Method and system for analyzing power consumption data of a notebook computer

Abstract

The application relates to the field of data analysis, and discloses a notebook computer power consumption data analysis method and system, which are used for solving the problem that different power source activities are ignored when power management is performed, collecting user usage habit data, obtaining the predicted use time of the user through user manual input and a machine learning algorithm, monitoring the power consumption state of each hardware component of the notebook computer in real time, calculating the predicted endurance time of the current device according to the monitored hardware power consumption state and the usage habit data of the user, collecting relevant data of the computer, calculating an endurance adjustment factor according to the relevant data, calculating the actual endurance time according to the adjustment factor, and dynamically adjusting the power source according to the actual endurance time and the predicted use time, so that the accuracy of battery condition detection is effectively improved, and the rationality of power management is improved.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of data analysis, more particularly to a method and system for analyzing power consumption data of a notebook computer. BACKGROUND

[0002] As notebook computers continue to become thinner and lighter, the space available for installing power supplies is limited, which greatly limits the endurance of notebook computers. Currently, power management systems are becoming increasingly important, by monitoring the power consumption state of a notebook computer, and optimizing power usage strategies based on power consumption data, it is expected to provide longer endurance time under the same power capacity. For example, by reducing CPU frequency or turning off certain unnecessary hardware components to reduce power consumption.

[0003] Currently, existing power management systems mainly focus on the overall power consumption of a notebook computer, ignoring the influence of different target power activities on overall power consumption, and during the use of a notebook computer, external factors may affect the performance of the computer. If the past evaluation criteria are still used, the detection result will be inaccurate and the power management will be unreasonable.

[0004] To solve the above problems, the present application provides a solution. SUMMARY

[0005] In order to overcome the above-mentioned defects of the prior art, the present application provides a method and system for analyzing power consumption data of a notebook computer to solve the problems in the background art.

[0006] To achieve the above purpose, the present application provides the following technical solutions:

[0007] The method for analyzing power consumption data of a notebook computer comprises the following steps:

[0008] Step 1: Collecting user usage habit data, the usage habit data including the usage frequency of all application programs, the single usage time, the number of programs opened simultaneously during usage, and the usage period;

[0009] Step 2: Obtaining the user's estimated usage time, which is manually input by the user, if the user does not input, the usage habit model is generated by using machine learning algorithm to analyze the user's usage habit according to the user usage habit data, and the usage time is predicted by using the usage habit model;

[0010] Step 3: Real-time monitoring of the power consumption state of each hardware component of the notebook computer, including but not limited to CPU, GPU, memory, display screen and other peripheral devices;

[0011] Step 4: According to the monitored hardware power consumption state and the user's usage habit data, the expected battery life of the current device is calculated;

[0012] Collecting relevant data of the computer, including battery-related data, temperature data, and user usage data, including the total duration of the user using the device and the number of simultaneously running programs since the start, and calculating the battery life adjustment factor according to the relevant data;

[0013] According to the adjustment factor, the actual battery life is calculated, and the formula is BF 实际 = RA × T ct , where BF 实际 represents the actual battery life, RA represents the battery life adjustment factor, and T ct represents the expected battery life;

[0014] Step 5: Dynamic power adjustment according to the actual battery life and the expected usage time.

[0015] Preferably, the step of collecting user usage habit data is:

[0016] Design a background process to continuously monitor and record user behavior, and mark the specific timestamp and related information;

[0017] Each time the user starts an application, record the application name, start time, and application start count; and each time the user closes the application, record the closing time, calculate the total duration of this use according to the application start time and closing time, and record it down;

[0018] Record the user's application start situation at different time periods in a day, including the specific time point and the corresponding usage duration, and classify the data by time period;

[0019] Store all collected data in chronological order.

[0020] Preferably, the step of using machine learning algorithm to analyze user's usage habit, generating usage habit model, and predicting usage time through usage habit model is:

[0021] Use the mean imputation method to fill in the missing values in the data, standardize the data to eliminate the scale influence of different features, and for each time point, construct a feature vector;

[0022] Use the support vector regression algorithm to build a prediction model, and divide the data set into training set and test set;

[0023] Use the training set data to train the selected model, use cross-validation technology to evaluate the performance of the model, and get the final usage habit model;

[0024] The predicted use time is predicted using the end use habit model.

[0025] Preferably, the step of evaluating the performance of the model using cross-validation technique is:

[0026] The data set is divided into k non-overlapping subsets, called folds;

[0027] Each time, one fold is selected from the k folds as the validation set, and the remaining k-1 folds are used as the training set. The model is trained on the k-1 folds, and the model performance is evaluated on the validation set. The error between the predicted result and the actual result is recorded;

[0028] For each fold, the mean square error, mean absolute error, and root mean square error are calculated;

[0029] The standard deviation of all fold performance indicators is calculated, and the cross-validation results of different models are compared to select the model with the smallest standard deviation.

[0030] Preferably, the step of calculating the endurance adjustment factor is:

[0031] Battery-related data of the notebook computer is obtained, and the battery aging degree is calculated based on the battery-related data;

[0032] Temperature data during use of the notebook computer is obtained, and the heat dissipation efficiency is calculated based on the temperature data;

[0033] The total duration of use of the computer and the number of maximum simultaneous running programs after the computer is turned on are obtained, and the use degree is calculated based on the total duration of use of the computer and the number of maximum simultaneous running programs. The calculation formula is US = DU x NUM P , wherein US represents the use degree, DU represents the total duration of use of the computer, and NUM P represents the number of maximum simultaneous running programs;

[0034] The endurance adjustment factor is comprehensively evaluated based on the battery aging degree, the heat dissipation efficiency, and the use degree. The calculation formula is , wherein RA represents the endurance adjustment factor, AG represents the battery aging degree, DH represents the heat dissipation efficiency, US represents the use degree, a1, a2, and a3 represent the weight coefficients of the battery aging degree, the heat dissipation efficiency, and the use degree.

[0035] Preferably, the step of calculating the battery aging degree based on the battery-related data is:

[0036] The number of charge and discharge times is extracted from the battery report of the battery management system of the notebook computer. The number of charge and discharge times is the total number of complete charge and discharge cycles experienced by the battery;

[0037] acquiring the current internal resistance of the battery from the data provided by the battery management system;

[0038] collecting the new internal resistance of the battery, calculating the internal resistance change rate according to the current internal resistance and the new internal resistance, and the calculation formula is wherein ir represents the internal resistance change rate, R new represents the new internal resistance, R battery represents the current internal resistance;

[0039] normalizing the internal resistance change rate and the number of charge and discharge, calculating the battery aging degree according to the normalized internal resistance change rate and the number of charge and discharge, and the calculation formula is wherein AG represents the battery aging degree, ir represents the internal resistance change rate, and NUM c represents the number of charge and discharge.

[0040] Preferably, the step of calculating the heat dissipation efficiency according to the temperature data is:

[0041] recording the temperature before using the computer as the initial temperature, and reading the real-time temperature data of the current CPU and GPU of the notebook computer through the temperature sensor;

[0042] measuring the real-time total power consumption of the notebook computer;

[0043] calculating the temperature rise value by difference between the real-time temperature data and the initial temperature, and calculating the heat dissipation efficiency according to the total power consumption and the temperature rise value, and the calculation formula is wherein DH represents the heat dissipation efficiency, P total represents the total power consumption, and ΔW represents the temperature rise value.

[0044] Preferably, the step of dynamically adjusting the power supply according to the actual endurance time and the expected use time is:

[0045] comparing the actual endurance time with the expected use time, if the actual endurance time is greater than the expected use time, no adjustment is made; if the actual endurance time is less than the expected use time, analyzing the calling probability of each program in the computer, and preferentially terminating the program with the smallest calling probability, and continuing to evaluate until the actual endurance time is greater than the expected use time;

[0046] The step of analyzing the calling probability of each program in the computer is that the number of starts of each program in the computer use in the detection period is obtained through the system log, the number of starts of each program is accumulated and added to obtain the total program start number, and the number of starts of each program is compared with the total program start number to obtain the calling probability.

[0047] Preferably, the power consumption data analysis system of the notebook computer comprises:

[0048] The user habit collection module is used for collecting habit data of the user in daily use of the notebook computer.

[0049] The predicted use time acquisition module is used for providing an interface for the user to manually input the predicted use time, and the user inputs the predicted use time by himself / herself; if the user does not input the predicted use time, a use habit model is generated by using a machine learning algorithm to analyze the use habit of the user according to the habit data of the user, and the use time is predicted by using the use habit model.

[0050] The power consumption monitoring module is used for monitoring the power consumption state of the notebook computer components in real time, the computer components including a CPU, a GPU, a memory and a display screen, and generating current power consumption data.

[0051] The endurance time calculation module is used for calculating the actual endurance time according to the current power consumption data.

[0052] The intelligent process management module is used for dynamically adjusting the power supply according to the actual endurance time and the predicted use time.

[0053] Technical effects and advantages of the present application:

[0054] The use habit data of the user is collected, the predicted use time of the user is obtained by manual input and machine learning algorithm, the power consumption state of each hardware component of the notebook computer is monitored in real time, the predicted endurance time of the current device is calculated according to the monitored hardware power consumption state and the use habit data of the user, relevant data of the computer is collected, the endurance adjustment factor is calculated according to the relevant data, the actual endurance time is calculated according to the adjustment factor, and the power supply is dynamically adjusted according to the actual endurance time and the predicted use time, thereby effectively improving the accuracy of the battery condition detection and the rationality of the power management. BRIEF DESCRIPTION OF DRAWINGS

[0055] Figure 1 The present application is a whole flowchart. DETAILED DESCRIPTION

[0056] The technical solutions in the present application will be described clearly and completely in combination with the drawings in the present application, and the forms of each structure described in the following embodiments are only examples, and the power consumption data analysis method and system of the notebook computer involved in the present application are not limited to each structure described in the following embodiments, and all other embodiments obtained by those skilled in the art without creative labor belong to the protection scope of the present application.

[0057] The present application provides a power consumption data analysis method of a notebook computer, comprising the following steps:

[0058] Step 1: Collect user's usage habit data, including the frequency of use of all applications, single use duration, the number of programs opened simultaneously during use, and the use period;

[0059] In this embodiment, it needs to be specifically pointed out that the step of collecting user's usage habit data is:

[0060] Design a background process to continuously monitor and record user's operation behavior. Ensure that the process can run seamlessly and does not affect the normal operation of the user, record the start, close, foreground and background switching of each application, and mark the specific timestamp and related information;

[0061] Each time the user starts an application, record the application name, start time and application start count, which is used to calculate the usage frequency;

[0062] And each time the user closes the application, record the closing time, calculate the total duration of this use according to the start time and closing time of the application, and record it;

[0063] When the user uses the application, record all the programs running at each time point, regularly count the number of programs opened simultaneously in a certain period of time, and associate it with the currently used application;

[0064] Record the user's application start situation at different time periods in a day, including the specific time point and the corresponding use duration, classify the data by time period (such as morning, afternoon, evening, etc.), and analyze the usage habits at different time periods;

[0065] Store all collected data in chronological order, ensure data integrity and consistency, regularly organize the collected data, remove abnormal or incomplete records, and prepare for subsequent analysis.

[0066] Step 2: Obtain the user's estimated use time, which can be manually input by the user, if the user does not input, use machine learning algorithm to analyze the user's usage habits, generate usage habit model, and predict the use time through the usage habit model;

[0067] Machine learning algorithm is a kind of calculation method that analyzes and learns the patterns and rules in data, and then automatically makes predictions, classifications or decisions. Unlike traditional rule-based programming, machine learning algorithm does not need to explicitly write all instructions, but through "training" from a large amount of data, it adjusts the model parameters by itself, and improves the performance on new data.

[0068] In this embodiment, it needs to be specifically pointed out that the step of using the machine learning algorithm to analyze the use habits of the user, generating a use habit model, and predicting the use time through the use habit model is:

[0069] Fill in or delete missing values. Common methods include mean filling or interpolation, and data standardization to eliminate the scale effect of different features, and the calculation formula is Where X norm represents the standardized data, X is the original data, μ represents the mean of the data, and σ represents the standard deviation of the data.

[0070] For each time point, a feature vector is constructed, and its expression is X i =[F i (t),D i (t),N i (t),T d (t)] where X i represents the feature vector of the i-th application at time t, F i (t) represents the use frequency of the i-th application at time t, D i (t) represents the single use duration of the i-th application at time t, N i (t) represents the number of programs simultaneously opened by the i-th application at time t, and T d (t) represents the use period of the i-th application at time t, such as morning, afternoon, evening, etc.

[0071] A support vector regression algorithm is used to construct a prediction model, and the data set is divided into a training set and a test set. The commonly used split ratio is 70% training set and 30% test set.

[0072] The support vector regression is a regression analysis method based on support vector machine, which is used to predict continuous numerical values. In a high-dimensional feature space, a regression function as smooth as possible is found, so that the error between the predicted value and the true value does not exceed the set tolerance, and it has strong robustness to noise and outliers in the training data. This goal is achieved by minimizing the complexity of the regression function and the combined loss function of the training sample error, thereby improving the prediction accuracy and preventing overfitting.

[0073] The selected model is trained using the training set data, the performance of the model is evaluated using cross-validation techniques to ensure the generalization ability of the model, and the final use habit model is obtained.

[0074] The cross-validation is a technique for evaluating the performance of machine learning models by dividing the dataset into multiple non-overlapping subsets called folds for model training and validation. Specifically, cross-validation divides the dataset into k folds, and each time uses k-1 folds as the training set and the remaining one fold as the validation set. The model is repeatedly trained and validated, and each time a different fold is selected as the validation set. Finally, by calculating the average value of the performance indicators (such as mean square error, accuracy, etc.) of k validations, cross-validation can provide a reliable estimate of the model's generalization ability on unseen data, effectively reducing the evaluation error caused by data division randomness.

[0075] The predicted usage time is obtained by using the final usage habit model for prediction.

[0076] In this embodiment, it needs to be specifically pointed out that the step of evaluating the performance of the model using the cross-validation technique is:

[0077] The dataset is divided into k non-overlapping subsets called folds, usually selected as 5 or 10, but other values can also be selected according to the size and characteristics of the dataset;

[0078] Each time, one fold is selected from the k folds as the validation set, and the remaining k-1 folds are used as the training set. The model is trained on the k-1 folds, and the model performance is evaluated on the validation set. The error between the predicted result and the actual result is recorded;

[0079] For each fold, the mean square error is calculated, and its expression is where MSE represents the mean square error, Y i is the actual value, is the predicted value, and n is the sample size; For each fold, the mean absolute error is calculated, and its expression is where MAE represents the mean absolute error; For each fold, the root mean square error is calculated, and its expression is where RMSE represents the root mean square error;

[0080] The standard deviation of all fold performance indicators is calculated, and the cross-validation results of different models are compared to select the model with the smallest standard deviation.

[0081] Step 3: Real-time monitoring of the power consumption state of each hardware component of the notebook computer, including but not limited to CPU, GPU, memory, display screen and other peripheral devices;

[0082] Step 4: According to the monitored hardware power consumption state and the user's usage habit data, the predicted endurance time of the current device is calculated;

[0083] Collecting computer-related data, the computer-related data including battery-related data, temperature data and user usage data, the user usage data including total duration of the user using the device since the start and the number of simultaneously running programs, and calculating a battery life adjustment factor according to the computer-related data;

[0084] Calculating an actual battery life time according to the battery life adjustment factor, the calculation formula being BF 实际 = RA x T ct , wherein BF 实际 represents the actual battery life time, RA represents the battery life adjustment factor, and T ct represents the estimated battery life time;

[0085] The battery performance is greatly affected by temperature. High temperature can accelerate the aging of the battery and reduce its effective capacity, and low temperature can affect the discharge efficiency of the battery. By considering the temperature data to adjust the estimated battery life time, the actual performance of the battery under different environmental conditions can be more accurately reflected; the usage duration of the device and the number of simultaneously running programs will directly affect the power consumption. For example, long-time high-load use will increase the power consumption and reduce the battery life time. The battery life adjustment factor can correct the estimated battery life time according to these usage conditions, so as to more accurately reflect the actual usage conditions; by calculating the battery life adjustment factor, the health status and performance change of the battery can be monitored. The power consumption in actual use and environmental factors (such as temperature) often do not match the theoretical values. The battery life adjustment factor helps to reduce such errors and makes the battery life time prediction closer to the actual situation.

[0086] In the embodiment, it needs to be specifically explained that the estimated battery life time calculation step of the current device is:

[0087] Monitoring the real-time power consumption of the device, including the power consumption of each hardware component such as CPU, GPU, screen and hard disk, and calculating the total power consumption by adding the power consumption of each component;

[0088] Obtaining the remaining energy of the battery, and calculating the estimated battery life time according to the total power consumption and the remaining energy of the battery, the calculation formula being , wherein T ct represents the estimated battery life time, E b represents the remaining energy of the battery, and P total represents the total power consumption.

[0089] In the embodiment, it needs to be specifically explained that the battery life adjustment factor calculation step is:

[0090] Obtaining the battery-related data of the notebook computer, and calculating the battery aging degree according to the battery-related data;

[0091] Obtaining the temperature data when the notebook computer is used, and calculating the heat dissipation efficiency according to the temperature data;

[0092] After obtaining the total duration of the user using the computer and the maximum number of simultaneously running programs after the computer is turned on, the usage degree is calculated according to the total duration of using the computer and the maximum number of simultaneously running programs, and the calculation formula is US = DU x NUM P , wherein US represents the usage degree, DU represents the total duration of using the computer, NUM P represents the maximum number of simultaneously running programs.

[0093] The endurance adjustment factor is comprehensively evaluated according to the battery aging degree, the heat dissipation efficiency and the usage degree, and the calculation formula is , wherein RA represents the endurance adjustment factor, AG represents the battery aging degree, as the battery aging degree increases, the actual endurance capability of the battery will decrease accordingly. Specifically, a higher battery aging degree will lead to a higher endurance adjustment factor, thereby shortening the predicted endurance time of the device. This means that when the battery aging degree increases, the system will adjust the predicted endurance time to reflect the decline in battery capacity and energy efficiency. As the heat dissipation efficiency improves, the endurance adjustment factor of the device will decrease, thereby reducing the impact of external factors on the predicted endurance time of the battery. When the device can dissipate heat more effectively, the overall energy consumption is optimized, especially the additional energy consumption and possible performance limitations caused by overheating are reduced. The improvement of heat dissipation efficiency reduces the situation of high temperature inside the device, thereby reducing the power consumption demand of the cooling system and reducing the energy waste caused by the processor reducing the frequency due to temperature rise, finally prolonging the battery endurance time. With the increase of the usage intensity of the computer, the predicted endurance time of the battery will be significantly shortened. When the user runs more application programs, performs more complex calculations or uses the device in a higher performance mode, the energy consumption of the computer will increase significantly, causing the battery to consume power faster. Therefore, the endurance adjustment factor will increase accordingly, reflecting the decline in the actual endurance capability of the battery. This means that high-intensity usage scenarios, such as multitasking or running high-performance applications, will all lead to a decrease in device endurance time. a1, a2, a3 represent the weight coefficients of the battery aging degree, the heat dissipation efficiency and the usage degree, and the specific values of a1, a2, a3 are determined by professionals according to the actual situation, and a1+a2+a3=1. For example, a1, a2, a3 can be 0.4, 0.4, 0.2.

[0094] In this embodiment, it needs to be specifically pointed out that the step of calculating the battery aging degree according to the battery related data is:

[0095] The charge and discharge times are extracted from the battery report of the battery management system of the notebook computer, and the charge and discharge times are the total number of complete charge and discharge cycles experienced by the battery, denoted as the charge and discharge times;

[0096] The current internal resistance of the battery is obtained using data provided by the battery management system, which reflects the electrical energy loss of the battery during charging and discharging.

[0097] The new battery internal resistance of the battery is collected, which represents the initial internal resistance of the battery when it leaves the factory. The internal resistance change rate is calculated based on the current internal resistance and the new battery internal resistance, and the calculation formula is Where ir represents the internal resistance change rate, R new represents the new battery internal resistance, R battery represents the current internal resistance.

[0098] The internal resistance change rate and the number of charge and discharge times are normalized, and the battery aging degree is calculated based on the normalized internal resistance change rate and the number of charge and discharge times, and the calculation formula is Where AG represents the battery aging degree, ir represents the internal resistance change rate, and NUM c represents the number of charge and discharge times.

[0099] In this embodiment, it needs to be specifically pointed out that the heat dissipation efficiency calculated according to the temperature data is:

[0100] The real-time temperature data of the key components (such as CPU, GPU) of the notebook computer is read by the sensor, which can usually be obtained through the temperature monitoring tool or API in the system; record the temperature before the start of heat dissipation, as the initial temperature;

[0101] Measure the real-time total power consumption of the notebook computer, which usually includes the power consumption of components such as CPU, GPU, which can be obtained through the power management system or special software;

[0102] The temperature rise value is calculated by subtracting the real-time temperature data from the initial temperature, and the heat dissipation efficiency is calculated based on the total power consumption and the temperature rise value, and the calculation formula is Where DH represents the heat dissipation efficiency, P total represents the total power consumption, and ΔW represents the temperature rise value.

[0103] Step 5: Dynamic power adjustment according to actual endurance time and estimated use time;

[0104] In this embodiment, it needs to be specifically pointed out that the dynamic power adjustment according to the actual endurance time and the estimated use time is:

[0105] Compare the actual endurance time with the estimated use time. If the actual endurance time is greater than the estimated use time, no adjustment is made; if the actual endurance time is less than the estimated use time, analyze the call probability of each program in the computer, and preferentially terminate the program with the smallest call probability, and continue to evaluate until the actual endurance time is greater than the estimated use time.

[0106] The analysis computer program call probability is that the starting number of each program in the computer use in the detection time period is obtained through the system log, the starting number of each program is accumulated to obtain the total program starting number, and the starting number of each program is calculated by the total program starting number to obtain the call probability, and the detection time period can be changed according to the situation, which can be one month or two months.

[0107] In this embodiment, it needs to be specifically pointed out that the power consumption data analysis system of the notebook computer, the system comprises:

[0108] The user usage habit acquisition module is used for collecting the habit data of the user using the notebook computer in daily life, and the habit data includes the use frequency of each application, the single use time length, the number of programs opened simultaneously during use, and the use period, etc.

[0109] The predicted use time acquisition module is used for providing an interface for the user to manually input the predicted use time, and the user can input the predicted time length of the current use period according to the demand; if the user does not input the predicted use time, the use habit model is generated by using the machine learning algorithm to analyze the use habit of the user according to the user habit data, and the use time is predicted by the use habit model;

[0110] The power consumption monitoring module is used for monitoring the power consumption state of the notebook computer components in real time, and the computer components include CPU, GPU, memory and display screen, and the current power consumption data is generated;

[0111] The endurance time calculation module is used for calculating the predicted endurance time supported by the remaining power according to the current power consumption data;

[0112] The intelligent process management module is used for dynamically adjusting the power supply according to the actual endurance time and the predicted use time.

[0113] Finally: the above only for the preferred embodiments of the present application, and not for limiting the present application, any modification, equivalent replacement, improvement, etc. made within the spirit and principles of the present application, should be included in the protection scope of the present application.

[0114] The above is only a specific embodiment of the present application, but the protection scope of the present application is not limited to this, any skilled person in the art can easily think of changes or replacements within the technical range disclosed in the present application, which should be included in the protection scope of the present application. Therefore, the protection scope of the present application should be subject to the protection scope of the claims.

Claims

1. A method for analyzing power consumption data of a laptop computer, characterized in that: The following steps are involved: Step 1: Collecting user usage habit data, including the usage frequency of all applications, the duration of a single session, the number of applications opened simultaneously, and the usage period; Step 2: Obtain the user's estimated usage time. The estimated usage time is manually input by the user. If the user does not input it, use a machine learning algorithm to analyze the user's usage habits based on the user's usage habit data, generate a usage habit model, and predict the usage time based on the usage habit model; Step 3: Monitor the power consumption status of various hardware components of the laptop in real time, including but not limited to the CPU, GPU, memory, display, and other peripheral devices; Step 4: Calculate the estimated battery life of the current device based on the monitored hardware power consumption status and user usage habits data; Collecting computer-related data, including battery-related data, temperature data, and user usage data, including the total time the user has used the device and the number of programs running simultaneously since the start of the current session, and calculating a battery life adjustment factor based on the relevant data; The actual endurance time is calculated based on the adjustment factor, and the calculation formula is BF 实际 =RA×T ct , where BF 实际 It represents the actual endurance time, RA represents the endurance adjustment factor, T ct Indicates the estimated battery life; Step 5: Dynamically adjust the power supply based on the actual battery life and the estimated usage time.

2. The method for analyzing power consumption data of a laptop computer according to claim 1, wherein: The steps for collecting user usage habit data are: Design a background process to continuously monitor and record user operations and mark specific timestamps and related information; Every time a user launches an application, the name of the application, the launch time, and the number of times the application is launched are recorded; Each time a user closes an application, the closing time is recorded. Based on the application's startup and closing times, the total duration of the session is calculated and recorded. Record the user's application launch situation at different time periods throughout the day, including the specific time points and corresponding usage duration, and classify the data by time period; All collected data is stored in date and time order.

3. The method for analyzing power consumption data of a laptop computer according to claim 1, wherein: The steps of using a machine learning algorithm to analyze the user's usage habits, generate a usage habit model, and predict usage time by using the habit model are as follows: Use the mean imputation method to fill missing values ​​in the data, standardize the data to eliminate the scale effect of different features, and construct a feature vector for each time point; Use the support vector regression algorithm to build a prediction model and divide the data set into training and test sets; Use the training set data to train the selected model, use cross-validation technology to evaluate the performance of the model, and obtain the final usage habit model; The estimated usage time is predicted using the final usage habit model.

4. The method for analyzing power consumption data of a laptop computer according to claim 3, wherein: The steps for evaluating the performance of the model using cross-validation techniques are: Divide the dataset into k non-overlapping subsets, called folds; Each time, one fold is selected from the k folds as the validation set, and the remaining k-1 folds are used as the training set. The model is trained on the k-1 folds, and the model performance is evaluated on the validation set. The error between the predicted result and the actual result is recorded. For each fold, calculate the mean squared error, mean absolute error, and root mean squared error; Calculate the standard deviation of the performance metrics across all folds, compare the cross-validation results of different models, and select the model with the smallest standard deviation.

5. The method for analyzing power consumption data of a notebook computer according to claim 1, wherein: The steps for calculating the endurance adjustment factor are as follows: Obtain battery-related data of the laptop computer and calculate the battery aging degree based on the battery-related data; Obtain temperature data of the laptop when it is in use, and calculate the heat dissipation efficiency based on the temperature data; Obtain the total time the user has used the computer and the maximum number of programs running simultaneously since the computer was turned on. Calculate the usage level based on the total time the user has used the computer and the maximum number of programs running simultaneously. The calculation formula is US = DU × NUM P , where US represents the degree of use, DU represents the total time of using the computer, and NUM P Expressed as the maximum number of simultaneously running programs; The battery life adjustment factor is obtained by comprehensive evaluation of battery aging, heat dissipation efficiency and usage level. The calculation formula is: Among them, RA represents the battery life adjustment factor, AG represents the battery aging, DH represents the heat dissipation efficiency, US represents the usage level, and a1, a2, and a3 represent the weight coefficients of battery aging, heat dissipation efficiency, and usage level.

6. The method for analyzing power consumption data of a notebook computer according to claim 5, wherein: The steps of calculating the battery aging degree based on the battery-related data are as follows: Extracting the charge and discharge count from a battery report of a battery management system of the laptop computer, where the charge and discharge count is the total number of complete charge and discharge cycles the battery undergoes; Obtain the current internal resistance of the battery from the data provided by the battery management system; Collect the new battery internal resistance of the battery, and calculate the internal resistance change rate based on the current internal resistance and the new battery internal resistance. The calculation formula is: Where ir is the internal resistance change rate, R new Expressed as the internal resistance of a new battery, R battery Expressed as current internal resistance; The internal resistance change rate and the number of charge and discharge times are normalized, and the battery aging degree is calculated based on the normalized internal resistance change rate and the number of charge and discharge times. The calculation formula is: Where AG is the battery aging degree, ir is the internal resistance change rate, NUM c Expressed as the number of charge and discharge cycles.

7. The method for analyzing power consumption data of a notebook computer according to claim 5, wherein: The steps for calculating the heat dissipation efficiency based on the temperature data are as follows: Record the temperature before the computer is used as the initial temperature, and read the real-time temperature data of the laptop's current CPU and GPU through the temperature sensor; Measure the total power consumption of a laptop in real time; The temperature rise value is obtained by calculating the difference between the real-time temperature data and the initial temperature. The heat dissipation efficiency is calculated based on the total power consumption and the temperature rise value. The calculation formula is: Where DH is the heat dissipation efficiency, P total It is expressed as total power consumption, and ΔW is expressed as temperature rise value.

8. The method for analyzing power consumption data of a notebook computer according to claim 1, wherein: The steps for dynamically adjusting the power supply based on the actual battery life and the expected usage time are as follows: Compare the actual battery life with the estimated usage time. If the actual battery life is longer than the estimated usage time, no adjustment will be made. If the actual battery life is shorter than the estimated usage time, analyze the probability of each program being called in the computer and terminate the program with the lowest probability of being called first. Continue the evaluation until the actual battery life is longer than the estimated usage time. The calling probability of each program in the analysis computer is obtained by obtaining the number of times each program is started on the computer during the detection time period through the system log, adding up the number of times each program is started to obtain the total number of program starts, and calculating the ratio of the number of times each program is started to the total number of program starts to obtain the calling probability.

9. A system for analyzing power consumption data of a notebook computer, for implementing the method for analyzing power consumption data of a notebook computer according to claims 1 to 8, characterized in that: The system comprises: User usage habit collection module, used to collect data on users' daily laptop usage habits; The estimated usage time acquisition module is used to provide an interface for users to manually enter the estimated usage time. If the user does not enter the estimated usage time, the module uses a machine learning algorithm to analyze the user's usage habits based on the user's habit data, generates a usage habit model, and predicts the usage time based on the usage habit model. A power consumption monitoring module is used to monitor the power consumption status of laptop components in real time, including the CPU, GPU, memory, and display, and generate current power consumption data; The battery life calculation module is used to calculate the actual battery life based on the current power consumption data; The intelligent process management module is used to dynamically adjust the power supply based on the actual battery life and the expected usage time.

Images