Efficient and energy-saving power management method for communication terminal

Abstract

The invention belongs to the technical field of power management, and particularly relates to a high-efficiency energy-saving power management method for a communication terminal, which comprises the following steps of: 1, identifying a current scene of a user, initializing monitoring parameters of multi-dimensional state data according to different scenes, and configuring different monitoring parameters corresponding to different scenes; 2, collecting multi-dimensional state data, and judging whether the data is abnormal or not based on double conditions: if the battery voltage is instantaneously lower than a current UVLO threshold value, and the voltage drop rate exceeds a specified value or the voltage change acceleration exceeds a preset value, entering a dynamic power supply regulation mode if the two conditions are both met, or entering a dynamic power supply regulation mode if the two conditions are met; by identifying the current scene of the user and initializing the adaptive multi-dimensional state data monitoring parameters, the problem of excessive monitoring or insufficient monitoring caused by adopting unified monitoring configuration is avoided, high-frequency monitoring parameters do not need to be configured for a low-power-consumption scene, low-frequency monitoring parameters do not need to be configured for a high-dynamic scene, and thus the user experience is improved. And invalid power consumption of a monitoring link is reduced from the source.

Description

Technical Field

[0001] This invention belongs to the field of power management technology, and in particular relates to a method for efficient and energy-saving power management of communication terminals. Background Technology

[0002] As the functions of communication terminals such as smartphones and portable tablets become increasingly sophisticated, high-frequency applications such as high-definition video, multitasking, and real-time navigation have led to a significant increase in terminal power consumption. Battery life has become a core bottleneck restricting user experience, making efficient power management technology a crucial aspect of terminal design. However, current power management solutions for communication terminals have many technical shortcomings, making it difficult to meet the dual requirements of high efficiency and energy saving while ensuring functionality.

[0003] First, existing solutions lack scenario-based monitoring design, employing a uniform configuration of multi-dimensional status data monitoring parameters. Whether it's a low-power sleep scenario, a high-dynamic commuting scenario, or a high-load application operation scenario, the same sampling frequency and threshold standard are used. This leads to over-monitoring in low-power scenarios, increasing unnecessary energy consumption, and insufficient monitoring in high-dynamic scenarios, failing to detect power supply anomalies in a timely manner, thus limiting power management efficiency from the source.

[0004] Secondly, the power supply anomaly detection relies on single-condition triggering (such as only monitoring whether the battery voltage is below a fixed threshold), without considering dynamic characteristics such as voltage drop rate and acceleration. This is susceptible to false triggering due to instantaneous voltage fluctuations, frequently interrupting normal terminal use; it may also cause missed triggering by ignoring dynamic voltage change patterns, missing real power supply anomalies, and failing to balance energy saving effect and functional stability.

[0005] Furthermore, power supply regulation strategies and emergency mechanisms lack scenario adaptability. Power reduction often adopts a one-size-fits-all approach, failing to design tiered strategies based on the core functional requirements of different scenarios. Emergency power supply is not linked to scenario characteristics; for example, in sleep scenarios, non-essential functions are still powered, resulting in wasted power, and in high-load scenarios, the current task is not prioritized for power supply, leading to data loss. It is difficult to match the differentiated needs of users. Therefore, there is an urgent need for a highly efficient and energy-saving power management method that takes into account scenario adaptability, judgment accuracy, dynamic regulation, and parameter self-optimization. Summary of the Invention

[0006] The purpose of this invention is to address the aforementioned technical problems by providing a highly efficient and energy-saving power management method for communication terminals.

[0007] In view of this, the present invention provides a high-efficiency energy-saving power management method for communication terminals, comprising the following steps: Step 1: Identify the user's current scenario and initialize monitoring parameters for multi-dimensional status data according to different scenarios. Different scenarios correspond to different monitoring parameter configurations. Step 2: Collect multi-dimensional status data and determine whether the data is abnormal based on two conditions: If the battery voltage is momentarily lower than the current UVLO threshold and the voltage drop rate exceeds the specified value or the voltage change acceleration exceeds the preset value, if both conditions are met, enter the dynamic power supply adjustment mode; otherwise, return to Step 2 to continue collecting data. Step 3: After entering the dynamic power supply adjustment mode, start timing according to the preset timing cycle of the current scenario, and at the same time match the preset tiered power supply strategy according to the current scenario, and output the scenario-based silent prompt. Step 4: During the timing process, monitor the battery voltage recovery status according to different scenarios: for commuting scenarios, collect the voltage every 2 seconds; for sleep scenarios, collect the voltage every 10 seconds; for high-load scenarios, collect the voltage every 1 second. Step 5: If voltage recovery is detected before the end of the preset timing period, restore the power consumption of the communication terminal according to the gradient corresponding to the current scenario; if voltage recovery is not detected after the end of the preset timing period, trigger the scenario-related emergency power supply mechanism. Step 6: Periodically analyze the multi-dimensional status data collected in history, predict the battery life of the communication terminal, and dynamically optimize the monitoring parameters in Step 1 based on the analysis results.

[0008] Furthermore, the specific method for identifying the user's current scenario in step 1 is as follows: if the user's movement speed is detected to be greater than 3m / s and the navigation application is running in the background, it is determined to be a commuting scenario; if the communication terminal screen is detected to be off for more than 10 minutes and there are no call or video tasks running, it is determined to be a hibernation scenario; if the communication terminal CPU utilization rate is detected to be greater than 70%, it is determined to be a high load scenario.

[0009] Furthermore, the specific content of the tiered power supply strategy in step 3 is as follows: if the current scenario is a commuting scenario, turn off mobile data, retain 2G call function and GPS navigation function, and reduce the screen brightness of the communication terminal to 30~45%; if the current scenario is a sleep scenario, turn off all wireless functions and data acquisition functions, and only maintain the minimum standby power consumption of the core chip of the communication terminal; if the current scenario is a high load scenario, close background running applications, limit the multi-core operation of the CPU, and reduce the screen brightness of the communication terminal to 25~35%.

[0010] Furthermore, the preset timing period in step 3 is as follows: the timing period for commuting scenarios is set to 30 minutes, the timing period for hibernation scenarios is set to 2 hours, and the timing period for high-load scenarios is set to 15 minutes.

[0011] Furthermore, the specific method of the scene-related emergency power supply mechanism in step 5 is as follows: If the current scene is a hibernation scene, trigger deep hibernation power supply: cut off all power supply circuits except for the alarm clock function and power monitoring function. When the alarm clock is triggered, temporarily restore the screen power supply circuit. If the current scene is a high-load scene, trigger task protection power supply: prioritize the allocation of power supply resources to the currently running task, and automatically compress the data of the currently running task and save it to local storage. When the battery power is lower than the emergency threshold, cut off the power supply circuits corresponding to non-current tasks. If the current scene is a commuting scene, trigger basic communication power supply: only retain the power supply circuits corresponding to the call function, SMS function and GPS positioning function, and turn off the power supply circuits corresponding to the screen display until the battery power is exhausted.

[0012] Furthermore, in step 6, the period for periodically analyzing historical data is every day at dawn; the specific method for dynamically optimizing monitoring parameters is as follows: for scenarios with high frequency of abnormal battery voltage in historical data, the frequency is increased by 5%-10%; for communication terminals with battery capacity below 80%, when the scenario is determined to be a sleep scenario, the frequency is reduced to 60%-80% of the original frequency.

[0013] Furthermore, the multi-dimensional state data collected in step 2 includes battery voltage data, communication terminal motion acceleration data, and user location data; when initializing monitoring parameters in step 1, UVLO threshold, voltage drop rate specified value, and voltage change acceleration preset value are set for different scenarios.

[0014] Furthermore, in step 4, when monitoring voltage recovery, if the voltage values ​​collected three times consecutively are higher than the UVLO threshold corresponding to the current scenario, it is determined that the voltage has recovered; in step 5, the emergency threshold is set according to the scenario: the emergency threshold for high load scenario is set to 10% of the total battery capacity, the emergency threshold for commuting scenario is set to 8% of the total battery capacity, and the emergency threshold for hibernation scenario is set to 5% of the total battery capacity.

[0015] Furthermore, in step 6, when predicting the battery life of the communication terminal, a prediction model based on an LSTM neural network is used, which combines historical battery voltage change data, power consumption data under different scenarios, and battery capacity decay data. The training data includes scenario-power consumption-voltage correlation data for the past 30 days, and outputs the predicted battery life for the next 24 hours. After dynamically optimizing the monitoring parameters, the optimized parameters are used to initialize the monitoring configuration when step 1 is executed next time.

[0016] The beneficial effects of this invention are: By identifying the user's current scenario and initializing adapted multi-dimensional state data monitoring parameters, the problem of over-monitoring or under-monitoring caused by using a unified monitoring configuration is avoided. There is no need to configure high-frequency monitoring parameters for low-power scenarios or low-frequency monitoring parameters for high-dynamic scenarios, thus reducing the ineffective power consumption in the monitoring process from the source.

[0017] By using dual-condition anomaly detection, the accuracy of the response is improved. Based on the dual-condition anomaly detection of battery voltage instantaneously falling below the current UVLO threshold and voltage drop rate exceeding the specified value, the problem of false triggering or missed triggering caused by single-condition detection can be effectively avoided. This ensures that the adjustment is only initiated when there is a real battery power supply anomaly, thus ensuring a balance between normal terminal use and energy saving.

[0018] By combining dynamic power supply adjustment mode with scenario-preset tiered strategies, non-core power consumption can be reduced in a targeted manner when power supply is abnormal; the scenario-related emergency power supply mechanism further ensures the core functions of the scenario in extreme cases, avoiding power waste and ensuring that users' critical needs are not affected.

[0019] By regularly analyzing historical multi-dimensional data and dynamically optimizing monitoring parameters, the power management strategy can adapt to changes in the terminal's state during long-term use, avoiding the decline in management efficiency caused by the solidification of initial parameters, and achieving continuous iterative improvement of power management capabilities. Attached Figure Description

[0020] Figure 1 This is a flowchart of the high-efficiency energy-saving power management method for communication terminals proposed in this invention; Detailed Implementation

[0021] The technical solutions of the embodiments of this application will be clearly described below with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of this application. All other embodiments obtained by those skilled in the art based on the embodiments of this application are within the scope of protection of this application.

[0022] In the description of this application, it should be noted that the terminology used herein is for the purpose of describing particular embodiments only and is not intended to limit the exemplary embodiments according to this application. For ease of description, the dimensions of the various parts shown in the drawings are not drawn to actual scale. Techniques, methods, and devices known to those skilled in the art may not be discussed in detail, but where appropriate, such techniques, methods, and devices should be considered part of the specification. In all examples shown and discussed herein, any specific values ​​should be interpreted as merely exemplary and not as limitations. Therefore, other examples of exemplary embodiments may have different values. It should be noted that similar reference numerals and letters in the following drawings denote similar items; therefore, once an item is defined in one drawing, it need not be further discussed in subsequent drawings.

[0023] It should be noted that the terms "first," "second," etc., used in the specification and claims of this application are used to distinguish similar objects and are not used to describe a specific order or sequence. It should be understood that such data can be interchanged where appropriate so that embodiments of this application can be implemented in orders other than those illustrated or described herein, and the objects distinguished by "first," "second," etc., are generally of the same class, and the number of objects is not limited; for example, a first object can be one or more. Furthermore, in the specification and claims, "and / " indicates at least one of the connected objects, and the character " / " generally indicates that the preceding and following objects are in an "or" relationship.

[0024] It should be noted that in the description of this application, the directional terms such as front, back, up, down, left, right, horizontal, vertical, perpendicular, horizontal, top, bottom, etc., indicate the orientation or positional relationship based on the orientation or positional relationship shown in the accompanying drawings. They are only for the convenience of describing this application and simplifying the description. Unless otherwise stated, these directional terms do not indicate or imply that the device or element referred to must have a specific orientation or be constructed and operated in a specific orientation, and therefore should not be construed as a limitation on the scope of protection of this application. The directional terms "inner" and "outer" refer to the inner and outer contours relative to the outline of each component itself.

[0025] It should be noted that in this application, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that includes a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus. Without further limitations, an element defined by the phrase "comprising a…" does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes that element. Furthermore, it should be noted that the scope of the methods and apparatuses in the embodiments of this application is not limited to performing functions in the order shown or discussed, but may also include performing functions substantially simultaneously or in the reverse order, depending on the functions involved. For example, the described methods may be performed in a different order than described, and various steps may be added, omitted, or combined. Additionally, features described with reference to certain examples may be combined in other examples.

[0026] Reference Figure 1 A method for efficient and energy-saving power management of communication terminals includes the following steps: Step 1: Identify the user's current scenario and initialize monitoring parameters for multi-dimensional status data according to different scenarios. Different scenarios correspond to different monitoring parameter configurations. Step 2: Collect multi-dimensional status data and determine whether the data is abnormal based on two conditions: If the battery voltage is momentarily lower than the current UVLO threshold and the voltage drop rate exceeds the specified value or the voltage change acceleration exceeds the preset value, if both conditions are met, enter the dynamic power supply adjustment mode; otherwise, return to Step 2 to continue collecting data. Step 3: After entering the dynamic power supply adjustment mode, start timing according to the preset timing cycle of the current scenario, and at the same time match the preset tiered power supply strategy according to the current scenario, and output the scenario-based silent prompt. Step 4: During the timing process, monitor the battery voltage recovery status according to different scenarios: for commuting scenarios, collect the voltage every 2 seconds; for sleep scenarios, collect the voltage every 10 seconds; for high-load scenarios, collect the voltage every 1 second. Step 5: If voltage recovery is detected before the end of the preset timing period, restore the power consumption of the communication terminal according to the gradient corresponding to the current scenario; if voltage recovery is not detected after the end of the preset timing period, trigger the scenario-related emergency power supply mechanism. Step 6: Periodically analyze the multi-dimensional status data collected in history, predict the battery life of the communication terminal, and dynamically optimize the monitoring parameters in Step 1 based on the analysis results.

[0027] The communication terminal of this invention first activates the scene recognition module, collecting user behavior and terminal status data through sensors and the system backend. It matches preset scene types (commuting, sleep, high load) and initializes monitoring parameters for multi-dimensional status data based on scene characteristics. Subsequently, the terminal collects multi-dimensional data such as battery voltage and motion acceleration in real time through the battery management module and motion sensors, entering a dual-condition anomaly judgment stage: first, it checks whether the battery voltage is momentarily lower than the UVLO threshold of the current scene; then, it judges whether the voltage drop rate exceeds the scene's specified value or whether the voltage change acceleration exceeds the scene's preset value. Only when both conditions are met simultaneously is the dynamic power supply adjustment mode triggered; otherwise, it returns to the data acquisition stage for continuous monitoring and enters the dynamic power supply adjustment stage. After entering power regulation mode, the terminal starts timing according to the preset timing cycle based on the current scenario. At the same time, it matches the tiered power supply strategy corresponding to the scenario and outputs scenario-specific silent prompts through system pop-ups or indicator lights. During the timing process, the terminal monitors the battery voltage at a scenario-differentiated frequency. If a voltage recovery signal is continuously collected before the timing ends, the power consumption is restored according to the scenario gradient. If the voltage does not recover before the timing ends, the scenario-related emergency power supply mechanism is triggered. Finally, the terminal retrieves historically collected multi-dimensional data at fixed intervals, and uses the data analysis module to statistically analyze information such as the frequency of scenario anomalies and changes in battery capacity. It dynamically adjusts the monitoring parameters in step 1 and stores the optimized parameters in the configuration library to provide an updated initialization basis for the next power management cycle.

[0028] The accelerometer sensor selected is the LIS3DH triaxial accelerometer (range ±2g, resolution 12-bit, power consumption ≤10μA) which is matched with the STM32L4 series. VCC is connected to the 3.3V power supply terminal, GND is connected to the system ground, and SDA / SCL are connected to the PB7 / PB6 pins of the main controller (such as STM32H743) (I2C1 interface, communication rate 400kHz). The positioning module is a Ublox NEO-6M GPS module (positioning accuracy 1-3m, update rate 1Hz, power consumption ≤50mA). VCC is connected to 3.3V, GND is connected to system ground, and TX / RX are connected to the PA3 / PA2 pins of the main controller (UART2 interface, baud rate 9600bps). The system background process monitoring module communicates with the operating system kernel through the main controller: the Android system calls the ActivityManager interface, and the iOS system calls the NSWorkspace framework, with a monitoring frequency of 500ms / time.

[0029] The mobile speed calculation method adopts a fusion algorithm: first, the acceleration data collected by LIS3DH is zero-drift calibrated, and then the instantaneous speed is calculated by trapezoidal integration; at the same time, the latitude and longitude data output by the GPS module every second is read, the distance difference between two adjacent positioning is calculated (based on the Haversine formula), and divided by 1 second to obtain the positioning speed; finally, Kalman filtering (filter coefficient 0.7) is used to fuse the two speeds, filtering out integral drift and GPS noise. If the GPS signal is lost (such as in an underground parking garage), only the acceleration integral speed is relied upon, and a 5-minute confidence time limit is set. If there is still no GPS signal after the time limit, it is judged by default as a non-commuting scenario.

[0030] The system resource management module reads the CPU_CNT register of the main controller and calculates the percentage of running cycles every 100ms, which is the CPU utilization rate. Only if the statistical value is greater than 70% for 3 consecutive times is it determined to be a high load scenario, so as to avoid misjudgment of instantaneous peak values.

[0031] In the example of this application, the specific method for identifying the user's current scenario in step 1 is as follows: if the user's movement speed is detected to be greater than 3m / s and the navigation application is running in the background, it is determined to be a commuting scenario; if the communication terminal screen is detected to be off for more than 10 minutes and there are no call or video tasks running, it is determined to be a hibernation scenario; if the communication terminal CPU utilization rate is detected to be greater than 70%, it is determined to be a high load scenario.

[0032] As a preferred example of the present invention, the communication terminal collects the quantitative data required for scene determination through multi-module collaborative acquisition. The movement speed data is jointly calculated by the accelerometer and the positioning module, and the navigation application running status is obtained by the system background process monitoring module. When the movement speed is detected to be >3m / s and the navigation application is running in the background, it is determined to be a commuting scene. The screen off time is recorded by the screen control module, and the call task running status is fed back by the communication module and the multimedia module. When the screen off time is detected to be >10min and there is no call task process, it is determined to be a sleep scene. The CPU utilization rate is statistically analyzed in real time by the system resource management module. When the CPU utilization rate is detected to be >70%, it is determined to be a high load scene. Through the above-mentioned quantitative threshold and multi-condition combination determination logic, the terminal can accurately distinguish ambiguous scenes, avoid misjudgment based on single data, and provide a reliable scene basis for subsequent power supply strategies and monitoring schemes.

[0033] The screen control module uses an SSD1963 display controller and communicates with the main controller via an SPI interface. The screen off duration is recorded by the controller's PWR_DOWN_TIMER register with an accuracy of 1 second. If a touch signal or incoming call signal is received after the screen is off, the timer is immediately reset.

[0034] The communication module (such as Qualcomm SDX55) queries the call status via AT command ("AT+CLCC") and reports back to the main controller every 1 second; the multimedia module outputs a level signal through the AUDIO_ACTIVE pin, and the main controller detects this signal through GPIO interrupt.

[0035] If "moving speed > 3m / s but navigation application not running" is detected, it is determined as "temporary moving scenario". The monitoring parameters of the previous scenario are used (if the previous scenario is sleep, the sleep parameters are maintained and re-determined after 5 minutes). If "CPU usage > 70% and screen is off" is detected, it is determined as "high background load scenario". The power supply strategy is executed according to the high load scenario, but the screen brightness adjustment logic is skipped (because the screen is off).

[0036] In the example of this application, the specific content of the tiered power supply strategy in step 3 is as follows: if the current scenario is a commuting scenario, turn off mobile data, retain 2G call function and GPS navigation function, and reduce the screen brightness of the communication terminal to 30~45%; if the current scenario is a sleep scenario, turn off all wireless functions and data acquisition functions, and only maintain the minimum standby power consumption of the core chip of the communication terminal; if the current scenario is a high load scenario, close background running applications, limit the multi-core operation of the CPU, and reduce the screen brightness of the communication terminal to 25~35%.

[0037] As a preferred example of the present invention, when the terminal determines the current scenario, it triggers a customized tiered power supply operation. In the commuting scenario, the terminal shuts down the mobile data transmission channel through the wireless module, maintains low-power operation of the GPS navigation function through the positioning module, and adjusts the screen brightness to 40% through the display control module to reduce unnecessary power consumption of the wireless module and the screen. In the sleep scenario, the terminal shuts down the network search and data transmission functions of the wireless module through system commands, disables the data acquisition process of the sensor, and maintains the minimum standby voltage and current of the core chip only through the power management chip to ensure that the terminal does not lose power and that power consumption is reduced to the minimum. In the high-load scenario, the terminal forcibly closes non-currently running applications through the background process management module, limits multi-core operation through the CPU control module, and adjusts the screen brightness to 30% to ensure the computing power requirements of the current high-load task while reducing redundant background power consumption and excessive hardware power consumption.

[0038] In the example of this application, the preset timing period in step 3 is as follows: the timing period for commuting scenario is set to 30 minutes, the timing period for hibernation scenario is set to 2 hours, and the timing period for high load scenario is set to 15 minutes.

[0039] As a preferred example of the present invention, after the terminal enters the dynamic power supply adjustment mode, the corresponding preset cycle parameter is retrieved from the timing cycle configuration library according to the determined scenario type and the timer is started. In high-load scenarios, because the CPU, GPU and other hardware are in a high-power operation state, the battery power consumption rate is fast, so the terminal sets a short timing cycle of 15 minutes; the battery voltage is collected once every 1 second during the timing period. If the voltage recovery is not detected within 15 minutes, the emergency mechanism can be quickly triggered to avoid continuous power waste. In sleep scenarios, the terminal core chip only maintains the minimum standby power consumption and the battery state is stable, so a long timing cycle of 2 hours is set; the battery voltage is collected once every 10 seconds during the timing period to reduce the extra power consumption caused by high-frequency monitoring and extend the battery life in sleep mode. In commuting scenarios, the terminal needs to balance the user's temporary use needs and power consumption control, so a medium timing cycle of 30 minutes is set; the battery voltage is collected once every 2 seconds during the timing period, which can respond to voltage recovery in a timely manner and avoid power consumption caused by frequent monitoring due to too short a timing period or power waste caused by too long a timing period.

[0040] In the example of this application, the specific method of the scene-related emergency power supply mechanism in step 5 is as follows: If the current scene is a hibernation scene, trigger deep hibernation power supply: cut off all power supply circuits except for the alarm clock function and power monitoring function, and temporarily restore the screen power supply circuit when the alarm clock is triggered. If the current scene is a high load scene, trigger task protection power supply: prioritize the allocation of power supply resources to the currently running task, and automatically compress the data of the currently running task and save it to local storage. When the battery power is lower than the emergency threshold, cut off the power supply circuit corresponding to the non-current task. If the current scene is a commuting scene, trigger basic communication power supply: only retain the power supply circuits corresponding to the call function, SMS function and GPS positioning function, and turn off the power supply circuit corresponding to the screen display until the battery power is exhausted.

[0041] As a preferred example of the present invention, when the preset timing period of the dynamic power supply adjustment mode ends and the terminal does not detect the battery voltage recovery, a scenario-related emergency power supply mechanism is triggered. Emergency power supply logic is invoked according to the scenario. In a sleep scenario, the terminal initiates a deep sleep power supply program: the power management chip cuts off all power supply circuits except for the alarm clock module and the power monitoring module; when the alarm is triggered, the screen power supply circuit is temporarily connected, and immediately disconnected after the alarm ends, ensuring that the remaining power is used only for core functions. In a high-load scenario, the terminal initiates a task protection power supply program: first, the resource scheduling module allocates battery power priority to the currently running task, and simultaneously triggers the data compression module to compress the temporary data of the current task and save it to local storage; then, the battery level is monitored in real time, and when the level is below 10%, the power supply circuits corresponding to non-current tasks are cut off, maintaining only the power supply for the current task and power monitoring. In a commuting scenario, the terminal initiates a basic communication power supply program: the communication module retains the power supply circuit for call and SMS functions, the positioning module retains the power supply circuit for GPS positioning functions, and simultaneously cuts off the power supply circuit for the screen display module; thereafter, the battery level is continuously monitored until the power is exhausted, ensuring that the user can still use communication and positioning functions in emergency situations.

[0042] Deep sleep power supply loop control: The power management chip employs a "path switching" function to cut off the power supply circuits for the screen, wireless module, and sensors (the main controller's PC13, PC14, and PC15 pins output a low level to control the MOSFET to turn off). Retain alarm clock function: The RTC timer continues to run. If the alarm trigger time (stored in RAM0x20005000) is detected, the main controller temporarily turns on the screen power supply circuit (PC13 outputs a high level). The alarm clock is turned off immediately after it ends. Battery monitoring function: The BMS chip maintains power supply (VCC=3.3V) and sends the remaining battery data to the main controller every 30 seconds. If the battery level is less than 2%, the alarm clock function is turned off, and only the battery monitoring function is retained.

[0043] Data processing for power supply protection in the task: Data compression: The LZ4 lightweight compression algorithm is used (compression ratio of approximately 2:1, compression speed > 400MB / s), which is suitable for embedded environments; the current task data path is obtained through the operating system's "file handle" (such as FileDescriptor in Android and NSFileHandle in iOS). Storage path: The compressed file will be named as follows: Save "emergency_backup_YYYYMMDDHHMMSS.lz4" to the local Flash memory (such as W25Q64JV) in the address range 0x08020000-0x08100000 (capacity 8MB). If the storage is full, delete the oldest backup file. Power allocation: Through the "priority power supply" function of the power management chip, 80% of the battery current is allocated to the current task (such as video editing and gaming), and 20% is allocated to data storage and power monitoring.

[0044] Basic communication power supply circuit configuration: Reserved power supply circuit: Calls / SMS: Powered by the 2G band of the wireless module (current ≤100mA). GPS positioning: Low-power power supply for GPS module (current ≤25mA); Power monitoring: Powered by BMS chip (current ≤1mA); Power supply circuit closed: Screen display: Disconnect the screen controller from the backlight power supply (current saving ≥200mA); Background applications: Disable CPU scheduling for all background processes (current saving ≥50mA); If an incoming call is detected, the screen power supply circuit is temporarily activated (for 30 seconds) to allow the user to answer the call, and the power is immediately turned off after the call ends.

[0045] Power consumption quantification of deep sleep power supply: Power consumption of each module in deep sleep mode: Module Power consumption (μA) Main controller (standby) 1 BMS chip 5 RTC timer 1 total ≤7 If the remaining battery capacity is 200mAh (emergency capacity in hibernation mode), the battery can last for approximately 2857 hours (about 119 days) in deep hibernation mode (200mAh / 0.007A).

[0046] Abnormal handling of power supply for task protection: If data compression fails (e.g., file corruption), the main controller will retry 3 times. If it still fails, the original data (uncompressed) will be saved to storage to ensure that data is not lost. If the current task is "real-time call", then no data saving will be performed (no temporary data), only the background power supply circuit will be cut off to ensure that the call is not interrupted.

[0047] User tips for basic communication power supply: When basic communication power is triggered, if the screen is already off, the indicator light will remain solid red to prompt the user to "enter emergency communication mode"; The indicator light flashes green three times every 30 minutes to remind the user of the current remaining battery level (e.g., three flashes indicate 30% remaining, and one flash indicates less than 10% remaining).

[0048] In the example of this application, the period for periodically analyzing historical data in step 6 is every morning; the specific method for dynamically optimizing monitoring parameters is as follows: for scenarios with high frequency of abnormal battery voltage in historical data, the frequency is increased by 5%-10%; for communication terminals with battery capacity below 80%, when the scenario is determined to be a sleep scenario, the frequency is reduced to 60%-80% of the original frequency.

[0049] As a preferred example of the present invention, the terminal has a built-in timed analysis module that automatically wakes up every morning at midnight to retrieve historical multi-dimensional data from the past 24 hours for statistical analysis. For scenarios with high frequency of abnormal battery voltage, the analysis module calculates the abnormal risk coefficient for the scenario, automatically increases the monitoring frequency by 5%-10%, and updates it to the scenario monitoring parameter mapping library. For terminals with battery capacity below 80%, the analysis module identifies the low power consumption characteristics of the sleep scenario and reduces the monitoring frequency in the scenario to 60%-80% of the original frequency, reducing the monitoring energy consumption of aging batteries during sleep. The optimized monitoring parameters are stored in the configuration file. When the terminal executes step 1 again, it directly reads the updated parameter configuration to avoid ineffective monitoring energy consumption.

[0050] Storage and management of historical data: Historical data storage format: CSV format, each record contains 12 fields: "Timestamp (YYYY-MM-DDHH:MM:SS), Scene type, Battery voltage (V), Remaining capacity (mAh), Discharge current (mA), Battery temperature (°C), CPU utilization (%), Screen brightness (%), GPS status (0=off, 1=on), Wireless module status (0=2G, 1=4G), Voltage drop rate (V / s), Number of anomalies (times)"; Storage location: Address range 0x08100000-0x087FFFFF (capacity 7MB) of local Flash memory (W25Q64JV), stored cyclically in time order, and the oldest data is overwritten when the memory is full; Data Upload: If the terminal is connected to the Internet, the historical data of the previous 24 hours will be uploaded to the cloud server at 3:00 AM every day (via the MQTT protocol) for subsequent big data analysis and optimization.

[0051] Software triggers for periodic analysis: The main controller is set to trigger the "historical data analysis task" at 2:00 AM every day via the RTC timer, with the highest priority (higher than other application tasks). The analysis task takes about 5 minutes to execute, during which other high-load tasks are restricted (CPU utilization ≤ 50%) to avoid affecting the accuracy of battery life prediction.

[0052] The specific logic for monitoring parameter optimization: Anomaly frequency statistics: Count the number of voltage anomalies in each scenario in the past 7 days. If the anomaly frequency of a certain scenario is >5 times / day, increase the monitoring frequency of that scenario by 5%-10% (e.g., increase the frequency of commuting scenarios from once every 2 seconds to once every 1.8 seconds). Battery capacity degradation judgment: If the current capacity monitored by BMS is less than 80% of the initial capacity (the initial capacity is stored at Flash address 0x08000000), then the monitoring frequency of the sleep mode will be reduced to 60%-80% of the original frequency (e.g., from once every 10 seconds to once every 12-16 seconds). Optimized parameter storage: The updated monitoring parameters (sampling interval, UVLO threshold, rate / acceleration threshold) are stored in the "optimized parameter area" (0x08010000-0x0801FFFF) of Flash. The parameters in this area will be read first during the next initialization in step 1.

[0053] In the example of this application, the multi-dimensional state data collected in step 2 includes battery voltage data, communication terminal motion acceleration data and user location data; when initializing the monitoring parameters in step 1, the UVLO threshold, voltage drop rate specified value and voltage change acceleration preset value are set for different scenarios.

[0054] As a preferred example of the present invention, the terminal collects multi-dimensional status data through the collaborative efforts of multiple hardware modules. Battery voltage data is collected in real time by the battery management system (BMS), motion acceleration data is collected by a triaxial accelerometer, and user location data is collected by a GPS / BeiDou positioning module. In step 1, the terminal calls a preset parameter library according to the scenario type to configure differentiated monitoring and judgment standards for different scenarios. In the sleep scenario, due to low battery power consumption and small voltage fluctuations, a lower UVLO threshold and a more lenient voltage drop rate and voltage change acceleration are set to avoid false triggering of regulation by slight voltage fluctuations. In the high load scenario, due to fast battery power consumption and high risk of voltage drop, a higher UVLO threshold and a stricter voltage drop rate and voltage change acceleration are set to ensure rapid response to power supply anomalies. Through scenario-based parameter configuration, the anomaly judgment standards are made to fit the power supply characteristics of each scenario, thereby improving the accuracy of anomaly identification.

[0055] In the example of this application, when monitoring voltage recovery in step 4, if the voltage values ​​collected three times consecutively are higher than the UVLO threshold corresponding to the current scenario, it is determined that the voltage has recovered; in step 5, the emergency threshold is set according to the scenario: the emergency threshold for high load scenario is set to 10% of the total battery capacity, the emergency threshold for commuting scenario is set to 8% of the total battery capacity, and the emergency threshold for hibernation scenario is set to 5% of the total battery capacity.

[0056] As a preferred example of the present invention, during the timing process of the dynamic power supply adjustment mode, after the terminal collects the battery voltage at a scene-differentiated frequency, it initiates the voltage recovery judgment logic: only when the voltage value collected three times consecutively is higher than the UVLO threshold corresponding to the current scene is it determined that the voltage has recovered. Through multiple sampling verifications, single voltage fluctuations are filtered out, avoiding repeated switching of power supply modes due to false recovery judgments, thus ensuring the stability of the terminal's power supply. At the same time, the terminal has preset an emergency threshold according to the scene during the initialization phase, and this threshold is stored in the configuration area of ​​the battery management module. When the scene-related emergency power supply mechanism is triggered, the battery management module calculates the current remaining battery power in real time. When the remaining power is lower than the emergency threshold of the corresponding scene, it automatically executes subsequent power supply control to ensure that the remaining power prioritizes meeting the core functional requirements of the scene and avoids premature power depletion or excessive power retention.

[0057] The interval between the three consecutive data collections is consistent with the scene sampling interval: Commuting scenario: Every 2 seconds once, for 3 consecutive times, that is, the average of the 3 times within 6 seconds is greater than the UVLO threshold (3.4V). Sleep scenario: Once every 10 seconds, for 3 consecutive times, i.e., the average of the 3 values ​​within 30 seconds is greater than the UVLO threshold (3.2V). High load scenario: Once every 1 second, for 3 consecutive times, that is, the average of the 3 times within 3 seconds is greater than the UVLO threshold (3.5V). If any of the intermediate collection values ​​is less than or equal to the UVLO threshold, immediately reset the "Recovery Count" and start counting again.

[0058] Specific calculation and calibration of emergency thresholds: The emergency threshold is calculated based on the total battery capacity (Ccurr, the current capacity monitored in real time by the BMS): Emergency capacity for high-load scenarios = 0.1 × Ccurr (e.g., Ccurr = 4000mAh, emergency capacity = 400mAh). Emergency capacity for commuting scenarios = 0.08 × Ccurr (e.g., Ccurr = 4000mAh, emergency capacity = 320mAh). Emergency capacity for hibernation scenarios = 0.05 × Ccurr (e.g., Ccurr = 4000mAh, emergency capacity = 200mAh). Automatic emergency threshold calibration once a month: The capacity calibration parameters of the BMS are updated by a full charge and discharge (from 0% to 100%) to ensure the accuracy of emergency capacity calculation.

[0059] Monitoring logic for emergency thresholds: The main controller reads the current remaining capacity Crem of the BMS every 5 seconds and compares it with the emergency capacity Cemerg. If Crem ≤ Cemerg, set the bit to trigger the next step of the emergency power supply mechanism.

[0060] If, under high load conditions, a voltage reading is 3.55V (>UVLO3.5V) but the next reading is 3.48V (<3.5V), then the "recovery count" is reset to zero and the voltage is not considered recovered. If, during a commuting scenario, the three voltage values ​​are 3.42V, 3.45V, and 3.43V respectively (all > 3.4V), then it is determined to be a recovery, triggering power consumption gradient recovery.

[0061] Storage of emergency threshold configuration area: The emergency threshold ratio parameters (10%, 8%, 5%) are stored in the EEPROM of the BMS chip (address 0x01-0x03). The main controller reads this ratio via I2C and calculates the emergency capacity in combination with the current capacity. If the BMS fails, the main controller uses a preset fixed emergency capacity (400mAh for high load, 320mAh for commuting, and 200mAh for hibernation) to ensure that the emergency mechanism does not fail.

[0062] Accuracy assurance of remaining battery power monitoring: The remaining power of the BMS is calibrated using the "coulomb counter" algorithm: the main controller integrates the discharge current (I×t) every second, accumulates the discharge capacity, and compares it with the remaining capacity of the BMS. If the error is >5%, the capacity parameters of the BMS are corrected. If the battery voltage is detected to be ≥4.2V (full charge voltage) during charging, the remaining power will be immediately reset to 100% and the capacity calculation benchmark will be calibrated.

[0063] In the example of this application, when predicting the battery life of the communication terminal in step 6, a prediction model based on LSTM neural network is used, which combines historical battery voltage change data, power consumption data under different scenarios and battery capacity decay data. The training data includes scenario-power consumption-voltage correlation data for the past 30 days, and outputs the battery life prediction result for the next 24 hours. After dynamically optimizing the monitoring parameters, the optimized parameters are used to initialize the monitoring configuration when step 1 is executed next time.

[0064] As a preferred example of the present invention, the terminal's battery life prediction function is implemented based on an LSTM neural network prediction model. During the model training phase, the terminal periodically retrieves historical data from the past 30 days, including battery voltage change curves, power consumption data, and battery capacity decay data under various scenarios. This data is organized into training samples according to dimensions and input into the LSTM model for training. The model parameters are optimized to enable the model to predict the correlation between scenarios and battery life. During the battery life output phase, when analyzing historical data every morning, the trained model combines the current battery capacity status and scenario predictions for the next 24 hours to output the battery life under different scenarios within the next 24 hours. This output is then fed back to the user through the system interface. After dynamically optimizing the monitoring parameters, the terminal stores the optimized parameters in the parameter configuration library. When step 1 is executed next time, the system directly reads the updated parameters from the configuration library to replace the original initial parameters and complete the monitoring configuration. This eliminates the need to wait for system version updates or manual settings, enabling the optimized parameters to take effect immediately and ensuring real-time adaptation between power management strategies and terminal status.

[0065] Training and deployment details of LSTM models: Model input feature processing: Scenario type coding: Commuting=1, Hibernation=2, High Load=3; Numerical feature normalization: Features such as voltage (2.7-4.5V), capacity (0-4000mAh), and power consumption (0-2000mA) are normalized to the [0,1] range (using Min-Max normalization: x_norm=(x-x_min) / (x_max-x_min)); Model training environment: Cloud training: The training was conducted on a server (CPU i7-12700K, GPU RTX3090) using the Python + TensorFlow 2.10 framework. The training dataset consisted of nearly 30 days of historical data from 100 terminals (approximately 86,400 samples). Terminal deployment: The trained model is converted into TensorFlowLite format (.tflite, about 500KB in size) and stored in the 0x08020000 address range of the terminal Flash. It is then loaded and run by the NPU of the main controller (such as Huawei Kirin 9000SNPU), with an inference time of <100ms.

[0066] Battery life prediction output: The model outputs the battery life for the next 24 hours (in hours), and breaks it down by scenario (e.g., "8 hours for commuting, 48 hours for hibernation, and 3 hours for high load"). The predicted battery life is displayed through the "Power Management APP" on the terminal. If the predicted battery life is less than 4 hours, a pop-up window will prompt the user that "the estimated battery life is less than 4 hours, and it is recommended to charge".

[0067] The mechanism for dynamically optimizing parameters to take effect: Parameter activation trigger: Before scene recognition in step 1, the main controller reads the flag bit of the Flash "optimized parameter area" (0 ​​= not updated, 1 = updated). If it is 1, the optimized parameters are loaded; otherwise, the default parameters are loaded. Parameter rollback mechanism: If, after loading the optimization parameters, an abnormal judgment is triggered three times in a row (e.g., the adjustment mode is triggered even though there is no abnormality), the parameters will be automatically rolled back to the default parameters, the "Optimization Parameter Area" flag will be set to 0, and the reason for the rollback will be recorded in the log.

[0068] Model update and calibration: When analyzing historical data at 2 AM every day, if the amount of new data exceeds 1000 records, the terminal will automatically download the updated .tflite model from the cloud (via 4G / 5G or WiFi) to overwrite the old model. A "model calibration" is performed monthly: the actual battery life data is obtained by fully charging and discharging the terminal and compared with the model prediction results. If the error is greater than 10%, the model's weight parameters are adjusted (through local fine-tuning) to improve prediction accuracy.

[0069] The embodiments of this application have been described above with reference to the accompanying drawings. Unless otherwise specified, the embodiments and features in the embodiments of this application can be combined with each other. This application is not limited to the specific embodiments described above. The specific embodiments described above are merely illustrative and not restrictive. Those skilled in the art can make many other forms under the guidance of this application without departing from the spirit and scope of the claims, and all of these forms are within the protection scope of this application.

Claims

1. An efficient power management method for a communication terminal, characterized by, The method comprises the following steps: Step 1: identifying the current scene of the user, initializing the monitoring parameters of multi-dimensional state data according to different scenes, and different scenes corresponding to different monitoring parameter configurations; Step 2: collecting multi-dimensional state data, and judging whether the data is abnormal based on double conditions: if the instantaneous battery voltage is lower than the current UVLO threshold value, and the voltage drop rate exceeds the specified value or the voltage change acceleration exceeds the preset value, both conditions are met, and the dynamic power supply adjustment mode is entered, otherwise, return to step 2 to continue collecting data; Step 3: after entering the dynamic power supply adjustment mode, start timing according to the preset timing period of the current scene, match the preset step-by-step power supply strategy according to the current scene, and output a scene-based silent prompt; Step 4: during the timing process, monitor the battery voltage recovery according to the scene difference: if it is a commuting scene, collect the voltage every 2 seconds; if it is a sleep scene, collect the voltage every 10 seconds; if it is a high-load scene, collect the voltage every 1 second; Step 5: if the voltage recovery is monitored before the end of the preset timing period, restore the communication terminal power consumption according to the gradient corresponding to the current scene; if the voltage recovery is not monitored at the end of the preset timing period, trigger the scene-associated emergency power supply mechanism; Step 6: regularly analyze the multi-dimensional state data collected in the history, predict the communication terminal endurance, and dynamically optimize the monitoring parameters in step 1 according to the analysis result.

2. The method of power management for a communication terminal according to claim 1, wherein, The specific way of identifying the current scene of the user in step 1 is: if the user's moving speed is greater than 3 m / s and the navigation application is in the background running state, it is determined as a commuting scene; if the communication terminal screen is closed for more than 10 minutes and there is no call or video task running, it is determined as a sleep scene; if the communication terminal CPU occupancy rate is greater than 70%, it is determined as a high-load scene.

3. The method of power management for a communication terminal according to claim 2, wherein, The specific content of the step-by-step power supply strategy in step 3 is: if the current scene is a commuting scene, turn off the mobile data, keep the 2G call function and GPS navigation function, and reduce the communication terminal screen brightness to 30-45%; if the current scene is a sleep scene, turn off all wireless functions and data collection functions, and only maintain the minimum standby power consumption of the communication terminal core chip; if the current scene is a high-load scene, close the background running application, limit the CPU multi-core running, and reduce the communication terminal screen brightness to 25-35%.

4. The method of power management for a communication terminal according to claim 3, wherein, The preset timing period according to the scene in step 3 is: the commuting scene timing period is set to 30 minutes, the sleep scene timing period is set to 2 hours, and the high-load scene timing period is set to 15 minutes.

5. The method of power management for a communication terminal according to claim 4, wherein, The specific way of the scene-associated emergency power supply mechanism in step 5 is: if the current scene is a sleep scene, trigger the deep sleep power supply: cut off all power supply circuits except the alarm function and power monitoring function, and temporarily restore the screen power supply circuit when the alarm is triggered; if the current scene is a high-load scene, trigger the task protection power supply: preferentially allocate power supply resources to the current running task, automatically compress the data of the current running task and save it to the local storage, and cut off the power supply circuit corresponding to the non-current task when the battery power is lower than the emergency threshold value; If the current scenario is a commuting scenario, trigger basic communication power supply: only keep the power supply circuit corresponding to the call function, SMS function and GPS positioning function, turn off the power supply circuit corresponding to the screen display, until the battery power is consumed.

6. The method of power management for a communication terminal according to claim 5, wherein, The period of periodically analyzing the historical collected data in step 6 is every morning; and the specific manner of dynamically optimizing the monitoring parameters is: For the scenarios with high abnormal frequency of battery voltage in the historical data, increase 5%-10%; For the communication terminal with battery capacity lower than 80%, when it is determined to be in the sleep scenario, reduce to 60%-80% of the original frequency.

7. The method of power management for a communication terminal according to claim 6, wherein, The multi-dimensional state data collected in step 2 includes battery voltage data, communication terminal motion acceleration data and user location data; and when the monitoring parameters are initialized in step 1, the UVLO threshold, voltage drop rate specified value and voltage change acceleration preset value are set for different scenarios respectively.

8. The method of power management for a communication terminal according to claim 7, wherein, When the voltage is monitored to recover in step 4, if the voltage values collected continuously for three times are all higher than the UVLO threshold corresponding to the current scenario, it is determined that the voltage recovers; and the emergency threshold is set according to the scenario in step 5: the emergency threshold of the high load scenario is set to 10% of the total capacity of the battery, the emergency threshold of the commuting scenario is set to 8% of the total capacity of the battery, and the emergency threshold of the sleep scenario is set to 5% of the total capacity of the battery.

9. The method of power management for a communication terminal according to claim 8, wherein, When the communication terminal endurance capability is predicted in step 6, a prediction model based on LSTM neural network is used, which combines historical battery voltage change data, power consumption data and battery capacity attenuation data under different scenarios, the training data includes scenario-power-voltage associated data in the last 30 days, and the output is the prediction result of the endurance time in the next 24 hours; after the monitoring parameters are dynamically optimized, the optimized parameters are used to initialize the monitoring configuration when step 1 is executed next time.

Images