Computer energy-saving state control method and system

Abstract

The invention relates to the technical field of energy-saving control, in particular to a computer energy-saving state control method and system, and the method comprises the following steps: constructing a task graph model based on a computer task dependency relationship, calculating a dependency intensity value between tasks, recognizing a task resource scheduling sequence, and analyzing the association degree between the tasks; tasks with high dependency are preferentially distributed to shared computing resources, and a task relevance scheduling scheme is generated. In the invention, through dynamic scheduling based on the task load and the dependency relationship, the system can accurately adjust the allocation and the power state of the computing resources under different loads, improve the resource use efficiency when the load is high, monitor the power consumption and the temperature change of the computing unit in real time, and dynamically adjust the working mode and the frequency of the computing resources; excessive energy consumption and heat accumulation are effectively avoided, the system can adjust the power state according to the actual demand of a task, efficient energy use of the system is kept, unnecessary resource consumption and temperature fluctuation are reduced, and the overall energy efficiency is improved.

Description

Technical Field

[0001] The present invention relates to the technical field of energy-saving control, and particularly to a method and system for controlling the energy-saving state of a computer. Background Art

[0002] The technical field of computer energy-saving technology includes technologies related to energy-saving control and energy management in computer hardware, software, and their operating environments. The core content of this technical field mainly focuses on how to achieve efficient use of energy in a computer system, by controlling the working state, power consumption, and heat emission of the computer system to achieve the purpose of energy saving. With the popularization of information technology and computer applications, the energy consumption of computers has gradually become a key issue of social concern. The research and application of energy-saving technologies aim to achieve effective management of computer resources through intelligent and automated means, reduce unnecessary energy consumption, and improve the working efficiency of the system. This technical field covers energy-saving control means from hardware design to the operating system level, from computer chips to external devices, and involves multiple aspects such as system scheduling, status monitoring, and power consumption optimization.

[0003] Among them, the computer energy-saving state control method refers to a technology that optimizes the power consumption control of a computer system in different working states. This method mainly solves the problem of energy waste in the computer during idle, low-load, and other non-high-efficiency working states. It intelligently identifies the working load state of the computer and dynamically adjusts the working mode of the computer or the start-stop state of devices according to the load situation, reducing unnecessary energy consumption. Specifically, this method sets energy-saving control parameters in the computer system and automatically switches the working state of the system or adjusts the power consumption of each device according to real-time workload and state changes to achieve the purpose of energy saving.

[0004] Most of the existing technologies rely on static power consumption modes or fixed scheduling mechanisms and cannot flexibly adapt to fluctuations in task loads in different working states. For example, in low-load situations, traditional technologies often fail to effectively reduce the power consumption of computing resources, resulting in energy waste in the idle state. In addition, the existing technologies lack the function of dynamically adjusting based on real-time data, and the management of the working state and temperature of computing resources is mostly preset or simple adjustment strategies, which cannot accurately respond to the energy efficiency loss caused by temperature fluctuations, resulting in the system being unable to adjust the frequency in a timely manner at high temperatures, increasing unnecessary power consumption and temperature loss. Summary of the Invention

[0005] The purpose of the present invention is to solve the deficiencies existing in the prior art and to propose a method and system for controlling the energy-saving state of a computer.

[0006] To achieve the above purpose, the present invention adopts the following technical solutions: A method for controlling the energy-saving state of a computer, comprising the following steps:

[0007] S1: Based on the computer task dependencies, construct a task graph model, calculate the dependency strength values between tasks, identify the task resource scheduling order, analyze the association degree between tasks, preferentially allocate tasks with strong dependencies to shared computing resources, and generate a task association scheduling scheme;

[0008] S2: Based on the task association scheduling scheme, monitor the load of computing units in real time, calculate the resource usage of each computing unit, adjust the allocation of computing resources according to task priorities and resource requirements, and optimize the resource configuration to generate a resource allocation optimization scheme;

[0009] S3: According to the resource allocation optimization scheme, analyze the power requirements of computing units, dynamically adjust the power states of computing units, and adjust the CPU frequency and memory power consumption to control the power of each computing resource, and adjust the power state of resources based on load changes to generate a power adjustment scheme;

[0010] S4: Based on the power adjustment scheme, analyze the load of computing resources in real time, calculate the power supply requirements, and dynamically switch the power consumption mode according to load changes and the working states of computing units to generate a power supply scheduling configuration;

[0011] S5: According to the power supply scheduling configuration, monitor the temperature changes of computing units, analyze the temperature fluctuations, set and determine whether the temperature threshold is exceeded. If the temperature is overloaded, adjust the computing unit frequency to optimize the power consumption configuration to generate a thermal management and power consumption optimization configuration.

[0012] As a further solution of the present invention, the task association scheduling scheme includes task dependency strength, task scheduling priority, task association degree, and shared resource allocation scheme. The resource allocation optimization scheme includes computing unit load assessment, resource utilization rate, task priority weight, and resource allocation adjustment strategy. The power adjustment scheme includes power requirement assessment, computing unit power state, CPU frequency adjustment, and memory power consumption control. The power supply scheduling configuration includes power supply requirement calculation, power consumption mode selection, load adaptability adjustment, and power mode switching. The thermal management and power consumption optimization configuration includes temperature change monitoring, temperature fluctuation judgment, temperature threshold setting, and computing unit frequency adjustment.

[0013] As a further solution of the present invention, the specific steps for constructing a task graph model based on computer task dependencies, calculating the dependency strength values between tasks, identifying the task resource scheduling order, analyzing the association degree between tasks, and preferentially allocating tasks with strong dependencies to shared computing resources to generate a task association scheduling scheme are as follows:

[0014] S101: Based on the dependency graph of the computing tasks, collect the input and output data of the tasks, identify the resource constraints and time requirements between the tasks one by one, and perform sorting in combination with the priorities, resource consumption, and time constraints of the tasks to generate a task dependency matrix;

[0015] S102: Based on the task dependency matrix, calculate the dependency strength values between the tasks. By comparing the resource occupancy, completion time, and priorities of the tasks, and conducting quantitative analysis, identify the strength of the mutual influence between the tasks to generate a task dependency strength graph;

[0016] S103: Based on the task dependency strength graph, preferentially select the tasks with higher dependency strength, and make allocations according to the resource consumption and operation cycle of the tasks, in combination with the scheduling strategy, to generate a task correlation scheduling plan.

[0017] As a further solution of the present invention, the specific formula for the dependency strength value is:

[0018]

[0019] Among them, I ij represents the dependency strength value between task i and task j, R i represents the resource occupancy value of task i, R j represents the resource occupancy value of task j, T ij represents the time difference between task i and task j, P i represents the priority of task i, T i represents the completion time of task i, T j represents the completion time of task j.

[0020] As a further solution of the present invention, based on the task correlation scheduling plan, the specific steps for real-time monitoring of the computing unit load, calculating the resource usage of each computing unit, adjusting the allocation of computing resources according to the task priorities and resource requirements, and optimizing the resource configuration to generate a resource allocation optimization plan are as follows:

[0021] S201: Based on the task correlation scheduling plan, real-time monitor the load status of the computing units, collect the resource consumption and processing progress of the computing units, analyze the computing capabilities and task execution situations of each unit to generate computing unit load data;

[0022] S202: Based on the computing unit load data, compare the resource consumption of each unit with the task priorities and resource requirements, judge whether each computing unit meets the task requirements, and perform adaptation adjustments to generate resource requirement and allocation adjustment data;

[0023] S203: Based on the resource requirement and allocation adjustment data, analyze the priorities and resource consumption among tasks, adjust the computing resource allocation, and generate an optimized resource allocation plan.

[0024] As a further solution of the present invention, according to the optimized resource allocation plan, analyze the power requirements of the computing units, dynamically adjust the power states of the computing units, and regulate the CPU frequency and memory power consumption to control the power of each computing resource. The specific steps for generating a power regulation plan based on the load change to adjust the power state of the resource are as follows:

[0025] S301: Based on the optimized resource allocation plan, collect the power requirement data of each computing unit, analyze the load conditions and power consumption changes of each module, calculate the power resource values required for each computing unit, and generate the computing unit power requirement data;

[0026] S302: Based on the computing unit power requirement data, analyze the relationship between the power state of the computing unit and the task execution, dynamically adjust the CPU frequency and memory power consumption, and generate the power state adjustment data;

[0027] S303: Based on the power state adjustment data, control the power of each computing resource according to the task load change situation, dynamically adjust the power state of the resource based on the load change, and generate the power regulation plan.

[0028] As a further solution of the present invention, the specific formula for the power resource value is as follows:

[0029]

[0030] where P m represents the power resource value required for computing unit m, R m represents the resource occupancy value of computing unit m, L m represents the load condition of computing unit m, V m represents the voltage of computing unit m, and T m represents the execution time of computing unit m.

[0031] As a further solution of the present invention, based on the power regulation plan, analyze the computing resource load in real time, calculate the power supply demand, and dynamically switch the power consumption mode according to the load change and the working state of the computing unit. The specific steps for generating the power supply scheduling configuration are as follows:

[0032] S401: Based on the power regulation plan, monitor the load change of the computing resources in real time, collect the working state and power supply demand data of the computing units, calculate the total power supply value required, and generate the computing resource power supply demand data;

[0033] S402: Based on the power demand data of the computing resources, compare the load fluctuations of the computing resources, and in combination with the current working state of the computing unit, dynamically adjust the power mode to generate power mode adjustment data;

[0034] S403: Based on the power mode adjustment data, switch the power consumption mode according to the load changes of the computing resources, adjust the power configuration, and generate a power scheduling configuration.

[0035] As a further solution of the present invention, according to the power scheduling configuration, monitor the temperature change of the computing unit, analyze the temperature fluctuation, set and determine whether it exceeds the temperature threshold. If the temperature is overloaded, adjust the computing unit frequency to optimize the power consumption configuration. The specific steps for generating the thermal management and power consumption optimization configuration are as follows:

[0036] S501: Based on the power scheduling configuration, monitor the temperature change of the computing unit in real time, collect the temperature sensor data, analyze the temperature fluctuation trend, and generate the computing unit temperature data;

[0037] S502: Based on the computing unit temperature data, analyze the temperature change range, determine whether it exceeds the set threshold, and generate a temperature fluctuation analysis result;

[0038] S503: Based on the temperature fluctuation analysis result, if the temperature is overloaded, adjust the computing unit frequency and optimize the power consumption configuration to generate the thermal management and power consumption optimization configuration.

[0039] A computer energy-saving state control system includes:

[0040] The computing task scheduling module constructs a task graph model based on the computer task dependency relationship, calculates the dependency strength value between tasks, identifies the task resource scheduling order, preferentially allocates tasks with strong dependencies to shared computing resources, and generates a task correlation scheduling plan;

[0041] The resource allocation optimization module, based on the task correlation scheduling plan, monitors the load of the computing unit in real time, calculates the resource usage of each computing unit, and adjusts the allocation of computing resources according to the task priority and resource requirements to generate a resource allocation optimization plan;

[0042] The power regulation module analyzes the power demand of the computing unit according to the resource allocation optimization plan, dynamically adjusts the power state of the computing unit, adjusts the CPU frequency and memory power consumption, controls the power of each computing resource, and generates a power regulation plan;

[0043] The power management module, based on the power regulation plan, analyzes the load of the computing resources in real time, calculates the power demand, dynamically switches the power consumption mode, and generates a power scheduling configuration;

[0044] The temperature control management module monitors the temperature change of the computing unit according to the power supply scheduling configuration, analyzes the temperature fluctuation, sets and determines whether it exceeds the temperature threshold, adjusts the computing unit frequency to optimize the power consumption configuration, and generates the thermal management and power consumption optimization configuration.

[0045] Compared with the prior art, the advantages and positive effects of the present invention are as follows:

[0046] In the present invention, through the dynamic scheduling based on the task load and dependency relationship, the system can accurately adjust the allocation of computing resources and power states under different loads, improve the resource utilization efficiency under high loads, monitor the power consumption and temperature changes of the computing unit in real time, dynamically adjust the working mode and frequency of the computing resources, effectively avoid excessive energy consumption and heat accumulation, the system can adjust the power state according to the actual needs of the task, maintain the efficient energy use of the system, reduce unnecessary resource consumption and temperature fluctuations, and improve the overall energy efficiency. BRIEF DESCRIPTION OF THE DRAWINGS

[0047] In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the following will briefly introduce the drawings required for the description of the embodiments. Obviously, the drawings in the following description are only some embodiments of the present invention. For those of ordinary skill in the art, without creative efforts, other drawings can be obtained according to these drawings.

[0048] Figure 1 is the schematic diagram of the step flow of the present invention;

[0049] Figure 2 is the flowchart of step S1 of the present invention;

[0050] Figure 3 is the flowchart of step S2 of the present invention;

[0051] Figure 4 is the flowchart of step S3 of the present invention;

[0052] Figure 5 is the flowchart of step S4 of the present invention;

[0053] Figure 6 is the flowchart of step S5 of the present invention;

[0054] Figure 7 is the system module diagram of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0055] The following will describe the technical solutions in the present invention with reference to the drawings.

[0056] In the embodiments of the present invention, words such as "exemplarily" and "for example" are used to give examples, illustrations or explanations. Any embodiment or design solution described as an "example" in the present invention should not be construed as being more preferred or having more advantages than other embodiments or design solutions. Rather, the use of the word "example" is intended to present concepts in a specific manner. In addition, in the embodiments of the present invention, the meaning expressed by "and / or" can be both, or either one of the two.

[0057] In the embodiments of the present invention, "image" and "picture" can sometimes be used interchangeably. It should be noted that when the difference is not emphasized, the meanings they express are the same. "Of", "corresponding", and "corresponding" can sometimes be used interchangeably. It should be noted that when the difference is not emphasized, the meanings they express are the same.

[0058] In the embodiments of the present invention, sometimes a subscript such as W1 may be written in a non-subscript form such as W1. When the difference is not emphasized, the meanings they express are the same.

[0059] To make the technical problems, technical solutions and advantages to be solved by the present invention clearer, the following will be described in detail with reference to the accompanying drawings and specific embodiments.

[0060] Please refer to Figure 1 , a computer energy-saving state control method, comprising the following steps:

[0061] S1: Based on the computer task dependency relationship, construct a task graph model, calculate the dependency strength value between tasks, identify the task resource scheduling order, analyze the association degree between tasks, preferentially allocate tasks with strong dependencies to shared computing resources, and generate a task association scheduling plan;

[0062] S2: Based on the task association scheduling plan, monitor the load of computing units in real time, calculate the resource usage of each computing unit, adjust the allocation of computing resources according to task priorities and resource requirements, and optimize resource allocation to generate a resource allocation optimization plan;

[0063] S3: According to the resource allocation optimization plan, analyze the power requirements of computing units, dynamically adjust the power state of computing units, and adjust the CPU frequency and memory power consumption to control the power of each computing resource, and adjust the power state of resources based on load changes to generate a power adjustment plan;

[0064] S4: Based on the power adjustment plan, analyze the load of computing resources in real time, calculate the power supply requirements, and dynamically switch the power consumption mode according to load changes and the working state of computing units to generate a power supply scheduling configuration;

[0065] S5: Monitor the temperature change of the computing unit according to the power supply scheduling configuration, analyze the temperature fluctuation, set and determine whether the temperature threshold is exceeded. If the temperature is overloaded, adjust the computing unit frequency to optimize the power consumption configuration, and generate the thermal management and power consumption optimization configuration.

[0066] The task correlation scheduling scheme includes task dependence intensity, task scheduling priority, task correlation degree, and shared resource allocation scheme. The resource allocation optimization scheme includes computing unit load assessment, resource utilization rate, task priority weight, and resource allocation adjustment strategy. The power adjustment scheme includes power demand assessment, computing unit power state, CPU frequency adjustment, and memory power consumption control. The power supply scheduling configuration includes power demand calculation, power consumption mode selection of the power supply, load adaptability adjustment, and power mode switching. The thermal management and power consumption optimization configuration includes temperature change monitoring, temperature fluctuation judgment, temperature threshold setting, and computing unit frequency adjustment.

[0067] Please refer to Figure 2 , the specific steps of S1 are as follows:

[0068] S101: Based on the dependency graph of the computing tasks, collect the input and output data of the tasks, and identify the resource constraints and time requirements between the tasks one by one. Combine the priority, resource consumption, and time constraints of the tasks to sort and generate the task dependency matrix.

[0069] First, identify the resource constraints and time requirements between the tasks, collect the input, output, and running environment information of each task, establish the association between the tasks and resources, and combine factors such as the resource consumption and execution time of the tasks to perform the allocation and scheduling analysis of the resources between the tasks to ensure that there will be no resource conflicts during the execution of each task. In particular, it is necessary to consider the priority scheduling of resources for tasks with higher priorities. Then, check the time requirements and constraints between the tasks one by one, set a reasonable task queuing mechanism based on the running duration and priority of the tasks, and calculate the appropriate task execution order in combination with the time constraints. Finally, convert the task dependencies and constraint conditions into a task dependency matrix to provide the necessary information support for subsequent task scheduling.

[0070] S102: Based on the task dependency matrix, calculate the dependency intensity value between the tasks. By comparing the resource occupancy, completion time, and priority of the tasks and performing quantitative analysis, identify the intensity of the mutual influence between the tasks and generate the task dependency intensity graph.

[0071] The specific formula for the dependency intensity value is:

[0072]

[0073] Among them, I ij represents the dependency intensity value between task i and task j, R i represents the resource occupancy value of task i, Rj Represents the resource occupancy value of task j, T ij Represents the time difference between task i and task j, P i Represents the priority of task i, T i Represents the completion time of task i, T j Represents the completion time of task j.

[0074] Parameter definition:

[0075] I ij : The dependency strength value between task i and task j, indicating the degree of mutual influence between tasks.

[0076] R i : The resource occupancy value of task i, usually measured in terms of the computing resources, memory, or storage required by task i.

[0077] R j : The resource occupancy value of task j, usually measured in terms of the computing resources, memory, or storage required by task j.

[0078] T ij : The time difference between task i and task j, calculated as the difference between the completion time of task i and the completion time of task j.

[0079] P i : The priority of task i, usually an integer value indicating the urgency of the task, with a larger value indicating a higher priority.

[0080] T i : The completion time of task i, i.e., the total time required for task i from start to end.

[0081] T j : The completion time of task j, i.e., the total time required for task j from start to end.

[0082] Parameter acquisition method:

[0083] R i and R j : Obtained by monitoring the resource usage during task execution, for example, using system monitoring tools (such as top, htop) or performance analysis tools (such as perf) to obtain resource occupancy data such as CPU, memory, and disk I / O of the task.

[0084] T ij : Calculate the time difference by recording the start and end times of the task. The timestamp function of the operating system can be used, or the scheduling and completion times of the task can be recorded in the task scheduling system.

[0085] P i: Manually evaluate according to the urgency and importance of the task or assign priorities based on preset rules. For example, priorities can be determined based on factors such as the task's due date and the degree of business impact.

[0086] T i and T j : Calculate the completion time by recording the start and end times of the task. The timestamp function of the operating system can be used, or the scheduling and completion times of the task can be recorded in the task scheduling system.

[0087] Formula calculation derivation process:

[0088] Suppose there are two tasks i and j with the following parameter values:

[0089] R i = 50 units of resources: Task i occupies 50 units of computing resources during execution.

[0090] R j = 30 units of resources: Task j occupies 30 units of computing resources during execution.

[0091] T ij = 10 units of time: The difference in completion time between task i and task j is 10 units of time.

[0092] P i = 5: The priority of task i is 5, indicating a relatively high level of task urgency.

[0093] T i = 100 units of time: The total time required for task i from start to end is 100 units of time.

[0094] T j = 80 units of time: The total time required for task j from start to end is 80 units of time.

[0095] Substitute the above parameter values into the formula:

[0096]

[0097] Calculation process:

[0098] Calculate the absolute value:

[0099] |50 - 30| = 20;

[0100] Calculate the numerator:

[0101] 20·10·5 = 1000;

[0102] Calculate the denominator:

[0103]

[0104] Calculate the dependency strength value:

[0105]

[0106] Result interpretation:

[0107] The calculated dependency strength value I ij ≈11.17 indicates the degree of mutual influence between task i and task j. The larger the value, the stronger the dependency relationship between tasks, and more coordination and resource sharing may be required.

[0108] This result shows that there is a strong dependency relationship between task i and task j, and more attention may need to be paid during scheduling and resource allocation to ensure the smooth execution of tasks.

[0109] S103: Based on the task dependency strength graph, preferentially select tasks with higher dependency strength, and allocate them according to the resource consumption and running cycle of the tasks, combined with the scheduling strategy, to generate a task correlation scheduling plan;

[0110] First, by analyzing the mutual influence between tasks shown in the dependency strength graph, determine which tasks need to be processed preferentially, especially those with strong dependency relationships. These tasks often play a key role in the overall scheduling and need to be allocated resources and started in a timely manner. Then, according to the preset scheduling strategy, allocate resources to these tasks and arrange the time to ensure that these highly dependent tasks can be completed in the shortest time, freeing up resources for the execution of other tasks. By comprehensively considering this priority and dependency relationship, a reasonable task scheduling priority order is finally generated.

[0111] Please refer to Figure 3 , the specific steps of S2 are as follows:

[0112] S201: Based on the task correlation scheduling plan, monitor the load status of the computing units in real time, collect the resource consumption and processing progress of the computing units, analyze the computing power and task execution status of each unit, and generate computing unit load data;

[0113] Monitor the load status of the computing units in real time, collect the resource consumption and processing progress of the computing units, analyze the computing power and task execution status of each unit, and calculate the load status of each computing unit according to the formula

[0114]

[0115] Calculate the load status of each computing unit.

[0116] In the formula, L i represents the load status of the i-th computing unit, C i represents the computing power of this unit, Ri represents the resource consumption of the unit, P i represents the task execution progress of the unit.

[0117] For the i-th computing unit, first monitor its computing power C i , for example, C1 = 80 represents the computing power of the computing unit, indicating the amount of computation completed per unit time; the resource consumption R i , for example, R1 = 50 units, represents the amount of resources used by the computing unit; the task execution progress P i , for example, P1 = 40%, represents the proportion of the task that has been completed.

[0118] Substitute these values into the formula for calculation:

[0119]

[0120] It is obtained that the load status of the computing unit L1 = 100. This result indicates that the load situation of the first computing unit meets the expectations and resource scheduling can continue.

[0121] S202: Based on the computing unit load data, compare the resource consumption of each unit with the task priority and resource requirements, determine whether each computing unit meets the task requirements, and perform adaptation and adjustment to generate resource requirement and allocation adjustment data;

[0122] First, compare the resource consumption situation of the computing unit with the priority and resource requirements of each task, and analyze the matching degree between the task and the resource. By evaluating the resource requirements of each computing unit and the priority of the task, determine whether the unit meets the execution requirements of the task. For units with resource mismatch or excessive load, perform adaptation and adjustment. The adjustment process may involve task rescheduling, dynamic allocation of resources, etc., to ensure that there will be no resource bottlenecks during the task execution of each computing unit. Finally, generate resource requirement and allocation adjustment data to provide a basis for resource allocation and task scheduling.

[0123] S203: Based on the resource requirement and allocation adjustment data, analyze the priority and resource consumption between tasks, adjust the computing resource allocation, and generate an optimized resource allocation plan;

[0124] First, evaluate the priority and resource consumption of each task, analyze the resource requirements and priority relationship between tasks, and determine the priority order of resource allocation. By analyzing the resource consumption between tasks, perform optimized allocation of computing resources, and allocate resources to tasks with higher priority and larger resource requirements. Combining the adjusted allocation data, optimize the allocation method of computing resources to ensure that each task can be executed in the optimal order, avoiding resource waste or scheduling delays. Finally, generate an optimized resource allocation plan to provide an optimized strategy for the resource scheduling of the computing unit.

[0125] Please refer to Figure 4 , the specific steps of S3 are as follows:

[0126] S301: Based on the resource allocation optimization scheme, collect the power demand data of each computing unit, analyze the load situation and power consumption changes of each module, calculate the power resource value required for each computing unit, and generate the computing unit power demand data;

[0127] The specific calculation formula for the power resource value is:

[0128]

[0129] where P m represents the power resource value required for computing unit m, R m represents the resource occupancy value of computing unit m, L m represents the load situation of computing unit m, V m represents the voltage of computing unit m, and T m represents the execution time of computing unit m.

[0130] Parameter definition and acquisition method:

[0131] P m : The power resource value required for computing unit m, usually measured in watts (W). The power value is obtained by comprehensively calculating the load, resource occupancy, voltage, and execution time of the computing unit. This value reflects the energy consumption demand of the module.

[0132] R m : The resource occupancy value of computing unit m, indicating the amount of resources required by the computing unit during execution, usually expressed as a percentage or relative unit. Obtain the resource occupancy of CPU, memory, or storage through performance monitoring tools (such as top, htop). Assume that in actual operation, module m uses 70% of the resources, then R m = 0.7.

[0133] L m : The load situation of computing unit m, indicating the workload level of the module. The load can be quantified by the number, complexity, or processing ability of tasks executed by the computing unit per unit time. Assume L m = 0.85, indicating that the load level of this module is 85%.

[0134] V m : The voltage of computing unit m, with the unit of volts (V), indicating the operating voltage of the module. Obtain the voltage value through hardware monitoring tools (such as sensors or power management tools). Assume that the voltage of module m is 5V, then V m = 5.

[0135] Tm : The execution time of computing unit m, in seconds (s). It represents the time required for a task to execute on this module. The execution time can be measured by a task scheduling system or a performance analysis tool. Assuming the execution time of the task is 120 seconds, then T m = 120.

[0136] Formula calculation and derivation process:

[0137] Given the following parameter values:

[0138] R m = 0.7;

[0139] L m = 0.85;

[0140] V m = 5;

[0141] T m = 120;

[0142] Substitute these values into the formula:

[0143]

[0144] Calculation process:

[0145] Calculate the numerator:

[0146] 0.7 · 0.85 = 0.595;

[0147] Calculate the denominator:

[0148]

[0149] Final calculated power:

[0150]

[0151] Result interpretation:

[0152] The calculated power resource value P m ≈0.0243W represents the power required by computing unit m during execution. This result indicates that module m has an energy consumption requirement of 0.0243 watts under the given resource occupancy, load, execution time, and voltage conditions. This value shows that the computing unit has relatively low power consumption, and based on this power value, further energy optimization and resource allocation adjustments can be made.

[0153] S302: Based on the computing unit power demand data, analyze the relationship between the power state of the computing unit and task execution, dynamically adjust the CPU frequency and memory power consumption, and generate power state adjustment data;

[0154] First, it is necessary to analyze the power state of the computing unit, detect the power requirements of different modules, and adjust the CPU frequency and the power consumption level of the memory based on the type of computing task of each module. This step may involve obtaining the power consumption data of each module through real-time monitoring and using optimization algorithms to dynamically adjust the resource usage of each module. When adjusting, the progress of task execution, the power consumption tolerance range of the module, and the resource requirements of priority tasks will be considered, so as to generate an optimized power adjustment plan.

[0155] S303: Based on the power state adjustment data, control the power of each computing resource according to the change of task load, dynamically adjust the power state of the resource based on the load change, and generate a power adjustment plan;

[0156] First, collect the power state adjustment data of all computing units, analyze it in combination with the load change of each task, and determine which computing resources need to increase power and which need to decrease power to adapt to the current computing requirements. For tasks with increased load, the power of their computing resources will be correspondingly increased, and vice versa, the power of the resources will be reduced. This adjustment process includes regular power detection and adjustment, and finally generates a power adjustment plan to ensure that each computing resource can flexibly respond to different load requirements and achieve more efficient power management.

[0157] Please refer to Figure 5 , the specific steps of S4 are as follows:

[0158] S401: Based on the power adjustment plan, real-time monitor the load change of the computing resources, collect the working state and power demand data of the computing unit, calculate the total power value required, and generate the power demand data of the computing resources;

[0159] Collect the working state and power demand data of the computing unit, calculate the total power value required, and calculate the total power demand of the computing resources according to the formula

[0160]

[0161] Calculate the total power demand of the computing resources.

[0162] In the formula, P total represents the total power required, S i represents the working state of the i-th computing unit, W i represents the power demand of the i-th computing unit, T i represents the task execution time of the i-th computing unit, and n represents the total number of computing units.

[0163] For each computing unit, first collect its working state S i , for example, S1 = 50, which represents the working state of the unit; the power demand W i, for example, W1 = 100 unit power, representing the power demand of the unit; task execution time T i , for example, T1 = 30 unit time, representing the duration of the task execution.

[0164] Substitute these values into the formula for calculation:

[0165]

[0166] Obtain the total required power P total = 166.67 unit power, indicating that the power demand of the computing resources has been calculated and can be used for subsequent resource allocation.

[0167] S402: Based on the computing resource power demand data, compare the load fluctuations of the computing resources, and combine with the current working state of the computing unit to dynamically adjust the power mode and generate power mode adjustment data;

[0168] First, it is necessary to monitor the load fluctuations of each computing resource, obtain the power demand of each unit in real time, and dynamically adjust the power mode according to its working state. When the load of the computing resource is high, the power mode can be selected to be improved to ensure the smooth execution of the task. On the contrary, it can be switched to a lower power mode to save energy. At the same time, considering the execution priority of the task, higher-priority tasks may require higher power support to ensure their stable execution. Finally, these adjustment data will be used to generate a power mode adjustment plan to ensure the optimal allocation of power resources in different states.

[0169] S403: Based on the power mode adjustment data, switch the power consumption mode according to the load changes of the computing resources, adjust the power configuration, and generate a power scheduling configuration;

[0170] It includes dynamically switching the power consumption mode of each computing resource by monitoring the changes in task load. As the load changes, the computing resource may need to switch from a lower power consumption mode to a higher power consumption mode, or vice versa, to meet the requirements of different loads. During the adjustment process, the optimization strategy of the power configuration needs to consider the execution time, priority, and availability of the task to ensure the reasonable allocation of power resources. In addition, according to the characteristics of different computing tasks, a personalized power scheduling plan is formulated to ensure the best match between the power demand of each computing resource and the load fluctuations, and then a final power scheduling configuration plan is generated.

[0171] Please refer to Figure 6 , the specific steps of S5 are as follows:

[0172] S501: Based on the power scheduling configuration, monitor the temperature changes of the computing unit in real time, collect temperature sensor data, analyze the temperature fluctuation trend, and generate computing unit temperature data;

[0173] The temperature data of the computing unit is calculated according to the formula

[0174]

[0175] Calculate the temperature data of the computing unit.

[0176] In the formula, T total represents the temperature data of the computing unit, T i represents the measured temperature of the i-th temperature sensor, and w represents the number of temperature sensors.

[0177] For each temperature sensor, collect its measured temperature T i , for example, T1 = 45°, T2 = 47°, T3 = 46°, which represents the temperature data collected from different sensors. Calculate the average temperature T total as the result of the temperature data:

[0178]

[0179] Obtain the temperature data T total = 46°, which indicates that the temperature fluctuation trend of the computing unit has been extracted and can be used for subsequent temperature management.

[0180] S502: Based on the temperature data of the computing unit, analyze the temperature change range, judge whether it exceeds the set threshold, and generate a temperature fluctuation analysis result;

[0181] First, it is necessary to monitor the temperature change of the computing unit, collect temperature data and calculate the temperature change range. Specifically, by comparing the temperature data, judge whether the temperature change range exceeds the set temperature threshold. For example, if the temperature change range is greater than a certain preset value, the system will judge that the temperature is too high and perform corresponding processing according to the result of the temperature data analysis. This process will ensure the safe operation of the device in a high-temperature environment and generate an analysis report of the temperature fluctuation for further optimizing the thermal management and power consumption configuration of the device.

[0182] S503: Based on the temperature fluctuation analysis result, if the temperature is overloaded, adjust the computing unit frequency and optimize the power consumption configuration to generate a thermal management and power consumption optimization configuration;

[0183] This process first needs to determine whether there is an overheating phenomenon in the computing unit based on the results of temperature fluctuation analysis. If the temperature exceeds the set safety threshold, the thermal management mechanism is activated to reduce the temperature by adjusting the frequency of the computing unit. At the same time, by adjusting the power consumption configuration and optimizing power management, the power consumption is reduced and the heat generation is decreased. This operation may involve adjusting the CPU frequency, memory usage, etc., to ensure that the device can operate stably and avoid overheating problems under temperature overload conditions. Finally, an optimized thermal management and power consumption configuration plan is generated to ensure the long-term stable operation of the device.

[0184] Please refer to Figure 7 , a computer energy-saving state control system, including:

[0185] The computing task scheduling module constructs a task graph model based on the computer task dependencies, calculates the dependency strength values between tasks, identifies the task resource scheduling order, preferentially allocates tasks with strong dependencies to shared computing resources, and generates a task correlation scheduling plan;

[0186] The resource allocation optimization module monitors the load of the computing unit in real time based on the task correlation scheduling plan, calculates the resource usage of each computing unit, and adjusts the allocation of computing resources according to the task priority and resource requirements to generate a resource allocation optimization plan;

[0187] The power adjustment module analyzes the power requirements of the computing unit according to the resource allocation optimization plan, dynamically adjusts the power state of the computing unit, and adjusts the CPU frequency and memory power consumption to control the power of each computing resource and generate a power adjustment plan;

[0188] The power management module analyzes the load of the computing resources in real time based on the power adjustment plan, calculates the power requirements, dynamically switches the power consumption mode, and generates a power supply scheduling configuration;

[0189] The temperature control management module monitors the temperature change of the computing unit according to the power supply scheduling configuration, analyzes the temperature fluctuation, sets and determines whether it exceeds the temperature threshold, adjusts the computing unit frequency to optimize the power consumption configuration, and generates a thermal management and power consumption optimization configuration.

[0190] As described above, it is only the specific implementation manner of the present invention, but the protection scope of the present invention is not limited thereto. Any person skilled in the art within the technical scope disclosed by the present invention can easily think of changes or substitutions, which should all be covered by the protection scope of the present invention. Therefore, the protection scope of the present invention should be subject to the protection scope of the claims.

Claims

1. A computer energy-saving state control method, characterized in that: The following steps are involved: S1: Based on the computer task dependency, a task graph model is constructed to calculate the dependency strength value between tasks, identify the task resource scheduling order, analyze the correlation between tasks, prioritize tasks with strong dependencies to shared computing resources, and generate a task correlation scheduling plan; S2: Based on the task correlation scheduling scheme, monitor the computing unit load in real time, calculate the resource usage of each computing unit, adjust the allocation of computing resources according to task priority and resource requirements, optimize resource configuration, and generate a resource allocation optimization scheme; S3: According to the resource allocation optimization plan, analyze the power requirements of the computing units, dynamically adjust the power state of the computing units, and adjust the CPU frequency and memory power consumption, control the power of each computing resource, adjust the power state of the resources based on load changes, and generate a power adjustment plan; S4: Based on the power regulation scheme, the computing resource load is analyzed in real time, the power supply demand is calculated, and according to the load change and the working state of the computing unit, the power consumption mode is dynamically switched to generate a power supply scheduling configuration; S5: According to the power scheduling configuration, monitor the temperature change of the computing unit, analyze the temperature fluctuation, set and determine whether the temperature threshold is exceeded, and if the temperature is overloaded, adjust the computing unit frequency to optimize the power consumption configuration, and generate thermal management and power consumption optimization configuration.

2. The computer energy-saving state control method according to claim 1, characterized in that: The task correlation scheduling scheme includes task dependency intensity, task scheduling priority, task correlation, and shared resource allocation scheme; the resource allocation optimization scheme includes computing unit load evaluation, resource utilization rate, task priority weight, and resource allocation adjustment strategy; the power regulation scheme includes power demand evaluation, computing unit power state, CPU frequency adjustment, and memory power consumption control; the power scheduling configuration includes power demand calculation, power power consumption mode selection, load adaptability adjustment, and power mode switching; the thermal management and power consumption optimization configuration includes temperature change monitoring, temperature fluctuation judgment, temperature threshold setting, and computing unit frequency adjustment.

3. The computer energy-saving state control method according to claim 1, characterized in that: Based on the computer task dependency, a task graph model is constructed to calculate the dependency strength value between tasks, identify the task resource scheduling order, analyze the correlation between tasks, and prioritize tasks with strong dependencies to shared computing resources. The specific steps for generating a task correlation scheduling plan are as follows: S101: Based on the dependency graph of computing tasks, collect input and output data of tasks, identify resource constraints and time requirements between tasks one by one, sort tasks based on their priorities, resource consumption, and time constraints, and generate a task dependency matrix; S102: Based on the task dependency matrix, calculate the dependency strength values ​​between tasks, compare the resource usage, completion time and priority of the tasks, and perform quantitative analysis to identify the strength of mutual influence between tasks, and generate a task dependency strength graph; S103: Based on the task dependency intensity graph, tasks with higher dependency intensity are preferentially selected, and the tasks are allocated according to resource consumption and operation cycle of the tasks in combination with scheduling strategies to generate a task correlation scheduling plan.

4. The computer energy-saving state control method according to claim 3, characterized in that: The calculation formula of the dependency strength value is specifically: Among them, I ij represents the dependency strength value between task i and task j, R i represents the resource usage value of task i, R j represents the resource usage value of task j, T ij represents the time difference between task i and task j, P i represents the priority of task i, T i represents the completion time of task i, T j represents the completion time of task j.

5. The computer energy-saving state control method according to claim 1, characterized in that: Based on the task correlation scheduling scheme, the computing unit load is monitored in real time, and the resource usage of each computing unit is calculated. According to the task priority and resource requirements, the allocation of computing resources is adjusted, and the resource configuration is optimized. The specific steps for generating the resource allocation optimization scheme are as follows: S201: Based on the task correlation scheduling scheme, monitor the load status of the computing unit in real time, collect the resource consumption and processing progress of the computing unit, analyze the computing capacity and task execution status of each unit, and generate computing unit load data; S202: Based on the computing unit load data, compare the resource consumption of each unit with the task priority and resource demand, determine whether each computing unit meets the task requirements, make adaptive adjustments, and generate resource demand and allocation adjustment data; S203: Based on the resource demand and allocation adjustment data, analyze the priorities and resource consumption between tasks, adjust the computing resource allocation, and generate a resource allocation optimization plan.

6. The computer energy-saving state control method according to claim 1, characterized in that: According to the resource allocation optimization scheme, the power requirements of the computing units are analyzed, the power state of the computing units is dynamically adjusted, and the CPU frequency and memory power consumption are adjusted to control the power of each computing resource. The power state of the resources is adjusted based on load changes. The specific steps for generating a power adjustment scheme are as follows: S301: Based on the resource allocation optimization solution, collect power demand data of each computing unit, analyze the load condition and power consumption change of each module, calculate the power resource value required by each computing unit, and generate computing unit power demand data; S302: Analyze the relationship between the power state of the computing unit and the task execution based on the computing unit power demand data, dynamically adjust the CPU frequency and memory power consumption, and generate power state adjustment data; S303: Based on the power state adjustment data, the power of each computing resource is controlled according to the task load change, the resource power state is dynamically adjusted based on the load change, and a power adjustment plan is generated.

7. The computer energy-saving state control method according to claim 6, characterized in that: The power resource numerical calculation formula is specifically: Among them, P m represents the power resource value required by computing unit m, R m represents the resource occupancy value of computing unit m, L m Represents the load of computing unit m, V m represents the voltage of computing unit m, T m Represents the execution time of computation unit m.

8. The computer energy-saving state control method according to claim 1, characterized in that: Based on the power regulation scheme, the computing resource load is analyzed in real time, the power supply demand is calculated, and the power consumption mode is dynamically switched according to the load change and the working state of the computing unit to generate the specific steps of power scheduling configuration are as follows: S401: Based on the power regulation scheme, monitor the load change of computing resources in real time, collect the working status and power demand data of the computing units, calculate the total power value required, and generate computing resource power demand data; S402: Based on the computing resource power demand data, comparing the load fluctuation of the computing resources, and combining the current working state of the computing unit, dynamically adjusting the power mode, and generating power mode adjustment data; S403: Based on the power mode adjustment data, according to the load change of the computing resources, switch the power consumption mode, adjust the power configuration, and generate a power scheduling configuration.

9. The computer energy-saving state control method according to claim 1, characterized in that: According to the power scheduling configuration, monitor the temperature change of the computing unit, analyze the temperature fluctuation, set and determine whether it exceeds the temperature threshold, if the temperature is overloaded, adjust the computing unit frequency to optimize the power consumption configuration, and generate the specific steps of thermal management and power consumption optimization configuration as follows: S501: Based on the power scheduling configuration, monitor the temperature change of the computing unit in real time, collect temperature sensor data, analyze the temperature fluctuation trend, and generate computing unit temperature data; S502: Analyze the temperature variation range based on the temperature data of the calculation unit, determine whether it exceeds a set threshold, and generate a temperature fluctuation analysis result; S503: Based on the temperature fluctuation analysis result, if the temperature is overloaded, adjust the frequency of the computing unit and optimize the power consumption configuration to generate a thermal management and power consumption optimization configuration.

10. A computer energy-saving state control system, characterized in that: According to a computer energy-saving state control method according to any one of claims 1 to 9, the system comprises: The computing task scheduling module builds a task graph model based on the computer task dependency, calculates the dependency strength value between tasks, identifies the task resource scheduling order, prioritizes tasks with strong dependencies to shared computing resources, and generates a task correlation scheduling plan; The resource allocation optimization module monitors the computing unit load in real time based on the task correlation scheduling scheme, calculates the resource usage of each computing unit, adjusts the allocation of computing resources according to task priority and resource requirements, and generates a resource allocation optimization scheme; The power regulation module analyzes the power requirements of the computing units according to the resource allocation optimization plan, dynamically adjusts the power state of the computing units, adjusts the CPU frequency and memory power consumption, controls the power of each computing resource, and generates a power regulation plan; The power management module analyzes computing resource loads in real time based on the power regulation scheme, calculates power requirements, dynamically switches power consumption modes, and generates power scheduling configurations; The temperature control management module monitors the temperature changes of the computing unit according to the power scheduling configuration, analyzes the temperature fluctuation, sets and determines whether the temperature threshold is exceeded, adjusts the computing unit frequency to optimize the power consumption configuration, and generates the thermal management and power consumption optimization configuration.

Images