Touch screen low power consumption mode switching method, device and equipment

Abstract

The present invention relates to a method, device and equipment for switching a touch screen to a low power consumption mode. The method: time-frequency analysis and feature extraction of touch signals of the touch screen in different usage scenarios to obtain power consumption characteristic parameters of the touch screen; real-time analysis to obtain real-time working status data of the touch screen; power consumption prediction through a first Bayesian probability model to obtain power consumption control parameters of the first touch area; regional division of the touch screen to obtain multiple independent power consumption control blocks, and power consumption switching data is collected; the power consumption switching data is input into a linear feedback control model for dynamic response analysis to obtain transition stage control parameters and stable stage control parameters for adaptive optimization to obtain power consumption control parameters of the second touch area, and the parameters are fed back to the first Bayesian probability model to obtain a second Bayesian probability model. While ensuring the touch response performance, the present invention effectively reduces the overall power consumption of the touch screen.

Description

Technical Field

[0001] The present invention relates to the technical field of touch screen power consumption management, and in particular to a method, device and equipment for switching a touch screen into a low power consumption mode. Background Art

[0002] When the touch screen is used for a long time, high power consumption not only affects the battery life of the device, but also causes problems such as local heating of the touch screen.

[0003] The current touch screen power consumption control method mainly adopts a unified power consumption management strategy, which cannot perform differentiated control according to the usage characteristics of different areas. In actual applications, this method has problems such as low power consumption control efficiency and unstable touch response performance. Especially in multi-touch and complex operation scenarios, it is difficult to balance the relationship between power consumption reduction and touch experience. Traditional touch screen power consumption control technology lacks in-depth analysis and precise modeling of touch usage patterns, and cannot accurately predict and dynamically adjust power consumption parameters. At the same time, during the power consumption mode switching process, due to the lack of effective feedback mechanism and optimization strategy, problems such as large power consumption fluctuations and slow switching response are prone to occur, affecting the normal user experience. Summary of the invention

[0004] The main purpose of the present invention is to provide a method, device and equipment for switching a touch screen into a low power consumption mode. The present invention effectively reduces the overall power consumption of the touch screen while ensuring the touch response performance.

[0005] To achieve the above object, the present invention provides a method for switching a touch screen to a low power consumption mode, comprising the following steps:

[0006] Perform time-frequency analysis and feature extraction on the touch signals of the touch screen in different usage scenarios to obtain the power consumption characteristic parameters of the touch screen;

[0007] Collecting touch position data, touch pressure data and touch time data through multiple touch sensors in the touch screen, and performing real-time analysis to obtain real-time working status data of the touch screen;

[0008] According to the power consumption characteristic parameters of the touch screen and the real-time working state data of the touch screen, power consumption prediction is performed through a first Bayesian probability model to obtain a first touch area power consumption control parameter;

[0009] Based on the first touch area power consumption control parameter, the touch screen is divided into regions to obtain a plurality of independent power consumption control blocks, and the power supply voltage parameters, sampling frequency parameters and sensitivity parameters of the independent power consumption control blocks are adjusted by region, and power consumption switching data of each independent power consumption control block is collected;

[0010] Inputting the power consumption switching data into a linear feedback control model for dynamic response analysis to obtain transition phase control parameters and stable phase control parameters;

[0011] The transition phase control parameter and the stable phase control parameter are adaptively optimized to obtain a second touch area power consumption control parameter, and the second touch area power consumption control parameter is fed back to the first Bayesian probability model to obtain a second Bayesian probability model.

[0012] The present invention also provides a touch screen low power consumption mode switching device, comprising:

[0013] A feature extraction module is used to perform time-frequency analysis and feature extraction on the touch signal of the touch screen in different usage scenarios to obtain the power consumption characteristic parameters of the touch screen;

[0014] A real-time analysis module, used to collect touch position data, touch pressure data and touch time data through multiple touch sensors in the touch screen, and perform real-time analysis to obtain real-time working status data of the touch screen;

[0015] A power consumption prediction module, configured to predict the power consumption through a first Bayesian probability model according to the power consumption characteristic parameters of the touch screen and the real-time working state data of the touch screen, and obtain a power consumption control parameter of a first touch area;

[0016] A regional adjustment module, configured to divide the touch screen into regions based on the first touch area power consumption control parameter to obtain a plurality of independent power consumption control blocks, and to adjust the supply voltage parameters, sampling frequency parameters and sensitivity parameters of the independent power consumption control blocks in different regions, and to collect power consumption switching data of each independent power consumption control block;

[0017] A dynamic response analysis module, used for inputting the power consumption switching data into a linear feedback control model for dynamic response analysis to obtain transition phase control parameters and stable phase control parameters;

[0018] The adaptive optimization module is used to adaptively optimize the transition phase control parameters and the stable phase control parameters to obtain the second touch area power consumption control parameters, and feed back the second touch area power consumption control parameters to the first Bayesian probability model to obtain the second Bayesian probability model.

[0019] The present invention also provides a computer device, comprising a memory and a processor, wherein the memory stores a computer program, and the processor implements the steps of any one of the above methods when executing the computer program.

[0020] The present invention also provides a computer-readable storage medium on which a computer program is stored. When the computer program is executed by a processor, the steps of any of the above-mentioned methods are implemented.

[0021] In summary, the technical solution provided by the present invention establishes a touch screen power consumption characteristic model through time-frequency analysis and feature extraction technology, realizes an accurate description of the touch screen usage characteristics, adopts a Bayesian probability model to predict power consumption, and establishes a dynamic power consumption prediction mechanism in combination with real-time working status data, so that power consumption control is more forward-looking; based on multi-region division and independent power consumption control strategy, differentiated power consumption management of different regions of the touch screen is realized, avoiding the waste of resources caused by global unified control; the introduction of linear feedback control model and dynamic response analysis mechanism makes the power consumption mode switching process smoother and reduces the power consumption fluctuation during the switching process; by establishing an adaptive optimization mechanism and a feedback update mechanism, the power consumption control parameters can be continuously optimized according to actual usage conditions, ensuring the continuous improvement of the control effect; by utilizing multi-level data processing and parameter optimization strategies, while ensuring the touch response performance, the overall power consumption of the touch screen is effectively reduced. BRIEF DESCRIPTION OF THE DRAWINGS

[0022] Figure 1 1 is a schematic diagram of steps of a method for switching a touch screen to a low power consumption mode in one embodiment of the present invention;

[0023] Figure 2 This is a structural block diagram of a touch screen low power consumption mode switching device in one embodiment of the present invention;

[0024] Figure 3 It is a schematic block diagram of the structure of a computer device according to an embodiment of the present invention.

[0025] The realization of the purpose, functional features and advantages of the present invention will be further explained in conjunction with embodiments and with reference to the accompanying drawings. DETAILED DESCRIPTION

[0026] In order to make the purpose, technical solution and advantages of the present invention more clearly understood, the present invention is further described in detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not intended to limit the present invention.

[0027] Reference Figure 1 This embodiment provides a method for switching a touch screen to a low power consumption mode, comprising the following steps:

[0028] S1, performing time-frequency analysis and feature extraction on the touch signals of the touch screen in different usage scenarios to obtain power consumption characteristic parameters of the touch screen;

[0029] Among them, the touch position data, touch pressure data and touch time data collected by the touch screen in different usage scenarios are subjected to data standardization processing to eliminate the differences in data scales in different scenarios, so that the subsequent analysis process is not affected by various types of noise, and a standardized touch signal sequence is obtained. The standardized touch signal sequence is decomposed in the time domain to extract the time domain feature vector of the touch signal. The time domain feature vector includes information such as touch duration, touch interval data and touch intensity. These features can reflect the touch behavior pattern of the user in different usage scenarios. For example, the touch duration indicates the time that the user keeps in contact with the touch screen in one operation, the touch interval indicates the time interval between two operations, and the touch intensity describes the pressure applied by the user to the touch screen. The time domain feature vector is input into the spectrum analysis model, and the frequency domain feature vector of the touch signal is obtained by Fourier transform. Fourier transform is the process of converting a signal from the time domain to the frequency domain, thereby extracting the frequency component of the signal. The frequency domain feature vector includes the main frequency data and frequency distribution data of the touch signal, which reflect the main frequency components of the touch operation and their energy distribution. Through frequency domain analysis, the main mode of user touch frequency in a specific scenario is identified, such as frequent fast touch and long-term static contact. The time domain feature vector is feature-correlated with the frequency domain feature vector to obtain the touch screen's natural frequency parameters. The touch screen's natural frequency parameters are used to characterize the touch screen's natural working characteristics in different usage scenarios. The touch screen's natural frequency reflects the most natural response characteristics of the touch screen based on user operation behavior when there is no external interference. Combining the time domain and frequency domain features, a more comprehensive understanding of the touch screen's behavior patterns in different scenarios is achieved. Dynamic threshold calculation is performed on the touch screen's natural frequency parameters to obtain dynamic power consumption parameters. According to different usage scenarios and user operation modes, the reasonable range of power consumption is adaptively determined to reduce unnecessary energy consumption. The dynamic power consumption parameters are the basis for dynamic management of the touch screen's power consumption characteristics under different times and operating conditions. Coefficient calculation is performed based on the dynamic power consumption parameters to obtain the time-frequency coefficient. The time-frequency coefficient is used to describe the relationship between touch signals in the time domain and frequency domain. The touch screen's natural frequency parameters, dynamic power consumption parameters, and time-frequency coefficients are data fused to obtain initial power consumption characteristic parameters, reflecting the touch screen's power consumption behavior pattern in typical operation scenarios. The initial power consumption characteristic parameters are calibrated based on historical data and actual measurement data to obtain the final touch screen power consumption characteristic parameters.

[0030] S2, collecting touch position data, touch pressure data and touch time data through multiple touch sensors in the touch screen, and performing real-time analysis to obtain real-time working status data of the touch screen;

[0031] Specifically, the original data such as the touch position, pressure and time collected by the touch sensor are classified to obtain an initial data set, which includes touch position data, touch pressure data and touch time data. The initial data set is cleaned and normalized to remove incomplete, unreasonable or noisy abnormal data to ensure that the subsequent analysis results have high reliability and accuracy. Normalization is to eliminate the differences caused by different data scales so that the data of each sensor can be compared and analyzed under the same dimension. Through data cleaning and normalization, a standardized data set is obtained. Based on the standardized data set, spatial distribution calculation is performed to obtain a touch area distribution matrix. It reflects the degree of spatial aggregation of user operations on the touch screen, that is, which areas are frequently touched and which areas are relatively rarely touched. Time series analysis is performed on the standardized data set to obtain a touch frequency matrix, including single-point touch frequency and multi-point touch frequency, which reflects the operation frequency characteristics of users when using the touch screen. Single-point touch frequency describes the frequency of a user's single contact with the screen, while multi-point touch frequency is used to describe the frequency of multiple touch operations performed simultaneously. Time series analysis can effectively capture the time-varying characteristics of user operations, thereby identifying the user's operation mode, such as fast continuous clicks, long presses, or multi-point gesture operations. Pattern analysis is performed based on the touch area distribution matrix and the touch frequency matrix to obtain touch pattern feature data. The touch pattern feature data can describe the user's behavior pattern when using the touch screen. The touch pattern feature data is divided into time windows to obtain touch feature sequences of multiple time segments. The user operation is segmented on the time axis so that the touch features in each time segment are analyzed and processed independently. Real-time state evaluation is performed based on the touch feature sequence to obtain the initial working state data of the touch screen, reflecting the power consumption requirements and response states of the touch screen in different time segments. In order to adapt to the dynamic changes of user operations, the initial working state data needs to be continuously and dynamically updated to ensure that the real-time working state data of the touch screen can accurately reflect the current user operation behavior and system state. In this way, the touch screen adaptively adjusts the power consumption strategy according to the user's operating habits and current working state, such as maintaining high sensitivity and response speed in areas with high user operation frequency, and reducing power consumption in low-frequency areas to achieve energy saving.

[0032] S3, predicting the power consumption through a first Bayesian probability model according to the power consumption characteristic parameters of the touch screen and the real-time working status data of the touch screen, and obtaining a power consumption control parameter of the first touch area;

[0033] It should be noted that the touch screen power consumption characteristic parameters and real-time working state data are input into the feature extraction layer of the Bayesian model. In the feature extraction layer, the input data undergoes a feature decomposition process to obtain two important feature outputs: a power consumption feature vector and a working state vector. The power consumption feature vector represents the energy consumption characteristics of the touch screen under different usage conditions, while the working state vector reflects the current real-time operation of the touch screen. These feature vectors together constitute important input information for power consumption prediction. The power consumption feature vector and the working state vector are input into the conditional probability calculation layer of the first Bayesian probability model for joint probability analysis. In this layer, the model calculates the prior probability matrix of the power consumption state through the Bayesian joint probability formula, reflecting the possible power consumption state distribution of the touch screen under different working conditions. The maximum a posteriori probability is calculated for the prior probability matrix of the power consumption state to obtain the a posteriori probability matrix of the power consumption state. The a posteriori probability matrix reflects the possibility of different power consumption states given real-time data. The maximum a posteriori probability is calculated by the Bayesian theorem. This process improves the accuracy of power consumption prediction by combining the prior probability with the real-time input data, so that the model can better adapt to the dynamic changes in actual work. The posterior probability matrix is ​​input into the parameter inference layer of the first Bayesian probability model to calculate the power consumption parameters and obtain the regional power consumption control vector. The regional power consumption control vector is a specific quantitative description of the power consumption control of different regions of the touch screen, indicating the power consumption requirements and power supply strategies of each touch region under a specific state. The regionalized power consumption control vector enables the system to independently manage the power consumption of different regions, thereby more effectively utilizing energy resources and reducing unnecessary energy consumption. Based on the regional power consumption control vector, a power consumption control parameter mapping matrix is ​​constructed to map the regionalized power consumption control vector to specific power consumption control parameters to facilitate overall power consumption management at the system level. These parameters include power supply voltage, response time, and sensitivity settings. In order to ensure that these power consumption control parameters meet the overall requirements of system operation, the mapping matrix is ​​input into the constraint optimization layer of the first Bayesian probability model for parameter constraint processing. A series of system constraints and target conditions are applied to adjust the power consumption control parameters to ensure that these parameters can effectively control energy consumption in actual applications without affecting the normal user experience. The power consumption control parameters after constraint optimization are smoothed in time to reduce the fluctuation of the power consumption parameters in the time dimension and avoid frequent power consumption state switching that may cause system instability or a decrease in user experience. The smoothed power consumption control parameters can make the touch screen perform more smoothly when the working state changes, thereby improving the user's operating experience and helping to extend the life of the touch screen and its related circuit components. The smoothed power consumption control parameters are input into the control strategy generation layer of the first Bayesian probability model for parameter integration to obtain the power consumption control parameters of the first touch area.

[0034] S4, dividing the touch screen into regions based on the power consumption control parameter of the first touch area to obtain a plurality of independent power consumption control blocks, and adjusting the power supply voltage parameters, sampling frequency parameters and sensitivity parameters of the independent power consumption control blocks by region, and collecting power consumption switching data of each independent power consumption control block;

[0035] Specifically, based on the first touch area power consumption control parameter, the similarity of the touch area is calculated to obtain a regional power consumption correlation matrix. The similarity calculation of the touch area is performed based on multiple dimensions, such as power consumption characteristics, usage frequency, and touch intensity, and the calculated power consumption correlation matrix is ​​used to describe the mutual correlation and characteristic similarity between the areas. Dynamic clustering analysis is performed based on the regional power consumption correlation matrix to obtain the initial area division result. Areas with similar power consumption characteristics are classified into one category, so that the areas with similar characteristics are uniformly adjusted in the subsequent power consumption management to improve the efficiency and accuracy of control. After the initial area division, a preliminary area division scheme is obtained, in which each area has a high internal similarity in space. The initial area division result is optimized by boundary. Through boundary optimization, unreasonable area division is eliminated, so that the boundary of each independent power consumption control block is smoother and more natural, and multiple independent power consumption control blocks are obtained. According to the independent power consumption control blocks, the first touch area power consumption control parameter is parameter deconstructed according to these blocks to obtain a block control parameter set. The block control parameter set includes the power supply voltage range, sampling frequency range, and sensitivity range of each independent power consumption control block. The setting of the supply voltage range affects the power consumption level of each block, the sampling frequency range determines the system's response frequency to touch operations, and the sensitivity range affects the ability to identify touch operations of different intensities. The setting of the control parameter set is to ensure that the corresponding energy consumption control is carried out in different areas according to the actual usage, so as to strike a balance between power consumption and user experience. In order to improve the accuracy and optimization effect of power management, the block control parameter set is optimized with multiple objectives. An optimal control parameter combination is found among multiple objectives, including the lowest power consumption, the highest sensitivity, and the best user experience. In the process of multi-objective optimization, a block optimization control parameter that takes into account both energy consumption and performance is obtained by weighing different objectives. The optimization process considers the specific needs of each block and finds the optimal solution between these needs through mathematical models to ensure that the final block optimization control parameters can maximize the use of system resources while meeting the user's operation requirements. Based on the block optimization control parameters, power consumption adjustment is carried out in different regions to obtain real-time power consumption data of the block. Each independent power control block performs real-time power consumption adjustment according to its optimized supply voltage, sampling frequency, and sensitivity parameters. For example, for areas where users frequently operate, maintain a high sampling frequency and sensitivity, while for areas that are less frequently operated, reduce the power supply voltage and sampling frequency to save energy. By adjusting each block by region, the overall power consumption can be effectively reduced and the battery life of the device can be extended. Perform timing analysis on the real-time power consumption data of the block to obtain the power consumption switching data of each independent power control block. Capture the power consumption changes of each block at different times, so as to model and predict the dynamic changes of power consumption. Through timing analysis, identify the changing trend of power consumption, such as the increase or decrease of power consumption of a block in a specific time period.

[0036] S5, inputting the power consumption switching data into a linear feedback control model for dynamic response analysis to obtain transition phase control parameters and stable phase control parameters;

[0037] The power consumption switching data is input into the state observer layer of the linear feedback control model for state decomposition. The complex power consumption switching data is converted into a power consumption state vector through the processing of the state observer layer. In the power consumption state vector, the energy consumption change characteristics of each control block of the touch screen during different power consumption switching processes are recorded. The power consumption state vector is input into the error calculation layer of the linear feedback control model for error analysis. The target power consumption state is compared with the current actual power consumption state, and the power consumption error matrix is ​​calculated to reflect the degree and direction of the current system power consumption deviation from the target value. The goal of the control system is to minimize the power consumption error by adjusting the control parameters so that the system tends to the predetermined power consumption target state. By analyzing the power consumption error, it is identified in which aspects of the system there is excessive energy consumption, and the corresponding improvement direction. The feedback gain matrix is ​​calculated for the power consumption error matrix to obtain the control gain parameter. The control gain parameter is a key factor used to amplify or reduce the control signal in the linear feedback control model, and determines the response strength of the system to the power consumption error. By calculating the feedback gain matrix, the amplification ratio of the system to the power consumption error is determined, so that the system can respond quickly and effectively and adjust the power consumption state. The control gain parameter plays the role of an "accelerator" in the system, which determines the response speed and control accuracy of the control system. The control gain parameter and the power consumption error matrix are input into the compensation control layer of the linear feedback control model for error compensation. In the compensation control layer, the power consumption error is compensated according to the control gain parameter to generate a compensation control signal. The compensation control signal is a specific measure to correct the power consumption error in the system. By adjusting the parameters such as the power supply voltage, sampling frequency and sensitivity of each power control block in real time, the power consumption state of the system tends to the target state. The generation of the compensation control signal is a dynamic process, which is constantly adjusted as the system state changes, so as to ensure that the system can maintain a relatively stable power consumption level under changing external conditions. Based on the compensation control signal, the phase margin analysis is performed to obtain the stability parameters of the system. The phase margin is an important indicator for measuring the stability of the control system. By performing the phase margin analysis on the compensation control signal, the stability parameters of the system under different power consumption states are obtained. These stability parameters are used to evaluate whether the system has the possibility of instability during the power consumption control process, such as excessive power consumption oscillation or system response delay. The stability parameters of the system are classified into state to obtain the transition state parameter set and the stable state parameter set. The transition state parameter set is used to describe the transient behavior of the system during the power switching process, such as how the system switches from a high power state to a low power state, while the stable state parameter set is used to describe the steady-state behavior of the system after reaching the target power state. The transition state parameter set is input into the dynamic response optimization layer of the linear feedback control model for parameter optimization to obtain the control parameters of the transition stage.In the dynamic response optimization layer, the optimal control strategy is found by analyzing and optimizing the transition state parameter set to ensure that the system can respond quickly during the power consumption switching process and smoothly transition to the target power consumption state. By optimizing the control parameters in the transition phase, the oscillation and overshoot of the system during the power consumption switching process can be effectively reduced, and the response speed and stability of the system can be improved. Similarly, the stable state parameter set is input into the steady-state control optimization layer of the linear feedback control model for parameter optimization to obtain the control parameters of the stable phase. In the steady-state control optimization layer, the stable state parameter set is analyzed to find the optimal control parameters in the stable phase of the system to ensure that the system can remain stable after reaching the target power consumption state without significant fluctuations in power consumption, while reducing energy consumption as much as possible to achieve the goal of energy saving.

[0038] S6, adaptively optimizing the transition phase control parameters and the stable phase control parameters to obtain the second touch area power consumption control parameters, and feeding the second touch area power consumption control parameters back to the first Bayesian probability model to obtain the second Bayesian probability model.

[0039] Specifically, the control parameters of the transition phase and the control parameters of the stable phase are fused and analyzed to obtain a control parameter fusion matrix, which includes a dynamic adjustment coefficient and a steady-state maintenance coefficient. The dynamic adjustment coefficient is used to describe how the system quickly adjusts the power consumption parameters during the transition phase, while the steady-state maintenance coefficient is used to ensure that the system can operate stably after reaching the target power consumption state. Through phased coefficient fusion, the system can maintain a fast response during the power consumption switching process while ensuring the energy efficiency in the stable phase. The control parameter fusion matrix is ​​constructed with an objective function to obtain an optimization objective function. The optimization objective function includes a dynamic response constraint term and a steady-state performance constraint term, wherein the dynamic response constraint term is used to ensure the rapid responsiveness of the system during the state switching process, and the steady-state performance constraint term is used to maintain the best balance between the performance and power consumption of the system during stable operation. Iterative calculation is performed based on the optimization objective function to obtain an optimization parameter sequence. Through iterative calculation, the optimal solution is gradually approached to find the best control parameter combination to achieve refined management of power consumption. In each iteration, the next optimization direction is calculated according to the current parameter value, and the parameters are continuously adjusted until convergence to an optimal state. The optimization parameter sequence is parameter smoothed to obtain smooth optimization parameters. Eliminate the short-term fluctuations caused by iterative calculations, make the final control parameters smoother and more stable, reduce the frequent fluctuations caused by the power consumption control process, and improve the overall stability of the system. Map the control parameters based on the smooth optimization parameters to obtain the second touch area power consumption control parameters. The second touch area power consumption control parameters include optimized power supply voltage values, sampling frequency values, and touch sensitivity values. Through the optimization of these parameters, it is ensured that in different touch areas, the power supply voltage, sampling frequency, and touch sensitivity can be dynamically adjusted according to the user's operating habits and current usage scenarios, thereby achieving more efficient energy management. Update the model parameters based on the second touch area power consumption control parameters to obtain the Bayesian model update parameters. The Bayesian model update parameters include a priori probability update value and a conditional probability update value. The priori probability update value is used to adjust the model's prior cognition of the system state, while the conditional probability update value is used to update the probability distribution of the model's power consumption under different states. By updating these parameters, the Bayesian model can better adapt to the current system state and usage scenarios, and improve the accuracy of power consumption prediction and the effectiveness of control. Based on the Bayesian model update parameters, the probability distribution of the first Bayesian probability model is updated to obtain an updated conditional probability distribution, so that the model can better reflect the current system status and user operation mode, thereby providing more accurate reference data for future power consumption prediction. The updated conditional probability distribution is reconstructed to obtain a second Bayesian probability model. By introducing a new probability distribution and updated model parameters, the Bayesian model can more accurately reflect the dynamic characteristics and power consumption status of the system.

[0040] In one example, touch signals of a touch screen in different usage scenarios are subjected to time-frequency analysis and feature extraction to obtain power consumption characteristic parameters of the touch screen, including:

[0041] Standardize the touch position data, touch pressure data, and touch time data collected by the touch screen in different usage scenarios to obtain a standardized touch signal sequence;

[0042] Decomposing the standardized touch signal sequence in the time domain to obtain a time domain feature vector of the touch signal, wherein the time domain feature vector includes touch duration data, touch interval data, and touch intensity data;

[0043] Input the time domain feature vector into the spectrum analysis model for Fourier transform to obtain the frequency domain feature vector of the touch signal, where the frequency domain feature vector includes the signal main frequency data and frequency distribution data;

[0044] Performing feature correlation on the time domain feature vector and the frequency domain feature vector to obtain the natural frequency parameters of the touch screen, where the natural frequency parameters of the touch screen represent the inherent working characteristics of the touch screen in different usage scenarios;

[0045] Performing dynamic threshold calculation on the natural frequency parameters of the touch screen to obtain dynamic power consumption parameters, and performing coefficient calculation based on the dynamic power consumption parameters to obtain time-frequency coefficients;

[0046] The inherent frequency parameters, dynamic power consumption parameters and time-frequency coefficients of the touch screen are subjected to data fusion processing to obtain initial power consumption characteristic parameters, and the initial power consumption characteristic parameters are subjected to parameter correction to obtain the power consumption characteristic parameters of the touch screen.

[0047] In this example, the touch position data, touch pressure data, and touch time data collected by the touch screen in different usage scenarios are normalized to eliminate the scale differences of the data in different scenarios, so that they have the same statistical characteristics and form a standardized touch signal sequence. Data normalization uses the mean-variance normalization method to subtract the mean from each data and then divide it by the standard deviation, so that the mean of the data is 0 and the standard deviation is 1. For example, for touch pressure data , its standardized representation is:

[0048] ;

[0049] in, is the standardized touch pressure data, is the mean touch pressure, is the standard deviation of the touch pressure data. Similarly, the touch position and touch time data are standardized to obtain the standardized touch position and time data. The standardized touch signal sequence is decomposed in the time domain to obtain the time domain feature vector of the touch signal. The touch signal is analyzed from the perspective of time to extract time-related features, including touch duration data, touch interval data, and touch intensity data. For example, the touch duration It refers to the duration of time that the user touches the screen in one operation; touch interval Refers to the interval between two consecutive operations; touch intensity Indicates the intensity of pressure applied by the user to the screen. The time domain feature vector is expressed in mathematical form as follows:

[0050] ;

[0051] Time domain feature vector Contains the time characteristics of user operations. By analyzing these characteristics, we can understand the user's operating habits, such as frequent clicks or long presses. Perform frequency domain analysis on the time domain feature vector and input it into the spectrum analysis model for Fourier transform to obtain the frequency domain feature vector of the touch signal. The touch signal is converted from the time domain to the frequency domain in order to identify the energy distribution of different frequency components in the signal. The Fourier transform is expressed as:

[0052] ;

[0053] in, represents the Fourier transform, is the frequency domain eigenvector. Including signal frequency data and frequency distribution data ,in represents the dominant frequency of the touch signal, Indicates the energy distribution of each frequency component. Through frequency domain analysis, the main frequency components of touch operation are identified, such as whether the user has a fast click operation mode or a stable gesture operation. The time domain feature vector and the frequency domain feature vector are feature-correlated to obtain the natural frequency parameters of the touch screen. Combining time domain and frequency domain information, the characteristics of the touch screen in different operation scenarios are fully reflected. Natural frequency parameters It is calculated through the correlation between time domain and frequency domain features, and is used to characterize the inherent working characteristics of the touch screen in different usage scenarios. It is calculated using the following formula:

[0054] ;

[0055] in, and is the weight coefficient, which is used to adjust the influence of time domain and frequency domain features in calculating the natural frequency parameters. Through feature association, the natural frequency parameters of the touch screen are obtained. , characterizes the response characteristics of the touch screen in typical usage scenarios. Perform dynamic threshold calculation on the natural frequency parameters of the touch screen to obtain dynamic power consumption parameters. Determine the power consumption range of the touch screen based on the current touch frequency to ensure that energy consumption is minimized while responding to user operations. Dynamic power consumption parameters Calculated as follows:

[0056] ;

[0057] in, and is the regulation coefficient of the system, is the natural frequency parameter obtained above. The dynamic power consumption parameter indicates the energy demand of the touch screen in different states. Based on the dynamic power consumption parameter, the time-frequency coefficient is calculated to better adjust and control the power consumption. It is used to describe the energy distribution characteristics of the touch screen in the time domain and frequency domain. The calculation formula is as follows:

[0058] ;

[0059] in, is the adjustment coefficient, It is a very small positive number used to avoid the denominator being zero. It is used to describe the comprehensive characteristics of touch signals in both time and frequency dimensions, which helps to perform refined power consumption control in different usage scenarios. The natural frequency parameters, dynamic power consumption parameters and time-frequency coefficients of the touch screen are fused to obtain the initial power consumption characteristic parameters. The initial power consumption characteristic parameters are expressed as:

[0060] ;

[0061] in, and is the fusion coefficient, which is used to adjust the influence of different features on the power consumption characteristic parameters. Through data fusion, the initial power consumption characteristic parameters are obtained. , reflecting the power consumption characteristics of the touch screen in typical usage scenarios. Perform parameter correction on the initial power consumption characteristic parameters to obtain the final power consumption characteristic parameters of the touch screen. Adjust the initial power consumption characteristic parameters based on historical data or real-time measurement data so that they can more accurately reflect the energy consumption status of the touch screen. Corrected touch screen power consumption characteristic parameters It is expressed by the following formula:

[0062] ;

[0063] in, It is a correction value, calculated based on historical power consumption data or real-time feedback from the system.

[0064] In one example, multiple touch sensors in a touch screen collect touch position data, touch pressure data, and touch time data, and perform real-time analysis to obtain real-time working status data of the touch screen, including:

[0065] Classifying the touch data collected by the multiple touch sensors in the touch screen to obtain an initial data set, where the initial data set includes touch position data, touch pressure data, and touch time data;

[0066] Performing data cleaning and normalization processing on the initial data set to obtain a standardized data set, and performing spatial distribution calculation based on the standardized data set to obtain a touch area distribution matrix, which reflects the degree of spatial aggregation of touch operations;

[0067] Perform time series analysis on the standardized data set to obtain a touch frequency matrix, which includes single-point touch frequency and multi-point touch frequency;

[0068] Performing pattern analysis based on the touch area distribution matrix and the touch frequency matrix to obtain touch pattern feature data, and dividing the touch pattern feature data into time windows to obtain touch feature sequences of multiple time segments;

[0069] A real-time status evaluation is performed based on the touch feature sequence to obtain initial working status data, and the initial working status data is dynamically updated to obtain real-time working status data of the touch screen.

[0070] In this example, the touch data is classified and processed to obtain an initial data set. Multiple sensors in the touch screen collect different types of data, including touch position data, touch pressure data, and touch time data. Touch position data Indicates the coordinate position when the user touches the screen, where and They are horizontal and vertical coordinates, touch pressure data Reflects the pressure applied by the user to the screen, while the touch time data It describes the time when the user touches the screen. By classifying these raw data, we get the initial data set containing touch position, pressure and time, recorded as ,in Indicates the number of touch data points collected. Perform data cleaning and normalization. Remove outliers and noise data, such as incorrectly collected extreme data points or invalid data caused by physical interference. and are the maximum and minimum values ​​of touch pressure, respectively. Data cleaning is achieved by filtering out data points that exceed a certain threshold range, for example, only retaining The data points are normalized during data cleaning so that the data have the same scale. Normalization is achieved through the following formula:

[0071] ;

[0072] in, Represents the normalized touch pressure data. Similarly, the position data and time data are normalized to obtain a standardized data set, which is recorded as , so as to ensure that the data is analyzed and processed at a unified scale. Based on the standardized data set, the spatial distribution calculation is performed to obtain the touch area distribution matrix. By analyzing the distribution characteristics of the user's touch on the screen, the user's operating habits and the frequency of use of each area are understood. Assume that the touch screen is divided into The grid area of ​​the touch area distribution matrix is ​​expressed as ,in Indicates Line The number of touches in the column grid. If the normalized touch position Falling in the grid area , then the corresponding element Increase by 1. By traversing each data point in the standardized data set, the complete touch area distribution matrix is ​​obtained. , reflecting the frequency and spatial concentration of user touches in different screen areas. For example, if The values ​​are much larger than those in other areas, indicating that these areas are high-frequency areas where users frequently operate and require higher sensitivity and response speed. Time series analysis is performed on the standardized data set to obtain the touch frequency matrix. By analyzing the time characteristics of touch operations, the user's operation mode is captured. The touch frequency matrix includes single-point touch frequency and multi-point touch frequency. Assume Indicates at time The single-point touch frequency at the moment, and Indicates the frequency of multi-touch mode, which is estimated by counting the number of touches within a period of time. For example, in the time interval The number of single touches in , then the single-point touch frequency is expressed as:

[0073] ;

[0074] Similarly, calculate the multi-touch frequency By traversing all time intervals, we can get the touch frequency matrix , reflecting the characteristics of user operations in the time dimension, including quick clicks, long presses, and multi-point gestures. Based on the touch area distribution matrix and touch frequency matrix Perform pattern analysis to obtain touch pattern feature data. Touch pattern feature data is used to describe the user's operation mode on the screen, such as whether the user tends to perform frequent operations in certain areas or whether there is a preference for multi-touch. Pattern analysis is implemented through clustering algorithms, such as the KMeans algorithm. Assume that the result after clustering is ,in is the number of cluster categories, each Represents a specific operation mode. Based on the clustering results, the touch pattern feature data is divided into time windows to obtain touch feature sequences of multiple time segments. The user's operation is divided into multiple segments according to time, and each time segment contains one or more touch patterns, so as to perform more detailed real-time analysis. For example, the entire operation process is divided into time windows, and the touch feature sequence of each window is expressed as ,in is the number of touch patterns in the current time window. Based on the obtained touch feature sequence, a real-time state evaluation is performed to obtain the initial working state data. The working state of the touch screen is determined by analyzing the touch features of the current time window, for example, whether the touch screen is in a high-frequency response state or a low-frequency standby state. Assume Indicates The working status data of a time window is calculated using the following formula:

[0075] ;

[0076] in, It means each touch mode The weight of the working state. By calculating all time windows, the complete initial working state data is obtained. , used to describe the power consumption requirements and response status of the touch screen in different time periods. The initial working status data is dynamically updated to obtain the real-time working status data of the touch screen. According to the changes in user operations, the working status of the touch screen is adjusted in real time. For example, if it is detected that the user changes from frequent operations to less frequent operations, the power consumption is reduced and the low-frequency standby mode is entered. The dynamic update is achieved through the exponentially weighted moving average method, and the update formula is as follows:

[0077] ;

[0078] in, It is Real-time working status data of a time window, It is a smoothing factor that controls the degree of dependence on historical data. Through dynamic updates, the working status data of the touch screen can reflect the user's operation changes in real time, thereby achieving flexible power consumption management.

[0079] In one example, according to the power consumption characteristic parameters of the touch screen and the real-time working state data of the touch screen, power consumption prediction is performed through a first Bayesian probability model to obtain the power consumption control parameters of the first touch area, including:

[0080] Inputting the power consumption characteristic parameters of the touch screen and the real-time working state data of the touch screen into the feature extraction layer of the first Bayesian probability model for feature decomposition to obtain a power consumption feature vector and a working state vector;

[0081] The power consumption feature vector and the working state vector are input into the conditional probability calculation layer of the first Bayesian probability model for joint probability analysis to obtain a power consumption state prior probability matrix;

[0082] Performing maximum a posteriori probability calculation on the power consumption state prior probability matrix to obtain the power consumption state posterior probability matrix;

[0083] The power consumption state posterior probability matrix is ​​input into the parameter inference layer of the first Bayesian probability model to calculate the power consumption parameters, and obtain the regional power consumption control vector;

[0084] Constructing a power consumption control parameter mapping matrix based on the regional power consumption control vector, and inputting the power consumption control parameter mapping matrix into the constraint optimization layer of the first Bayesian probability model for parameter constraint processing to obtain constrained power consumption control parameters;

[0085] The constrained power consumption control parameters are subjected to time series smoothing to obtain smoothed power consumption control parameters, and the smoothed power consumption control parameters are input into the control strategy generation layer of the first Bayesian probability model for parameter integration to obtain the first touch area power consumption control parameters.

[0086] In this example, the power consumption characteristic parameters and real-time working status data of the touch screen are input into the feature extraction layer of the first Bayesian probability model, and feature decomposition is performed to obtain the power consumption feature vector and the working status vector. Describes the energy consumption characteristics of the touch screen under various conditions, such as operating frequency, operation intensity, etc. Real-time working status data It reflects the current user's interaction with the touch screen, such as the frequency of the current touch screen response area, the size of the touch pressure, etc. These parameters and data are used as input to the feature extraction layer. After feature decomposition, two independent feature vectors are obtained, namely, the power consumption feature vector and the working state vector . Power consumption characteristic vector It is expressed as:

[0087] ;

[0088] in Representative in the The power consumption characteristics in a specific scenario. Working state vector It is expressed as:

[0089] ;

[0090] in Indicates that the current touch screen is in The performance of each working state, such as the number of frequent touches, pressure intensity, etc. and the working state vector Input the conditional probability calculation layer of the first Bayesian probability model, perform joint probability analysis, and obtain the prior probability matrix of power consumption state The conditional probability calculation layer of the Bayesian probability model uses the Bayesian formula to calculate the joint probability under different states. The calculation formula is:

[0091] ;

[0092] in, Indicates power consumption characteristics The prior probability of Indicates that at a given power consumption characteristic In the case of By calculating the combination of all power consumption characteristics and working states, the power consumption state prior probability matrix is ​​obtained: , whose elements Indicates power consumption characteristics and working status The probability of joint occurrence. Prior probability matrix for power consumption state Perform maximum a posteriori probability calculation to obtain the power consumption state a posteriori probability matrix The calculation of the posterior probability is based on the Bayesian formula, which combines the observed data and prior information to obtain a more accurate probability distribution. The specific formula is:

[0093] ;

[0094] in, Indicates working status The likelihood function reflects the system's adaptability to power consumption characteristics under the currently observed working state. The denominator is the weighted sum of the likelihoods of all possible working states, which is used to normalize the posterior probability. Through this process, the posterior probability matrix of power consumption state is obtained. Represents the probability of various power consumption characteristics and working state combinations under current conditions. The power consumption state posterior probability matrix Input to the parameter inference layer of the Bayesian model to calculate the power consumption parameters and obtain the regional power consumption control vector The regional power consumption control vector is used to describe the specific power consumption control requirements of the touch screen in different areas. The formula is:

[0095] ;

[0096] in Indicates The power consumption control parameters of each independent touch area, such as supply voltage, sampling frequency and sensitivity, are calculated based on the posterior probability matrix The probability value of each area in the inference is used to reasonably allocate power consumption according to the usage frequency and status of different areas. , construct the power consumption control parameter mapping matrix . Mapping matrix The regional control vector is mapped to specific control parameters to adapt to different areas of the touch screen. The matrix is ​​represented as:

[0097] ;

[0098] in Indicates Control area control parameters, such as power supply voltage, sampling frequency, etc. The constructed power control parameter mapping matrix is ​​input into the constraint optimization layer of the Bayesian model to perform parameter constraint processing to obtain the constrained power control parameters, ensuring that the power control parameters meet the hardware constraints and operating conditions of the system. For example, the voltage range cannot exceed a certain threshold, and the sampling frequency should not exceed the processing capacity of the device. Through constraint optimization, the constrained power control parameters are obtained. The constrained power control parameters are time-smoothed to obtain smoothed power control parameters, reducing the fluctuations of the control parameters between different time points to ensure the stability of the system and the user's operating experience. The exponential smoothing method is used for processing, and the formula is:

[0099] ;

[0100] in, Represents the smoothed Power consumption control parameters, is a smoothing factor used to control the weight of the current value and the previous smoothed value. Through time series smoothing, the system instability caused by rapid power consumption changes is effectively eliminated, ensuring the smooth transition of the touch screen during power consumption adjustment. The smoothed power consumption control parameters are input into the control strategy generation layer of the first Bayesian probability model for parameter integration to obtain the power consumption control parameters of the first touch area. The control strategy generation layer generates a unified power consumption management strategy for the entire system based on the smoothed power consumption control parameters. This includes a comprehensive evaluation of the power consumption control parameters of different areas to ensure that the touch screen can achieve optimal power consumption efficiency during the overall operation. For example, if the touch frequency of a certain area is significantly reduced, the power supply voltage and sampling frequency of the area are reduced; on the contrary, if a certain area is frequently operated, the sensitivity and response speed are increased to ensure user experience.

[0101] In one example, the touch screen is divided into regions based on the power consumption control parameter of the first touch area to obtain multiple independent power consumption control blocks, and the power supply voltage parameters, sampling frequency parameters and sensitivity parameters of the independent power consumption control blocks are adjusted by region, and the power consumption switching data of each independent power consumption control block is collected, including:

[0102] Performing touch area similarity calculation based on the first touch area power consumption control parameter to obtain an area power consumption correlation matrix;

[0103] Based on the regional power consumption correlation matrix, dynamic clustering analysis is performed to obtain the initial regional division result, and the boundary of the initial regional division result is optimized to obtain multiple independent power consumption control blocks;

[0104] Deconstructing the power consumption control parameter of the first touch area according to a plurality of independent power consumption control blocks to obtain a block control parameter set, wherein the block control parameter set includes a power supply voltage range, a sampling frequency range, and a sensitivity range of each block;

[0105] Perform multi-objective optimization on the block control parameter set to obtain the block optimized control parameters;

[0106] Based on the block optimization control parameters, the power consumption of each area is adjusted to obtain the real-time power consumption data of the block, and the real-time power consumption data of the block is analyzed in time series to obtain the power consumption switching data of each independent power consumption control block.

[0107] In this example, the touch area similarity calculation is performed based on the first touch area power consumption control parameter, the similarity of different areas in power consumption characteristics is identified, and the regional power consumption correlation matrix is ​​obtained. Assume that the touch screen is divided into The power consumption control parameters of each area are ,in Indicates the supply voltage, represents the sampling frequency, Represents the sensitivity parameter. By calculating the similarity between different regions, the regional power consumption correlation matrix is ​​obtained: ,in Indicates Regions and The power consumption similarity between the regions. The similarity calculation is based on the Euclidean distance formula:

[0108] ;

[0109] in, The closer the value is to 1, the more similar the power consumption characteristics of the two regions are. Perform dynamic clustering analysis to obtain the initial regional division results. Group similar areas together so that these areas can be managed uniformly. Use the K-Means clustering algorithm to achieve the initial regional division. Assume that the result after clustering is ,in is the number of blocks after clustering, each Represents a group of regions with similar power consumption. The initial region division results are optimized to ensure that each independent power control block is as compact as possible in physical location to reduce unnecessary energy consumption caused by cross-region power consumption adjustment. The boundary optimization method based on gradient descent is used to adjust the boundary of each region so that the total power consumption of each block is minimized and the boundary is as smooth as possible. After boundary optimization, multiple independent power control blocks are obtained. , each block contains one or more initial touch areas. The power consumption control parameters of the first touch area are deconstructed according to the independent power consumption control blocks to obtain a block control parameter set. The block control parameter set includes the power supply voltage range, sampling frequency range and sensitivity range of each block, which are recorded as ,in , and Respectively represent The range of the supply voltage, sampling frequency and sensitivity parameters of each block. The deconstruction of these parameters helps to perform fine-grained power management for each block, ensuring that the power consumption control of different blocks is in line with their actual usage. Multi-objective optimization is performed on the block control parameter set to obtain the block optimization control parameters. To strike a balance between power consumption, performance and user experience, the optimization objective function is expressed as:

[0110] ;

[0111] in, They are the weight coefficients of supply voltage, sampling frequency and sensitivity, which are used to control the importance of each objective in the overall optimization. Through multi-objective optimization, the optimal control parameters of each block are determined , these parameters can reduce power consumption as much as possible while satisfying user experience, thus achieving the goal of energy saving. Based on the block optimization control parameters, the power consumption of each block is adjusted in different regions to obtain the real-time power consumption data of the block. The real-time power consumption data of each block is dynamically adjusted according to factors such as the current touch frequency and touch pressure. For example, if the user's touch frequency is detected in a certain block If it increases significantly, the sampling frequency of this block will be increased. This improves response speed and ensures user experience. Real-time power consumption data is expressed as:

[0112] ;

[0113] in, Indicates time, , and Respectively represent Blocks in time The power supply voltage, sampling frequency and sensitivity of each independent power control block are analyzed in time series to obtain power switching data. The power consumption change trend of each block is captured so that power switching can be performed at the appropriate time. For example, the power consumption change rate can be calculated by sliding window averaging the real-time power consumption data of each block. , the formula is:

[0114] ;

[0115] in, Indicates at time Moment, The power consumption change rate of each block, is the length of the sliding window. If If the value of is greater than a preset threshold, it means that the power consumption of the block is changing rapidly and further power consumption optimization is required, such as reducing the supply voltage or reducing the sampling frequency to balance energy consumption and performance.

[0116] In one example, the power switching data is input into a linear feedback control model for dynamic response analysis to obtain transition phase control parameters and stable phase control parameters, including:

[0117] The power consumption switching data is input into the state observer layer of the linear feedback control model for state decomposition to obtain a power consumption state vector;

[0118] The power consumption state vector is input into the error calculation layer of the linear feedback control model for error analysis to obtain the power consumption error matrix;

[0119] Perform feedback gain matrix calculation on the power consumption error matrix to obtain control gain parameters;

[0120] Input the control gain parameter and the power consumption error matrix into the compensation control layer of the linear feedback control model to perform error compensation, and obtain a compensation control signal;

[0121] Performing phase margin analysis based on the compensation control signal to obtain system stability parameters, and classifying the system stability parameters into states to obtain a transition state parameter set and a stable state parameter set;

[0122] Input the transition state parameter set into the dynamic response optimization layer of the linear feedback control model for parameter optimization to obtain the control parameters of the transition stage;

[0123] The steady-state parameter set is input into the steady-state control optimization layer of the linear feedback control model for parameter optimization to obtain the control parameters of the stable stage.

[0124] In this example, the power switching data is input into the state observer layer of the linear feedback control model for state decomposition to obtain the power state vector. The power switching data refers to the energy consumption changes when the touch screen switches between different working states. These data contain real-time power consumption information about the touch screen. The state observer layer decomposes and processes these complex power switching data, extracts the state characteristics of the system, and forms a power state vector. The power state vector is expressed in mathematical form as:

[0125] ;

[0126] in, Indicates The power consumption state of the touch screen at a certain moment, such as the energy consumption of the touch area, the power supply voltage, etc. The error calculation layer is input into the linear feedback control model for error analysis. The error calculation layer calculates the power consumption error of the system by comparing the current power consumption state with the target power consumption state. The target power consumption state is usually the ideal state predetermined by the system during design to ensure that the power consumption of the touch screen fluctuates within a reasonable range, thereby achieving the purpose of energy saving. The formula for error calculation is expressed as:

[0127] ;

[0128] in, represents the target power consumption state vector, Represents the power consumption error vector. Each element Indicates the deviation between the current system power consumption state and the target state. If the power consumption of a certain area is too high or too low, the corresponding element in the error vector will show a large deviation. The feedback gain matrix is ​​calculated to obtain the control gain parameters. The feedback gain parameters are used to determine the response strength of the system to the power consumption error. The formula for calculating the feedback gain matrix is ​​expressed as:

[0129] ;

[0130] in, represents the feedback gain matrix, is the covariance matrix of the input signal, is the system input matrix, is the state covariance matrix. By solving the feedback gain matrix, the required response strength for each power consumption error term is obtained. The purpose of the feedback gain parameter is to enable the system to quickly correct the error by adjusting the control amount, so as to achieve the predetermined power consumption state. After that, the control gain parameters and the power consumption error matrix are input into the compensation control layer of the linear feedback control model for error compensation. The compensation control layer calculates the compensation control signal according to the feedback gain and the size of the error, so as to adjust the working state of the system to make it as close as possible to the target power consumption state. The calculation formula of the compensation control signal is:

[0131] ;

[0132] in, represents the compensation control signal vector, is the feedback gain matrix, is the power consumption error vector. The negative sign indicates that the control is negative feedback, which reduces the power consumption error of the system by correcting the power consumption deviation. For example, if the power consumption of a certain area is significantly higher than the target value, the compensation signal will reduce the supply voltage or sampling frequency of the area to reduce power consumption. Phase margin analysis is performed based on the compensation control signal to obtain the stability parameters of the system. Evaluate the stability of the system to ensure that the system can operate stably without oscillation under different operating conditions. Phase margin is an important indicator to measure the system's ability to resist external disturbances, and is calculated by the following formula:

[0133] ;

[0134] in, represents the open-loop transfer function of the system, is the transfer function of the feedback link, is the phase margin. By calculating the phase margin, the stability parameters of the system during the power consumption control process are obtained. The stability parameters of the system are classified into states to obtain a transition state parameter set and a stable state parameter set. According to the stability of the system in different time periods, the parameters of the transition stage and the stable stage are divided. For example, when the system has just received a new touch operation, it is in a transition state. At this time, the control goal is to respond quickly to the user's operation. After the system enters a stable operation, the control goal becomes to maintain a low power consumption state. Through the analysis of the phase margin, the stability parameters are classified into a transition state parameter set. and the steady-state parameter set . Set the transition state parameter set Input to the dynamic response optimization layer of the linear feedback control model for parameter optimization to obtain the control parameters of the transition phase. The dynamic response optimization layer optimizes the control parameters based on the characteristics of the transition phase so that the system can quickly reach the target power consumption state in a shorter time. For example, the optimization goal is expressed as minimizing the transition time or reducing the power consumption oscillation amplitude, and the optimization problem is expressed as:

[0135] ;

[0136] in, represents the performance indicator function, represents the feedback gain in the transition phase, Indicates that the system is at time The state of the moment, is the target state. By solving this optimization problem, we can obtain the control parameters that make the system respond fastest and oscillate least in the transition phase. Similarly, we can set the stable state parameter set to Input to the steady-state control optimization layer of the linear feedback control model for parameter optimization to obtain the control parameters in the stable stage, ensuring that the system maintains the lowest power consumption after reaching a stable state while maintaining good response characteristics. The optimization problem is expressed as:

[0137] ;

[0138] in, Indicates that the system is at time The actual power consumption at the moment, Represents the target power consumption. By solving the steady-state optimization problem, the control parameters that minimize power consumption in long-term operation are obtained. .

[0139] In one example, adaptive optimization is performed on the transition phase control parameter and the stable phase control parameter to obtain the second touch area power consumption control parameter, and the second touch area power consumption control parameter is fed back to the first Bayesian probability model to obtain the second Bayesian probability model, including:

[0140] Based on the transition phase control parameters and the stable phase control parameters, a fusion analysis is performed to obtain a control parameter fusion matrix, which includes a dynamic adjustment coefficient and a steady-state maintenance coefficient.

[0141] The objective function is constructed for the control parameter fusion matrix to obtain the optimization objective function, which includes dynamic response constraint items and steady-state performance constraint items;

[0142] Iterative calculation is performed based on the optimization objective function to obtain an optimization parameter sequence, and parameter smoothing is performed based on the optimization parameter sequence to obtain smoothed optimization parameters;

[0143] Performing control parameter mapping on the smoothing optimization parameter to obtain a second touch area power consumption control parameter, wherein the second touch area power consumption control parameter includes an optimized power supply voltage value, a sampling frequency value, and a touch sensitivity value;

[0144] Based on the second touch area power consumption control parameter, the model parameters are updated to obtain the Bayesian model update parameters, where the Bayesian model update parameters include a priori probability update value and a conditional probability update value;

[0145] The probability distribution of the first Bayesian probability model is updated based on the Bayesian model update parameters to obtain an updated conditional probability distribution, and the updated conditional probability distribution is reconstructed to obtain a second Bayesian probability model.

[0146] In this example, a fusion analysis is performed based on the transition phase control parameters and the stable phase control parameters to obtain a control parameter fusion matrix. and stable phase control parameters Describe the control requirements of the system in different states respectively, where the transition phase focuses on fast response, while the stable phase focuses on maintaining low power consumption and stability. Combine the control strategies of these two stages to achieve balanced control goals in different usage scenarios. Control parameter fusion matrix It is expressed as:

[0147] ;

[0148] in, It represents the dynamic adjustment coefficient, which reflects the adjustment strength of the system in the transition stage. Represents the steady-state maintenance coefficient, which reflects the system's ability to maintain energy consumption in the stable stage. By integrating the parameters of the transition and stable stages, it ensures that the system strikes a balance between dynamic response and steady-state performance. The objective function of the control parameter fusion matrix is ​​constructed to obtain the optimization objective function. The optimization objective function consists of two main parts: dynamic response constraints and steady-state performance constraints. The goal of the dynamic response constraint is to minimize the response time of the system during state changes to ensure that the user's operations on the touch screen can respond quickly; while the goal of the steady-state performance constraint is to minimize the power consumption of the system in a stable state to extend the battery life of the device. The optimization objective function is expressed as:

[0149] ;

[0150] in, represents the optimization objective function, and is the weight coefficient of dynamic response and steady-state performance, Indicates that the system is at time The state of the moment, is the target state, Indicates the power consumption of the system in the stable stage. By adjusting the weight coefficient and , balance the relationship between response speed and power consumption. Perform iterative calculation based on the optimization objective function to obtain the optimization parameter sequence. Through multiple optimizations, the control parameters are continuously adjusted to gradually approach the optimal solution. Each iteration is adjusted based on the previous parameters, and the formula is expressed as:

[0151] ;

[0152] in, Indicates The control parameters for the iterations are is the learning rate, which controls the step size of each iteration, is the gradient of the optimization objective function with respect to the control parameters. Through multiple iterations, the optimization parameter sequence is obtained. , until the objective function converges to a minimum value. The result of the optimization is a control parameter sequence that gradually approaches the optimal value, ensuring that the system can achieve good energy efficiency performance in both the transition and stable stages. The optimized parameter sequence is smoothed to reduce the drastic changes in the control parameters at different times, so as to ensure the stability of the system operation and the user's operating experience, and obtain smoothed optimized parameters. The control parameters are smoothed using the exponential smoothing method, and the formula is as follows:

[0153] ;

[0154] in, represents the smoothed control parameter, is a smoothing factor that controls the weight between the current value and the previous value. Through smoothing, the short-term fluctuations caused by the iterative optimization process are effectively eliminated, making the control parameters smoother and more stable in time. The control parameters are mapped to the smoothing optimization parameters to obtain the power consumption control parameters of the second touch area. The optimized control parameters are converted into specific power consumption control strategies, including the optimized supply voltage value. , sampling frequency value and touch sensitivity value The control parameter mapping is performed in a linear mapping manner, and the formula is as follows:

[0155] ;

[0156] in, represents the power consumption control parameter vector, is the mapping matrix, and the smoothed control parameters Mapped to power supply voltage, sampling frequency and sensitivity value. Through the mapping process, the power consumption control parameters of the second touch area are obtained. The model parameters are updated based on the power consumption control parameters of the second touch area to obtain the Bayesian model update parameters. The Bayesian model update parameters include a priori probability update value and a conditional probability update value. The purpose of the priori probability update is to enable the Bayesian model to reflect the latest power consumption control strategy. The priori probability update formula is expressed as:

[0157] ;

[0158] in, represents the updated prior probability, is the old prior probability, is the observed power consumption state probability, is the update factor. Through the update method, the prior probability can reflect the actual power consumption state of the current system. The update of conditional probability is to adjust the conditional relationship between different states in the Bayesian model according to the newly observed data. The conditional probability update is performed by the following formula:

[0159] ;

[0160] in, represents the updated conditional probability, is the likelihood function of the observed data, is the updated prior probability. By updating the conditional probability, the Bayesian model can more accurately reflect the current system status and improve the accuracy of power consumption management. Based on the Bayesian model update parameters, the probability distribution of the first Bayesian probability model is updated to obtain an updated conditional probability distribution, and the updated conditional probability distribution is reconstructed to obtain a second Bayesian probability model. The purpose of model reconstruction is to make the Bayesian model more adaptable to the current system status and power consumption control strategy, ensuring that the model can more accurately reflect system changes in future predictions. Through model reconstruction, the second Bayesian probability model can combine the updated prior probability and conditional probability to provide more reliable power consumption prediction and control support.

[0161] Reference Figure 2 This embodiment provides a touch screen low power consumption mode switching device, including:

[0162] Feature extraction module 1, used to perform time-frequency analysis and feature extraction on touch signals of the touch screen in different usage scenarios to obtain power consumption characteristic parameters of the touch screen;

[0163] A real-time analysis module 2 is used to collect touch position data, touch pressure data and touch time data through multiple touch sensors in the touch screen, and perform real-time analysis to obtain real-time working status data of the touch screen;

[0164] The power consumption prediction module 3 is used to predict the power consumption through a first Bayesian probability model according to the power consumption characteristic parameters of the touch screen and the real-time working state data of the touch screen, and obtain the power consumption control parameters of the first touch area;

[0165] A regional adjustment module 4 is used to divide the touch screen into regions based on the power consumption control parameter of the first touch area to obtain multiple independent power consumption control blocks, and to adjust the power supply voltage parameters, sampling frequency parameters and sensitivity parameters of the independent power consumption control blocks in different regions, and to collect power consumption switching data of each independent power consumption control block;

[0166] A dynamic response analysis module 5 is used to input the power consumption switching data into a linear feedback control model for dynamic response analysis to obtain transition phase control parameters and stable phase control parameters;

[0167] The adaptive optimization module 6 is used to adaptively optimize the transition phase control parameters and the stable phase control parameters to obtain the second touch area power consumption control parameters, and feed the second touch area power consumption control parameters back to the first Bayesian probability model to obtain the second Bayesian probability model.

[0168] In this embodiment, for the specific implementation of each unit in the above device embodiment, please refer to the above method embodiment, which will not be repeated here.

[0169] Reference Figure 3 In an embodiment of the present invention, a computer device is also provided. The computer device may be a server, and its internal structure may be as follows: Figure 3 As shown. The computer device includes a processor, a memory, a display screen, an input device, a network interface and a database connected through a system bus. Among them, the processor designed by the computer is used to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system, a computer program and a database. The internal memory provides an environment for the operation of the operating system and the computer program in the non-volatile storage medium. The database of the computer device is used to store the corresponding data in this embodiment. The network interface of the computer device is used to communicate with an external terminal through a network connection. When the computer program is executed by the processor, the above method is implemented.

[0170] Those skilled in the art will understand that Figure 3 The structure shown in the figure is merely a block diagram of a portion of the structure related to the solution of the present invention, and does not constitute a limitation on the computer device to which the solution of the present invention is applied.

[0171] An embodiment of the present invention further provides a computer-readable storage medium on which a computer program is stored, and when the computer program is executed by a processor, the above method is implemented. It can be understood that the computer-readable storage medium in this embodiment can be a volatile readable storage medium or a non-volatile readable storage medium.

[0172] Those skilled in the art can understand that all or part of the processes in the above-mentioned embodiment methods can be completed by instructing the relevant hardware through a computer program, and the computer program can be stored in a non-volatile computer-readable storage medium. When the computer program is executed, it can include the processes of the embodiments of the above-mentioned methods. Among them, any reference to memory, storage, database or other media provided by the present invention and used in the embodiments may include non-volatile and / or volatile memory. Non-volatile memory may include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM) or flash memory. Volatile memory may include random access memory (RAM) or external cache memory. As an illustration and not limitation, RAM is available in a variety of forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), double-speed data rate SDRAM (SSRSDRAM), enhanced SDRAM (ESDRAM), synchronous link (Synchlink) DRAM (SLDRAM), memory bus (Rambus) direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM), and memory bus dynamic RAM.

[0173] It should be noted that, in this article, the terms "include", "comprises" or any other variations thereof are intended to cover non-exclusive inclusion, so that a process, device, article or method including a series of elements includes not only those elements, but also includes other elements not explicitly listed, or also includes elements inherent to such process, device, article or method. In the absence of further restrictions, an element defined by the sentence "includes a ..." does not exclude the presence of other identical elements in the process, device, article or method including the element.

[0174] The above description is only a preferred embodiment of the present invention, and does not limit the patent scope of the present invention. Any equivalent structure or equivalent process transformation made by using the contents of the present invention specification and drawings, or directly or indirectly applied in other related technical fields, are also included in the patent protection scope of the present invention.

Claims

1. A method for switching a touch screen to a low power consumption mode, characterized in that: The following steps are involved: Perform time-frequency analysis and feature extraction on the touch signals of the touch screen in different usage scenarios to obtain the power consumption characteristic parameters of the touch screen; Collecting touch position data, touch pressure data and touch time data through multiple touch sensors in the touch screen, and performing real-time analysis to obtain real-time working status data of the touch screen; According to the power consumption characteristic parameters of the touch screen and the real-time working state data of the touch screen, power consumption prediction is performed through a first Bayesian probability model to obtain a first touch area power consumption control parameter; Based on the first touch area power consumption control parameter, the touch screen is divided into regions to obtain a plurality of independent power consumption control blocks, and the power supply voltage parameter, sampling frequency parameter and sensitivity parameter of the independent power consumption control block are adjusted by region, and the power consumption switching data of each independent power consumption control block is collected; specifically, the method comprises: performing touch area similarity calculation based on the first touch area power consumption control parameter to obtain a regional power consumption correlation matrix; performing dynamic clustering analysis based on the regional power consumption correlation matrix to obtain an initial region division result, and performing boundary optimization on the initial region division result to obtain a plurality of independent power consumption control blocks; performing parameter deconstruction on the first touch area power consumption control parameter according to the plurality of independent power consumption control blocks to obtain a block control parameter set, and the block control parameter set includes a power supply voltage range, a sampling frequency range and a sensitivity range of each block; performing multi-objective optimization on the block control parameter set to obtain a block optimization control parameter; performing regional power consumption adjustment based on the block optimization control parameter to obtain real-time power consumption data of the block, and performing time series analysis on the real-time power consumption data of the block to obtain power consumption switching data of each independent power consumption control block; Inputting the power consumption switching data into a linear feedback control model for dynamic response analysis to obtain transition phase control parameters and stable phase control parameters; The transition phase control parameters and the stable phase control parameters are adaptively optimized to obtain a second touch area power consumption control parameter, and the second touch area power consumption control parameter is fed back to the first Bayesian probability model to obtain a second Bayesian probability model, wherein the second touch area power consumption control parameter includes an optimized supply voltage value, a sampling frequency value, and a touch sensitivity value.

2. The touch screen low power consumption mode switching method according to claim 1, characterized in that: The touch signal of the touch screen in different usage scenarios is subjected to time-frequency analysis and feature extraction to obtain the power consumption characteristic parameters of the touch screen, including: Standardize the touch position data, touch pressure data, and touch time data collected by the touch screen in different usage scenarios to obtain a standardized touch signal sequence; Performing time domain decomposition on the standardized touch signal sequence to obtain a time domain feature vector of the touch signal, wherein the time domain feature vector includes touch duration data, touch interval data, and touch intensity data; Inputting the time domain feature vector into a spectrum analysis model for Fourier transform to obtain a frequency domain feature vector of the touch signal, wherein the frequency domain feature vector includes signal main frequency data and frequency distribution data; Performing feature correlation on the time domain feature vector and the frequency domain feature vector to obtain a natural frequency parameter of the touch screen, wherein the natural frequency parameter of the touch screen represents the natural working characteristics of the touch screen in different usage scenarios; Performing dynamic threshold calculation on the natural frequency parameters of the touch screen to obtain dynamic power consumption parameters, and performing coefficient calculation based on the dynamic power consumption parameters to obtain time-frequency coefficients; Data fusion processing is performed on the touch screen natural frequency parameter, the dynamic power consumption parameter and the time-frequency coefficient to obtain initial power consumption characteristic parameters, and parameter correction is performed on the initial power consumption characteristic parameters to obtain touch screen power consumption characteristic parameters.

3. The touch screen low power consumption mode switching method according to claim 2, characterized in that: The touch position data, touch pressure data and touch time data are collected by multiple touch sensors in the touch screen, and real-time analysis is performed to obtain real-time working status data of the touch screen, including: Classifying the touch data collected by the multiple touch sensors in the touch screen to obtain an initial data set, wherein the initial data set includes touch position data, touch pressure data, and touch time data; Performing data cleaning and normalization processing on the initial data set to obtain a standardized data set, and performing spatial distribution calculation based on the standardized data set to obtain a touch area distribution matrix, wherein the touch area distribution matrix reflects the spatial aggregation degree of touch operations; Performing time series analysis on the standardized data set to obtain a touch frequency matrix, wherein the touch frequency matrix includes single-point touch frequency and multi-point touch frequency; Performing pattern analysis based on the touch area distribution matrix and the touch frequency matrix to obtain touch pattern feature data, and dividing the touch pattern feature data into time windows to obtain touch feature sequences of multiple time segments; A real-time status evaluation is performed based on the touch feature sequence to obtain initial working status data, and the initial working status data is dynamically updated to obtain real-time working status data of the touch screen.

4. The touch screen low power consumption mode switching method according to claim 3, characterized in that: The method of performing power consumption prediction by using a first Bayesian probability model according to the power consumption characteristic parameter of the touch screen and the real-time working state data of the touch screen to obtain the power consumption control parameter of the first touch area includes: Inputting the power consumption characteristic parameter of the touch screen and the real-time working state data of the touch screen into the feature extraction layer of the first Bayesian probability model for feature decomposition to obtain a power consumption feature vector and a working state vector; Inputting the power consumption feature vector and the working state vector into the conditional probability calculation layer of the first Bayesian probability model for joint probability analysis to obtain a power consumption state prior probability matrix; Performing maximum a posteriori probability calculation on the power consumption state prior probability matrix to obtain a power consumption state posterior probability matrix; Inputting the power consumption state posterior probability matrix into the parameter inference layer of the first Bayesian probability model to calculate the power consumption parameters, and obtaining a regional power consumption control vector; Constructing a power consumption control parameter mapping matrix based on the regional power consumption control vector, and inputting the power consumption control parameter mapping matrix into the constraint optimization layer of the first Bayesian probability model for parameter constraint processing to obtain constrained power consumption control parameters; The constrained power consumption control parameter is subjected to time series smoothing to obtain a smoothed power consumption control parameter, and the smoothed power consumption control parameter is input into the control strategy generation layer of the first Bayesian probability model for parameter integration to obtain a first touch area power consumption control parameter.

5. The touch screen low power consumption mode switching method according to claim 1, characterized in that: The step of inputting the power consumption switching data into a linear feedback control model for dynamic response analysis to obtain transition phase control parameters and stable phase control parameters includes: Inputting the power consumption switching data into the state observer layer of the linear feedback control model for state decomposition to obtain a power consumption state vector; Inputting the power consumption state vector into the error calculation layer of the linear feedback control model for error analysis to obtain a power consumption error matrix; Performing feedback gain matrix calculation on the power consumption error matrix to obtain control gain parameters; Inputting the control gain parameter and the power consumption error matrix into the compensation control layer of the linear feedback control model for error compensation to obtain a compensation control signal; Performing phase margin analysis based on the compensation control signal to obtain system stability parameters, and classifying the system stability parameters into states to obtain a transition state parameter set and a stable state parameter set; Inputting the transition state parameter set into the dynamic response optimization layer of the linear feedback control model for parameter optimization to obtain transition stage control parameters; The steady-state parameter set is input into the steady-state control optimization layer of the linear feedback control model for parameter optimization to obtain the steady-state control parameters.

6. The touch screen low power consumption mode switching method according to claim 5, characterized in that: The adaptively optimizing the transition phase control parameter and the stable phase control parameter to obtain a second touch area power consumption control parameter, and feeding back the second touch area power consumption control parameter to the first Bayesian probability model to obtain a second Bayesian probability model, includes: Perform fusion analysis based on the transition phase control parameters and the stable phase control parameters to obtain a control parameter fusion matrix, wherein the control parameter fusion matrix includes a dynamic adjustment coefficient and a steady-state maintenance coefficient; Constructing an objective function for the control parameter fusion matrix to obtain an optimization objective function, wherein the optimization objective function includes a dynamic response constraint term and a steady-state performance constraint term; Performing iterative calculation based on the optimization objective function to obtain an optimization parameter sequence, and performing parameter smoothing processing based on the optimization parameter sequence to obtain smoothed optimization parameters; Performing control parameter mapping on the smoothing optimization parameter to obtain a second touch area power consumption control parameter; Based on the second touch area power consumption control parameter, the model parameter is updated to obtain a Bayesian model update parameter, wherein the Bayesian model update parameter includes a priori probability update value and a conditional probability update value; The probability distribution of the first Bayesian probability model is updated based on the Bayesian model update parameters to obtain an updated conditional probability distribution, and the updated conditional probability distribution is reconstructed to obtain a second Bayesian probability model.

7. A touch screen low power consumption mode switching device, characterized in that: For implementing the steps of the method according to any one of claims 1 to 6, the device comprises: A feature extraction module is used to perform time-frequency analysis and feature extraction on the touch signal of the touch screen in different usage scenarios to obtain the power consumption characteristic parameters of the touch screen; A real-time analysis module, used to collect touch position data, touch pressure data and touch time data through multiple touch sensors in the touch screen, and perform real-time analysis to obtain real-time working status data of the touch screen; A power consumption prediction module, configured to predict the power consumption through a first Bayesian probability model according to the power consumption characteristic parameters of the touch screen and the real-time working state data of the touch screen, and obtain a power consumption control parameter of a first touch area; The sub-region adjustment module is used to divide the touch screen into regions based on the first touch area power consumption control parameter to obtain multiple independent power consumption control blocks, and to adjust the power supply voltage parameter, sampling frequency parameter and sensitivity parameter of the independent power consumption control block by region, and collect the power consumption switching data of each independent power consumption control block; specifically including: performing touch area similarity calculation based on the first touch area power consumption control parameter to obtain a regional power consumption correlation matrix; performing dynamic clustering analysis based on the regional power consumption correlation matrix to obtain an initial region division result, and performing boundary optimization on the initial region division result to obtain multiple independent power consumption control blocks; performing parameter deconstruction on the first touch area power consumption control parameter according to the multiple independent power consumption control blocks to obtain a block control parameter set, and the block control parameter set includes a power supply voltage range, a sampling frequency range and a sensitivity range of each block; performing multi-objective optimization on the block control parameter set to obtain block optimization control parameters; performing sub-region power consumption adjustment based on the block optimization control parameters to obtain block real-time power consumption data, and performing time series analysis on the block real-time power consumption data to obtain power consumption switching data of each independent power consumption control block; A dynamic response analysis module, used for inputting the power consumption switching data into a linear feedback control model for dynamic response analysis to obtain transition phase control parameters and stable phase control parameters; An adaptive optimization module is used to adaptively optimize the transition phase control parameters and the stable phase control parameters to obtain a second touch area power consumption control parameter, and feed the second touch area power consumption control parameter back to the first Bayesian probability model to obtain a second Bayesian probability model, wherein the second touch area power consumption control parameter includes an optimized supply voltage value, a sampling frequency value, and a touch sensitivity value.

8. A computer device comprising a memory and a processor, wherein a computer program is stored in the memory, wherein: When the processor executes the computer program, the steps of the method according to any one of claims 1 to 6 are implemented.

9. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the steps of the method according to any one of claims 1 to 6 are implemented.

Images