A smart power-saving method, system, and smart terminal for intelligent terminals.

Abstract

This invention belongs to the field of smart terminal power management technology, and provides a smart terminal power-saving method, system, and smart terminal. The method acquires the smart terminal's historical communication records and future schedule events recorded in the smart terminal, generates communication data, and trains a prediction model. This model can predict the smart terminal's future communication status and calculate the safe power threshold required to complete the predicted communication task, thus determining the available surplus power when the smart terminal connects to external powered devices. The solution provided by this invention offers a proactive power planning strategy. By analyzing the user's unique historical communication behavior and future schedule, it can generate personalized dynamic safe power lines. Through accurate calculation of power consumption, it effectively avoids future risks and optimizes energy allocation at the system level, reducing overall energy waste, maximizing energy utilization efficiency, and optimizing the user experience.

Description

Technical Field

[0001] This invention belongs to the field of smart terminal power management technology, specifically relating to a smart terminal power protection method, system, and smart terminal. Background Technology

[0002] With the increasing power of smartphones, tablets and other smart terminals, various smart terminals have been deeply integrated into people's daily work and life, becoming indispensable communication and information processing tools. However, the improvement speed of the battery life of smart terminals has always been unable to fully meet the growing usage needs of users, and the problem of battery anxiety is common. Especially when it is necessary to reverse charge external devices such as Bluetooth headsets and smartwatches, users are often worried that the phone will affect its subsequent communication functions due to excessive power consumption.

[0003] To address this issue, common power-saving solutions in existing technologies often employ static threshold management strategies. The system presets a fixed battery percentage, such as 20% or 15%. When the phone's current battery level is detected to be below this threshold, it automatically limits or stops high-power applications and prohibits charging external devices, thereby ensuring that the phone can maintain standby for a period of time or make emergency calls.

[0004] However, this static, one-size-fits-all power-saving strategy has obvious drawbacks. For example, for business people who communicate frequently, 20% of the power may not be enough to support their important afternoon calls, failing to provide truly reliable power protection. On the other hand, for users with less communication needs, it may limit their ability to make better energy allocation decisions, resulting in wasted power. Summary of the Invention

[0005] To overcome the shortcomings of the prior art, this invention proposes an intelligent power-saving method for smart terminals, comprising:

[0006] The system acquires the historical communication records of the smart terminal and the future schedule events recorded in the smart terminal, and generates communication data of the smart terminal based on the historical communication records; the communication data includes historical schedule events, calling frequency, called frequency, calling duration, called duration, average communication duration, maximum communication duration, and active time periods;

[0007] The communication data is used as training data to train a time series prediction model, and the future schedule events are used as input data. The trained time series prediction model is used to quantitatively predict the communication demand in a specific future time period based on the input data, and outputs the possible communication duration and communication time distribution of the smart terminal in the specific future time period as the prediction result.

[0008] Obtain the battery status parameters of the smart terminal, and dynamically calculate the safe power threshold required to ensure the completion of the predicted communication task based on the battery status parameters and the prediction results;

[0009] When the smart terminal connects to an external power receiving device, it detects the current power level of the smart terminal. If the current power level is greater than the safe power threshold, the difference between the current power level and the safe power threshold is determined as the surplus power level. When the surplus power level is greater than the preset start threshold of the smart terminal, the smart terminal is allowed to charge the power receiving device with the surplus power level as the upper limit.

[0010] Specifically, the method for training the time series prediction model includes:

[0011] The communication data is cleaned to remove communication records whose communication duration is less than a preset time value;

[0012] The cleaned communication data is subjected to feature extraction and enhancement to generate a training dataset; each training data in the training dataset contains date type features based on the timestamp of the communication record, and the date type features include weekday type or holiday type;

[0013] The training dataset is input into a recurrent neural network that incorporates an attention mechanism for model training; wherein, the input features of the model include a historical communication duration sequence obtained by aggregating historical communication records in hours according to a preset time window, event calendar tags extracted from historical schedule events, location context information derived from historical location data, and the date type features; the output of the model includes a probability distribution of expected communication duration in hours for the next 24 to 72 hours.

[0014] The trained time series prediction model is deployed locally on the smart terminal, and the time series prediction model periodically acquires the communication data of the smart terminal. The model parameters are fine-tuned through incremental learning to achieve continuous optimization of prediction accuracy and adaptation to changes in communication habits.

[0015] Furthermore, the communication data also includes communication patterns with specific contacts, and each piece of training data also includes contact type features, which include general contacts or specific contacts. The input features of the model also include the specific contact features.

[0016] Specifically, the battery status parameters include current battery level, battery standby power consumption, and battery health status parameters. The smart terminal has a pre-configured power consumption configuration file that defines the typical operating current and power values ​​of the smart terminal's communication module under various network types in the form of a data table or function. Based on the battery status parameters and the prediction results, the safe power threshold required to ensure the completion of the predicted communication task is dynamically calculated, including:

[0017] Based on the communication period distribution output by the time series prediction model, a most likely network type is matched for each predicted communication period in the future specific time period, and the corresponding standby power consumption value and communication power consumption value per unit time are determined according to the power consumption configuration file.

[0018] A basic power value is obtained by using the overall duration of the future specific time period and the standby power consumption value per unit time, and a communication power value is obtained by using the communication duration predicted by the time series prediction model and the communication power consumption value. A basic power threshold is determined based on the basic power value and the communication power value.

[0019] The base power threshold is adjusted based on the battery health status parameters to obtain the safe power threshold required to ensure the completion of the predicted communication task.

[0020] Preferably, the method further includes:

[0021] During the charging process of the smart terminal for the powered device, the battery status parameters of the smart terminal are continuously monitored, and a real-time protection threshold is determined based on the battery status parameters and a dynamic buffer value; the real-time protection threshold is greater than the safe power threshold, and the dynamic buffer value is determined by the communication period distribution output by the time series prediction model and the current time value;

[0022] When the current battery level of the smart terminal is detected to be lower than the real-time protection threshold, the charging power of the smart terminal for the powered device is reduced accordingly, or the charging of the powered device by the smart terminal is stopped.

[0023] Preferably, the method further includes:

[0024] Acquire the location data, network data, and / or sensor data of the smart terminal, and determine whether the smart terminal is in a stable environment;

[0025] If the smart terminal is in a stable environment, the safe power threshold is lowered based on the first adjustment strategy;

[0026] If the smart terminal is in an unstable environment, the safe power threshold is increased based on the second adjustment strategy.

[0027] This invention also proposes an intelligent power-saving system for smart terminals, comprising:

[0028] The record acquisition module is used to acquire the historical communication records of the smart terminal and the future schedule events recorded in the smart terminal, and generate the communication data of the smart terminal based on the historical communication records; the communication data includes historical schedule events, calling communication frequency, called communication frequency, calling communication duration, called communication duration, average communication duration, maximum communication duration, and active time periods;

[0029] The communication prediction module is used to train a time series prediction model using the communication data as training data and the future schedule events as input data. The trained time series prediction model is used to quantitatively predict the communication demand in a specific future time period based on the input data, and outputs the possible communication duration and communication time distribution of the smart terminal in the specific future time period as the prediction result.

[0030] The threshold calculation module is used to obtain the battery status parameters of the smart terminal and dynamically calculate the safe power threshold required to ensure the completion of the predicted communication task based on the battery status parameters and the prediction results.

[0031] The reverse charging module is used to detect the current power value of the smart terminal when the smart terminal is connected to an external power receiving device. If the current power value is greater than the safe power threshold, the difference between the current power value and the safe power threshold is determined as the surplus power value. When the surplus power value is greater than the preset start threshold value of the smart terminal, the smart terminal is allowed to charge the power receiving device with the surplus power value as the upper limit power supply.

[0032] Specifically, the communication prediction module includes:

[0033] The data cleaning unit is used to clean the communication data and remove communication records whose communication duration is less than a preset time value.

[0034] The data generation unit is used to extract and enhance features from the cleaned communication data and generate a training dataset; each training data in the training dataset contains date type features based on the timestamp of the communication record, and the date type features include weekday type or holiday type;

[0035] The model training unit is used to input the training dataset into a recurrent neural network that incorporates an attention mechanism for model training; wherein, the input features of the model include a historical communication duration sequence obtained by aggregating historical communication records into hours according to a preset time window, event calendar tags extracted from historical schedule events, location context information derived from historical location data, and the date type features; the output of the model includes a probability distribution of expected communication duration in hours for the next 24 to 72 hours.

[0036] The model optimization unit is used to deploy the trained time series prediction model locally on the smart terminal, and to enable the time series prediction model to periodically acquire the communication data of the smart terminal, and to fine-tune the model parameters through incremental learning, so as to achieve continuous optimization of prediction accuracy and adaptation to changes in communication habits.

[0037] The present invention also proposes a smart terminal that applies the smart power-saving method for smart terminals as described above.

[0038] The present invention also proposes a computer-readable storage medium storing computer instructions that, when executed by a processor, implement the steps of the intelligent power-saving method for a smart terminal as described above.

[0039] The present invention has at least the following beneficial effects:

[0040] The proposed solution can break the limitations of fixed thresholds, freeing users from complex manual decision-making. It introduces the concept of dynamically changing safe power thresholds, providing each user with a personalized power protection strategy that fits their actual lifestyle. This avoids energy waste caused by excessive conservatism and prevents the risk of communication failure due to excessive power consumption, thus maximizing energy utilization efficiency while ensuring absolute communication security.

[0041] Furthermore, this solution can improve the accuracy and reliability of the time series prediction model through data cleaning and feature enhancement. Introducing date type, location context, and specific contact features provides rich information. Attention mechanism enables the model to focus on key historical information, improving the targeting of predictions. Incremental learning mechanism allows the model to continuously fine-tune using the latest data, ensuring long-term prediction accuracy. By introducing power consumption profile and correcting it according to battery health status, the calculation results of this solution can be more accurate, strengthening adaptive communication protection. The introduction of a real-time protection threshold for dynamic buffer values ​​can ensure that the power reserve of smart terminals is not accidentally depleted.

[0042] In addition, this solution can determine the environment of the smart terminal based on location data, network data, and sensor data, and adopt the first strategy of lowering the safe power threshold and the second strategy of raising the safe power threshold accordingly. This improves the user's convenience and energy efficiency, realizes decision optimization in complex scenarios, and enhances the robustness of communication security.

[0043] Therefore, this invention proposes an intelligent power protection method, system, and intelligent terminal for smart terminals. The solution provided by this invention offers a proactive power planning strategy. By analyzing the user's unique historical communication behavior and future schedule, it can generate a personalized dynamic safe power line. By accurately calculating power consumption, it effectively avoids future risks and optimizes energy allocation at the system level, reducing overall energy waste, maximizing energy utilization efficiency, and optimizing the user experience. Attached Figure Description

[0044] To more clearly illustrate the technical solutions in the embodiments of this application, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0045] Figure 1 A schematic diagram of the intelligent power-saving method for a smart terminal provided in Example 1;

[0046] Figure 2 This is a flowchart illustrating the method for training the time series prediction model.

[0047] Figure 3 This is a flowchart illustrating the method for calculating the safe power threshold.

[0048] Figure 4 This is a schematic diagram of the module structure of the intelligent power protection system for the smart terminal provided in Example 2. Detailed Implementation

[0049] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of the present invention, and not all of them. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative effort are within the scope of protection of the present invention.

[0050] Various embodiments of the invention will be described more fully below. The invention may have various embodiments, and adjustments and changes may be made therein. However, it should be understood that there is no intention to limit the various embodiments of the invention to the specific embodiments disclosed herein, but rather the invention should be understood to cover all modifications, equivalents, and / or alternatives falling within the spirit and scope of the various embodiments of the invention.

[0051] In the following, the terms “comprising” or “may include” as used in various embodiments of the invention indicate the presence of the disclosed functions, operations, or elements, and do not limit the addition of one or more functions, operations, or elements. Furthermore, as used in various embodiments of the invention, the terms “comprising,” “having,” and their cognates are intended only to indicate a specific feature, number, step, operation, element, component, or combination of the foregoing, and should not be construed as primarily excluding the presence of one or more other features, numbers, steps, operations, elements, components, or combinations of the foregoing, or the possibility of adding one or more combinations of the foregoing.

[0052] In various embodiments of the invention, the expression "or" or "at least one of A and / or B" includes any combination or all combinations of the words listed simultaneously. For example, the expression "A or B" or "at least one of A and / or B" may include A, may include B, or may include both A and B.

[0053] The expressions used in the various embodiments of the present invention (such as "first," "second," etc.) may modify various constituent elements in the various embodiments, but do not limit the corresponding constituent elements. For example, the above expressions do not limit the order and / or importance of the elements. The above expressions are only used for the purpose of distinguishing one element from other elements. For example, a first user device and a second user device refer to different user devices, although both are user devices. For example, a first element may be referred to as a second element without departing from the scope of the various embodiments of the present invention, and similarly, a second element may also be referred to as a first element.

[0054] It should be noted that, in this invention, unless otherwise explicitly specified and defined, terms such as "installation," "connection," and "fixation" should be interpreted broadly. For example, they can refer to a fixed connection, a detachable connection, or an integral connection; they can refer to a mechanical connection or an electrical connection; they can refer to a direct connection or an indirect connection through an intermediate medium; and they can refer to the internal connection of two components. Those skilled in the art can understand the specific meaning of the above terms in this invention according to the specific circumstances.

[0055] In this invention, those skilled in the art should understand that the terms indicating orientation or positional relationship in the text are based on the orientation or positional relationship shown in the accompanying drawings, and are only for the purpose of facilitating the description of this invention and simplifying the description, and are not intended to 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 of this invention.

[0056] The terminology used in the various embodiments of the invention is for the purpose of describing particular embodiments only and is not intended to limit the various embodiments of the invention. As used herein, the singular form is intended to include the plural form as well, unless the context clearly indicates otherwise. Unless otherwise defined, all terms used herein (including technical and scientific terms) have the same meaning as commonly understood by one of ordinary skill in the art to which the various embodiments of the invention pertain. The terms (such as those defined in a generally used dictionary) are to be interpreted as having the same meaning as in the context of the relevant technical field and are not to be interpreted as having an idealized or overly formal meaning, unless clearly defined in the various embodiments of the invention.

[0057] Example 1

[0058] This embodiment proposes an intelligent power-saving method for smart terminals. The method dynamically calculates a personalized safe power limit based on the user's behavior patterns and future plans, ensuring that charging external devices will not affect important communication needs during specific future periods, while maximizing the use of the smart terminal's surplus power. Please refer to [link to relevant documentation]. Figure 1 The methods specifically include:

[0059] S100: Obtain historical communication records and future schedule events recorded in the smart terminal, and generate communication data for the smart terminal based on the historical communication records.

[0060] In this embodiment, communication data may include, but is not limited to, historical schedule events, calling frequency, called frequency, calling duration, called duration, average communication duration, maximum communication duration, and active time periods. Among them, historical schedule events can reflect the user's past communication behavior related to the schedule; calling frequency and called frequency can reflect the frequency with which the user actively makes and receives calls; calling duration and called duration can reflect the time spent by the user actively making and receiving calls; average communication duration and maximum communication duration can reflect the user's calling habits and possible extreme situations; and active time periods can help identify the user's peak calling times.

[0061] S200: The time series prediction model is trained using communication data as training data, and future schedule events are used as input data. The trained time series prediction model is used to quantitatively predict the communication demand in a specific future time period based on the input data, and outputs the possible communication duration and communication period distribution of the smart terminal in the specific future time period as the prediction result.

[0062] It should be noted that the method proposed in this embodiment treats communication behavior as sequential data that changes over time. Therefore, in step S200, a time series prediction model is used to process this type of data. The time series prediction model learns the patterns and rules of user communication behavior over time through the communication data obtained in step S100. After inputting future schedule events into the already trained model, the model will comprehensively analyze historical patterns and future special events to output a quantitative prediction result. This prediction result includes the possible communication duration and communication time distribution within a specific future time period. For example, the prediction result may include "there is a high probability of needing a 60-minute call duration from 9-11 am tomorrow" or "10-11 am tomorrow is a high-demand period for video calls".

[0063] S300: Obtains the battery status parameters of the smart terminal, and dynamically calculates the safe power threshold required to ensure the completion of the predicted communication task based on the battery status parameters and prediction results.

[0064] In this embodiment, battery status parameters may include, but are not limited to, current battery charge, current battery temperature, battery standby power consumption, battery discharge capacity, and battery health status parameters.

[0065] Preferably, the method proposed in this embodiment can also acquire the location data, network data and / or sensor data of the smart terminal, and determine whether the smart terminal is in a stable environment. When it is determined that the smart terminal is in a stable environment, the safe power threshold is lowered based on the first adjustment strategy; and when it is determined that the smart terminal is in an unstable environment, the safe power threshold is raised based on the second adjustment strategy.

[0066] For example, the characteristic information of a stable environment may include location data of being at a familiar location for a long time, network data of being continuously connected to a known and stable WiFi network, and sensor data of the device being in a stationary charging state. The characteristic information of an unstable environment may include location data of continuously changing location, network data of frequently switching between different cellular networks or large fluctuations in signal strength, and sensor data of the device being in a continuous state of motion.

[0067] In this embodiment, the first adjustment strategy can be implemented by multiplying the adjustment coefficients to lower the adjustment, by subtracting the absolute value to lower the adjustment, and by dynamically lowering the adjustment based on the charging probability. The second adjustment strategy can be implemented by multiplying the adjustment coefficients to raise the adjustment, by adding the adjustment to the absolute value to raise the adjustment, and by dynamically raising the adjustment based on environmental risk.

[0068] S400: When the smart terminal connects to an external powered device, the current power value of the smart terminal is detected. If the current power value is greater than the safe power threshold, the difference between the current power value and the safe power threshold is determined as the surplus power value. When the surplus power value is greater than the preset start threshold of the smart terminal, the smart terminal is allowed to charge the powered device with the surplus power value as the upper limit of the power supply.

[0069] Specifically, please see Figure 2 Methods for training time series prediction models include:

[0070] S210: Clean the communication data and remove communication records whose communication duration is less than a preset time value.

[0071] It should be noted that short communication records such as misdialed calls and nuisance calls usually do not reflect the user's real communication needs. Step S210 can eliminate these communication techniques to prevent the model from making incorrect predictions due to learning these accidental and meaningless communication records. In an optional implementation, the preset time can be set to 10 seconds.

[0072] S220: Extract and enhance features from the cleaned communication data and generate a training dataset.

[0073] In step S210, each training data point in the training dataset contains date type features based on timestamps of communication records. These date type features include weekday type or holiday type.

[0074] In this embodiment, step S220 aggregates the historical communication records of the past N days at an hourly granularity to form a [N] that can intuitively reflect the time pattern. 24] The sequence of dimensions makes it easier for the model to extract information such as "hour of the day" and "day of the week" from the timestamp, thereby capturing the periodic pattern of communication;

[0075] Information extracted from historical events can be represented as event types such as meetings, medical events, and travel using one-hot encoding or embedding. For example, “There is a multinational video conference on Tuesday morning from 9:30 to 11:00” can be marked as a specific event feature.

[0076] Communication data can also include historical location data, from which location information such as home, company, commute, and public places can be derived, and can help the model distinguish between communication data on weekdays and holidays.

[0077] S230: Input the training dataset into a recurrent neural network that incorporates an attention mechanism for model training.

[0078] In one optional implementation, the input features of the model include a historical communication duration sequence obtained by aggregating historical communication records in hours according to a preset time window, event calendar tags extracted from historical schedule events, location context information derived from historical location data, and date type features. The output of the model includes a probability distribution of expected communication duration in hours for the next 24 to 72 hours.

[0079] It should be noted that recurrent neural networks can effectively capture short-term dependencies and long-term periodic patterns in communication behavior. Furthermore, the addition of attention mechanisms to recurrent neural networks can enable the model to automatically evaluate and weight the importance of different time points in the historical sequence for predicting future moments.

[0080] For example, when the model predicts "communication needs at 10 a.m. next Monday", it may find that data from the past four consecutive Mondays at 10 a.m. is very important, information from a meeting memo with a client last Friday afternoon mentioning "follow-up on Monday" is relatively important, and information from a shopping call two weeks ago is not important. The attention mechanism can dynamically assign appropriate weights to this historical information, making the model prediction more interpretable and accurate.

[0081] The final output of the model is a probability distribution of expected communication duration in hours over the next 24 to 72 hours, rather than just a single value. For example, there is a 70% probability that the call will take 0-5 minutes between 10 and 11 a.m. tomorrow, a 20% probability that it will take 5-15 minutes, and a 10% probability that it will take more than 15 minutes. Therefore, the output of the model can provide richer information for risk assessment. After completing step S230, a time series prediction model that has been trained can be obtained.

[0082] S240: Deploy the trained time series prediction model locally on the smart terminal, and enable the time series prediction model to periodically acquire the communication data of the smart terminal. Fine-tune the model parameters through incremental learning to achieve continuous optimization of prediction accuracy and adaptation to changes in communication habits.

[0083] It should be noted that by implementing continuous training and inference of the model through step S240, the model can adapt to changes in users' communication habits. Periodic incremental learning can be set to be performed when the system is idle, avoiding the consumption of a large amount of computing resources during periods of user activity and ensuring the smoothness of the system. At the same time, since step S240 is completed locally, it can be ensured that users' sensitive communication data and schedule data will not be uploaded to the cloud, thus protecting user privacy.

[0084] In an optional implementation, the time series prediction model used by the method proposed in this embodiment includes a sequence-to-sequence (Seq2Seq) model based on an encoder-decoder architecture. The encoder can be composed of two stacked layers of long short-term memory (LSTM) networks. The input of the encoder includes preprocessed and feature-enhanced historical communication sequence data. The decoder can also be composed of two stacked layers of long short-term memory (LSTM) networks. At each decoding time step of the decoder, the attention mechanism dynamically assigns different weights to the hidden states of all time steps of the encoder according to the current state of the decoder. Thus, when predicting communication needs at a certain future moment, the model can focus on the most relevant historical time period, rather than treating all historical information equally.

[0085] Furthermore, the communication data also includes communication patterns with specific contacts. Each training data set also includes contact type features, which include general contacts or specific contacts. The model's input features also include specific contact features, such as classifying contacts into types like family, friends, colleagues, and important clients. Among these, communication with important clients has higher priority and different time patterns. The model can learn things like "the probability and duration of calls with important clients will increase during working hours on weekdays."

[0086] Specifically, the smart terminal has a pre-installed power consumption configuration file that defines the typical operating current and power values ​​of the smart terminal's communication module under various network types in the form of a data table or function; please refer to Figure 3 The step S300, which involves dynamically calculating the safe power threshold required to ensure the completion of the predicted communication task based on battery state parameters and prediction results, includes:

[0087] S310: Based on the communication period distribution output by the time series prediction model, it matches the most likely network type for each predicted communication period within a specific future time period, and determines the corresponding standby power consumption and communication power consumption per unit time according to the power consumption profile.

[0088] For example, step S310 analyzes the network type that the user usually uses based on the communication situation in each predicted communication period. For example, when the time title includes video conferencing, the network type can be determined as WiFi or 5G network; when the location is the company, the network type can be determined as WiFi; and when the location is commuting, the network type can be determined as 3G network or 4G network.

[0089] The power consumption unit corresponding to the standby power consumption value and communication power consumption value determined by the power consumption configuration file in step S310 can be mA / h or mW / h. In this embodiment, the power consumption configuration file uses mA / h as the power consumption unit. The power consumption configuration file can include the power consumption configuration table shown in Table 1.

[0090] Table 1 Power Consumption Configuration Table

[0091]

[0092] S320: Obtain the basic power value by the overall duration and standby power consumption per unit time of a specific future time period, and obtain the communication power value by the communication duration and communication power consumption value predicted by the time series prediction model, and determine the basic power threshold based on the basic power value and the communication power value.

[0093] S330: Adjusts the base power threshold based on battery health status parameters to obtain the safe power threshold required to ensure the completion of the predicted communication task.

[0094] Specifically, step S330 can introduce a safety factor that is dynamically determined based on the battery health status parameter. When the battery health status parameter indicates that the battery capacity is decaying, the safety factor is set to a value greater than 1 to compensate for the risk of increased internal resistance and decreased usable capacity caused by battery aging. By multiplying the basic power threshold by the safety factor, the basic power threshold can be corrected, thereby obtaining the safe power threshold required to ensure the completion of the predicted communication task.

[0095] Calculate the safe power threshold The formula can include:

[0096]

[0097] in, This represents the safety factor determined based on the battery's state of health (SOH). This represents the standby power consumption per unit of time. The power consumption of communication per unit time. Represents the overall duration of a specific time period in the future. Represents the predicted communication duration within a specific future time period; in an optional implementation, when SOH > 90%. When 80% < SOH ≤ 90% When SOH≤80 .

[0098] For example, when the basic power threshold is 20% of the power, if the battery health status parameters indicate that the battery has a certain degree of aging, the system can determine the safety factor to be 1.1 accordingly. At this time, the value obtained by multiplying the basic power threshold of 20% by the safety factor of 1.1 can be determined as the safe power threshold, that is, the power value of 22% can be determined as the safe power threshold.

[0099] Preferably, the method proposed in this embodiment further includes:

[0100] S500: During the charging process of the smart terminal for the powered device, the battery status parameters of the smart terminal are continuously monitored, and the real-time protection threshold is determined based on the battery status parameters and dynamic buffer value.

[0101] In this embodiment, the real-time protection threshold is greater than the safe power threshold, and the dynamic buffer value is determined by the communication period distribution output by the time series prediction model and the current time value;

[0102] In an optional implementation, the dynamic buffer value can be obtained from an uncertain buffer component, an urgent buffer component, a first weighting coefficient, and a second weighting coefficient, wherein the first weighting coefficient is used to adjust the uncertain buffer component, and the second weighting coefficient is used to adjust the urgent buffer component. The uncertain buffer component represents the expected communication duration probability distribution at the hourly granularity based on the output of a time series prediction model, while the urgent buffer component represents the event attributes extracted from future schedule events.

[0103] For example, when a period of high communication demand is predicted within the next 6 hours, the dynamic buffer value is determined to be 8%, at which point the safe power threshold is set. The real-time protection threshold can be adjusted upwards by 8% accordingly; however, when it is predicted that the next 24 hours will be periods of low communication demand, the dynamic buffer value is set at 3%, at which point the safe power threshold is determined. The real-time protection threshold can be adjusted upwards by 3%.

[0104] S600: When the current battery level of the smart terminal is detected to be lower than the real-time protection threshold, the power of the smart terminal charging the powered device is reduced accordingly, or the charging of the powered device by the smart terminal is stopped.

[0105] In one specific embodiment, communication data from User A's smart terminal indicates that User A is active in calls between 9:00 AM and 12:00 PM on weekdays. Simultaneously, the smart terminal's calendar records a multinational video conference scheduled for 9:30 AM to 11:00 AM next Tuesday. Based on this video conference event and historical data, the time series prediction model predicts that 9:00 AM to 11:00 AM next Tuesday will be a high-demand period, with an estimated communication duration of 90 minutes and a high probability of using a 5G network. According to the power consumption profile, the power consumption of a 5G video call is approximately 450 mA / h. Considering the slightly aging battery of the smart terminal, the safe battery threshold for Tuesday morning is calculated to be 35%. When User A connects to headphones for charging before 9:00 AM on Tuesday, if the smart terminal's battery is at 50%, the surplus battery value is calculated to be 15%. This 15% battery value can be used as the upper limit for reverse charging of the headphones.

[0106] In another embodiment, communication data from user B's smart terminal indicates that user B communicates very little during holidays, and the smart terminal's calendar does not contain any scheduled events. Based on the smart terminal's network data, it can be determined that user B is at home, i.e., in a stable environment. The time series prediction model can combine historical data to predict that Saturday will be a period of low communication demand, calculate a safe power threshold of 20%, and further lower the safe power threshold to 12% based on the first adjustment strategy. When user B connects to charge the smartwatch via the smart terminal before Saturday morning, if the smart terminal's power is 40%, the surplus power value can be calculated as 28%. This 28% power value can be used as the upper limit for reverse charging of the smartwatch.

[0107] To verify the feasibility of the method proposed in this embodiment, communication data from 20 users over a period of 3 months was collected. A 20% fixed threshold power supply scheme and the method proposed in this embodiment were used as two different power supply methods. A comparative experiment was conducted using communication failure rate and surplus power utilization rate as evaluation indicators. The results are shown in Table 2. The communication failure rate is used to represent the percentage of communication interruptions due to power failure during the predicted communication demand period, and the surplus power utilization rate is used to represent the percentage of electricity successfully used to charge external devices out of the total surplus power.

[0108] Table 2 Comparison and Evaluation Table of Traditional Solution and Intelligent Power Protection Algorithm

[0109]

[0110] Compared to the traditional approach that guarantees a fixed 20% battery capacity, the method proposed in this embodiment significantly reduces the communication guarantee failure rate from 4.5% to 0.3% through accurate prediction of special events and a compensation mechanism for battery aging. At the same time, the surplus battery utilization rate reaches 79%, which is more than twice that of the control group. This demonstrates that the algorithm is extremely efficient in identifying safe battery capacity, greatly reducing idle and wasted power, and the experimental group has significantly higher user satisfaction.

[0111] Example 2

[0112] Please see Figure 4 This embodiment proposes an intelligent power-saving system for smart terminals. The system proposed in this embodiment can implement the intelligent power-saving method for smart terminals proposed in Embodiment 1. The system specifically includes:

[0113] The record acquisition module 10 is used to acquire the historical communication records of the smart terminal and the future schedule events recorded in the smart terminal, and generate communication data of the smart terminal based on the historical communication records; the communication data includes historical schedule events, calling communication frequency, called communication frequency, calling communication duration, called communication duration, average communication duration, maximum communication duration, and active time period;

[0114] The communication prediction module 20 is used to train a time series prediction model using communication data as training data and future schedule events as input data. By completing the training of the time series prediction model, it quantitatively predicts the communication demand in a specific future time period based on the input data and outputs the possible communication duration and communication period distribution of the smart terminal in a specific future time period as the prediction result.

[0115] The threshold calculation module 30 is used to obtain the battery status parameters of the smart terminal and dynamically calculate the safe power threshold required to ensure the completion of the predicted communication task based on the battery status parameters and the prediction results.

[0116] The reverse charging module 40 is used to detect the current power value of the smart terminal when the smart terminal is connected to an external power receiving device. If the current power value is greater than the safe power threshold, the difference between the current power value and the safe power threshold is determined as the surplus power value. When the surplus power value is greater than the preset start threshold value of the smart terminal, the smart terminal is allowed to charge the power receiving device with the surplus power value as the upper limit.

[0117] The status detection module 50 is used to continuously detect the battery status parameters of the smart terminal during the charging process of the smart terminal for the powered device, and determine the real-time protection threshold based on the battery status parameters and dynamic buffer value.

[0118] The charging adjustment module 60 is used to reduce the charging power of the smart terminal to the powered device or stop the smart terminal from charging the powered device when the current power value of the smart terminal is detected to be lower than the real-time protection threshold.

[0119] In this embodiment, communication data may include, but is not limited to, historical schedule events, calling communication frequency, called communication frequency, calling communication duration, called communication duration, average communication duration, maximum communication duration, and active time periods; battery status parameters may include, but are not limited to, current battery level, current battery temperature, battery standby power consumption, battery discharge capacity, and battery health status parameters.

[0120] Preferably, the threshold calculation module 30 can also acquire the location data, network data and / or sensor data of the smart terminal, and determine whether the smart terminal is in a stable environment. When the smart terminal is determined to be in a stable environment, the safe power threshold is lowered based on the first adjustment strategy; and when the smart terminal is determined to be in an unstable environment, the safe power threshold is raised based on the second adjustment strategy.

[0121] Specifically, the communication prediction module 20 includes:

[0122] Data cleaning unit 21 is used to clean communication data and remove communication records whose communication duration is less than a preset time value.

[0123] The data generation unit 22 is used to extract and enhance features from the cleaned communication data and generate a training dataset. Each training data in the training dataset contains date type features based on the timestamp of the communication record. The date type features include weekday type or holiday type.

[0124] The model training unit 23 is used to input the training dataset into a recurrent neural network that incorporates an attention mechanism for model training. The input features of the model include a sequence of historical communication durations obtained by aggregating historical communication records into hours according to a preset time window, event calendar tags extracted from historical schedule events, location context information derived from historical location data, and date type features. The output of the model includes the probability distribution of expected communication durations in hours for the next 24 to 72 hours.

[0125] The model optimization unit 24 is used to deploy the trained time series prediction model locally on the smart terminal, and to enable the time series prediction model to periodically acquire the communication data of the smart terminal, and to fine-tune the model parameters through incremental learning, so as to achieve continuous optimization of prediction accuracy and adaptation to changes in communication habits.

[0126] Each training data point in the training dataset contains date type features based on timestamps of communication records. These date type features include weekday type or holiday type. Furthermore, the communication data also includes communication patterns with specific contacts. Each training data point also includes contact type features, which include general contacts or specific contacts. The model's input features also include specific contact features.

[0127] The threshold calculation module 30 also includes:

[0128] The power matching unit 31 is used to match the most likely network type for each predicted communication period in a specific future time period based on the communication period distribution output by the time series prediction model, and to determine the corresponding standby power consumption value and communication power consumption value per unit time according to the power consumption configuration file.

[0129] The basic computing unit 32 is used to obtain a basic power value by the overall duration and standby power consumption per unit time of a specific future time period, and to obtain a communication power value by the communication duration and communication power consumption value predicted by the time series prediction model, and to determine a basic power threshold based on the basic power value and the communication power value.

[0130] The threshold correction unit 33 is used to correct the basic power threshold based on the battery health status parameters to obtain the safe power threshold required to ensure the completion of the predicted communication task.

[0131] Example 3

[0132] This embodiment also proposes a smart terminal, which can be used to implement the smart power protection method of the smart terminal proposed in Embodiment 1 above.

[0133] Example 4

[0134] This embodiment also proposes a computer-readable storage medium storing computer instructions, which, when executed by a processor, implement the steps of the intelligent power-saving method for the smart terminal proposed in Embodiment 1 above.

[0135] It should be noted that computer-readable media, including both permanent and non-permanent, removable and non-removable media, can store information using any method or technology. Information can be computer-readable instructions, data structures, program modules, or other data. Examples of computer storage media include, but are not limited to, phase-change memory (PRAM), static random access memory (SRAM), dynamic random access memory (DRAM), other types of random access memory (RAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), flash memory or other memory technologies, CD-ROM, digital versatile optical disc (DVD) or other optical storage, magnetic tape, disk storage, quantum memory, graphene-based storage media or other magnetic storage devices, or any other non-transferable medium that can be used to store information accessible by a computing device. As defined herein, computer-readable media does not include transient computer-readable media, such as modulated data signals and carrier waves.

[0136] In summary, this invention proposes an intelligent power-saving method, system, and intelligent terminal for smart terminals. The solution provided by this invention offers a proactive power planning strategy. By analyzing a user's unique historical communication behavior and future schedule, it can generate a personalized dynamic power safety line. By accurately calculating power consumption, it effectively avoids future risks and optimizes energy allocation at the system level, reducing overall energy waste, maximizing energy utilization efficiency, and optimizing the user experience.

[0137] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A method for intelligent power protection in a smart terminal, characterized in that, include: The system acquires the historical communication records of the smart terminal and the future schedule events recorded in the smart terminal, and generates communication data of the smart terminal based on the historical communication records; the communication data includes historical schedule events, calling frequency, called frequency, calling duration, called duration, average communication duration, maximum communication duration, and active time periods; The communication data is used as training data to train a time series prediction model, and the future schedule events are used as input data. The trained time series prediction model is used to quantitatively predict the communication demand in a specific future time period based on the input data, and outputs the possible communication duration and communication time distribution of the smart terminal in the specific future time period as the prediction result. Obtain the battery status parameters of the smart terminal, and dynamically calculate the safe power threshold required to ensure the completion of the predicted communication task based on the battery status parameters and the prediction results; When the smart terminal connects to an external power receiving device, it detects the current power value of the smart terminal. If the current power value is greater than the safe power threshold, the difference between the current power value and the safe power threshold is determined as the surplus power value. When the surplus power value is greater than the preset start threshold of the smart terminal, the smart terminal is allowed to charge the power receiving device with the surplus power value as the upper limit of the power supply. The methods for training the time series prediction model include: The communication data is cleaned to remove communication records whose communication duration is less than a preset time value; The cleaned communication data is subjected to feature extraction and enhancement to generate a training dataset; each training data in the training dataset contains date type features based on the timestamp of the communication record, and the date type features include weekday type or holiday type; The training dataset is input into a recurrent neural network that incorporates an attention mechanism for model training; wherein, the input features of the model include a historical communication duration sequence obtained by aggregating historical communication records in hours according to a preset time window, event calendar tags extracted from historical schedule events, location context information derived from historical location data, and the date type features; the output of the model includes a probability distribution of expected communication duration in hours for the next 24 to 72 hours. The trained time series prediction model is deployed locally on the smart terminal, and the time series prediction model periodically acquires the communication data of the smart terminal. The model parameters are fine-tuned through incremental learning to achieve continuous optimization of prediction accuracy and adaptation to changes in communication habits. The battery status parameters include current battery level, battery standby power consumption, and battery health status parameters. The smart terminal has a pre-configured power consumption configuration file that defines the typical operating current and power values ​​of the smart terminal's communication module under various network types in the form of a data table or function. Based on the battery status parameters and the prediction results, the safe power threshold required to ensure the completion of the predicted communication task is dynamically calculated, including: Based on the communication period distribution output by the time series prediction model, a most likely network type is matched for each predicted communication period in the future specific time period, and the corresponding standby power consumption value and communication power consumption value per unit time are determined according to the power consumption configuration file. A basic power value is obtained by using the overall duration of the future specific time period and the standby power consumption value per unit time, and a communication power value is obtained by using the communication duration predicted by the time series prediction model and the communication power consumption value. A basic power threshold is determined based on the basic power value and the communication power value. The base power threshold is adjusted based on the battery health status parameters to obtain the safe power threshold required to ensure the completion of the predicted communication task.

2. The intelligent power-saving method for a smart terminal according to claim 1, characterized in that, The communication data also includes communication patterns with specific contacts, and each piece of training data also includes contact type features, which include general contacts or specific contacts. The input features of the model also include the specific contact features.

3. The intelligent power-saving method for a smart terminal according to claim 1, characterized in that, Also includes: During the charging process of the smart terminal for the powered device, the battery status parameters of the smart terminal are continuously monitored, and a real-time protection threshold is determined based on the battery status parameters and the dynamic buffer value. The real-time protection threshold is greater than the safe power threshold, and the dynamic buffer value is determined by the communication period distribution output by the time series prediction model and the current time value; When the current battery level of the smart terminal is detected to be lower than the real-time protection threshold, the charging power of the smart terminal for the powered device is reduced accordingly, or the charging of the powered device by the smart terminal is stopped.

4. The intelligent power-saving method for a smart terminal according to claim 1, characterized in that, Also includes: Acquire the location data, network data, and / or sensor data of the smart terminal, and determine whether the smart terminal is in a stable environment; If the smart terminal is in a stable environment, the safe power threshold is lowered based on the first adjustment strategy; If the smart terminal is in an unstable environment, the safe power threshold is increased based on the second adjustment strategy.

5. A smart power-saving system for a smart terminal, characterized in that, The intelligent power-saving method for a smart terminal as described in any one of claims 1-4 includes: The record acquisition module is used to acquire the historical communication records of the smart terminal and the future schedule events recorded in the smart terminal, and generate the communication data of the smart terminal based on the historical communication records; the communication data includes historical schedule events, calling communication frequency, called communication frequency, calling communication duration, called communication duration, average communication duration, maximum communication duration, and active time periods; The communication prediction module is used to train a time series prediction model using the communication data as training data and the future schedule events as input data. The trained time series prediction model is used to quantitatively predict the communication demand in a specific future time period based on the input data, and outputs the possible communication duration and communication time distribution of the smart terminal in the specific future time period as the prediction result. The threshold calculation module is used to obtain the battery status parameters of the smart terminal and dynamically calculate the safe power threshold required to ensure the completion of the predicted communication task based on the battery status parameters and the prediction results. The reverse charging module is used to detect the current power level of the smart terminal when the smart terminal is connected to an external power receiving device. If the current power level is greater than the safe power threshold, the difference between the current power level and the safe power threshold is determined as the surplus power level. When the surplus power level is greater than the preset start threshold of the smart terminal, the smart terminal is allowed to charge the power receiving device with the surplus power level as the upper limit.

6. The intelligent power protection system for a smart terminal according to claim 5, characterized in that, The communication prediction module includes: The data cleaning unit is used to clean the communication data and remove communication records whose communication duration is less than a preset time value. The data generation unit is used to extract and enhance features from the cleaned communication data and generate a training dataset; each training data in the training dataset contains date type features based on the timestamp of the communication record, and the date type features include weekday type or holiday type; The model training unit is used to input the training dataset into a recurrent neural network that incorporates an attention mechanism for model training; wherein, the input features of the model include a historical communication duration sequence obtained by aggregating historical communication records into hours according to a preset time window, event calendar tags extracted from historical schedule events, location context information derived from historical location data, and the date type features; the output of the model includes a probability distribution of expected communication duration in hours for the next 24 to 72 hours. The model optimization unit is used to deploy the trained time series prediction model locally on the smart terminal, and to enable the time series prediction model to periodically acquire the communication data of the smart terminal, and to fine-tune the model parameters through incremental learning, so as to achieve continuous optimization of prediction accuracy and adaptation to changes in communication habits.

7. A smart terminal, characterized in that, The intelligent power-saving method for smart terminals as described in any one of claims 1-4 is applied.

8. A computer-readable storage medium, characterized in that, It stores computer instructions that, when executed by a processor, implement the steps of the intelligent power-saving method for the smart terminal as described in any one of claims 1-4.

Images