each time period within a cycle obtained in step 3.1) and the peak characteristics of the effective wake-up distribution in each time period obtained in step 3.2), the fast wake-up start-up demand score for each time period within a cycle is obtained.
[0045] In this embodiment, the method for fusing the density and peak characteristics of the effective wake-up distribution across different time periods within a cycle is weighted summation. The formula is expressed as follows: In the formula, For the first time in a period The score for rapid wake-up and startup requirements in each time period. Density features The weight, Peak characteristics The weight.
[0046] and The values are based on density features and peak characteristics The correlation between the frequency and the overall wake-up frequency ratio in different time periods within a cycle is determined. Specific methods include: Calculate the overall wake-up frequency percentage for each time period within a cycle: In the formula, For the first time in a period The overall wake-up frequency percentage for each time period In this embodiment, in order to set the quantity, =7.
[0047] Calculate the density characteristics for each time period within a period. The constructed sequence (with density features) (Arranged chronologically by time period) and the percentage of total wake-up frequency for each time period within a cycle. The sequence constitutes the proportion of overall wake-up frequency. The degree of relevance (arranged chronologically by time period) And the peak characteristics of each period within a cycle. The constructed sequence (with peak features) (Arranged chronologically by time period) and the percentage of total wake-up frequency for each time period within a cycle. The degree of correlation of the constructed sequences This leads to the density characteristics. weight and peak characteristics weight : In this embodiment, the correlation coefficient is used to determine the degree of correlation. Since density and peak characteristics are positively correlated with arousal intensity, and the overall arousal frequency percentage is the mean of the arousal frequency percentages, and the arousal frequency percentage is also positively correlated with arousal intensity, it can be concluded that density and peak characteristics are also positively correlated with the overall arousal frequency percentage. and The values of are all greater than or equal to 0. When hour, and For the set fixed weights, The specific values for both can be set based on the emphasis placed on density and peak features, such as setting... .
[0048] Step 4: Perform cluster analysis on the rapid wake-up demand scores for each time period within a cycle to determine the hibernation level of smart terminal devices for each time period within a cycle.
[0049] In this embodiment, there are three sleep levels for smart terminal devices: Level 1 sleep (lowest sleep level, high frequency, high demand mode), Level 2 sleep (medium frequency, medium demand mode), and Level 3 sleep (highest sleep level, low frequency, low demand mode). The sleep levels increase sequentially from Level 1 to Level 2 to Level 3, and the specific details of each sleep level are as follows: Level 1 hibernation – ultra-fast wake-up (wake-up latency ≤ 20ms, and hibernation power consumption is controllable): The main CPU does not enter a deep power-down state, but only shuts down the extended computing module (floating-point operation) to maintain a "lightweight operating state" (sleep power consumption ≤40μA) without needing to be re-powered and initialized; the core modules corresponding to the high-frequency wake-up source (such as the communication module and the sensing module) enter a "standby state" (power is always on, only data processing is turned off, startup time ≤5ms), and non-core modules (such as the display module) are deeply powered down; the communication address of the communication module and the configuration information of the sensing module in the storage module are stored in the memory module, and can be read from the storage module without waking up. The storage module is powered down, but the memory module is not powered down.
[0050] Level 2 Sleep – Power Consumption-Speed Balance (Wake-up latency ≤ 50ms, Sleep power consumption ≤ 20μA): The main CPU is deeply powered down (only the interrupt wake-up pin of the main CPU is powered down, and all other pins are powered down), but the "CPU fast power-up configuration" is stored in advance in the guard domain RAM (the guard domain RAM is a specific area in the memory module that does not lose power during hibernation to support the technical effect of "fast power-up"); the core modules (such as the communication module and the sensing module) enter "semi-hibernation state", and the non-core modules (such as the display module) are deeply powered down; the configuration information of the sensing module in the storage module is stored in the memory module, and can be read from the storage module without waking up. The storage module is powered down, but the memory module is not powered down.
[0051] Level 3 Sleep – Extremely Low Power Consumption (Wake-up latency ≤100ms, Sleep power consumption ≤10μA): The main CPU and all hardware modules are completely and deeply powered down (only all non-essential power links are disconnected), achieving a sleep power consumption of ≤10μA.
[0052] Therefore, through cluster analysis, the rapid wake-up demand scores for each time period within a cycle are divided into three categories: high, medium, and low, corresponding to three sleep levels. The cluster centers of the three clusters are extracted and sorted from high to low. The cluster with the highest rapid wake-up demand score is designated as high wake-up demand, the intermediate cluster as medium wake-up demand, and the lowest cluster as low wake-up demand. The high wake-up demand cluster corresponds to Level 1 sleep (high frequency, high demand mode), the medium wake-up demand cluster corresponds to Level 2 sleep (medium frequency, medium demand mode), and the low wake-up demand cluster corresponds to Level 3 sleep (low frequency, low demand mode).
[0053] In this embodiment, the K-means clustering algorithm is used, with K=3. During the clustering analysis, the initial centroid is taken as... The 25%, 50%, and 75% quantiles of the sequence are clustered using existing techniques, which will not be elaborated upon in this embodiment.
Claims
1. A fast wake-up and startup method for a low-power smart terminal device, characterized in that, Includes the following steps: 1) Based on the wake-up data of a set number of historical periods of smart terminal devices, calculate the wake-up frequency ratio of each time period of each period and the complexity of each effective wake-up of each time period of each period, and combine the wake-up frequency ratio and the complexity to obtain the wake-up intensity of each effective wake-up of each time period of each period of the smart terminal devices. 2) Based on the wake-up intensity of each effective wake-up in each time period of each cycle of the smart terminal device, calculate the effective wake-up distribution characteristics in each time period within a cycle, obtain the fast wake-up start-up demand score of the smart terminal device in each time period within a cycle, and then determine the sleep level of the smart terminal device in each time period within a cycle.
2. The fast wake-up and startup method for low-power intelligent terminal devices according to claim 1, characterized in that, The effective wake-up distribution characteristics of each time period within a cycle include the density characteristics and peak characteristics of the effective wake-up distribution of each time period within a cycle; Methods for obtaining the fast wake-up and startup demand scores of smart terminal devices at various time periods within a cycle include: By integrating the density and peak characteristics of the effective wake-up distribution in each time period within a cycle, a rapid wake-up start-up demand score is obtained for each time period within a cycle.
3. The fast wake-up and startup method for low-power intelligent terminal devices according to claim 1, characterized in that, The calculation methods for the complexity of each effective wake-up in each time period of each cycle include: Based on a set number of historical wake-up data from smart terminal devices, filter out the valid wake-up data for the set number of historical periods. By integrating the percentage of wake-up modules, the execution time after wake-up, and the amount of data processed after wake-up in each effective wake-up period of each cycle, the complexity of each effective wake-up in each effective wake-up period of each cycle can be obtained.
4. The fast wake-up and startup method for low-power intelligent terminal devices according to claim 1, characterized in that, Methods for determining the sleep level of a smart terminal device at different times within a cycle include: Cluster analysis is performed on the fast wake-up start-up demand scores for each time period within a cycle to obtain each cluster. Based on the fast wake-up start-up demand scores in each cluster, the hibernation level for each time period is determined. The higher the fast wake-up start-up demand score in a cluster, the lower the determined hibernation level.
5. The fast wake-up and startup method for low-power intelligent terminal devices according to claim 2, characterized in that, Methods for calculating the density characteristics of the effective arousal distribution at any given time interval within a cycle include: For a given period, the average wake-up intensity of all valid wake-ups during that period and its adjacent periods is calculated. Then, the average wake-up intensity of a set number of historical periods is averaged to obtain the density characteristics of the valid wake-up distribution during that period within a given period.
6. The fast wake-up and startup method for low-power intelligent terminal devices according to claim 2, characterized in that, Methods for calculating the peak characteristics of the effective wake-up distribution at any given time interval within a cycle include: Find the maximum value of the wake-up intensity of all valid wake-ups in a given historical period, and obtain the peak characteristics of the valid wake-up distribution in that period within a period.
7. The fast wake-up and startup method for low-power intelligent terminal devices according to claim 2, characterized in that, Methods for obtaining the fast wake-up demand score of a smart terminal device in each time period within a cycle by integrating the density and peak characteristics of the effective wake-up distribution in each time period within a cycle include: The density and peak characteristics of each time period within a cycle are weighted and summed to obtain the fast wake-up and startup demand score of the smart terminal device in each time period within a cycle. The weights for the weighted summation are derived from the correlation between the density and peak characteristics of each time period within a cycle and the proportion of the overall wake-up frequency of each time period within a cycle. The percentage of total wake-up frequency in any given period within a cycle is the average percentage of wake-up frequency in any given period across a set number of historical cycles.
8. The fast wake-up and startup method for a low-power intelligent terminal device according to claim 1, characterized in that, Wake-up data includes at least the amplitude of the wake-up signal, the rising edge slope, the falling edge slope, and the duration. Methods for filtering effective wake-up data include: Wake-up data with an amplitude greater than or equal to the noise threshold, a rising edge slope within the set rising edge slope range, a falling edge slope within the set falling edge slope range, and a duration greater than the set duration are selected as valid wake-up data.
9. The fast wake-up and startup method for a low-power intelligent terminal device according to claim 1, characterized in that, If the number of historical cycles for a smart terminal device is less than the set number, the sleep level of the smart terminal device will be configured to the preset sleep level.
10. The fast wake-up and startup method for a low-power intelligent terminal device according to claim 4, characterized in that, The clustering analysis method used is K-means clustering analysis.