Dynamic power consumption adjustment method and device for multi-mode SOC main control chip

Abstract

The present invention relates to the technical field of chip control, and discloses a dynamic power consumption regulation method and device for a multi-mode SOC main control chip. The present invention obtains chip load characteristics of the SOC main control chip, analyzes the load characteristics according to a preset power consumption regulation scheme, determines a corresponding power consumption mode and adjusts the power consumption mode, and transmits the chip load characteristics to a cloud computing platform through a network. The cloud platform analyzes chip user portraits based on the load characteristics, analyzes expected behaviors of current load characteristics according to the user portraits, predicts the load requirements of the SOC main control chip, matches the expected load characteristics based on a cloud strategy library, obtains an optimal regulation strategy and adjusts the power consumption mode. The power consumption mode can be dynamically adjusted according to actual load characteristics to improve energy efficiency, and accurate regulation of different usage scenarios can be achieved through user portrait analysis. The cloud platform ensures timely updating and precise execution of strategies, thereby solving the problem that the power consumption regulation mode in the prior art cannot achieve the optimal effect.

Description

Technical Field

[0001] The present invention relates to the technical field of chip control, and in particular to a method and device for dynamic power consumption regulation of a multi-mode SOC main control chip. Background Art

[0002] With the increasing popularity of smart devices and the continuous development of SOC (system-on-chip) technology, especially in applications such as mobile devices, embedded systems and IoT terminals, SOC main control chips, as core computing units, undertake complex computing tasks and multiple functions. The high performance and low power consumption of SOC main control chips have become key requirements in the design of modern electronic products.

[0003] Currently, most power consumption adjustment methods for SOC main control chips rely on static load characteristics and preset power consumption modes. This method is difficult to cope with dynamically changing workloads. Different application scenarios and user needs will affect the load characteristics and power consumption modes of the SOC main control chip. Traditional power consumption adjustment methods often ignore the personalized needs of users, resulting in poor universality of adjustment strategies. Summary of the invention

[0004] The object of the present invention is to provide a method and device for dynamic power consumption regulation of a multi-mode SOC main control chip, aiming to solve the problem that the power consumption regulation mode in the prior art cannot achieve the optimal effect.

[0005] The present invention is implemented in this way. In a first aspect, the present invention provides a dynamic power consumption adjustment method for a multi-mode SOC main control chip, comprising:

[0006] Collecting chip load data of the SOC main control chip to obtain chip load characteristics of the SOC main control chip;

[0007] Analyze the chip load characteristics according to a preset power consumption adjustment scheme to obtain a chip power consumption mode corresponding to the chip load characteristics, and adjust the power consumption mode of the SOC main control chip according to the chip power consumption mode;

[0008] The chip load characteristics are transmitted to a cloud computing platform through a network channel, so that the cloud computing platform can analyze the user profile of the SOC main control chip according to the chip load characteristics to obtain the user profile of the SOC main control chip;

[0009] Receiving a user profile from a cloud computing platform through a network channel, and analyzing the expected behavior of the chip load characteristics at the current moment according to the user profile to obtain the expected load characteristics of the SOC main control chip;

[0010] The expected load characteristics are matched based on a cloud policy library to obtain an optimal adjustment strategy corresponding to the expected load characteristics, and the power consumption mode of the SOC main control chip is adjusted according to the optimal adjustment strategy.

[0011] In a second aspect, the present invention provides a dynamic power consumption adjustment device for a multi-mode SOC main control chip, which is used to implement a dynamic power consumption adjustment method for a multi-mode SOC main control chip as described in any one of the first aspects, including:

[0012] A data acquisition module is used to collect chip load data of the SOC main control chip to obtain the chip load characteristics of the SOC main control chip;

[0013] A mode adjustment module, used for analyzing the chip load characteristics according to a preset power consumption adjustment scheme to obtain a chip power consumption mode corresponding to the chip load characteristics, and adjusting the power consumption mode of the SOC main control chip according to the chip power consumption mode;

[0014] A portrait analysis module, used to transmit the chip load characteristics to the cloud computing platform through a network channel, so that the cloud computing platform can analyze the user portrait of the SOC main control chip according to the chip load characteristics to obtain the user portrait of the SOC main control chip;

[0015] An expected analysis module, used to receive a user profile from a cloud computing platform through a network channel, and analyze the expected behavior of the chip load characteristics at the current moment according to the user profile, so as to obtain the expected load characteristics of the SOC main control chip;

[0016] The strategy matching module is used to perform strategy matching on the expected load characteristics based on the cloud strategy library to obtain the optimal adjustment strategy corresponding to the expected load characteristics, and adjust the power consumption mode of the SOC main control chip according to the optimal adjustment strategy.

[0017] The present invention provides a method for dynamic power consumption adjustment of a multi-mode SOC main control chip, which has the following beneficial effects:

[0018] The present invention collects real-time load data of the SOC main control chip, obtains the chip load characteristics, analyzes the load characteristics according to a preset power consumption adjustment scheme, determines the corresponding power consumption mode, and adjusts the power consumption mode. The chip load characteristics are transmitted to the cloud computing platform through the network. The cloud platform analyzes the chip user portrait based on the load characteristics, analyzes the expected behavior of the current load characteristics according to the user portrait, predicts the load demand of the SOC main control chip, matches the expected load characteristics based on the cloud policy library, obtains the optimal adjustment strategy, and adjusts the power consumption mode. The power consumption mode can be dynamically adjusted according to the actual load characteristics to improve energy efficiency. Through user portrait analysis, accurate adjustment for different usage scenarios is achieved. The cloud platform ensures timely updating and accurate execution of the strategy, which solves the problem that the power consumption adjustment mode in the prior art cannot achieve the optimal effect. BRIEF DESCRIPTION OF THE DRAWINGS

[0019] Figure 1 It is a schematic diagram of the steps of a dynamic power consumption adjustment method of a multi-mode SOC main control chip provided by an embodiment of the present invention;

[0020] Figure 2 It is a structural schematic diagram of a dynamic power consumption adjustment device for a multi-mode SOC main control chip provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0021] In order to make the purpose, technical solution and advantages of the present invention more clearly understood, the present invention is further described in detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not intended to limit the present invention.

[0022] The implementation of the present invention is described in detail below in conjunction with specific embodiments.

[0023] Reference Figure 1 , Figure 2 As shown, a preferred embodiment of the present invention is provided.

[0024] In a first aspect, the present invention provides a method for dynamic power consumption adjustment of a multi-mode SOC main control chip, comprising:

[0025] S1: collecting chip load data of the SOC main control chip to obtain chip load characteristics of the SOC main control chip;

[0026] S2: analyzing the chip load characteristics according to a preset power consumption adjustment scheme to obtain a chip power consumption mode corresponding to the chip load characteristics, and adjusting the power consumption mode of the SOC main control chip according to the chip power consumption mode;

[0027] S3: transmitting the chip load characteristics to the cloud computing platform through a network channel, so that the cloud computing platform can analyze the user portrait of the SOC main control chip according to the chip load characteristics to obtain the user portrait of the SOC main control chip;

[0028] S4: receiving a user profile from a cloud computing platform through a network channel, and performing an expected behavior analysis on the chip load characteristics at the current moment according to the user profile to obtain an expected load characteristic of the SOC main control chip;

[0029] S5: Performing strategy matching on the expected load characteristics based on the cloud strategy library to obtain an optimal adjustment strategy corresponding to the expected load characteristics, and adjusting the power consumption mode of the SOC main control chip according to the optimal adjustment strategy.

[0030] Specifically, in step S1 of the embodiment provided by the present invention, chip load data of the SOC main control chip is collected to obtain the chip load characteristics of the SOC main control chip.

[0031] It should be noted that the SOC (System on Chip) main control chip is an integrated circuit that integrates almost all the functional modules required by the computing system into a single chip. It can contain various different components, such as: Central Processing Unit (CPU): used to execute software instructions and perform data processing, Graphics Processing Unit (GPU): used for graphics rendering and image processing, etc. The core advantage of the SOC main control chip lies in its high integration, which can significantly reduce the physical size, power consumption and cost of the system, while improving the performance and reliability of the system. It can be widely used in various devices and fields, including but not limited to: smartphones and tablets, Internet of Things devices, and embedded systems.

[0032] It is understandable that in order to achieve requirements such as extending equipment life, reducing power consumption, and heat dissipation management, the SOC main control chip can set multiple power consumption modes to cope with different working scenarios, thereby achieving a balance between performance requirements and equipment life, power consumption index, heat dissipation management, and other aspects.

[0033] Furthermore, in order to adjust the power consumption mode of the SOC main control chip, it is first necessary to obtain the scenario conditions faced by the SOC main control chip. Therefore, in the technical solution of the present invention, the chip load data of the SOC main control chip is first collected to obtain the chip load characteristics of the SOC main control chip, so as to subsequently adjust the power consumption mode of the SOC main control chip according to this chip load characteristic.

[0034] Specifically, in step S2 of the embodiment provided by the present invention, the chip load characteristics are analyzed according to a preset power consumption adjustment scheme to obtain a chip power consumption mode corresponding to the chip load characteristics, and the power consumption mode of the SOC main control chip is adjusted according to the chip power consumption mode.

[0035] More specifically, in the technical solution of the present invention, there are two methods for adjusting chip load characteristics. The first is to analyze the chip load characteristics based on the existing preset power consumption adjustment scheme to obtain the chip power consumption mode corresponding to the chip load characteristics. The second is to predict the expected behavior of the chip load characteristics and match it with the optimal adjustment strategy according to the user portrait of the terminal where the SOC main control chip is located in a subsequent step.

[0036] More specifically, the difference between the first method and the second method is that the first method has fewer power consumption adjustment modes and the differences between each power consumption adjustment mode are large, that is, it makes approximate adjustments to different scenarios and cannot achieve the optimal response to the current chip load. The second method predicts the expected behavior of the chip load characteristics according to the user portrait, and calls the most optimized adjustment strategy based on the prediction results. Taking into account the possible duration and possible conversion direction of the existing chip load characteristics in the future, the most adaptable and detailed power consumption mode is called.

[0037] It should be noted that the second method requires a certain amount of testing time, and the use effect of the first method in the past usage records is analyzed to obtain reference materials for the second method, so the first method needs to be executed before the second method.

[0038] Specifically, in step S3 of the embodiment provided by the present invention, the chip load characteristics are transmitted to the cloud computing platform through a network channel, so that the cloud computing platform can analyze the user portrait of the SOC main control chip according to the chip load characteristics to obtain the user portrait of the SOC main control chip.

[0039] More specifically, the cloud computing platform simultaneously receives chip load characteristics from each SOC main control chip, and the cloud computing platform stores the chip load characteristics of each SOC main control chip in the past usage records, and performs interactive, sequential, and comprehensive analysis based on the chip load characteristics of each SOC main control chip in the past usage records to obtain a user portrait of the terminal where each SOC main control chip is located.

[0040] More specifically, the role of user portraits is to describe the usage tendencies of different types of users for the SOC main control chip. Based on user portraits, future changes in the chip load characteristics at the current moment can be predicted. For example, if the chip load characteristic executed at the current time node is a certain game, how long the game is usually used by users of the terminal, and the tendency to use other applications after the game is completed. The prediction results based on user portraits can be used to fine-tune the functions of the SOC main control chip to optimize the required functions and reduce the inadaptable functions in the process.

[0041] Specifically, in step S4 of the embodiment provided by the present invention, a user profile is received from a cloud computing platform through a network channel, and the expected behavior of the chip load characteristics at the current moment is analyzed based on the user profile to obtain the expected load characteristics of the SOC main control chip.

[0042] More specifically, the SOC main control chip establishes a stable connection with the cloud computing platform through a network channel to ensure the reliability and real-time performance of data transmission, receives user portrait data from the cloud computing platform through a predetermined transmission protocol, decrypts the received user portrait data to ensure the integrity and security of the data, parses the user portrait data, extracts key information such as user usage habits, performance requirements, power consumption preferences, etc., and stores the parsed user portrait data in a local storage unit for easy access at any time.

[0043] More specifically, based on user portrait data and current load characteristic data, an expected behavior analysis model is constructed. This model can predict user behavior patterns in different scenarios based on machine learning, deep learning and other technologies. The expected behavior analysis model is used to analyze the chip load characteristics at the current moment, predict possible load changes and behavior patterns in the future, and fuse the user portrait data with the current load characteristic data to comprehensively analyze user needs and current system status. Based on the fusion analysis results, the expected load characteristics of the SOC main control chip are generated to predict load changes in the future.

[0044] More specifically, according to the expected load characteristics, the resource scheduling and power consumption management strategies of the SOC main control chip are adjusted to provide the best performance and energy efficiency balance under different load conditions. According to the real-time monitoring data and changes in expected load characteristics, the prediction model and management strategy are continuously optimized to improve the system's adaptability and intelligence level.

[0045] Specifically, in step S5 of the embodiment provided by the present invention, strategy matching is performed on the expected load characteristics based on the cloud policy library to obtain the optimal adjustment strategy corresponding to the expected load characteristics, and the power consumption mode of the SOC main control chip is adjusted according to the optimal adjustment strategy.

[0046] More specifically, the locally generated expected load characteristic data is uploaded to the cloud through a secure network channel. During the upload process, the data is verified for integrity and accuracy to ensure that the data is correct. The data is connected to the policy library in the cloud to retrieve policies related to the uploaded expected load characteristics. The policy matching algorithm (such as rule matching, machine learning models, etc.) is used to find the policy that best matches the expected load characteristics.

[0047] More specifically, there may be multiple matching strategies in the strategy library. According to the preset priority or weighted scoring mechanism, the optimal adjustment strategy is screened out, and the selected optimal adjustment strategy is evaluated to ensure its effectiveness and reliability under the current expected load characteristics. The selected optimal adjustment strategy is sent to the SOC main control chip through the network channel. After receiving the adjustment strategy, the SOC main control chip parses it to obtain specific adjustment parameters and operation instructions.

[0048] More specifically, according to the analyzed adjustment strategy, the power consumption mode of the SOC main control chip is switched. For example, the CPU / GPU frequency, voltage, memory status, etc. are adjusted. According to the specific parameters in the strategy, the power consumption of each module of the SOC main control chip is accurately adjusted. After the adjustment strategy is applied, the power consumption and performance status of the SOC main control chip are monitored in real time to ensure that the adjustment effect meets expectations. According to the real-time monitoring data, the strategy is fine-tuned or re-matched when necessary to ensure stable and efficient operation of the system.

[0049] It should be noted that the cloud policy library stores preset adjustment strategies, which are pre-set based on a large amount of detailed collection of strategy execution effects. At the same time, the cloud policy library also has an update and optimization mechanism for the strategies, that is, after the optimal adjustment strategy is issued according to the expected load characteristics, the effect of the SOC main control chip executing the optimal adjustment strategy will be tested to obtain the usage effect between the optimal adjustment strategy and the SOC main control chip, and the cloud policy library will be updated and optimized according to the usage effect to further improve the performance of the cloud policy library.

[0050] The present invention provides a method for dynamic power consumption adjustment of a multi-mode SOC main control chip, which has the following beneficial effects:

[0051] The present invention collects real-time load data of the SOC main control chip, obtains the chip load characteristics, analyzes the load characteristics according to a preset power consumption adjustment scheme, determines the corresponding power consumption mode, and adjusts the power consumption mode. The chip load characteristics are transmitted to the cloud computing platform through the network. The cloud platform analyzes the chip user portrait based on the load characteristics, analyzes the expected behavior of the current load characteristics according to the user portrait, predicts the load demand of the SOC main control chip, matches the expected load characteristics based on the cloud policy library, obtains the optimal adjustment strategy, and adjusts the power consumption mode. The power consumption mode can be dynamically adjusted according to the actual load characteristics to improve energy efficiency. Through user portrait analysis, accurate adjustment for different usage scenarios is achieved. The cloud platform ensures timely updating and accurate execution of the strategy, which solves the problem that the power consumption adjustment mode in the prior art cannot achieve the optimal effect.

[0052] Preferably, the step of collecting chip load data of the SOC main control chip to obtain the chip load characteristics of the SOC main control chip includes:

[0053] S11: collecting data of program driving instructions on the SOC main control chip to obtain the load object characteristics of the SOC main control chip; wherein the load object characteristics are used to describe the chip load status brought by the program to be started after the SOC main control chip receives the program driving instructions;

[0054] S12: collecting data on chip load performance of the SOC main control chip to obtain load performance parameters when the SOC main control chip executes the load object characteristics; wherein the load performance parameters include temperature parameters, current parameters, and voltage parameters;

[0055] S13: arranging the load performance parameters of the SOC main control chip at each moment in a preset timing cycle to obtain a load performance sequence, and extracting sequence features of the load performance sequence to obtain the load performance features of the SOC main control chip;

[0056] S14: combining the load object characteristic and the load performance characteristic to obtain a chip load characteristic of the SOC main control chip.

[0057] Specifically, through built-in or external monitoring tools, the program-driven instructions received by the SOC main control chip are monitored and recorded in real time, the program-driven instructions are parsed, and the loaded applications or services and their corresponding load object characteristics, such as application type, required resources, running time, etc. are identified. The temperature, current, voltage and other parameters of the chip are collected in real time using the built-in or external sensors of the SOC main control chip. The collected temperature parameters, current parameters, voltage parameters and other data are recorded and stored locally or transmitted to the cloud for further processing.

[0058] More specifically, within a preset timing period (such as every second or every millisecond), the load performance parameters at each moment are recorded, the collected load performance parameters are arranged in time sequence to form a continuous load performance sequence, the load performance sequence is analyzed, and key features such as average value, peak value, rate of change, etc. are extracted. Algorithms (such as statistical analysis, machine learning, etc.) are used to extract characteristic parameters that can reflect the load performance of the SOC main control chip.

[0059] More specifically, the load object characteristics and the load performance characteristics are combined to form a comprehensive chip load characteristic, and the comprehensive chip load characteristic is stored locally or uploaded to the cloud for strategy matching and performance optimization.

[0060] Preferably, the chip load characteristics are analyzed according to a preset power consumption adjustment scheme to obtain a chip power consumption mode corresponding to the chip load characteristics, and the step of adjusting the power consumption mode of the SOC main control chip according to the chip power consumption mode includes:

[0061] S21: locating the chip operation resource demand of the load object characteristic of the chip load characteristic according to the preset power consumption adjustment scheme, so as to determine whether the load object characteristic belongs to a low load state, a medium load state or a high load state;

[0062] S22: matching the judgment result with the power consumption adjustment strategy according to the preset power consumption adjustment scheme to obtain a chip power consumption mode corresponding to the judgment result, and adjusting the power consumption mode of the SOC main control chip according to the matched chip power consumption mode, so that the SOC main control chip switches to the corresponding chip power consumption mode;

[0063] S23: performing adaptability analysis of the power consumption mode fed back by the load performance characteristics on the chip load characteristics to obtain adaptability characteristics of the SOC main control chip corresponding to the chip power consumption mode, and modifying specific parameters of the chip power consumption mode of the SOC main control chip according to the adaptability characteristics;

[0064] S24: When the judgment result shows that the load object characteristic belongs to a low load state, the matching chip power consumption mode is a standby current LDO mode;

[0065] S25: When the judgment result shows that the load object characteristic belongs to the medium load state, the matching chip power consumption mode is the high-efficiency DC / DC mode;

[0066] S26: When the judgment result shows that the load object characteristics belong to a high load state, the matching chip power consumption mode is a full performance mode.

[0067] Specifically, according to the preset power consumption adjustment scheme, the load object characteristics in the chip load characteristics are analyzed to determine the required operating resource requirements, and the load object characteristics are classified into a low load state, a medium load state or a high load state. According to the preset power consumption adjustment scheme, the power consumption adjustment strategy is matched with the above judgment results to determine the corresponding chip power consumption mode.

[0068] More specifically, according to the matching results, the corresponding chip power consumption mode is selected, the low load state matches the standby current LDO mode, the medium load state matches the high efficiency DC / DC mode, and the high load state matches the full performance mode. According to the matched chip power consumption mode, the SOC main control chip is adjusted accordingly to switch the chip to the corresponding power consumption mode.

[0069] More specifically, the load performance characteristics in the chip load characteristics are analyzed, the adaptability of the current power consumption mode is checked, the adaptability characteristics are extracted, and it is evaluated whether the current power consumption mode matches the load characteristics. According to the adaptability characteristics, the specific parameters of the current power consumption mode of the SOC main control chip are corrected to optimize the efficiency and stability of the power consumption mode.

[0070] More specifically, in low-load state, it matches the standby current LDO mode to significantly reduce power consumption and extend the battery life of the device. In medium-load state, it matches the high-efficiency DC / DC mode to provide efficient energy conversion and improve energy efficiency. In high-load state, it matches the full-performance mode to give full play to the chip performance and meet high-performance requirements.

[0071] It can be understood that by real-time analysis of load object characteristics and load performance characteristics, the chip power consumption mode can be dynamically adjusted to achieve precise power consumption management, locate resource requirements according to load status, ensure reasonable allocation of resources under different load conditions, and avoid resource waste.

[0072] Preferably, the chip load characteristics are transmitted to a cloud computing platform through a network channel, so that the cloud computing platform can analyze the user portrait of the SOC main control chip according to the chip load characteristics. The step of obtaining the user portrait of the SOC main control chip includes:

[0073] S31: transmitting the chip load characteristics to a cloud computing platform through a network channel, and allowing the cloud computing platform to mark and classify the chip load characteristics to obtain a data set to be analyzed corresponding to each SOC main control chip;

[0074] S32: classifying and summarizing chip load characteristics of the data sets to be analyzed of each of the SOC main control chips to obtain a chip load list based on each of the data sets to be analyzed;

[0075] S33: performing list correlation analysis on the chip load lists of the data sets to be analyzed, so as to obtain list correlation feature distribution between the chip load lists;

[0076] S34: performing list classification on each of the chip load lists under multiple standards according to the list association feature distribution between each of the chip load lists, and generating a corresponding user tag for each of the chip load lists according to the list classification under multiple standards;

[0077] S35: Comprehensively analyze each of the user tags in the chip load list to obtain a user portrait of the SOC main control chip.

[0078] Specifically, the load characteristic data of the SOC main control chip is collected, which may include CPU usage, memory occupancy, I / O operation frequency, etc., and the collected chip load characteristic data is transmitted to the cloud computing platform through the network channel.

[0079] More specifically, on the cloud computing platform, the received chip load feature data is marked, and the marking content may include timestamp, device ID, application type, etc. The marked chip load feature data is classified and included in the data set to be analyzed. The classification basis of the data set may be device model, application scenario, user behavior, etc. Feature extraction is performed on the data set to be analyzed of each SOC main control chip to form a detailed chip load list, and the extracted chip load features are classified and summarized into a structured chip load list.

[0080] More specifically, a correlation analysis is performed on the chip load lists in each data set to be analyzed to find out the commonalities and differences between the load lists. Based on the results of the correlation analysis, a correlation feature distribution map between the chip load lists is drawn. The distribution map can show the distribution of different load features among various chips. According to business needs and analysis objectives, multiple classification standards are formulated, such as usage frequency, response time, energy consumption, etc. Based on these standards, each chip load list is classified under multiple standards and classification labels are generated. According to the list classification results under multiple standards, corresponding user labels are generated for each chip load list. User labels can include usage habits, performance requirements, preferences and other information.

[0081] More specifically, the generated user tags are comprehensively analyzed and combined with multi-dimensional data to build a comprehensive user portrait, including user behavior characteristics, usage preferences, performance requirements, etc., to ultimately form a user portrait of the SOC main control chip, which specifically describes the chip's performance and requirements in different user scenarios.

[0082] It is understandable that through detailed analysis and classification of chip load characteristic data, the user portrait of each SOC main control chip can be accurately portrayed, the usage habits and needs of different users can be identified, and the analysis based on big data and cloud computing can ensure the accuracy and comprehensiveness of the user portrait. By understanding different user portraits, the resources of the SOC main control chip can be optimized to improve resource utilization and overall system performance. According to the user portrait, personalized services and functions can be provided to enhance user experience. Through user portrait analysis, the system performance in different scenarios can be predicted, and optimization and adjustment can be carried out in advance. According to the user portrait, intelligent system adjustment can be realized to dynamically adapt to user needs and ensure system stability and efficient operation.

[0083] Preferably, the step of classifying and summarizing chip load characteristics of the data sets to be analyzed of each of the SOC main control chips to obtain a chip load list based on each of the data sets to be analyzed includes:

[0084] S321: Analyze the load pressure and load adaptability of each chip load feature of the data set to be analyzed of the SOC main control chip to obtain the basic positioning feature of each chip load feature of the data set to be analyzed;

[0085] S322: performing preliminary classification processing on each of the chip load features according to the basic positioning features of each of the chip load features to obtain a basic list level of the chip load list; wherein the basic list level is used to classify the load type of each of the chip load features in the data set to be analyzed;

[0086] S323: Analyze the continuous load duration, overall load duration, and load corresponding time of each type of chip load characteristics on the data set to be analyzed according to the basic list level, so as to obtain the time list level of the chip load list; wherein the time list level is used to describe the load time characteristics of each type of chip load characteristics of the basic list level;

[0087] S324: Analyze the type proportion distribution and the time characteristic proportion distribution of the basic list level and the time list level respectively, so as to obtain the distribution list level of the chip load list;

[0088] S325: The basic list level, the time list level, and the distribution list level together constitute the chip load list.

[0089] Specifically, data collection is performed on the data set to be analyzed of the SOC main control chip, the operating data of the chip under different load conditions is obtained, and a load stress test is performed to measure the performance of the chip under high load conditions, such as CPU occupancy, memory usage, I / O operation frequency, etc. The adaptability of the chip under different load levels is tested, and its responsiveness and stability under different load conditions are evaluated. Based on the results of the stress test and adaptability test, the basic positioning features of each chip load characteristic are extracted, such as load capacity, response time, stability, etc.

[0090] More specifically, based on the basic positioning features, each chip load feature is preliminarily classified into different load types (such as high load, low load, I / O intensive, compute intensive, etc.), forming a basic list level of the chip load list, which is used to classify the load type of each chip load feature in the data set to be analyzed.

[0091] More specifically, according to the basic list level, the time characteristic measurements are performed on the load characteristics of each type of chip, including continuous load duration, overall load duration, load corresponding time, etc. Based on the time characteristic measurement results, the time list level of the chip load list is formed to describe the load time characteristics of each type of chip load characteristics in the basic list level.

[0092] More specifically, a type proportion distribution analysis is performed on the basic list level to calculate the proportion of each type of chip load feature in the entire data set. A time characteristic proportion distribution analysis is performed on the time list level to calculate the proportion of each time characteristic in the entire data set. A distribution list level of the chip load list is formed to display the proportion distribution of the basic list level and the time list level.

[0093] More specifically, the basic list level, the time list level, and the distribution list level are comprehensively organized to form a complete chip load list, and the formed chip load list is verified to ensure accurate classification, complete description of time characteristics, and reasonable proportion distribution.

[0094] It can be understood that through load pressure and adaptability analysis, the basic positioning features of each chip load feature can be accurately identified to ensure the accuracy of classification. The preliminary classification processing is based on the accurately identified basic positioning features and can accurately classify the chip load features.

[0095] More specifically, by analyzing the load time characteristics, we can comprehensively describe the time attributes of various types of chip load characteristics, provide detailed load time performance, and form a time list hierarchy, so that the time attributes of chip load characteristics can be systematically classified and described.

[0096] More specifically, the analysis of type proportion distribution and time characteristic proportion distribution can clearly show the distribution of various types of load characteristics and time characteristics. Through the distribution list hierarchy, the proportion of chip load characteristics in the entire data set can be intuitively displayed, which is convenient for further analysis and optimization.

[0097] More specifically, the common composition of the basic list level, the time list level and the distribution list level makes the chip load list systematic, complete and operational. The formed chip load list provides a solid data foundation for subsequent user portrait analysis, resource optimization configuration, performance prediction, etc.

[0098] Preferably, the steps of classifying each of the chip load lists under multiple standards according to the list association feature distribution between the chip load lists, and generating corresponding user tags for each of the chip load lists according to the list classification under multiple standards include:

[0099] S341: performing similarity matching and identification on each of the chip load lists according to the list association feature distribution between each of the chip load lists, so as to obtain a plurality of similar matching features of each of the chip load lists;

[0100] S342: Summarizing the similar matching parts of each chip load list based on the similar matching features of each chip load list to obtain a summary description of each similar matching part; wherein each similar matching part is a classification basis of multiple standards of each chip load list;

[0101] S343: Based on the user information database of the SOC main control chip, user information data is retrieved, content mapped and weighted analyzed for the summary description of each similar matching part to obtain a user label for each similar matching part.

[0102] Specifically, the detailed data of each chip load list is collected, including load type, time characteristics, proportion distribution and other features, and the correlation features between the chip load lists are extracted, such as common load type, time distribution law, load frequency, etc. The correlation feature distribution between the chip load lists is analyzed, and the feature patterns that meet the similarity are identified. The similarity between the chip load lists is calculated using a similarity algorithm (such as cosine similarity, Euclidean distance, etc.), and the chip load lists with higher similarity are identified. The similar matching features are extracted, and the similar matching features are clustered to form several similar matching parts. The features of each similar matching part are summarized, and its core feature description is extracted, such as "high load-high frequency-short duration" or "low load-low frequency-long duration". Based on the results of the feature summary extraction, a summary description of each similar matching part is formed as the basis for classification of multiple standards.

[0103] More specifically, the user information database of the SOC main control chip is accessed to retrieve relevant user information data, the summary description of the similar matching part is mapped with the user information, the user behavior and needs related to each similar matching part are identified, and weighted analysis is performed. The mapping results are weighted in combination with factors such as the frequency and importance of user behavior to ensure the accuracy and rationality of the analysis results, and the rules and standards for generating user tags are defined, such as "high-frequency users", "low-frequency users", "computationally intensive users", etc. According to the results of the weighted analysis, corresponding user tags are generated for each similar matching part, and the generated user tags are verified to ensure the accuracy and applicability of the tags.

[0104] It is understandable that through similarity calculation and feature clustering, the multi-dimensional similarities between the chip load lists can be accurately identified, the accuracy of classification can be improved, and the formation of feature summaries and summary descriptions makes the basis for multi-standard classification more comprehensive and systematic. Through access to the user information database and content mapping, the association between user behavior and chip load characteristics can be accurately identified. The use of weighted analysis ensures the accuracy and rationality of the mapping results and improves the accuracy of user tag generation. Through the generation of user tags, personalized user descriptions can be generated for different chip load characteristics, which helps to understand user behavior and needs. The verification step after tag generation ensures the actual applicability and accuracy of user tags. Through list classification and user tag generation under multiple standards, the resource allocation and performance of the SOC main control chip can be further optimized. Based on accurate user tags, better personalized services can be provided to enhance user experience.

[0105] Preferably, the step of receiving a user profile from a cloud computing platform through a network channel, and analyzing the expected behavior of the chip load characteristics at the current moment according to the user profile to obtain the expected load characteristics of the SOC main control chip includes:

[0106] S41: receiving a user portrait from a cloud computing platform through a network channel;

[0107] S42: Parameters of expected load duration, expected load pressure, and expected load adaptability are predicted and combined for the chip load characteristics at the current moment according to the user portrait to obtain the expected load characteristics of the SOC main control chip.

[0108] Specifically, a communication connection is established with the cloud computing platform through a network channel (such as HTTP, WebSocket, etc.) to ensure the reliability and real-time nature of data transmission, and user portrait data is received from the cloud computing platform, which usually includes user historical behavior, preference settings, device usage patterns, usage frequency, resource requirements and other information. The received user portrait data is parsed and synchronized to the local system to provide the necessary basic data for subsequent analysis and prediction.

[0109] More specifically, based on the device usage patterns and historical behaviors involved in the user portrait, the possible load duration of the SOC main control chip at the current moment is predicted. For example, if the user usually frequently uses computing-intensive applications in a specific time period, it is predicted that the load duration of the SOC main control chip during this period is longer. Combined with the user's usage habits, the load requirements of the device and the cloud computing platform's scheduling strategy for the device, the load pressure of the SOC main control chip at the current moment (such as CPU, memory, network bandwidth, etc.) is predicted. If the user is running a task with high computing requirements, the load pressure will be relatively high. Based on the user's historical load adaptability (such as the stability and responsiveness of the chip under high load conditions), the load adaptability of the SOC main control chip at the current moment is evaluated. Adaptability prediction can be analyzed based on the hardware characteristics of the chip and the resource scheduling strategy of the cloud computing platform.

[0110] More specifically, based on the hardware performance of the SOC main control chip, the usage behavior in the user profile, and the load characteristics prediction, a parameterized model is established, which comprehensively considers factors such as load duration, pressure, and adaptability. This model will become the basis for analyzing the expected load characteristics, combining parameters such as load duration, load pressure, and load adaptability to form a comprehensive load characteristic description. For example, if the user profile shows that it performs high-frequency computing tasks in a certain period of time, and the chip has good adaptability and scalability, the model will predict that the chip's load pressure is high during this period, but it can adapt to high loads to a certain extent.

[0111] More specifically, by combining the above-mentioned prediction parameters (load duration, load pressure, load adaptability), the expected load characteristics of the SOC main control chip at the current moment are obtained. These characteristics may include expected CPU load, memory occupancy, network bandwidth requirements, etc. The expected load characteristics are output through a graphical interface or log system to facilitate operation and maintenance personnel to monitor the load of the SOC main control chip and make resource scheduling and optimization in advance.

[0112] Preferably, the steps of constructing the cloud policy library include:

[0113] S51: constructing a strategy reserve area, and constructing a strategy analysis area consisting of a chip performance analysis dimension and a task actual situation analysis dimension, wherein the strategy analysis area and the strategy reserve area together constitute the cloud strategy library;

[0114] S52: continuously updating the reserve of load regulation strategies in the strategy reserve area, and deploying mapping relationship parameters of the reserve updated load regulation strategies relative to the strategy analysis area, so that each load regulation strategy in the strategy reserve area has a mapping relationship with the user portrait analysis dimension, the chip performance analysis dimension, and the task actual situation analysis dimension in the strategy analysis area;

[0115] S53: Construct a policy calling unit having a data interaction relationship with the policy reserve area and the policy analysis area. When the policy calling unit obtains a policy matching request of the SOC main control chip, it performs a multi-dimensional analysis of the chip performance analysis dimension and the task actual situation analysis dimension on the policy matching request of the SOC main control chip to obtain the policy matching multi-dimensional characteristics of the SOC main control chip, and calls the load adjustment strategy corresponding to the policy matching multi-dimensional characteristics according to the mapping relationship between the policy reserve area and the policy analysis area; wherein the policy matching request includes chip performance information, expected load characteristics and terminal actual situation information, the chip performance information is used to describe the specification performance corresponding to the model of the SOC main control chip, the terminal actual situation information is used to describe the power information of the smart terminal where the SOC main control chip is located, the chip performance information is used to be substituted into the chip performance analysis dimension, and the expected load characteristics and the terminal actual situation information are used to be substituted into the task actual situation analysis dimension;

[0116] S54: The policy reserve area, the policy analysis area and the policy retrieval unit together constitute the cloud policy library.

[0117] Specifically, a policy reserve area is established to store and manage different load adjustment strategies. This area contains load adjustment strategies for the performance of the SOC main control chip and the actual task status. The strategies will be classified and stored according to different load conditions and requirements so that they can be called according to different scenarios. A policy analysis area consisting of chip performance analysis dimensions and task actual situation analysis dimensions is constructed. The chip performance analysis dimension includes the processing capability, resource usage, and processing load of the SOC main control chip. The task actual situation analysis dimension takes into account the requirements, duration, priority, etc. of the current task. The policy analysis area analyzes the load through these two dimensions to provide a basis for the selection of load adjustment strategies. Data interaction is achieved between the policy reserve area and the policy analysis area to ensure real-time updating and analysis matching of the strategies.

[0118] More specifically, the load regulation strategies in the strategy reserve area are continuously updated. By collecting and analyzing the performance data of the SOC main control chip, task execution status and user behavior data in real time, the existing strategies are optimized or new strategies are added to maintain the effectiveness of the strategies. The reserve updated load regulation strategies are mapped with the chip performance analysis dimension, task actual analysis dimension and user portrait analysis dimension of the strategy analysis area to ensure that each load regulation strategy can match the specific chip and task load characteristics so as to enable intelligent scheduling based on real-time data.

[0119] More specifically, a policy calling unit is constructed, which is responsible for data interaction with the policy reserve area and the policy analysis area. When the SOC main control chip issues a policy matching request, the policy calling unit will perform a multi-dimensional analysis based on the chip performance analysis dimension and the task actual situation analysis dimension to evaluate the current chip load and task requirements. When the policy calling unit receives the policy matching request from the SOC main control chip, it performs the following analysis: Chip performance information: analyze the model, computing power, processing load and other performance indicators of the SOC main control chip; expected load characteristics: analyze the expected load characteristics of the SOC main control chip at the current moment, such as load duration, load pressure, load adaptability, etc.; terminal real-time information: analyze the operating status of the terminal device, task type and user needs.

[0120] More specifically, based on the above analysis results, the policy calling unit generates matching policy features, and according to the mapping relationship between the policy reserve area and the policy analysis area, calls the corresponding load regulation strategy to adjust the load of the SOC main control chip to optimize performance and extend the service life of the equipment.

[0121] It can be understood that through precise strategy analysis and dynamic adjustment, the load of the SOC main control chip can be adjusted in real time according to the chip performance and task status, thereby improving the chip's adaptability and efficiency to complex loads. The precise deployment of mapping relationships ensures that each task can obtain the most appropriate load adjustment strategy according to its needs, thereby improving the stability and smoothness of task execution. By dynamically adjusting the chip load, excessive load and excessive energy consumption can be reduced, resource utilization can be optimized, and the service life of the SOC main control chip and terminal equipment can be extended. The policy calling unit can quickly respond to the request of the SOC main control chip, and dynamically adjust the strategy according to the current chip operating status and terminal task requirements to ensure real-time performance optimization.

[0122] Preferably, the step of continuously updating the reserve of the load regulation strategy for the strategy reserve area comprises:

[0123] S521: Initializing and deploying a preset load adjustment strategy for the strategy reserve area, so that the strategy reserve area has a load adjustment strategy that can be directly called;

[0124] S522: Sending adjustment effect detection information to the SOC main control chip that sends a policy matching request to the cloud policy library, so as to collect data on the adjustment effect after the SOC main control chip receives the optimal adjustment strategy through the adjustment effect detection information, so as to obtain a matching index between the optimal adjustment strategy and the SOC main control chip;

[0125] S523: performing strategy update optimization on the load regulation strategy in the strategy reserve area according to the matching index to generate a new load regulation strategy, thereby achieving continuous reserve update of the load regulation strategy for the strategy reserve area.

[0126] Specifically, the policy reserve area is first initialized and deployed, and a series of preset load adjustment policies are set. These policies can be designed based on existing chip performance models, common task load requirements, user behavior analysis and other factors. These initial policies provide a basic set of load adjustment policies for the cloud policy library, ensuring that these preset policies have the ability to be directly called. In other words, each policy can form a clear mapping with the performance analysis dimension and task requirement dimension defined in the policy analysis area, which is convenient for subsequent real-time policy retrieval.

[0127] More specifically, when the SOC main control chip issues a strategy matching request, in addition to requesting the load adjustment strategy, it will also send adjustment effect detection information. This information includes the current performance data, load conditions, task requirements, and the chip's response after applying the initial adjustment strategy (such as temperature, battery consumption, processing power changes, etc.). The main purpose of the adjustment effect detection information is to collect the actual response data of the chip after executing the preset adjustment strategy, understand the actual effect of the strategy, and through real-time feedback data, determine whether the strategy has achieved the expected effect and whether further optimization is needed.

[0128] More specifically, the adjustment effect detection information is used to collect data. These data may include performance changes, resource consumption changes, response speed changes, etc. of the SOC main control chip. By monitoring these indicators, feedback data on the applied strategy can be obtained. Based on these feedback data, the matching index between the strategy and the SOC main control chip is calculated. The matching index is a quantitative evaluation of the effect of the strategy, usually based on multiple factors, such as performance improvement, energy efficiency optimization, heat management, etc. A high matching index indicates that the strategy is more effective in a specific situation, while a low matching index indicates that the strategy may have room for optimization.

[0129] More specifically, based on the obtained matching index, the load regulation strategies in the strategy reserve area are optimized and updated. If the matching index of some strategies is low, indicating that their effects are not good, these strategies can be adjusted or completely replaced; while strategies with high matching indexes can be retained or slightly adjusted. Based on the analysis results of the matching index, new load regulation strategies are formulated, updated and optimized. The new strategies will combine the advantages of existing strategies and enhance the actual effects to adapt to different chip and task load requirements.

[0130] More specifically, the entire process is dynamic. The strategy reserve area will continuously update and optimize the strategy based on the feedback and data collection results of the SOC main control chip each time. As time goes by, new data will continue to accumulate, and the effectiveness of the load regulation strategy will continue to improve, ensuring that the strategy always remains efficient and adaptable.

[0131] It can be understood that through the feedback of adjustment effect detection information and the calculation of matching index, the actual effect of the load adjustment strategy can be evaluated in real time. This data-driven feedback mechanism makes each strategy update more accurate and can continuously optimize the strategy according to actual conditions. By continuously optimizing the load adjustment strategy, it can better adapt to the actual operation needs of the SOC main control chip. After the strategy is updated, the performance of the chip is continuously improved, and it can handle various tasks more efficiently and reduce the waste of resources caused by excessive or light load.

[0132] Reference Figure 2 As shown, in a second aspect, the present invention provides a dynamic power consumption adjustment device for a multi-mode SOC main control chip, which is used to implement a dynamic power consumption adjustment method for a multi-mode SOC main control chip as described in any one of the first aspects, including:

[0133] A data acquisition module is used to collect chip load data of the SOC main control chip to obtain the chip load characteristics of the SOC main control chip;

[0134] A mode adjustment module, used for analyzing the chip load characteristics according to a preset power consumption adjustment scheme to obtain a chip power consumption mode corresponding to the chip load characteristics, and adjusting the power consumption mode of the SOC main control chip according to the chip power consumption mode;

[0135] A portrait analysis module, used to transmit the chip load characteristics to the cloud computing platform through a network channel, so that the cloud computing platform can analyze the user portrait of the SOC main control chip according to the chip load characteristics to obtain the user portrait of the SOC main control chip;

[0136] An expected analysis module, used to receive a user profile from a cloud computing platform through a network channel, and analyze the expected behavior of the chip load characteristics at the current moment according to the user profile, so as to obtain the expected load characteristics of the SOC main control chip;

[0137] The strategy matching module is used to perform strategy matching on the expected load characteristics based on the cloud strategy library to obtain the optimal adjustment strategy corresponding to the expected load characteristics, and adjust the power consumption mode of the SOC main control chip according to the optimal adjustment strategy.

[0138] In this embodiment, for the specific implementation of each module in the above-mentioned device embodiment, please refer to the above-mentioned method embodiment, which will not be repeated here.

[0139] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions and improvements made within the spirit and principles of the present invention should be included in the protection scope of the present invention.

Claims

1. A method for dynamic power consumption adjustment of a multi-mode SOC main control chip, characterized in that: include: Collecting chip load data of the SOC main control chip to obtain chip load characteristics of the SOC main control chip; Analyze the chip load characteristics according to a preset power consumption adjustment scheme to obtain a chip power consumption mode corresponding to the chip load characteristics, and adjust the power consumption mode of the SOC main control chip according to the chip power consumption mode; The chip load characteristics are transmitted to a cloud computing platform through a network channel, so that the cloud computing platform can analyze the user profile of the SOC main control chip according to the chip load characteristics to obtain the user profile of the SOC main control chip; Receiving a user profile from a cloud computing platform through a network channel, and analyzing the expected behavior of the chip load characteristics at the current moment according to the user profile to obtain the expected load characteristics of the SOC main control chip; Performing strategy matching on the expected load characteristics based on the cloud strategy library to obtain an optimal adjustment strategy corresponding to the expected load characteristics, and adjusting the power consumption mode of the SOC main control chip according to the optimal adjustment strategy; The chip load characteristics are transmitted to a cloud computing platform through a network channel, so that the cloud computing platform can analyze the user portrait of the SOC main control chip according to the chip load characteristics. The step of obtaining the user portrait of the SOC main control chip includes: Transmitting the chip load characteristics to a cloud computing platform through a network channel, and allowing the cloud computing platform to mark and classify the chip load characteristics to obtain a data set to be analyzed corresponding to each SOC main control chip; Classifying and summarizing chip load characteristics of the data sets to be analyzed of each of the SOC main control chips to obtain a chip load list based on each of the data sets to be analyzed; Performing list correlation analysis on the chip load lists of each of the data sets to be analyzed to obtain list correlation feature distribution between each of the chip load lists; Classifying each of the chip load lists under multiple standards according to the list association feature distribution between each of the chip load lists, and generating a corresponding user tag for each of the chip load lists according to the list classification under multiple standards; Comprehensively analyzing each of the user tags in the chip load list to obtain a user profile of the SOC main control chip; The step of classifying and summarizing chip load characteristics of the data sets to be analyzed of each of the SOC main control chips to obtain a chip load list based on each of the data sets to be analyzed includes: Performing load pressure and load adaptability analysis on the load characteristics of each chip in the data set to be analyzed of the SOC main control chip to obtain basic positioning characteristics of the load characteristics of each chip in the data set to be analyzed; Preliminary classification processing is performed on each of the chip load features according to the basic positioning features of each of the chip load features to obtain a basic list level of the chip load list; wherein the basic list level is used to classify the load type of each of the chip load features in the data set to be analyzed; According to the basic list level, the continuous load duration, the overall load duration, and the load corresponding time of each type of chip load characteristics are analyzed for the data set to be analyzed, so as to obtain the time list level of the chip load list; wherein the time list level is used to describe the load time characteristics of each type of chip load characteristics of the basic list level; Analyzing the type proportion distribution and the time characteristic proportion distribution of the basic list level and the time list level respectively, so as to obtain the distribution list level of the chip load list; The basic list level, the time list level, and the distribution list level together constitute the chip load list.

2. The method for dynamic power consumption adjustment of a multi-mode SOC main control chip according to claim 1, characterized in that: The step of collecting chip load data of the SOC main control chip to obtain the chip load characteristics of the SOC main control chip includes: Performing data collection of program-driven instructions on the SOC main control chip to obtain the load object characteristics of the SOC main control chip; wherein the load object characteristics are used to describe the chip load status brought by the program to be started by the SOC main control chip after receiving the program-driven instructions; Collecting data on chip load performance of the SOC main control chip to obtain load performance parameters when the SOC main control chip executes the load object characteristics; wherein the load performance parameters include temperature parameters, current parameters, and voltage parameters; Arranging the load performance parameters of the SOC main control chip at each moment in a preset timing cycle to obtain a load performance sequence, and extracting sequence features of the load performance sequence to obtain the load performance features of the SOC main control chip; The load object characteristics and the load performance characteristics are combined to obtain the chip load characteristics of the SOC main control chip.

3. The dynamic power consumption adjustment method of the multi-mode SOC main control chip according to claim 2, characterized in that: The steps of analyzing the chip load characteristics according to a preset power consumption adjustment scheme to obtain a chip power consumption mode corresponding to the chip load characteristics, and adjusting the power consumption mode of the SOC main control chip according to the chip power consumption mode include: Positioning the chip operation resource demand of the load object characteristic of the chip load characteristic according to the preset power consumption adjustment scheme to determine whether the load object characteristic belongs to a low load state, a medium load state or a high load state; According to the preset power consumption adjustment scheme, the judgment result is matched with the power consumption adjustment strategy to obtain the chip power consumption mode corresponding to the judgment result, and the power consumption mode of the SOC main control chip is adjusted according to the matched chip power consumption mode, so that the SOC main control chip switches to the corresponding chip power consumption mode; Performing adaptability analysis of the power consumption mode fed back by the load performance characteristics on the chip load characteristics to obtain adaptability characteristics of the SOC main control chip corresponding to the chip power consumption mode, and performing specific parameter correction of the chip power consumption mode on the SOC main control chip according to the adaptability characteristics; When the judgment result shows that the load object characteristic belongs to a low load state, the matching chip power consumption mode is a standby current LDO mode; When the judgment result shows that the load object characteristic belongs to the medium load state, the matching chip power consumption mode is the high-efficiency DC / DC mode; When the judgment result shows that the load object characteristic belongs to a high load state, the matching chip power consumption mode is a full performance mode.

4. The method for dynamic power consumption adjustment of a multi-mode SOC main control chip according to claim 1, characterized in that: The steps of classifying each of the chip load lists under multiple standards according to the list association feature distribution between the chip load lists, and generating corresponding user tags for each of the chip load lists according to the list classification under multiple standards include: Performing similarity matching and identification on each of the chip load lists according to the list association feature distribution between each of the chip load lists, so as to obtain a plurality of similar matching features of each of the chip load lists; Based on each similar matching feature of each chip load list, a similar matching part of each chip load list is summarized to obtain a summary description of each similar matching part; wherein each similar matching part is a classification basis of multiple standards of each chip load list; Based on the user information database of the SOC main control chip, the user information data is retrieved, the content is mapped and weighted analysis is performed on the summary description of each similar matching part to obtain the user label of each similar matching part.

5. The method for dynamic power consumption adjustment of a multi-mode SOC main control chip according to claim 1, characterized in that: The steps of receiving a user profile from a cloud computing platform through a network channel, and analyzing the expected behavior of the chip load characteristics at the current moment according to the user profile to obtain the expected load characteristics of the SOC main control chip include: Receive user portraits from the cloud computing platform through a network channel; According to the user portrait, the expected load duration, expected load pressure, and expected load adaptability of the chip load characteristics at the current moment are predicted and combined with the parameters to obtain the expected load characteristics of the SOC main control chip.

6. The method for dynamic power consumption adjustment of a multi-mode SOC main control chip according to claim 1, characterized in that: The steps of constructing the cloud policy library include: Constructing a strategy reserve area, and constructing a strategy analysis area consisting of a chip performance analysis dimension and a task actual situation analysis dimension, wherein the strategy analysis area and the strategy reserve area together constitute the cloud strategy library; Continuously updating the reserve of load regulation strategies in the strategy reserve area, and deploying mapping relationship parameters of the reserve updated load regulation strategies relative to the strategy analysis area, so that each load regulation strategy in the strategy reserve area has a mapping relationship with the user portrait analysis dimension, chip performance analysis dimension, and task actual situation analysis dimension in the strategy analysis area; Constructing a policy retrieval unit having a data interaction relationship with the policy reserve area and the policy analysis area, wherein the policy reserve area, the policy analysis area and the policy retrieval unit together constitute the cloud policy library; The strategy retrieval unit is used to perform a multi-dimensional analysis of the chip performance analysis dimension and the task actual situation analysis dimension on the strategy matching request of the SOC main control chip to obtain the multi-dimensional characteristics of the strategy matching of the SOC main control chip, and retrieve the load regulation strategy corresponding to the multi-dimensional characteristics of the strategy matching according to the mapping relationship between the strategy reserve area and the strategy analysis area; The policy matching request includes chip performance information, expected load characteristics and terminal real-time information. The chip performance information is used to describe the specification performance corresponding to the model of the SOC main control chip, and the terminal real-time information is used to describe the power information of the smart terminal where the SOC main control chip is located. The chip performance information is used to be substituted into the chip performance analysis dimension, and the expected load characteristics and the terminal real-time information are used to be substituted into the task real-time analysis dimension.

7. The method for dynamic power consumption adjustment of a multi-mode SOC main control chip according to claim 6, characterized in that: The step of continuously updating the reserve of the load regulation strategy for the strategy reserve area includes: Initializing and deploying a preset load adjustment strategy for the strategy reserve area, so that the strategy reserve area has a load adjustment strategy that can be directly called; Sending adjustment effect detection information to the SOC main control chip that sends a strategy matching request to the cloud strategy library, so as to collect data on the adjustment effect after the SOC main control chip receives the optimal adjustment strategy through the adjustment effect detection information, so as to obtain a matching index between the optimal adjustment strategy and the SOC main control chip; The load regulation strategy in the strategy reserve area is updated and optimized according to the matching index to generate a new load regulation strategy, thereby achieving continuous reserve update of the load regulation strategy in the strategy reserve area.

8. A dynamic power consumption adjustment device for a multi-mode SOC main control chip, characterized in that: A method for dynamic power consumption adjustment of a multi-mode SOC main control chip for implementing any one of claims 1 to 7, comprising: A data acquisition module is used to collect chip load data of the SOC main control chip to obtain the chip load characteristics of the SOC main control chip; A mode adjustment module, used for analyzing the chip load characteristics according to a preset power consumption adjustment scheme to obtain a chip power consumption mode corresponding to the chip load characteristics, and adjusting the power consumption mode of the SOC main control chip according to the chip power consumption mode; A portrait analysis module, used to transmit the chip load characteristics to the cloud computing platform through a network channel, so that the cloud computing platform can analyze the user portrait of the SOC main control chip according to the chip load characteristics to obtain the user portrait of the SOC main control chip; An expected analysis module, used to receive a user profile from a cloud computing platform through a network channel, and analyze the expected behavior of the chip load characteristics at the current moment according to the user profile, so as to obtain the expected load characteristics of the SOC main control chip; The strategy matching module is used to perform strategy matching on the expected load characteristics based on the cloud strategy library to obtain the optimal adjustment strategy corresponding to the expected load characteristics, and adjust the power consumption mode of the SOC main control chip according to the optimal adjustment strategy.

Images