Mobile phone power consumption optimization method and device, equipment and storage medium

Abstract

The invention relates to a mobile phone power consumption optimization method and device, equipment and a storage medium, and the method comprises the following steps: carrying out the multi-dimensional real-time collection of the operation data of a target mobile phone, and obtaining the power consumption scene data; performing operation time sequence analysis on the target mobile phone based on the power consumption scene data to obtain a power consumption state transition chain; performing weight calculation on the power consumption state transition chain through a mixed entropy weight method to obtain a scene priority table; performing power consumption planning solution on the target mobile phone based on the scene priority table to obtain a power consumption scheduling strategy; analyzing the power consumption scheduling strategy step by step to obtain a control instruction sequence, and dynamically adjusting a control unit of the target mobile phone in real time based on the control instruction sequence to realize optimal balance between power consumption and performance. The technical problem that an existing power consumption optimization strategy often neglects relevance between different application scenes and dynamic changes of the application scenes along with time lapse, and consequently efficiency is possibly low is solved.

Description

Technical Field

[0001] The present invention relates to the technical field of mobile phones, and particularly to a method, device, equipment and storage medium for optimizing the power consumption of mobile phones. Background Art

[0002] With the rapid development of mobile technology, smart phones have become an indispensable part of people's daily lives. However, although the functions of mobile phones are becoming increasingly rich and powerful, their battery life has not kept up with the improvement of hardware performance, which has become an important factor restricting the user experience. The development of battery technology is relatively slow, while users' demands for device performance and function diversity are constantly increasing, resulting in the urgent problem of how to improve the energy efficiency ratio of mobile phones without increasing the battery volume or weight. Therefore, it has become particularly crucial to study a method that can effectively optimize the power consumption of mobile phones.

[0003] Most of the power consumption management methods of current mobile phones on the market are relatively simple, usually relying on preset modes or adjusting based on a limited number of parameters, such as screen brightness, CPU frequency, etc. Although this method can reduce power consumption to a certain extent, it is not flexible enough in complex application scenarios and cannot accurately match the changing demands in actual use. In addition, existing power consumption optimization strategies often ignore the relevance between different application scenarios and their dynamic changes over time, which may lead to low efficiency or damaged user experience. To overcome these problems, it is necessary to explore a more intelligent and adaptable power consumption optimization method.

[0004] For this reason, a method for optimizing the power consumption of mobile phones based on multi-dimensional real-time collection of operation data and combined with operation time series analysis is proposed. This method not only considers the immediate power consumption state, but also predicts the possible future change trends by constructing a power consumption state transition chain, and determines the priorities in different scenarios using the hybrid entropy weight method, finally forming a set of accurate power consumption scheduling strategies. The purpose of this method is to maximize the battery life of mobile phones through intelligent power consumption management without affecting or even improving the user experience, and provide more stable and reliable services for users. At the same time, this also represents an important transformation of mobile phone power consumption management technology from static to dynamic and from single to comprehensive. Summary of the Invention

[0005] The main object of the present invention is to provide a method, device, equipment and storage medium for optimizing the power consumption of mobile phones, which solves the technical problem that existing power consumption optimization strategies often ignore the relevance between different application scenarios and their dynamic changes over time, which may lead to low efficiency.

[0006] To achieve the above object, the present invention provides a method for optimizing the power consumption of mobile phones, including the following steps: Multidimensionally and real-time collect the running data of the target mobile phone to obtain power consumption scenario data; Based on the power consumption scenario data, perform a running timing analysis on the target mobile phone to obtain a power consumption state migration chain; Calculate the weights of the power consumption scenario data in the power consumption state migration chain through the hybrid entropy weight method to obtain a scenario priority table; Based on the scenario priority table, perform a power consumption planning solution on the target mobile phone to obtain a power consumption scheduling strategy; Perform a step-by-step analysis on the power consumption scheduling strategy to obtain a control instruction sequence, and based on the control instruction sequence, dynamically adjust the control unit of the target mobile phone in real time to achieve the optimal balance between power consumption and performance.

[0007] Further, the multidimensional real-time collection of the running data of the target mobile phone to obtain power consumption scenario data includes: Real-time collect the hardware parameters of the target mobile phone through a preset sensor group to obtain an original power consumption data set; wherein, the original power consumption data set includes CPU power consumption data, GPU power consumption data, memory power consumption data, and battery temperature; Based on the original power consumption data set, perform a hardware resource tracking analysis on the target mobile phone to obtain a resource dynamic load vector; Extract the power consumption scenario features of the resource dynamic load vector through a multidimensional data fusion algorithm to obtain power consumption scenario data; wherein, the power consumption scenario data includes game scenarios, video scenarios, web browsing scenarios, and standby scenarios.

[0008] Further, the running timing analysis on the target mobile phone based on the power consumption scenario data to obtain a power consumption state migration chain includes: Based on discrete wavelet analysis, perform timing segmentation on the power consumption scenario data to obtain a scenario time series, and perform a Fourier transform on the scenario time series to obtain scenario spectrum features; Perform state recognition on the scenario spectrum features through adaptive edge detection to obtain a scenario state sequence, and perform hierarchical clustering on the scenario state sequence to obtain a state classification result; wherein, the state classification result includes a high-frequency switching state, a stable running state, and a low-frequency switching state; Perform a time-dependence analysis on the state classification result based on a recurrent neural network to obtain a state transition probability, and perform Bayesian state inference on the target mobile phone based on the state transition probability to obtain a state prediction sequence; wherein, the state prediction sequence includes scenario duration, switching time points, and transition directions; Calculate the similarity of the state prediction sequence through the dynamic time warping algorithm to obtain a state distance matrix, and perform spectral clustering on the state distance matrix to obtain a set of state clusters, where the set of state clusters includes a high power consumption cluster, a medium power consumption cluster, and a low power consumption cluster; Analyze the transition rules of the set of state clusters based on the Markov decision process to obtain a state transition map, and perform path extraction and path optimization based on the state transition map to obtain an optimized transition path; where the optimized transition path includes the shortest path, the optimal switching point, and the transition cost; Construct a link based on the optimized transition path through a graph neural network to obtain a power consumption state migration chain.

[0009] Furthermore, calculate the weights of the power consumption scenario data in the power consumption state migration chain through the hybrid entropy weight method to obtain a scenario priority table, including: Extract the features of the power consumption scenario data in the power consumption state migration chain based on the information entropy algorithm to obtain a state entropy value sequence, and perform normalization processing on the state entropy value sequence to obtain a normalized entropy value matrix; Allocate weights to the normalized entropy value matrix through a multi-criteria decision analysis method to obtain an initial weight vector, and perform constraint optimization on the initial weight vector to obtain a tuned weight set; Perform hierarchical analysis on the power consumption state migration chain based on the tuned weight set to obtain a scenario priority table; Calculate the grey correlation degree of the scenario priority table through grey correlation analysis to obtain a scenario correlation degree sequence, and perform orthogonal decomposition on the scenario correlation degree sequence to obtain scenario feature weights; where the scenario feature weights include scenario switching cost, resource consumption ratio, and performance impact factor; Rank the scenario feature weights based on the coefficient of variation method to obtain a scenario priority sequence, and detect whether there is a circular dependency conflict in the scenario priority sequence. If so, exclude the circular dependency conflict in the scenario priority sequence to obtain a conflict-free scenario set; Perform priority mapping on the power consumption scenario data based on the conflict-free scenario set through a multi-objective optimization algorithm to obtain a scenario priority table.

[0010] Furthermore, solve the power consumption planning of the target mobile phone based on the scenario priority table to obtain a power consumption scheduling strategy, including: Analyze the dependency relationship of the scenario priority table through a topological sorting algorithm to obtain a scenario scheduling map, and perform constraint propagation on the scenario scheduling map to obtain a scheduling constraint set; where the scheduling constraint set includes resource capacity constraints, scenario switching constraints, and timing dependency constraints; Solve and analyze the scheduling constraint set based on mixed integer programming to obtain an initial scheduling plan, and perform a spatial search on the initial scheduling plan through the branch and bound method to obtain a set of candidate plans; where the set of candidate plans includes a power consumption threshold interval, a switching time window, and a resource allocation ratio; Perform temperature regulation calculations on the set of candidate plans through the simulated annealing algorithm to obtain a cooling strategy sequence, and perform a solution space exploration based on the cooling strategy sequence to obtain a locally optimal solution set; where the locally optimal solution set includes power consumption control parameters, scheduling time points, and resource allocation plans; Perform a neighborhood traversal on the locally optimal solution set based on the tabu search algorithm to obtain a global search path, and perform a state transition calculation on the global search path through dynamic programming techniques to obtain an optimal scheduling sequence; where the optimal scheduling sequence includes a scenario switching order, a resource allocation strategy, and a power consumption control instruction; Perform a multi-objective trade-off on the optimal scheduling sequence through a cooperative optimization algorithm to obtain a scheduling evaluation index, and perform a plan screening based on the scheduling evaluation index to obtain an optimal scheduling plan; where the optimal scheduling plan includes a scenario execution duration, a resource usage efficiency, and a power consumption balance coefficient; Perform a population evolution calculation on the optimal scheduling plan based on the genetic algorithm to obtain an evolution plan sequence, and perform a non-dominated sorting on the evolution plan sequence through the Pareto optimal set to obtain a power consumption scheduling strategy; where the power consumption scheduling strategy includes a scenario scheduling time sequence table, a resource allocation matrix, and a set of power consumption control parameters.

[0011] Furthermore, the step-by-step analysis of the power consumption scheduling strategy to obtain a control instruction sequence, and the real-time dynamic adjustment of the control unit of the target mobile phone based on the control instruction sequence to achieve the optimal balance between power consumption and performance includes: Perform instruction parsing on the power consumption scheduling strategy through a lexical analyzer to obtain an instruction token sequence, and construct a syntax tree for the instruction token sequence to obtain a scheduling parse tree; Perform a dependency analysis on the scheduling parse tree based on a semantic analyzer to obtain a control dependency graph, and perform an instruction division on the control dependency graph to obtain an atomic instruction set; where the atomic instruction set includes a CPU frequency regulation instruction, a GPU voltage regulation instruction, and a memory bandwidth allocation instruction; Perform a parallelism analysis on the atomic instruction set to obtain an instruction scheduling table; where the instruction scheduling table includes critical path instructions, parallel execution instructions, and serial execution instructions; Perform a timing sorting on the instruction scheduling table based on a scheduling orchestrator to obtain an execution timing chain, and perform a priority division on the execution timing chain to obtain a hierarchical control sequence; Detect whether there is an abnormality in the hierarchical control sequence through a fault-tolerant processor. If so, extract the abnormal hierarchical control sequence from the hierarchical control sequence, remove the abnormal hierarchical control sequence from the hierarchical control sequence to obtain a hierarchical control sequence after removal, and perform sequence reconstruction on the hierarchical control sequence after removal to obtain a corrected hierarchical control sequence; Based on the corrected hierarchical control sequence, perform hardware mapping on the control unit of the target mobile phone to obtain a control instruction sequence, and dynamically adjust the control unit of the target mobile phone in real time based on the control instruction sequence to achieve the optimal balance between power consumption and performance.

[0012] Further, constructing a syntax tree for the instruction tag sequence to obtain a scheduling parsing tree includes: Perform syntax unit division on the instruction tag sequence through enhanced two-way syntax analysis to obtain a syntax unit set, and perform semantic annotation on the syntax unit set to obtain an annotated syntax unit table; Perform syntax rule matching on the annotated syntax unit table based on a bottom-up reduction algorithm to obtain a reduction sequence diagram; Perform semantic attribute calculation on the reduction sequence diagram through attribute grammar derivation to obtain a semantic attribute table, and perform dependency analysis on the semantic attribute table to obtain an attribute dependency network, where the attribute dependency network includes instruction execution attributes, resource consumption attributes, and timing constraint attributes; Perform network parsing and subtree generation on the attribute dependency network based on a recursive descent parser to obtain a subtree set, and perform pruning optimization on the subtree set to obtain an optimized subtree group; Perform tree structure merging on the optimized subtree group through a syntax forest synthesis algorithm to obtain a syntax forest, and perform state transition analysis on the syntax forest based on a tree automaton to obtain a scheduling parsing tree.

[0013] The present invention also provides a mobile phone power consumption optimization device, including: An acquisition module for performing multi-dimensional real-time acquisition on the operating data of the target mobile phone to obtain power consumption scenario data; An analysis module for performing operating timing analysis on the target mobile phone based on the power consumption scenario data to obtain a power consumption state transition chain; A calculation module for calculating the weights of the power consumption scenario data in the power consumption state transition chain through a hybrid entropy weight method to obtain a scenario priority table; A solution module for performing power consumption planning and solution on the target mobile phone based on the scenario priority table to obtain a power consumption scheduling strategy; A parsing module is used to perform step-by-step parsing on the power consumption scheduling policy to obtain a control instruction sequence, and based on the control instruction sequence, dynamically adjust the control unit of the target mobile phone in real time to achieve the optimal balance between power consumption and performance.

[0014] The present invention also provides a computer device, including a memory and a processor. A computer program is stored in the memory, and when the processor executes the computer program, the steps of the method described in any one of the above are implemented.

[0015] 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 the method described in any one of the above are implemented.

[0016] The mobile phone power consumption optimization method provided by the present invention includes the following steps: performing multi-dimensional real-time collection on the operation data of the target mobile phone to obtain power consumption scenario data; performing operation timing analysis on the target mobile phone based on the power consumption scenario data to obtain a power consumption state migration chain; calculating the weights of the power consumption scenario data in the power consumption state migration chain by the hybrid entropy weight method to obtain a scenario priority table; performing power consumption planning and solution on the target mobile phone based on the scenario priority table to obtain a power consumption scheduling policy; performing step-by-step parsing on the power consumption scheduling policy to obtain a control instruction sequence, and dynamically adjusting the control unit of the target mobile phone in real time based on the control instruction sequence to achieve the optimal balance between power consumption and performance. This solves the technical problem that existing power consumption optimization strategies often ignore the correlation between different application scenarios and their dynamic changes over time, which may lead to low efficiency. It realizes power consumption planning and solution based on the scenario priority table, and finally obtains a specific power consumption scheduling policy. This systematic process from data collection, analysis to strategy formation ensures the effectiveness and pertinence of the power consumption management scheme, and can minimize energy consumption while meeting the user's performance requirements. Description of the Drawings

[0017] Figure 1 is a schematic diagram of the steps of the mobile phone power consumption optimization method in an embodiment of the present invention; Figure 2 is a block diagram of the structure of the mobile phone power consumption optimization device in an embodiment of the present invention; Figure 3 is a schematic block diagram of the structure of a computer device in an embodiment of the present invention.

[0018] The realization, functional characteristics and advantages of the purpose of the present invention will be further described with reference to the embodiments and the accompanying drawings. Detailed Embodiments

[0019] In order to make the objectives, technical solutions and advantages of the present invention more clear and understandable, the present invention will be further described in detail below with reference to 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 used to limit the present invention.

[0020] As Figure 1 shown, Figure 1 is a schematic diagram of the steps of a mobile phone power consumption optimization method in an embodiment of the present invention; An embodiment of the present invention provides a mobile phone power consumption optimization method, including the following steps: Step S1, perform multi-dimensional real-time collection of the operation data of the target mobile phone to obtain power consumption scenario data.

[0021] Specifically, the process of performing multi-dimensional real-time collection of the operation data of the target mobile phone to obtain power consumption scenario data is an important basis for implementing the mobile phone power consumption optimization method. In this step, the system will carry out a comprehensive data collection activity for various operation parameters of the target mobile phone. These parameters cover multiple aspects from the hardware layer to the software layer, such as CPU usage rate, memory occupancy, screen brightness, and the status of the wireless communication module. Through monitoring components deployed at the bottom layer of the operating system or in specific applications, it can be ensured that these data are captured in real time and the impact on the performance of the mobile phone is minimized. In order to ensure the accuracy and timeliness of data collection, a series of advanced technical means are adopted in this process to enhance its implementation ability. For example, for the monitoring of core resources such as CPU and memory, relevant indicators can be directly read through kernel-level probes or dedicated API interfaces; while for variables related to the user interface such as screen brightness, it may rely on the callback mechanism provided by the system service to obtain the latest adjustment information. In addition, considering that the change in the wireless network connection status may significantly affect the power consumption level, the working modes of Wi-Fi, Bluetooth, and cellular data connections will also be particularly concerned in this process, including whether they are in the active transmission state or the standby mode, so as to provide detailed information support for subsequent analysis. For example, when the user is watching an online video, the power consumption scenario data in this application scenario will include the CPU and GPU loads of the video player, the memory allocation for decoding the video stream, the screen maintaining a high brightness to ensure a good visual experience, and the continuously working Wi-Fi module to maintain a smooth data stream transmission. By continuously collecting data from the above-mentioned various dimensions, we can construct a complete picture reflecting the power consumption characteristics of the mobile phone in the current video viewing scenario, and then lay a solid foundation for the next step of performing an operation timing analysis on the target mobile phone based on the power consumption scenario data to obtain a power consumption state migration chain. Such an approach not only helps to understand the energy consumption patterns of the mobile phone under different operating conditions, but also provides a key basis for formulating effective power management strategies.

[0022] Step S2: Based on the power consumption scenario data, perform an operating timing analysis on the target mobile phone to obtain a power consumption state transition chain.

[0023] Specifically, the process of performing an operating timing analysis on the target mobile phone based on the power consumption scenario data to obtain a power consumption state transition chain is a key step in implementing the mobile phone power consumption optimization method. At this stage, the system utilizes the power consumption scenario data collected in real time from multiple dimensions to deeply explore the time series characteristics of energy consumption changes under different operating states, so as to reveal the internal relationships between various parameters and the laws of evolution over time. To achieve this goal, it is first necessary to establish a mathematical model that can reflect the interaction and dependency relationships of various operating parameters, which usually involves the application of complex data processing and statistical analysis techniques. For example, in the application scenario of a user watching an online video, when the system receives data from multiple aspects such as CPU usage rate, memory occupancy, screen brightness, and the status of the wireless communication module, it will analyze these data through time series analysis methods. The system not only focuses on static parameters such as the CPU load or screen brightness setting at a certain moment, but also tracks the changing trends of these parameters over time. Considering various dynamic factors that may occur during video playback, such as Wi-Fi connection fluctuations caused by network conditions and resolution changes due to video content switching, the analysis tool will examine the correlations between these variables and the overall power consumption level. Through learning from a large number of sample data, the algorithm can identify the laws of power consumption increase or decrease under specific operating modes. For example, when the video player pauses or switches to a lower resolution, the corresponding hardware resource consumption will also change. Further, in order to construct the power consumption state transition chain, the system must quantify the transition probabilities between each state and their corresponding power consumption differences. The "state" mentioned here is a working mode defined by a series of power consumption-related indicators, such as high-load operation, low-brightness display, or sleep standby. Once the transition paths between different states are determined, it is possible to predict the state changes that the mobile phone may experience in the future. Continuing with the video playback example, if it is detected that an advertisement insertion period is about to enter, at this time the video stream bit rate decreases and the screen may briefly dim, then the system can react in advance and correspondingly reduce the allocation of unnecessary computing resources, thereby saving power. In this way, an effective prediction of future power consumption state changes is achieved, providing a solid foundation for formulating a reasonable power consumption scheduling strategy in the subsequent stage, and ensuring optimal power consumption management without affecting the user experience.

[0024] Step S3: Calculate the weights of the power consumption scenario data in the power consumption state transition chain through the hybrid entropy weight method to obtain a scenario priority table.

[0025] Specifically, calculating the weights of the power consumption scenario data in the power consumption state migration chain through the hybrid entropy weight method to obtain the scenario priority table is one of the key steps in realizing intelligent power consumption management. In this process, the system first extracts the features of the power consumption scenario data in the power consumption state migration chain based on the information entropy algorithm to quantify the uncertainty and complexity of power consumption changes in different states. For example, in the application scenario of video viewing, when the user switches from normal high-definition video playback to an advertisement insertion, parameters such as the CPU, GPU load, and screen brightness of the system will experience significant changes. By applying the information entropy algorithm to these change patterns, it can be revealed which state transitions have higher uncertainty, thereby providing a basis for subsequent weight allocation. Next, in order to reasonably allocate the relative importance between each state, the system uses a multi-criteria decision analysis method to normalize the extracted state entropy value sequence to obtain a normalized entropy value matrix, and based on this, weight allocation is performed to obtain an initial weight vector. This method comprehensively considers multiple evaluation criteria, such as power consumption level, user experience, and resource utilization, to ensure the scientificity and rationality of weight setting. For video playback, this may involve evaluating the balance between high-load operation states (such as playing high-definition videos) and low-power standby states (such as during advertisement insertions). Subsequently, the system will perform constraint optimization on the initial weight vector to obtain an optimized weight set. The constraint optimization process aims to ensure that the sum of all weights is equal to 1 while minimizing the influence of subjective factors, so that the weights more objectively reflect the actual importance of each state. With the optimized weight set, the system further performs hierarchical analysis on the power consumption state migration chain based on this to obtain the scenario priority table. The Analytic Hierarchy Process (AHP) is a structured multi-criteria decision-making tool that allows us to decompose complex decision-making problems into several levels and determine the importance of each factor through pairwise comparisons. For video playback, this means that a multi-level priority framework can be constructed according to factors such as user preferences and device performance requirements, thereby clarifying the priority relationship between different types of operation modes. For example, the state of normal high-definition video playback may be assigned a higher priority due to its importance to the user experience, while operations such as advertisement insertion or background caching may be ranked lower. In order to more accurately evaluate the degree of association between these priorities, the system calculates the grey correlation degree of the scenario priority table through grey correlation analysis to obtain the scenario correlation degree sequence. Grey correlation analysis is an effective means of dealing with uncertainty and small sample data, and it can measure the similarity and correlation between different sequences. In the video playback scenario, this analysis can help us understand the mutual influence between various operation modes, such as whether the network traffic fluctuation during advertisement insertion will affect the quality of the main video stream. Then, the system will perform orthogonal decomposition on the scenario correlation degree sequence to obtain the scenario feature weights.The scenario feature weights here include three main aspects: scenario switching cost, resource consumption ratio, and performance impact factor. Together, they determine the overall burden borne in a specific state and its impact on the system. Based on this, the system ranks the priorities of the scenario feature weights using the coefficient of variation method to obtain a scenario priority sequence. The coefficient of variation method measures the relative fluctuation amplitude by calculating the ratio of the standard deviation to the mean of each feature, thereby assigning corresponding weights to each scenario feature. For example, if it is found that the scenario switching cost is relatively large in a certain state, then it may occupy a more important position in the priority ranking. In addition, the system also detects whether there are circular dependency conflicts in the scenario priority sequence. If so, it eliminates the circular dependency conflicts in the scenario priority sequence to obtain a conflict-free scenario set. This is because circular dependencies may lead to deadlocks or inefficiencies in policy execution and must be avoided. Finally, the system performs priority mapping on the power consumption scenario data based on the conflict-free scenario set through a multi-objective optimization algorithm to obtain a scenario priority table. The multi-objective optimization algorithm can find a set of optimal solutions under the premise of satisfying multiple constraints, enabling each objective function to reach the best state simultaneously. In the video playback scenario, this means that while ensuring a smooth viewing experience, the overall power consumption can be reduced as much as possible to extend the battery life. Through the above series of steps, the system can not only identify the power consumption characteristics in different usage patterns but also predict the future state change trends and formulate reasonable optimization strategies accordingly to achieve intelligent and personalized power consumption management. In summary, by calculating the weights of the power consumption scenario data in the power consumption state migration chain using the hybrid entropy weight method, the system not only improves the depth of understanding of the mobile phone's energy consumption pattern but also provides solid technical support for realizing refined power consumption management. This method not only ensures flexible adjustment of resource allocation without affecting the user experience but also achieves the optimal balance between power consumption and performance in various application scenarios, significantly improving user satisfaction and the device's battery life. By organically combining the information entropy algorithm, multi-criteria decision analysis, analytic hierarchy process, grey relational analysis, coefficient of variation method, and multi-objective optimization algorithm, the system realizes an effective transition from complex data analysis to intuitive strategy formulation, opening up a new path for intelligent power consumption management.

[0026] Step S4: Based on the scenario priority table, perform power consumption planning and solution for the target mobile phone to obtain a power consumption scheduling strategy.

[0027] Specifically, the process of solving the power consumption plan for the target mobile phone based on the scene priority table to obtain the power consumption scheduling strategy is the key link in the entire mobile phone power consumption optimization method that transforms theoretical analysis into actual operations. At this stage, the system uses the scene priority table calculated through the hybrid entropy weight method before as the basis for formulating a specific power management plan. This table details the power consumption characteristics and their relative importance in different application scenarios, providing a clear direction and basis for the subsequent power consumption planning. Specifically, in the application scenario of video viewing, after the system obtains the weights of each state, it will determine which operating modes should be prioritized and which can be appropriately adjusted to save power without affecting the user experience. For example, for the high-priority state of normal high-definition video playback, the system may maintain a high CPU and GPU performance output and screen brightness setting to ensure a smooth visual experience; while for low-priority situations such as during advertisement inserts or when the user pauses to rest, measures such as reducing the resolution, decreasing the screen brightness, or entering the sleep mode can be taken to effectively reduce power consumption. In this way, not only can the immediate needs of users be met, but the system can also intelligently optimize resource allocation in the background to achieve a longer battery life. To implement the concretization of the power consumption scheduling strategy, the system needs to further consider how to transform the above principles into actual operation instructions. This involves the ability to precisely adjust various control units within the mobile phone (such as processor frequency, display parameters, network connection status, etc.). During the process of constructing the power consumption scheduling strategy, the algorithm comprehensively considers factors such as the current working state of the device, the remaining battery power, and the expected usage time, and dynamically generates a series of optimization suggestions. For example, if it is detected that the user is about to enter a long advertisement insert period, the system can plan in advance to gradually reduce the level of unnecessary hardware activities without affecting the subsequent video playback quality, thereby achieving the purpose of energy conservation. In addition, for possible emergencies, such as a sudden increase in network traffic requests, the system also reserves a certain amount of flexibility to quickly respond and maintain the overall performance stability. Finally, through in-depth analysis of the scene priority table, the system can formulate a comprehensive and flexible power consumption scheduling strategy. This strategy takes into account both the personalized needs of users in specific application scenarios and the overall efficiency of long-term use. Taking video playback as an example, the system can not only provide excellent audio-visual effects during high-definition video playback but also cleverly save power at other times, ensuring that users enjoy the best balance point throughout the viewing process - that is, the optimal configuration between power consumption and performance. Such a design enables the smart phone to maintain high-efficiency operation while extending the battery life in the face of complex and changing application environments, significantly enhancing user satisfaction and usage experience.

[0028] Step S5, perform step-by-step analysis on the power consumption scheduling policy to obtain a control instruction sequence, and based on the control instruction sequence, dynamically adjust the control unit of the target mobile phone in real time to achieve the optimal balance between power consumption and performance.

[0029] Specifically, the process of step-by-step analysis of the power consumption scheduling strategy to obtain a control instruction sequence and based on the control instruction sequence to dynamically adjust the control unit of the target mobile phone in real time to achieve the optimal balance between power consumption and performance is the key step to transform the optimization strategy into specific operations. After completing the power consumption planning solution based on the scenario priority table, the next task of the system is to carefully disassemble the generated power consumption scheduling strategy and convert it into a series of clear and executable control instructions. These instructions are designed to guide how various hardware and software components inside the mobile phone work together to ensure that the established energy consumption management goals can be achieved at any given moment. Specifically, in the application scenario of video viewing, when the system has formulated specific power consumption scheduling strategies for different states (such as high-definition playback, advertisement insertion, user pause, etc.), it will further refine these strategies to form a set of precise control instruction sequences. For example, for the state of normal high-definition video playback, the system may issue instructions to keep the working frequencies of the CPU and GPU at a relatively high level while maintaining the screen brightness to provide the best visual effect; during advertisement insertion or when the user pauses to rest, the system can issue instructions to reduce the resolution, decrease the screen brightness or enter the sleep mode, thus effectively saving power. In this way, not only can the immediate needs of users be met, but also the resource allocation can be intelligently optimized in the background to achieve a longer battery life. To ensure that these control instructions can be executed accurately and error-free, the system needs to have the ability to dynamically adjust in real time. This means that once it detects a change in the current application scenario, such as switching from normal playback to advertisement insertion, the system must respond quickly and adjust the corresponding hardware parameters according to the pre-set rules. For example, when the video player recognizes that an advertisement period is about to enter, it will notify the power management system in advance to prepare to implement energy-saving measures, such as gradually reducing the activity level of non-essential hardware. At the same time, the system will continuously monitor the device status and user behavior to make fine-tuning at any time to ensure the overall performance is stable. In addition, considering possible emergencies, such as a sudden increase in network traffic requests, the system will also reserve a certain amount of flexibility to ensure a quick response without affecting the user experience. Finally, through the step-by-step analysis of the power consumption scheduling strategy, the system can generate a series of precise control instructions, which are transmitted to each control unit of the mobile phone in real time, such as the processor frequency controller, display driver, wireless communication module, etc. Each instruction represents a specific action, and together they achieve the efficient operation of the entire system. Taking video playback as an example, the system can not only provide excellent audio-visual effects during high-definition video playback, but also cleverly save power at other times, ensuring that users enjoy the best balance point throughout the viewing process - that is, the optimal configuration between power consumption and performance. Such a design enables the smart phone to maintain high-performance operation while extending the battery life in the face of complex and changing application environments, significantly improving user satisfaction and usage experience.Through this intelligent dynamic adjustment mechanism, the system can optimize the energy utilization efficiency of the mobile phone to the greatest extent without affecting the user experience.

[0030] In a specific embodiment, the multi-dimensional real-time collection of the running data of the target mobile phone to obtain power consumption scenario data includes: Real-time collection of the hardware parameters of the target mobile phone through a preset sensor group to obtain an original power consumption data set; wherein, the original power consumption data set includes CPU power consumption data, GPU power consumption data, memory power consumption data, and battery temperature; Based on the original power consumption data set, perform hardware resource tracking analysis on the target mobile phone to obtain a resource dynamic load vector; Extract power consumption scenario features from the resource dynamic load vector through a multi-dimensional data fusion algorithm to obtain power consumption scenario data; wherein, the power consumption scenario data includes game scenarios, video scenarios, web browsing scenarios, and standby scenarios.

[0031] Specifically, the process of multi-dimensional real-time collection of the running data of the target mobile phone to obtain power consumption scenario data is the basic link for realizing intelligent power consumption management. In this process, the system first collects the hardware parameters of the target mobile phone in real time through a preset sensor group to obtain the original power consumption data set. These sensor groups cover key components inside the mobile phone, such as the CPU, GPU, memory, and battery temperature, etc., and can accurately capture the hardware state at each moment. For example, in the application scenario of video viewing, when the user starts the video player and selects a high-definition video to play, the system will immediately activate the sensor group to real-time track a series of important indicators such as CPU power consumption data, GPU power consumption data, memory power consumption data, and battery temperature. These data not only reflect the current working load of the device but also provide rich basic information for subsequent analysis. With the acquisition of the original power consumption data set, the next task is to conduct hardware resource tracking analysis on the target mobile phone to obtain the resource dynamic load vector. By deeply analyzing these hardware parameters, the actual working intensity and its change trend of each component under a specific operation mode can be revealed. For video playback, this may mean analyzing the load conditions of the CPU and GPU when decoding high-quality video frames, evaluating whether the memory bandwidth is sufficient to support smooth data transmission, and at the same time monitoring the battery temperature to ensure it is within a safe range. The resource dynamic load vector not only records the instantaneous state of each hardware component but also shows their dynamic characteristics evolving over time, providing a key perspective for understanding the overall behavior of the system. For example, during the process of playing a high-definition video, the loads of the CPU and GPU will increase significantly, the memory bandwidth will also increase accordingly, and the battery temperature needs to be maintained within a reasonable range to avoid overheating problems. These dynamic load characteristics together constitute a vector that comprehensively reflects the running status of the device, laying the foundation for further analysis. In order to further extract meaningful information, the system extracts power consumption scenario features from the resource dynamic load vector through a multi-dimensional data fusion algorithm, and finally obtains the power consumption scenario data. Multi-dimensional data fusion is a method of integrating multiple data points from different sources or types into a higher-level representation. Here, it is used to comprehensively consider the interactions between multiple factors such as CPU power consumption, GPU power consumption, memory power consumption, and battery temperature. Through the fusion processing of these data, the system can identify different usage scenarios and classify them into typical categories such as game scenarios, video scenarios, web browsing scenarios, and standby scenarios. For example, during video playback, if continuous high CPU and GPU loads, stable network traffic, and relatively fixed screen brightness settings are detected, the system can reasonably infer that the user is in a state of concentrated viewing; while when it is found that the network connection is suddenly interrupted or the screen brightness drops sharply, it may mean that the user has paused the viewing or switched to other applications.Specifically, in the application scenario of video viewing, when the user continuously plays multiple movies, the system will identify this as a typical video viewing scenario based on factors such as continuous high power consumption levels, stable network traffic, and relatively fixed screen settings. At this time, the power consumption scenario data will include all relevant hardware operating parameters and user behavior characteristics, providing detailed background information for subsequent power consumption optimization strategies. When the user occasionally browses social media or checks news, although the screen is still on, the CPU and GPU loads are low, and network access is more inclined to intermittent bursts within a short period. The system can then identify this as a web browsing scenario and adjust its power management strategy accordingly. In addition, when the phone is stationary and there is no obvious hardware activity, the system will classify it as a standby scenario and take minimal power consumption maintenance measures to ensure the maximization of battery life. To better understand this process, we can explore a specific example in detail: Suppose the user is watching a high-definition movie. During the movie viewing process, the system collects the following data in real-time through a preset sensor group: The CPU frequency is relatively high to handle video decoding requirements, the GPU load is also at a high level to support high-resolution image rendering, the memory occupancy rate increases significantly to store a large amount of temporary data, and the battery temperature rises slightly but remains within the safe range. These raw power consumption data sets constitute a description of the hardware operating state of the device during this period. Based on these data, the system conducts hardware resource tracking analysis to obtain a resource dynamic load vector, which includes the load conditions and their changing trends of each hardware component. For example, the loads of the CPU and GPU gradually stabilize at a relatively high level over time, while the memory occupancy rate shows periodic fluctuations. Next, the system extracts power consumption scenario features from the resource dynamic load vector through a multi-dimensional data fusion algorithm. This process involves combining data from different dimensions (such as CPU power consumption, GPU power consumption, memory power consumption, and battery temperature) to form a unified feature representation. For example, the system may find that during high-definition video playback, the power consumption of the CPU and GPU is always at a relatively high level, the memory occupancy rate fluctuates greatly, and the battery temperature remains within the safe range. Through this multi-dimensional data fusion, the system can accurately identify the current operation mode as the "video scenario". Similarly, when the user switches to an advertising break, the system will detect that the loads of the CPU and GPU decrease, the memory occupancy rate tends to be stable, and the battery temperature also decreases, thus identifying this as a low-power "advertising break scenario". In addition, when the user switches from video playback to web browsing, the system will detect that the loads of the CPU and GPU decrease significantly, the memory occupancy rate decreases, the network traffic becomes more intermittent and irregular, and the battery temperature further decreases. These features indicate that the current operation mode has changed to the "web browsing scenario".The system can dynamically adjust the power management strategy according to these recognition results. For example, it can moderately reduce the CPU frequency and screen brightness to save power while ensuring that the web page loading speed and user experience are not affected. Finally, when the user has not operated the mobile phone for a long time and the phone enters the standby state, the system will detect that the load of almost all hardware components has dropped to the lowest level and the battery temperature has also stabilized. At this time, the system will classify it as a "standby scenario" and take corresponding energy-saving measures, such as closing unnecessary background processes and services, to maximize the battery life. In summary, the process of obtaining power consumption scenario data by multi-dimensional real-time collection of the running data of the target mobile phone is a complex project integrating a variety of advanced technical means. From the precise measurement of hardware parameters to the detailed analysis of load characteristics, and then to the fusion processing of multi-dimensional data, each link contributes to the construction of a complete power management framework. This method based on comprehensive data analysis not only improves the depth of understanding of the mobile phone's energy consumption pattern but also provides a reliable basis for formulating scientific and reasonable energy-saving strategies. Through the above method, the system can intelligently adjust resource allocation without affecting the user experience, ensuring the optimal balance between power consumption and performance in various application scenarios. Whether facing daily use or special requirements, this mechanism can provide stable and reliable performance support for users while maximizing energy consumption savings. In the video playback scenario, this means that the system can not only provide a smooth audio-visual experience but also intelligently optimize resource allocation, extend the battery life, and allow users to enjoy more durable and high-quality multimedia services.

[0032] In a specific embodiment, the running time series analysis of the target mobile phone based on the power consumption scenario data to obtain a power consumption state transition chain includes: Performing time series segmentation on the power consumption scenario data based on discrete wavelet analysis to obtain a scenario time series, and performing Fourier transform on the scenario time series to obtain scenario spectrum features; Performing state recognition on the scenario spectrum features through adaptive edge detection to obtain a scenario state sequence, and performing hierarchical clustering on the scenario state sequence to obtain a state classification result; wherein, the state classification result includes a high-frequency switching state, a stable operation state, and a low-frequency switching state; Performing time-dependence analysis on the state classification result based on a recurrent neural network to obtain state transition probabilities, and performing Bayesian state inference on the target mobile phone based on the state transition probabilities to obtain a state prediction sequence; wherein, the state prediction sequence includes scenario duration, switching time points, and transition directions; Performing similarity calculation on the state prediction sequence through a dynamic time warping algorithm to obtain a state distance matrix, and performing spectral clustering on the state distance matrix to obtain a state cluster set, wherein the state cluster set includes a high-power consumption cluster, a medium-power consumption cluster, and a low-power consumption cluster; Analyze the transition law of the state cluster set based on the Markov decision process to obtain a state transition map, and perform path extraction and path optimization based on the state transition map to obtain an optimized transfer path; wherein, the optimized transfer path includes the shortest path, the optimal switching point, and the transfer cost; Based on the optimized transfer path, construct a link through a graph neural network to obtain a power consumption state migration chain.

[0033] Specifically, the process of performing an operating timing analysis on the target mobile phone based on the power consumption scenario data to obtain a power consumption state transition chain is a complex and delicate data processing and modeling process. In this process, the system first uses discrete wavelet analysis to segment the collected power consumption scenario data in time series, thereby obtaining a series of scenario time series. Discrete wavelet transform (DWT) is a powerful tool that can provide both time and frequency information at the same time. It allows us to decompose the original time series data into subsequences at different scales, which helps to capture the multi-scale characteristics in power consumption changes. For example, in the application scenario of video viewing, when the user starts playing a high-definition video, the hardware parameters such as the system's CPU, GPU, and screen brightness will quickly adjust to a high-load state within a short period of time; during the advertisement insertion or when the user pauses to rest, these parameters will correspondingly decrease. Through discrete wavelet analysis, the entire video playback process can be segmented into multiple time periods with similar characteristics according to different activity patterns. Next, in order to further reveal the internal laws within each time period, the system performs a Fourier transform on the scenario time series to obtain the scenario spectral characteristics. The Fourier transform converts the time-domain signal into a frequency-domain representation, enabling us to observe the energy distribution of each frequency component. For video playback, the high-frequency components may correspond to rapidly switching pictures or sudden fluctuations in network traffic, while the low-frequency components reflect relatively stable background operations, such as continuous audio output or slowly changing image frames. By studying the spectral characteristics, not only can different types of power consumption events be identified, but also their continuity and periodicity in time can be evaluated, providing an important basis for further state recognition. On this basis, the system uses adaptive edge detection technology to perform state recognition on the scenario spectral characteristics to obtain a scenario state sequence. Adaptive edge detection can automatically adjust the detection threshold to highlight important feature points while maintaining the original signal characteristics. In the video playback scenario, this means that it is possible to more accurately identify those critical moments that mark the change in user behavior, such as switching from normal playback to advertisement insertion, or from watching a video to browsing the web. Subsequently, the system performs hierarchical clustering on the scenario state sequence to obtain a state classification result. According to the different power consumption levels and their change rates, these states are divided into three categories: high-frequency switching state, stable operation state, and low-frequency switching state. The high-frequency switching state usually appears when the user interacts frequently or the application switches frequently. The stable operation state refers to maintaining a specific working mode for a long time, such as playing a high-definition video. The low-frequency switching state is between the two and may involve some light interactions or background task processing. After obtaining the above state classification results, the system performs a time-dependence analysis on these states based on a recurrent neural network (RNN) to obtain the state transition probability. RNN is good at processing data with time series properties, so it is very suitable for predicting possible future states and corresponding transition paths.Specifically, the system learns the conversion rules between different types of states and calculates the likelihood of each conversion. For example, during video playback, if it is detected that an advertisement insertion period is about to begin, the system can infer from historical data that the probability of converting from a high-power video playback state to a lower-power advertisement playback state is relatively high. Based on these state transition probabilities, the system further performs Bayesian state inference on the target mobile phone to obtain a state prediction sequence. This sequence contains key information such as the duration of the scenario, the switching time point, and the transfer direction, providing guidance for subsequent optimization strategies. To ensure the accuracy of state prediction, the system calculates the similarity of the state prediction sequence through the Dynamic Time Warping (DTW) algorithm to obtain a state distance matrix. DTW is a method for measuring the distance between two time series, which can effectively handle the problem of inconsistent lengths and is particularly suitable for comparing the power consumption change trends under different states. By performing spectral clustering on the state distance matrix, the system can group similar states together to form a set of state clusters, including high-power clusters, medium-power clusters, and low-power clusters. This clustering method not only simplifies the complexity of subsequent analysis but also facilitates the formulation of targeted power management measures. Finally, the system analyzes the transfer rules of the state cluster set based on the Markov Decision Process (MDP) to obtain a state transfer map. MDP provides a framework to describe the process of taking actions and receiving rewards in an environment and is very suitable for planning the optimal power consumption state migration path. By extracting and optimizing the paths from the state transfer map, the system can find the shortest path, the optimal switching point, and the scheme with the minimum transfer cost. For example, in a video playback scenario, when the system anticipates that the user is about to finish the current movie and prepare to launch the next application, it can plan the resource allocation in advance to minimize unnecessary energy consumption waste while ensuring that the new application can respond quickly. Through the Graph Neural Network (GNN), the system constructs a link based on the optimized transfer path to finally obtain a complete power consumption state migration chain. GNN can effectively capture the relationships between nodes and apply them to the modeling of complex network structures, which is crucial for constructing an accurate and efficient power consumption state migration model. In summary, through in-depth analysis of the running time series of power consumption scenario data, the system can not only identify the power consumption characteristics under different usage patterns but also predict future state change trends and formulate reasonable optimization strategies accordingly. This method not only improves the depth of understanding of the mobile phone energy consumption pattern but also provides a solid technical support for realizing intelligent and personalized power management. Through the above steps, the system can flexibly adjust the resource allocation without affecting the user experience, ensuring the optimal balance between power consumption and performance in various application scenarios and significantly improving the user satisfaction and the device's battery life.

[0034] In a specific embodiment, calculating weights for the power consumption scenario data in the power consumption state transition chain through the hybrid entropy weight method to obtain a scenario priority table, including: Extracting features from the power consumption scenario data in the power consumption state transition chain based on the information entropy algorithm to obtain a state entropy value sequence, and performing normalization processing on the state entropy value sequence to obtain a normalized entropy value matrix; Allocating weights to the normalized entropy value matrix through a multi-criteria decision analysis method to obtain an initial weight vector, and performing constraint optimization on the initial weight vector to obtain a tuned weight set; Performing hierarchical analysis on the power consumption state transition chain based on the tuned weight set to obtain a scenario priority table; Calculating the grey correlation degree of the scenario priority table through grey correlation analysis to obtain a scenario correlation degree sequence, and performing orthogonal decomposition on the scenario correlation degree sequence to obtain scenario feature weights; wherein, the scenario feature weights include scenario switching cost, resource consumption ratio, and performance impact factor; Performing priority sorting on the scenario feature weights based on the coefficient of variation method to obtain a scenario priority sequence, and detecting whether there is a cyclic dependency conflict in the scenario priority sequence. If so, excluding the cyclic dependency conflict in the scenario priority sequence to obtain a conflict-free scenario set; Performing priority mapping on the power consumption scenario data based on the conflict-free scenario set through a multi-objective optimization algorithm to obtain a scenario priority table.

[0035] Specifically, the process of calculating the weights of the power consumption scenario data in the power consumption state transition chain by the hybrid entropy weight method to obtain the scenario priority table is a complex and systematic analysis process. In this process, the system first extracts the features of the power consumption scenario data in the power consumption state transition chain based on the information entropy algorithm to obtain the state entropy value sequence, and normalizes the state entropy value sequence to obtain the normalized entropy value matrix. Information entropy is a statistical method for measuring uncertainty, and here it is used to quantify the randomness and complexity of power consumption changes in different states. For example, in the application scenario of video viewing, when the user switches from normal high-definition video playback to an advertisement insertion, parameters such as the CPU, GPU load, and screen brightness of the system will experience significant changes. By applying the information entropy algorithm to these change patterns, it can be revealed which state transitions have a high degree of uncertainty, thereby providing a basis for subsequent weight allocation. Next, in order to reasonably allocate the relative importance between each state, the system uses a multi-criteria decision analysis method to allocate weights to the normalized entropy value matrix to obtain the initial weight vector. This method comprehensively considers multiple evaluation criteria, such as power consumption level, user experience, and resource utilization, to ensure the scientificity and rationality of weight setting. In the video playback scenario, this may involve evaluating the balance between high-load operation states (such as playing high-definition videos) and low-power standby states (such as during advertisement insertions). Subsequently, the system will perform constraint optimization on the initial weight vector to obtain the optimized weight set. The constraint optimization process aims to ensure that the sum of all weights is equal to 1 while minimizing the influence of subjective factors as much as possible, so that the weights more objectively reflect the actual importance of each state. With the optimized weight set, the system further performs hierarchical analysis on the power consumption state transition chain based on this to obtain the scenario priority table. The Analytic Hierarchy Process (AHP) is a structured multi-criteria decision-making tool that allows us to decompose complex decision-making problems into several levels and determine the importance of each factor through pairwise comparisons. For video playback, this means that a multi-level priority framework can be constructed according to factors such as user preferences and device performance requirements, so as to clarify the priority relationship between different types of operation modes. For example, the state of normal high-definition video playback may be given a higher priority due to its importance to the user experience, while operations such as advertisement insertion or background caching may be ranked lower. In order to more accurately evaluate the degree of association between these priorities, the system calculates the grey correlation degree of the scenario priority table through grey correlation analysis to obtain the scenario correlation degree sequence. Grey correlation analysis is an effective means of dealing with uncertainty and small sample data, and it can measure the similarity and correlation between different sequences. In the video playback scenario, this analysis can help us understand the mutual influence between various operation modes, such as whether the network traffic fluctuation during advertisement insertion will affect the quality of the main video stream.Next, the system performs orthogonal decomposition on the scene correlation degree sequence to obtain scene feature weights. The scene feature weights here include three main aspects: scene switching cost, resource consumption ratio, and performance impact factor. Together, they determine the overall burden borne in a specific state and its impact on the system. On this basis, the system performs priority sorting on the scene feature weights based on the coefficient of variation method to obtain a scene priority sequence. The coefficient of variation method measures the relative fluctuation range by calculating the ratio of the standard deviation to the mean of each feature, thereby assigning corresponding weights to each scene feature. For example, if it is found that the scene switching cost is large in a certain state, then it may occupy a more important position in the priority sorting. In addition, the system also detects whether there is a circular dependency conflict in the scene priority sequence. If so, it eliminates the circular dependency conflict in the scene priority sequence to obtain a conflict-free scene set. This is because circular dependencies may lead to deadlocks or inefficiencies in policy execution and must be avoided. Finally, the system performs priority mapping on the power consumption scene data based on the conflict-free scene set through a multi-objective optimization algorithm to obtain a scene priority table. The multi-objective optimization algorithm can find a set of optimal solutions under the premise of meeting multiple constraint conditions, enabling each objective function to reach the best state simultaneously. In the video playback scenario, this means that while ensuring a smooth viewing experience, the overall power consumption can be reduced as much as possible to extend the battery life. Specifically, when the user switches from playing a high-definition video to an advertisement insertion, the system will dynamically adjust the configuration of hardware resources according to the pre-calculated scene priority table, such as moderately reducing the CPU frequency and screen brightness, to save power without affecting the display effect of the advertisement content. In this way, the system not only achieves accurate identification of power consumption characteristics under different usage modes but also formulates reasonable optimization strategies to ensure flexible adjustment of resource allocation without affecting the user experience, achieving the optimal balance between power consumption and performance. To sum up, by calculating the weights of the power consumption scene data in the power consumption state migration chain through the hybrid entropy weight method, the system not only improves the depth of understanding of the mobile phone energy consumption mode but also provides solid technical support for achieving refined power consumption management. This method not only ensures flexible adjustment of resource allocation without affecting the user experience but also achieves the optimal balance between power consumption and performance in various application scenarios, significantly improving user satisfaction and device battery life. By organically combining the information entropy algorithm, multi-criteria decision analysis, analytic hierarchy process, grey relational analysis, coefficient of variation method, and multi-objective optimization algorithm, the system realizes an effective transition from complex data analysis to intuitive strategy formulation, opening up a new path for intelligent power consumption management. In the video playback scenario, this mechanism can not only provide a smooth audio-visual experience but also intelligently optimize resource allocation, extend battery life, and allow users to enjoy more durable and high-quality multimedia services.

[0036] In a specific embodiment, performing power consumption planning and solution on the target mobile phone based on the scenario priority table to obtain a power consumption scheduling strategy, including: Performing dependency analysis on the scenario priority table through a topological sorting algorithm to obtain a scenario scheduling graph, and performing constraint propagation on the scenario scheduling graph to obtain a scheduling constraint set; wherein, the scheduling constraint set includes resource capacity constraints, scenario switching constraints, and timing dependency constraints; Performing solution analysis on the scheduling constraint set based on mixed integer programming to obtain an initial scheduling plan, and performing a spatial search on the initial scheduling plan through a branch and bound method to obtain a set of candidate plans; wherein, the set of candidate plans includes a power consumption threshold interval, a switching time window, and a resource allocation ratio; Performing temperature regulation calculation on the set of candidate plans through a simulated annealing algorithm to obtain a cooling strategy sequence, and performing solution space exploration based on the cooling strategy sequence to obtain a locally optimal solution set; wherein, the locally optimal solution set includes power consumption control parameters, scheduling time points, and resource allocation schemes; Performing neighborhood traversal on the locally optimal solution set based on a tabu search algorithm to obtain a global search path, and performing state transition calculation on the global search path through dynamic programming techniques to obtain an optimal scheduling sequence; wherein, the optimal scheduling sequence includes a scenario switching order, a resource allocation strategy, and a power consumption control instruction; Performing multi-objective trade-off on the optimal scheduling sequence through a cooperative optimization algorithm to obtain a scheduling evaluation index, and performing plan screening based on the scheduling evaluation index to obtain an optimal scheduling plan; wherein, the optimal scheduling plan includes scenario execution duration, resource utilization efficiency, and power consumption balance coefficient; Performing population evolution calculation on the optimal scheduling plan based on a genetic algorithm to obtain an evolution plan sequence, and performing non-dominated sorting on the evolution plan sequence through a Pareto optimal set to obtain a power consumption scheduling strategy; wherein, the power consumption scheduling strategy includes a scenario scheduling time sequence table, a resource allocation matrix, and a power consumption control parameter set.

[0037] Specifically, the process of solving the power consumption planning for the target mobile phone based on the scenario priority table to obtain a power consumption scheduling strategy is a complex decision-making process that combines multiple optimization algorithms and technologies. In this process, the system first analyzes the dependency relationships of the scenario priority table through a topological sorting algorithm to obtain a scenario scheduling graph, and performs constraint propagation on the scenario scheduling graph to obtain a scheduling constraint set. Topological sorting is a technique used to reveal the order of precedence between nodes. Here, it is used to identify the dependency relationships and execution order between different application scenarios. For example, in the video viewing application scenario, when the user switches from normal high-definition video playback to an advertisement insertion, the current video stream must be paused first, and then the new advertisement content loading is started. By applying topological sorting to these operation modes, a clear scenario scheduling graph can be constructed to ensure that all state transitions follow a reasonable logical order. Subsequently, the system performs constraint propagation on this graph to determine a series of specific scheduling constraint sets, including resource capacity constraints, scenario switching constraints, and temporal dependency constraints, etc. These constraint conditions limit the total amount of available resources in each state and the conversion rules between them, thus providing clear boundaries for subsequent solution analysis. Next, in order to find the best scheduling plan that meets the above constraint conditions, the system solves and analyzes the scheduling constraint set based on mixed integer programming (MIP) to obtain an initial scheduling plan. MIP is a powerful mathematical modeling tool that can consider the influence of continuous variables while dealing with discrete variables, and is very suitable for solving complex resource allocation problems. In the video playback scenario, this may involve evaluating how to minimize energy consumption while ensuring a smooth viewing experience. Then, the system performs a spatial search on the initial scheduling plan through the branch and bound method to obtain a set of candidate plans. The branch and bound method is a method that recursively decomposes the problem into smaller sub-problems. It can effectively explore every possibility in the solution space to ensure that no potential optimal solution is missed. The set of candidate plans here covers various parameter configurations, such as power consumption threshold intervals, switching time windows, and resource allocation ratios, etc. They jointly determine the operating efficiency and performance of the system in a specific state. With these candidate plans, the system further performs temperature control calculations on the set of candidate plans through the simulated annealing algorithm to obtain a sequence of cooling strategies, and performs solution space exploration based on the sequence of cooling strategies to obtain a set of locally optimal solutions. Simulated annealing is a heuristic global optimization algorithm that mimics the physical phenomenon during the metal cooling process. It allows the system to accept worse solutions in the early stage to jump out of local extreme value traps. As the "temperature" gradually decreases, it finally converges to a more ideal solution. In the video playback scenario, this means that different power consumption control parameter settings, such as CPU frequency adjustment or screen brightness change, can be tried in the initial stage. As the number of iterations increases, gradually approach the best power consumption and performance balance point.This dynamic adjustment process helps to overcome the problem that traditional optimization methods are prone to falling into local optima, and improves the quality of the final result. The obtained local optimal solution set includes multiple aspects such as power consumption control parameters, scheduling time points, and resource allocation schemes, providing a solid foundation for the next global search. On this basis, the system performs neighborhood traversal on the local optimal solution set based on the tabu search algorithm to obtain the global search path, and performs state transition calculations on the global search path through dynamic programming techniques to obtain the optimal scheduling sequence. The tabu search algorithm accelerates the speed of global search by introducing a memory mechanism to avoid repeatedly visiting solutions that have been explored. For video playback, this means that various possible combinations of scene switching sequences and resource allocation strategies can be explored more efficiently to find solutions that can save power without affecting the user experience. Dynamic programming provides an effective method to calculate the transition costs between different states, ensuring that the entire scheduling process is both coherent and economical. The finally obtained optimal scheduling sequence not only specifies the specific sequence of scene switching, but also details the resource allocation strategies and power consumption control instructions in each step, enabling the system to execute strictly according to the plan in actual operation. Finally, to ensure that the obtained scheduling scheme is both scientific and reasonable and can adapt to the changing application environment, the system performs multi-objective trade-offs on the optimal scheduling sequence through a collaborative optimization algorithm to obtain scheduling evaluation indicators, and performs scheme screening based on the scheduling evaluation indicators to obtain the preferred scheduling scheme. The collaborative optimization algorithm comprehensively considers multiple conflicting objective functions, such as scene execution duration, resource utilization efficiency, and power consumption balance coefficient, aiming to find a set of optimal solutions that can meet the requirements of all aspects at the same time. For video playback, this means not only ensuring the smoothness of movie playback, but also minimizing battery consumption and extending the device's battery life. After strict screening, the system will retain those solutions that best meet the expected effects as the preferred scheduling scheme. However, considering the many uncertainties and dynamic change factors in the real world, the system needs to further improve the robustness and adaptability of the scheduling strategy. Therefore, based on the genetic algorithm, population evolution calculations are performed on the preferred scheduling scheme to obtain an evolutionary scheme sequence, and non-dominated sorting is performed on the evolutionary scheme sequence through the Pareto optimal set to obtain the power consumption scheduling strategy. The genetic algorithm mimics the principle of biological evolution in nature and continuously optimizes the individuals in the population through operations such as selection, crossover, and mutation, finally generating a series of high-quality solutions. The Pareto optimal set is used to select those solutions that perform well in multiple objectives and have no obvious disadvantages. In the video playback scenario, this means that through continuous evolution and optimization, scheduling strategies that can achieve the best balance in both power consumption control and performance guarantee can be found. The finally formed power consumption scheduling strategy not only includes a detailed scene scheduling time sequence table, a resource allocation matrix, and a power consumption control parameter set, but also has good adaptive capabilities and can always operate efficiently in the face of complex and changing application environments.In summary, through the process of power consumption planning and solution for the scenario priority table to obtain the power consumption scheduling strategy, the system not only realizes the refined management of the mobile phone energy consumption mode, but also provides a solid technical support for formulating scientific and reasonable energy-saving measures. This method not only ensures flexible adjustment of resource allocation without affecting the user experience, but also achieves the optimal balance between power consumption and performance in various application scenarios, significantly improving the user satisfaction and the battery life of the device. By organically combining topological sorting, mixed integer programming, branch and bound method, simulated annealing, tabu search, dynamic programming, collaborative optimization and genetic algorithm, the system has opened up an effective path from complex data analysis to intuitive strategy formulation, injecting new vitality into intelligent power consumption management.

[0038] In a specific embodiment, the step-by-step analysis of the power consumption scheduling strategy to obtain a control instruction sequence and the real-time dynamic adjustment of the control unit of the target mobile phone based on the control instruction sequence to achieve the optimal balance between power consumption and performance include: Parse the instructions of the power consumption scheduling strategy through a lexical analyzer to obtain an instruction token sequence, and construct a syntax tree for the instruction token sequence to obtain a scheduling parse tree; Based on a semantic analyzer, perform dependency analysis on the scheduling parse tree to obtain a control dependency graph, and perform instruction partitioning on the control dependency graph to obtain an atomic instruction set; wherein, the atomic instruction set includes CPU frequency regulation instructions, GPU voltage regulation instructions, and memory bandwidth allocation instructions; Perform parallelism analysis on the atomic instruction set to obtain an instruction scheduling table; wherein, the instruction scheduling table includes critical path instructions, parallel execution instructions, and serial execution instructions; Based on a scheduling orchestrator, perform timing sorting on the instruction scheduling table to obtain an execution timing chain, and perform priority partitioning on the execution timing chain to obtain a hierarchical control sequence; Detect whether there is an abnormality in the hierarchical control sequence through a fault-tolerant processor. If so, extract the abnormal hierarchical control sequence in the hierarchical control sequence, remove the abnormal hierarchical control sequence from the hierarchical control sequence to obtain a hierarchical control sequence after removal, and perform sequence reconstruction on the hierarchical control sequence after removal to obtain a corrected hierarchical control sequence; Based on the corrected hierarchical control sequence, perform hardware mapping on the control unit of the target mobile phone to obtain a control instruction sequence, and based on the control instruction sequence, perform real-time dynamic adjustment on the control unit of the target mobile phone to achieve the optimal balance between power consumption and performance.

[0039] Specifically, the process of step-by-step analysis of the power consumption scheduling strategy to obtain a control instruction sequence and dynamically adjusting the control unit of the target mobile phone in real time based on the control instruction sequence to achieve the optimal balance between power consumption and performance is a complex process involving multiple levels of analysis and processing. This process first performs instruction parsing on the power consumption scheduling strategy through a lexical analyzer to obtain an instruction token sequence. In a video playback scenario, this can be understood as transforming a pre-determined power management strategy into a series of discrete operation steps. For example, when a user switches from watching a normal video to content with a higher resolution or frame rate, the system needs to adjust the working states of hardware resources such as the CPU and GPU to adapt to the new requirements. The task of the lexical analyzer is to identify the basic elements of these operation commands, such as "increase", "decrease", or "remain unchanged", and tokenize them for subsequent processing. Next, in order to further understand the relationship between each instruction and how they jointly act on the overall power consumption scheduling, the system constructs a syntax tree for the instruction token sequence to obtain a scheduling parse tree. In this example, the syntax tree can show the hierarchical structure and logical association between different operation commands, such as the sequential arrangement of changing the CPU frequency first and then adjusting the GPU voltage. This not only helps ensure that all necessary adjustments are executed in the correct order, but also helps identify which instructions are interdependent and which can run independently. By constructing this structured representation, the system can more accurately plan the cooperation mode between each component, thereby optimizing the working efficiency of the entire device. Subsequently, based on a semantic analyzer, dependency analysis is performed on the scheduling parse tree to obtain a control dependency graph, and instruction partitioning is performed on the control dependency graph to obtain an atomic instruction set. The semantic analyzer mentioned here is responsible for examining the actual meaning of each instruction and its impact on other instructions, and then drawing a graph showing the interaction mode between each instruction - that is, the control dependency graph. For video playback, this means clearly pointing out the causal relationship between the CPU frequency regulation instruction, the GPU voltage regulation instruction, and the memory bandwidth allocation instruction. For example, before starting a high-definition video stream, it may be necessary to first increase the CPU frequency to speed up the decoding speed, then correspondingly increase the GPU voltage to support higher rendering quality, and at the same time appropriately adjust the memory bandwidth to ensure unobstructed data transmission. Through careful partitioning of these instructions, a set of atomic instruction sets with the smallest units is finally formed, each of which is an indivisible basic operation. With these atomic instructions, the system performs parallelism analysis on the atomic instruction set to obtain an instruction schedule table. This stage mainly explores which instructions can be executed concurrently at the same time and which must be completed sequentially to determine the best task execution plan. During video playback, some operations may not have a direct dependency relationship with each other, so they can be performed simultaneously; while others need to wait until the previous task ends before they can start.In this way, the system can make the most of the advantages of multi-core processors as much as possible without affecting the results, improving the processing speed while reducing the waiting time. The resulting instruction scheduling table details the critical path instructions (those key tasks that determine the overall progress), parallel execution instructions (tasks that can be processed simultaneously), and serial execution instructions (tasks that need to be executed in sequence), providing a basis for the next scheduling and arrangement. On this basis, the scheduling orchestrator performs a timing sort on the instruction scheduling table to obtain an execution timing chain, and divides the execution timing chain into priorities to obtain a hierarchical control sequence. The role of the scheduling orchestrator is to formulate a clear schedule according to the order and importance of the instructions to guide the implementation of various tasks. Considering different situations that may occur during video playback, such as suddenly loading large files or receiving new input instructions, the system needs to flexibly adjust its work plan to cope with emergencies. By assigning corresponding priority tags to different instructions, it can be ensured that the most critical operations are always processed first, maintaining a good user experience even under limited resources. The hierarchical control sequence formed in this way not only reflects the relative urgency between the instructions, but also reflects their status and role in the entire power consumption management strategy. However, due to the often uncertain actual operating environment, unexpected situations may occur, resulting in the original plan being unable to be executed smoothly. Therefore, it is particularly important to detect whether there are abnormalities in the hierarchical control sequence through a fault-tolerant processor. Once any unexpected behavior or error is found, the system will immediately take measures to correct it. Specifically, if there is an abnormality, the abnormal hierarchical control sequence in the hierarchical control sequence is extracted, and the abnormal hierarchical control sequence is removed from the hierarchical control sequence to obtain a post-removal hierarchical control sequence, and the post-removal hierarchical control sequence is reconstructed to obtain a corrected hierarchical control sequence. For example, if a certain instruction fails to be completed on time or has an adverse effect, then it will be removed, and the relationship between the remaining instructions will be re-evaluated to ensure that the entire sequence is still coherent and effective. This mechanism greatly enhances the robustness and reliability of the system, enabling it to better cope with various challenges. Finally, based on the corrected hierarchical control sequence, a hardware mapping is performed on the control unit of the target mobile phone to obtain a control instruction sequence, and the control unit of the target mobile phone is dynamically adjusted in real time based on the control instruction sequence to achieve the optimal balance between power consumption and performance. Hardware mapping refers to converting abstract software instructions into specific physical operations so that they can directly act on various components inside the mobile phone. For the video playback scenario, this means implementing all power consumption management and performance optimization measures on specific CPU, GPU, and memory settings, such as adjusting parameters such as frequency, voltage, or bandwidth. By continuously updating and applying these control instruction sequences, the system can respond in real time to the user's operations and changes in the external environment, always maintaining an energy-saving and efficient operating state.Finally, this method not only effectively controls the power consumption of the mobile phone, but also ensures excellent performance in application scenarios such as video playback, greatly improving user satisfaction and experience. In summary, by deeply analyzing the power consumption scheduling strategy and converting it into an accurate control instruction sequence, the system successfully builds a bridge from high-level strategies to low-level hardware operations. This process fully considers the logical relationships, execution order, and possible problems between instructions, ensuring that each step can be executed accurately. More importantly, it provides a dynamic and flexible way to manage power consumption, allowing the system to respond quickly according to the actual situation, thus achieving the best energy efficiency ratio. Whether facing daily use or special requirements, this mechanism can provide users with stable and reliable performance support while minimizing energy consumption.

[0040] In a specific embodiment, the construction of a syntax tree for the instruction tag sequence to obtain a scheduling parse tree includes: Dividing the instruction tag sequence into syntax units through enhanced two-way syntax analysis to obtain a syntax unit set, and performing semantic annotation on the syntax unit set to obtain an annotated syntax unit table; Based on a bottom-up reduction algorithm, matching syntax rules to the annotated syntax unit table to obtain a reduction sequence graph; Performing semantic attribute calculation on the reduction sequence graph through attribute grammar derivation to obtain a semantic attribute table, and performing dependency analysis on the semantic attribute table to obtain an attribute dependency network, where the attribute dependency network includes instruction execution attributes, resource consumption attributes, and timing constraint attributes; Based on a recursive descent parser, performing network parsing and subtree generation on the attribute dependency network to obtain a subtree set, and performing pruning optimization on the subtree set to obtain an optimized subtree group; Merging the tree structures of the optimized subtree group through a syntax forest synthesis algorithm to obtain a syntax forest, and performing state transition analysis on the syntax forest based on a tree automaton to obtain a scheduling parse tree.

[0041] Specifically, the process of constructing a scheduling parse tree from the instruction token sequence is a complex and delicate semantic and structural analysis process. In this process, the system first divides the instruction token sequence into grammar units through enhanced bidirectional grammar analysis technology, obtains a grammar unit set, and performs semantic annotation on the grammar unit set to obtain an annotated grammar unit table. Enhanced bidirectional grammar analysis is an advanced natural language processing method, which is used here to accurately identify and classify each operation command and its parameters in the power consumption scheduling strategy. For example, in a video playback scenario, when the user switches from watching a video in normal resolution to higher-resolution content, the system needs to adjust the working states of hardware resources such as CPU frequency, GPU voltage, and memory bandwidth. Enhanced bidirectional grammar analysis can break down these complex operations into smaller grammar units, such as "increase", "decrease", or "remain unchanged", and associate them with specific hardware components to form a detailed grammar unit set. Subsequently, by performing semantic annotation on these grammar units, a clear meaning and role can be assigned to each unit to ensure the accuracy and consistency of subsequent processing. Next, based on the bottom-up reduction algorithm, grammar rule matching is performed on the annotated grammar unit table to obtain a reduction sequence diagram. The bottom-up reduction algorithm is a method of gradually constructing a high-level structure starting from the most basic elements at the bottom layer, which is used here to organize and combine each grammar unit according to predefined grammar rules. For video playback, this means arranging different types of instructions in a certain logical order, such as changing the CPU frequency first and then adjusting the GPU voltage. The reduction sequence diagram shows the hierarchical relationship and interdependence between these instructions, enabling the system to better understand the position and function of each operation in the entire scheduling strategy. In this way, not only is it ensured that all necessary adjustments can be executed in the correct order, but it also helps to identify which instructions are interdependent and which can run independently. To further explore the deep meaning behind these instructions, the system performs semantic attribute calculation on the reduction sequence diagram through attribute grammar derivation to obtain a semantic attribute table, and performs dependency analysis on the semantic attribute table to obtain an attribute dependency network. Attribute grammar derivation is a formal tool used to describe language structures and their related attributes, which is used here to quantify the operation effects of each instruction, such as instruction execution attributes (i.e., the functions of the instructions themselves), resource consumption attributes (the hardware resources involved and their usage amounts), and timing constraint attributes (the time requirements for instruction execution). In a video playback scenario, this may mean clearly indicating whether a certain instruction will cause an increase in power consumption, whether it will occupy a large amount of memory bandwidth, or whether it must be completed within a specific time. Through dependency analysis, the system can reveal the internal connections between these attributes and construct an attribute dependency network that comprehensively reflects the interaction patterns between instructions.This network not only reflects the causal relationships among the instructions but also provides information on how to coordinate them to achieve optimal performance and energy efficiency. Based on this, network parsing and subtree generation are performed on the attribute dependency network using a recursive descent parser to obtain a set of subtrees, and pruning optimization is carried out on the set of subtrees to obtain an optimized subtree group. A recursive descent parser is a commonly used syntax parsing method that parses the input data structure layer by layer through recursive function calls. Here, the parser is responsible for generating a series of subtrees according to the connection relationships in the attribute dependency network, and each subtree represents a group of related instructions and their attributes. For example, during video playback, there may be a subtree dedicated to handling all operations related to CPU frequency adjustment, and another subtree focused on changes in GPU voltage. By pruning and optimizing these subtrees, redundant parts can be removed, and the most core and effective information can be retained, thus simplifying the overall structure and improving the parsing efficiency. The finally formed optimized subtree group is not only more compact and concise but also easier to understand and apply. Finally, the tree structure of the optimized subtree group is merged using the syntax forest synthesis algorithm to obtain a syntax forest, and state transition analysis is performed on the syntax forest based on a tree automaton to obtain a scheduling parse tree. The syntax forest synthesis algorithm aims to integrate multiple independent subtrees into a complete syntax tree to represent the overall picture of the entire power consumption scheduling strategy. In this example, this means unifying all operation instructions related to video playback in a coherent structure, showing the sequence and mutual influence among them. The tree automaton provides an effective method to simulate and analyze the state change process of this structure, ensuring that the instructions at each node can be correctly executed according to the predefined rules. The finally obtained scheduling parse tree not only clearly presents the specific implementation steps of the power consumption management strategy but also reflects the logical relationships and timing requirements among these steps, laying a solid foundation for the subsequent generation of control instructions. In summary, the process of obtaining a scheduling parse tree by constructing a syntax tree from the instruction token sequence not only realizes a deep analysis of the power consumption scheduling strategy but also provides strong technical support for formulating scientific and reasonable energy-saving measures. This method not only ensures flexible adjustment of resource allocation without affecting the user experience but also achieves the optimal balance between power consumption and performance in various application scenarios, significantly improving user satisfaction and the battery life of the device. By combining enhanced two-way syntax analysis, bottom-up reduction algorithm, attribute grammar derivation, recursive descent parser, and syntax forest synthesis algorithm, the system has opened up an effective path from complex data analysis to intuitive strategy formulation, injecting new vitality into intelligent power management. At the same time, this set of mechanisms also has high flexibility and adaptability, and can always maintain efficient operation in the face of changing application environments, providing stable and reliable performance support for users while maximizing energy consumption savings.In a video playback scenario, this means that the system can not only provide a smooth audio-visual experience, but also intelligently optimize resource allocation, extend battery life, and enable users to enjoy more durable and high-quality multimedia services.

[0042] The mobile phone power consumption optimization method in the embodiments of the present invention has been described above. Next, the mobile phone power consumption optimization device in the embodiments of the present invention will be described. Please refer to Figure 2 , an embodiment of the mobile phone power consumption optimization device in the embodiments of the present invention includes: An acquisition module 21, configured to perform multi-dimensional real-time acquisition on the target mobile phone operation data to obtain power consumption scenario data; An analysis module 22, configured to perform an operation timing analysis on the target mobile phone based on the power consumption scenario data to obtain a power consumption state transition chain; A calculation module 23, configured to calculate the weights of the power consumption scenario data in the power consumption state transition chain by using the hybrid entropy weight method to obtain a scenario priority table; A solution module 24, configured to perform a power consumption planning solution on the target mobile phone based on the scenario priority table to obtain a power consumption scheduling strategy; An analysis module 25, configured to perform a step-by-step analysis on the power consumption scheduling strategy to obtain a control instruction sequence, and based on the control instruction sequence, dynamically adjust the control unit of the target mobile phone in real time to achieve the optimal balance between power consumption and performance.

[0043] In this embodiment, for the specific implementation of each unit in the above device embodiment, please refer to that described in the above method embodiment, and details will not be repeated here.

[0044] Refer to Figure 3 , the embodiments of the present invention also provide a computer device, and its internal structure can be as Figure 3 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 of the computer design 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.

[0045] Those skilled in the art can understand that Figure 3 the structure shown in only shows the block diagram of the part 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.

[0046] An embodiment of the present invention further provides a computer-readable storage medium, on which a computer program is stored. 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.

[0047] Those of ordinary skill in the art can understand that all or part of the processes of implementing the methods in the above embodiments can be completed by instructing relevant hardware through a computer program. 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 above method embodiments. Among them, any reference to a memory, storage, database, or other medium provided by the present invention and used in the embodiments can include non-volatile and / or volatile memories. Non-volatile memory can include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM), or flash memory. Volatile memory can include random access memory (RAM) or external cache memory. By way of illustration and not limitation, RAM is available in many forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), double data rate SDRAM (SSRSDRAM), enhanced SDRAM (ESDRAM), synchronous link (Synchlink) DRAM (SLDRAM), Rambus direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM), and memory bus dynamic RAM, etc.

[0048] It should be noted that in this article, the term "comprising", "including" or any other variant thereof is intended to cover non-exclusive inclusion, so that a process, device, article or method including a series of elements not only includes those elements, but also includes other elements not expressly listed, or further includes elements inherent to such process, device, article or method. Without further limitation, an element defined by the statement "including one..." does not exclude the existence of additional identical elements in the process, device, article or method including the element.

[0049] The above are only the preferred embodiments of the present invention, and do not limit the patent scope of the present invention accordingly. Any equivalent structure or equivalent process transformation made by using the specification and drawings of the present invention, or directly or indirectly applied to other related technical fields, shall be equally included in the patent protection scope of the present invention.

Claims

1. A method for optimizing the power consumption of a mobile phone, characterized in that, It includes the following steps: Perform multi-dimensional real-time collection on the running data of the target mobile phone to obtain power consumption scenario data; Based on the power consumption scenario data, perform running time sequence analysis on the target mobile phone to obtain a power consumption state migration chain; Calculate the weights of the power consumption scenario data in the power consumption state migration chain by the hybrid entropy weight method to obtain a scenario priority table; Based on the scenario priority table, perform power consumption planning and solution on the target mobile phone to obtain a power consumption scheduling strategy; Perform step-by-step analysis on the power consumption scheduling strategy to obtain a control instruction sequence, and based on the control instruction sequence, dynamically adjust the control unit of the target mobile phone in real time to achieve the optimal balance between power consumption and performance.

2. The mobile phone power consumption optimization method according to claim 1, characterized in that, The multi-dimensional real-time collection of the running data of the target mobile phone to obtain power consumption scenario data includes: Real-time collect the hardware parameters of the target mobile phone through a preset sensor group to obtain an original power consumption data set; wherein, the original power consumption data set includes CPU power consumption data, GPU power consumption data, memory power consumption data, and battery temperature; Based on the original power consumption data set, perform hardware resource tracking analysis on the target mobile phone to obtain a resource dynamic load vector; Extract power consumption scenario features from the resource dynamic load vector through a multi-dimensional data fusion algorithm to obtain power consumption scenario data; wherein, the power consumption scenario data includes game scenarios, video scenarios, web browsing scenarios, and standby scenarios.

3. The mobile phone power consumption optimization method according to claim 1, wherein The running time sequence analysis based on the power consumption scenario data on the target mobile phone to obtain a power consumption state migration chain includes: Based on discrete wavelet analysis, perform time sequence segmentation on the power consumption scenario data to obtain a scenario time series, and perform Fourier transform on the scenario time series to obtain scenario spectrum features; Perform state recognition on the scenario spectrum features through adaptive edge detection to obtain a scenario state sequence, and perform hierarchical clustering on the scenario state sequence to obtain a state classification result; wherein, the state classification result includes a high-frequency switching state, a stable operation state, and a low-frequency switching state; Based on a recurrent neural network, perform time dependence analysis on the state classification result to obtain a state transition probability, and based on the state transition probability, perform Bayesian state inference on the target mobile phone to obtain a state prediction sequence; wherein, the state prediction sequence includes scenario duration, switching time points, and transfer directions; Calculate the similarity of the state prediction sequence through the dynamic time warping algorithm to obtain a state distance matrix, and perform spectral clustering on the state distance matrix to obtain a state cluster set, wherein the state cluster set includes a high power consumption cluster, a medium power consumption cluster, and a low power consumption cluster; Based on the Markov decision process, perform transfer law analysis on the state cluster set to obtain a state transfer map, and based on the state transfer map, perform path extraction and path optimization to obtain an optimized transfer path; wherein, the optimized transfer path includes the shortest path, the optimal switching point, and transfer cost; Through a graph neural network, based on the optimized transfer path, perform link construction to obtain a power consumption state migration chain.

4. The mobile phone power consumption optimization method according to claim 1, wherein Calculating the weights of the power consumption scenario data in the power consumption state migration chain through the hybrid entropy weight method to obtain a scenario priority table, including: Extracting features from the power consumption scenario data in the power consumption state migration chain based on the information entropy algorithm to obtain a state entropy value sequence, and normalizing the state entropy value sequence to obtain a normalized entropy value matrix; Allocating weights to the normalized entropy value matrix through a multi-criteria decision analysis method to obtain an initial weight vector, and performing constraint optimization on the initial weight vector to obtain a tuned weight set; Performing hierarchical analysis on the power consumption state migration chain based on the tuned weight set to obtain a scenario priority table; Calculating the grey correlation degree of the scenario priority table through grey correlation analysis to obtain a scenario correlation degree sequence, and performing orthogonal decomposition on the scenario correlation degree sequence to obtain scenario feature weights; wherein, the scenario feature weights include scenario switching cost, resource consumption ratio, and performance impact factor; Performing priority sorting on the scenario feature weights based on the coefficient of variation method to obtain a scenario priority sequence, and detecting whether there is a circular dependency conflict in the scenario priority sequence. If so, excluding the circular dependency conflict in the scenario priority sequence to obtain a conflict-free scenario set; Performing priority mapping on the power consumption scenario data based on the conflict-free scenario set through a multi-objective optimization algorithm to obtain a scenario priority table.

5. The mobile phone power consumption optimization method according to claim 1, wherein Solving the power consumption planning of the target mobile phone based on the scenario priority table to obtain a power consumption scheduling strategy, including: Analyzing the dependency relationship of the scenario priority table through a topological sorting algorithm to obtain a scenario scheduling graph, and performing constraint propagation on the scenario scheduling graph to obtain a scheduling constraint set; wherein, the scheduling constraint set includes resource capacity constraint, scenario switching constraint, and timing dependency constraint; Solving and analyzing the scheduling constraint set based on mixed integer programming to obtain an initial scheduling plan, and performing a spatial search on the initial scheduling plan through the branch and bound method to obtain a candidate plan set; wherein, the candidate plan set includes a power consumption threshold interval, a switching time window, and a resource allocation ratio; Performing temperature regulation calculation on the candidate plan set through a simulated annealing algorithm to obtain a cooling strategy sequence, and performing a solution space exploration based on the cooling strategy sequence to obtain a locally optimal solution set; wherein, the locally optimal solution set includes power consumption control parameters, scheduling time points, and resource allocation schemes; Performing neighborhood traversal on the locally optimal solution set based on a tabu search algorithm to obtain a global search path, and performing state transition calculation on the global search path through dynamic programming technology to obtain an optimal scheduling sequence; wherein, the optimal scheduling sequence includes a scenario switching order, a resource allocation strategy, and a power consumption control instruction; Performing multi-objective trade-off on the optimal scheduling sequence through a collaborative optimization algorithm to obtain a scheduling evaluation index, and screening the plan based on the scheduling evaluation index to obtain an optimal scheduling plan; wherein, the optimal scheduling plan includes a scenario execution duration, a resource utilization efficiency, and a power consumption balance coefficient; Performing population evolution calculation on the optimal scheduling scheme based on a genetic algorithm to obtain an evolution scheme sequence, and performing non-dominated sorting on the evolution scheme sequence through a Pareto optimal set to obtain a power consumption scheduling strategy; wherein, the power consumption scheduling strategy includes a scenario scheduling time sequence table, a resource allocation matrix, and a power consumption control parameter set.

6. The mobile phone power consumption optimization method according to claim 1, wherein Performing step-by-step analysis on the power consumption scheduling strategy to obtain a control instruction sequence, and dynamically adjusting the control unit of the target mobile phone in real time based on the control instruction sequence to achieve the optimal balance between power consumption and performance, including: Performing instruction parsing on the power consumption scheduling strategy through a lexical analyzer to obtain an instruction token sequence, and constructing a syntax tree for the instruction token sequence to obtain a scheduling parse tree; Performing dependency analysis on the scheduling parse tree based on a semantic analyzer to obtain a control dependency graph, and performing instruction partitioning on the control dependency graph to obtain an atomic instruction set; wherein, the atomic instruction set includes CPU frequency regulation instructions, GPU voltage regulation instructions, and memory bandwidth allocation instructions; Performing parallelism analysis on the atomic instruction set to obtain an instruction scheduling table; wherein, the instruction scheduling table includes critical path instructions, parallel execution instructions, and serial execution instructions; Performing time sequence sorting on the instruction scheduling table based on a scheduling orchestrator to obtain an execution time sequence chain, and performing priority partitioning on the execution time sequence chain to obtain a hierarchical control sequence; Detecting whether there is an abnormality in the hierarchical control sequence through a fault-tolerant processor. If so, extracting the abnormal hierarchical control sequence in the hierarchical control sequence, removing the abnormal hierarchical control sequence from the hierarchical control sequence to obtain a hierarchical control sequence after removal, and performing sequence reconstruction on the hierarchical control sequence after removal to obtain a corrected hierarchical control sequence; Performing hardware mapping on the control unit of the target mobile phone based on the corrected hierarchical control sequence to obtain a control instruction sequence, and dynamically adjusting the control unit of the target mobile phone in real time based on the control instruction sequence to achieve the optimal balance between power consumption and performance.

7. The mobile phone power consumption optimization method according to claim 6, characterized in that The constructing a syntax tree for the instruction token sequence to obtain a scheduling parse tree includes: Performing syntax unit partitioning on the instruction token sequence through enhanced two-way syntax analysis to obtain a syntax unit set, and performing semantic annotation on the syntax unit set to obtain an annotated syntax unit table; Performing syntax rule matching on the annotated syntax unit table based on a bottom-up reduction algorithm to obtain a reduction sequence graph; Performing semantic attribute calculation on the reduction sequence graph through attribute grammar derivation to obtain a semantic attribute table, and performing dependency analysis on the semantic attribute table to obtain an attribute dependency network, wherein the attribute dependency network includes instruction execution attributes, resource consumption attributes, and timing constraint attributes; Performing network parsing and subtree generation on the attribute dependency network based on a recursive descent parser to obtain a subtree set, and performing pruning optimization on the subtree set to obtain an optimized subtree group; Performing tree structure merging on the optimized subtree group through a syntax forest synthesis algorithm to obtain a syntax forest, and performing state transition analysis on the syntax forest based on a tree automaton to obtain a scheduling parse tree.

8. A mobile phone power consumption optimization device, characterized in that, Including: A collection module, which is used to perform multi-dimensional real-time collection on the running data of the target mobile phone to obtain power consumption scenario data; An analysis module, which is used to perform running time sequence analysis on the target mobile phone based on the power consumption scenario data to obtain a power consumption state migration chain; A calculation module, which is used to calculate the weights of the power consumption scenario data in the power consumption state migration chain by the hybrid entropy weight method to obtain a scenario priority table; A solution module, which is used to perform power consumption planning solution on the target mobile phone based on the scenario priority table to obtain a power consumption scheduling strategy; An analysis module, which is used to perform step-by-step analysis on the power consumption scheduling strategy to obtain a control instruction sequence, and based on the control instruction sequence, dynamically adjust the control unit of the target mobile phone in real time to achieve the optimal balance between power consumption and performance.

9. A computer device, comprising a memory and a processor, wherein a computer program is stored in the memory, characterized in that, When the processor executes the computer program, it implements the steps of the method according to any one of claims 1 to 7.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by the processor, it implements the steps of the method according to any one of claims 1 to 7.

Images