Energy consumption optimization control method and device, server and storage medium
By processing server equipment operation data and applying load prediction models, an accurate energy consumption optimization control strategy is generated, which solves the problem of low load prediction accuracy in existing technologies and realizes efficient energy consumption management and resource optimization of servers.
Patent Information
- Application Number
- CN202510897204.8
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2025-06-30
- Publication Date
- 2025-09-16
AI Technical Summary
In the existing technology, server operation data is predicted through a fixed time series algorithm, resulting in low load prediction accuracy and the inability to achieve precise energy consumption optimization control.
By obtaining the operating data of multiple devices on the server, cleaning, removing outliers, reducing noise and format assimilation are performed, and then input into the pre-trained load prediction model for processing, the predicted load is determined, and the resource allocation strategy and target working mode are generated based on the predicted load. The resource allocation strategy is updated to generate an energy consumption optimization control strategy.
It improves the accuracy of server load prediction, realizes precise energy consumption optimization control, improves server performance and resource utilization efficiency, reduces operating costs, and complies with green operation requirements.
Smart Images
Figure CN120653094A_ABST
Abstract
Description
Technical Field
[0001] The present application relates to the field of server energy consumption technology, and in particular to an energy consumption optimization control method, device, server and storage medium. Background Art
[0002] Server energy optimization and control refers to a systematic approach to reducing power consumption during data center or server operation while ensuring stable performance through load forecasting and management strategies. Load forecasting, a core technology for server energy optimization and control, analyzes server operating indicators to predict future resource requirements. Therefore, accurate server load forecasting is crucial in server energy optimization and control.
[0003] Current energy consumption optimization and control methods in related technologies primarily use fixed time series algorithms to predict server operating data, obtain a predicted load, and then determine subsequent energy consumption optimization and control strategies based on the predicted load to optimize server energy consumption. However, this method of predicting server operating data using fixed time series algorithms in related technologies reduces the accuracy of server load predictions, making it impossible to subsequently accurately optimize server energy consumption. Summary of the Invention
[0004] The present application provides an energy consumption optimization control method, device, server and storage medium to at least solve the problem in the related art that the operating data in the server is predicted through a fixed time series algorithm, which reduces the accuracy of server load prediction and makes it impossible to subsequently perform accurate energy consumption optimization control on the server.
[0005] This application provides an energy consumption optimization control method, comprising:
[0006] Get the operating data of multiple devices in the server;
[0007] The operation data of each device is packaged and processed to obtain the operation data set;
[0008] Input the running data set into the pre-trained load prediction model for processing to obtain the predicted load of the server;
[0009] Perform resource allocation analysis and processing based on the predicted load to obtain a resource allocation strategy;
[0010] Determine a target operating mode from a plurality of preset operating modes according to the predicted load;
[0011] The resource allocation strategy is updated according to the target working mode to obtain an updated resource allocation strategy;
[0012] Generate an energy consumption optimization control strategy for the server based on the updated resource allocation strategy and target working mode;
[0013] According to the energy consumption optimization control strategy, the energy consumption of the server is optimized and controlled.
[0014] The present application also provides an energy consumption optimization control device, comprising:
[0015] A first acquisition module is used to acquire operation data of multiple devices in the server;
[0016] The first packaging module is used to package the operation data of each device to obtain an operation data set;
[0017] The input module is used to input the running data set into the pre-trained load prediction model for processing to obtain the predicted load of the server;
[0018] The analysis module is used to analyze and process resource allocation based on the predicted load and obtain the resource allocation strategy;
[0019] a determination module, configured to determine a target operating mode from a plurality of preset operating modes according to the predicted load;
[0020] An updating module is used to update the resource configuration strategy according to the target working mode to obtain an updated resource configuration strategy;
[0021] A generation module is used to generate an energy consumption optimization control strategy for the server based on the updated resource allocation strategy and target working mode;
[0022] The control module is used to optimize the energy consumption of the server according to the energy consumption optimization control strategy.
[0023] The present application also provides a server, comprising: a memory for storing a computer program; and a processor for implementing the steps of any of the above-mentioned energy consumption optimization control methods when executing the computer program.
[0024] The present application also provides a computer-readable storage medium, in which a computer program is stored. When the computer program is executed by a processor, the steps of any of the above-mentioned energy consumption optimization control methods are implemented.
[0025] The energy consumption optimization control method, device, server and storage medium provided in the embodiments of the present application obtain the operating data of multiple devices in the server; package the operating data of each device to obtain an operating data set; input the operating data set into a pre-trained load prediction model for processing to obtain the predicted load of the server; perform resource allocation analysis and processing based on the predicted load to obtain a resource allocation strategy; determine a target working mode from multiple preset working modes based on the predicted load; update the resource allocation strategy based on the target working mode to obtain an updated resource allocation strategy; generate an energy consumption optimization control strategy for the server based on the updated resource allocation strategy and the target working mode; perform energy consumption optimization control on the server based on the energy consumption optimization control strategy, and obtain the predicted load of the server by packaging each operating data into an operating data set and processing the operating data set through a trained load prediction model, so that the predicted load is more accurate, which is beneficial to the subsequent precise energy consumption optimization control of the server. BRIEF DESCRIPTION OF THE DRAWINGS
[0026] In order to more clearly illustrate the embodiments of the present application, the following is a brief introduction to the drawings required for use in the embodiments. Obviously, the drawings described below are only some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without any creative work.
[0027] Figure 1 A schematic diagram of an application scenario of the energy consumption optimization control method provided in an embodiment of the present application;
[0028] Figure 2 Schematic diagram of the process of energy consumption optimization control method provided in the embodiment of the present application Figure 1 ;
[0029] Figure 3 Schematic diagram of the process of energy consumption optimization control method provided in the embodiment of the present application Figure 2 ;
[0030] Figure 4 A schematic diagram of the structure of the energy consumption optimization control device provided in an embodiment of the present application;
[0031] Figure 5 A schematic diagram of the hardware structure of the server provided in an embodiment of the present application. DETAILED DESCRIPTION
[0032] The following will be combined with the accompanying drawings in the embodiments of this application to clearly and completely describe the technical solutions in the embodiments of this application. Obviously, the embodiments described are only part of the embodiments of this application, not all of them. Based on the embodiments in this application, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of this application.
[0033] It should be noted that, in the description of this application, the terms "comprises," "includes," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or device comprising a series of elements includes not only those elements, but also other elements not explicitly listed, or elements inherent to such process, method, article, or device. The terms "first," "second," etc., in this application are used to distinguish similar objects, and are not used to describe a particular order or sequence.
[0034] Energy consumption optimization and control of servers refers to a systematic method that uses load prediction and management strategies to reduce the power consumption of data centers or servers during operation while ensuring stable performance. Among them, load prediction, as the core technology of server energy consumption optimization and control, can understand the load change trend of the system in advance by analyzing the operating indicators of the server, so as to make corresponding resource allocation. Therefore, in the process of energy consumption optimization and control of servers, it is particularly important to accurately predict the load of the server. In the related art, the current energy consumption optimization and control method mainly uses a fixed time series algorithm to predict the operating data in the server, obtain the predicted load, and determine the subsequent energy consumption optimization and control strategy based on the predicted load to optimize the energy consumption of the server. However, in the related art, the method of predicting the operating data in the server through a fixed time series algorithm reduces the accuracy of the server load prediction, making it impossible to perform accurate energy consumption optimization and control of the server subsequently.
[0035] In order to solve the above technical problems, the embodiments of the present application propose the following technical concepts: the inventor takes into account the operating data of each device in the server, and determines an operating data set based on the operating data of each device; processes the operating data set based on a pre-trained load prediction model to obtain the predicted load of the server; uses the predicted load to determine the resource allocation strategy and the target working mode; updates the resource allocation strategy according to the target working mode, and generates an energy consumption optimization control strategy for the server based on the updated resource allocation strategy and the target working mode, which is used to optimize the energy consumption of the server. By determining each operating data as an operating data set and processing the operating data set through a trained load prediction model, the predicted load of the server is obtained, making the predicted load more accurate, which is beneficial to the subsequent precise energy consumption optimization control of the server.
[0036] In order to enable those skilled in the art to better understand the present application, the present application is further described in detail below with reference to the accompanying drawings and specific implementation methods.
[0037] In combination with the specific application environment architecture or specific hardware architecture on which the execution of the energy consumption optimization control method depends, the specific application environment architecture or specific hardware architecture is described here. Figure 1 , Figure 1 Schematic diagram of the application scenario of the energy consumption optimization control method provided in the embodiment of the present application.
[0038] like Figure 1 As shown, the scenario includes: a server 10.
[0039] The server 10 may be an independent server or a cluster composed of multiple servers, and the server 10 includes a controller 101 and multiple devices 102 .
[0040] The controller 101 obtains operating data from multiple devices 102 in the server 10; packages the operating data of each device 102 to obtain an operating data set; inputs the operating data set into a pre-trained load prediction model to obtain a predicted load of the server 10; performs resource allocation analysis based on the predicted load to obtain a resource allocation policy; determines a target operating mode from multiple preset operating modes based on the predicted load; updates the resource allocation policy based on the target operating mode to obtain an updated resource allocation policy; generates an energy consumption optimization control policy for the server 10 based on the updated resource allocation policy and the target operating mode; and performs energy consumption optimization control on the server 10 based on the energy consumption optimization control policy. A detailed embodiment will be used below for detailed description.
[0041] Figure 2 Schematic diagram of the process of energy consumption optimization control method provided in the embodiment of the present application Figure 1 ,like Figure 2 As shown, the embodiment of the present application provides an energy consumption optimization control method, which is described in detail as follows:
[0042] S201: Obtaining operation data of multiple devices in the server.
[0043] Specifically, the operation data of multiple devices in the server are obtained through a monitoring tool based on the Simple Network Management Protocol.
[0044] The operation data of each device may include the utilization rate and cache size of the central processing unit, the usage of the memory bar, the I / O data of the disk, the network bandwidth of the network adapter, and other operation data.
[0045] S202: Packaging the operation data of each device to obtain an operation data set.
[0046] Specifically, step S202 includes:
[0047] S2021: Clean the operating data of each device to obtain the cleaned operating data.
[0048] S2022: performing outlier removal processing on each cleaned operation data to obtain each cleaned operation data.
[0049] S2023: Perform noise reduction processing on each cleared operation data to obtain each processed operation data.
[0050] S2024: Performing format assimilation processing on the processed operation data to obtain operation data in the same format.
[0051] S2025: Sort the operation data in the same format in chronological order to obtain an operation data sequence.
[0052] S2026: Package the running data sequence into a running data set.
[0053] S203: Input the running data set into the pre-trained load prediction model for processing to obtain the predicted load of the server.
[0054] In this embodiment, the training process of the load prediction model in step S203 includes:
[0055] S2031: Obtain historical operation data of multiple devices in the server.
[0056] Exemplarily, the historical operation data of the plurality of devices are historical utilization and historical cache size of a central processing unit, historical usage of a memory bar, historical I / O data of a disk, and historical network bandwidth of a network adapter.
[0057] S2032: Packaging the historical operation data to obtain a historical operation data set.
[0058] Specifically, each historical operation data is cleaned, outliers are removed, noise is reduced, format assimilated, sorted and packaged to obtain a historical operation data set.
[0059] S2033: Extract features from the historical operation data set to obtain one or more key features.
[0060] For example, the key features are historical utilization of the central processing unit, historical usage of the memory bar, and historical I / O data of the disk.
[0061] S2034: Train the load prediction model according to the key features to obtain a pre-trained load prediction model.
[0062] In this embodiment, the load preset model is constructed using a machine learning algorithm; accordingly, step S2034 specifically includes:
[0063] S20341: Input each key feature into the machine learning algorithm for training to obtain the corresponding hidden layer state data.
[0064] In this embodiment, the machine learning algorithm may be a long short-term memory network or other algorithms.
[0065] In this embodiment, each key feature is input into a machine learning algorithm for training to obtain the corresponding hidden layer state data, and the calculation formula includes:
[0066]
[0067] Where, is the hidden layer state data at time t; σ is the activation function; is the connection weight matrix; is the hidden layer state data at time t-1; is the vector corresponding to the key feature at time t; is the bias term.
[0068] In addition, each key feature is input into the machine learning algorithm for training, and the back propagation algorithm is used to adjust the model parameters to reduce the prediction error.
[0069] S20342: Determine a pre-trained load prediction model based on the hidden layer state data.
[0070] In this embodiment, the calculation formula of the pre-trained load prediction model is determined based on the hidden layer state data, including:
[0071]
[0072] Where, To predict load; is the hidden layer state data at time t; T is the length of the prediction period.
[0073] S204: Perform resource allocation analysis and processing based on the predicted load to obtain a resource allocation strategy.
[0074] Specifically, step S204 includes:
[0075] S2041: Determine the required resource amount of the server based on the predicted load.
[0076] In this embodiment, the calculation formula for determining the required resource amount of the server according to the predicted load includes:
[0077]
[0078] Where, The amount of resources required for the server; To predict load; is the resource demand coefficient; is the basic resource demand coefficient.
[0079] S2042: Determine one or more corresponding services according to the running data set.
[0080] S2043: Determine the target required resource amount for each service based on the required resource amount.
[0081] S2044: Evaluate the priority of the corresponding services according to the resource requirements of each target to obtain the priority of each service.
[0082] In this embodiment, the priority of the corresponding services is evaluated according to the resource requirements of each target, and the calculation formula for obtaining the priority of each service includes:
[0083]
[0084] Where, The priority of the i-th service; The target resource requirement for the i-th service; The quality of service requirement for the i-th service; 、 and They are different adjustment coefficients respectively.
[0085] The QoS requirement for the i-th service refers to the quality standard that the i-th service is expected to achieve during its operation, covering multiple aspects, including response time, throughput, availability, fault tolerance, and user satisfaction, which jointly determine the performance of the service in meeting user needs.
[0086] For example, Specifically, the quantification method uses the five indicators of response time, throughput, availability, fault tolerance and user satisfaction, and the corresponding weights w1, w2, w3, w4 and w5 respectively to jointly calculate the service quality requirements of the i-th service. for:
[0087]
[0088] Where RTnorm is the normalized value of response time; TPnorm is the normalized value of throughput; Anorm is the normalized value of availability; FTnorm is the normalized value of fault tolerance; and USnorm is the normalized value of user satisfaction.
[0089] S2045: Determine the total required resources for each service based on the required resource amounts for each target and the priority of each service.
[0090] In this embodiment, the total required resources for each service are determined based on the target required resource amount and the priority of each service. The calculation formula includes:
[0091]
[0092] Where, Total resources required for each service; for the number of services; The priority of the i-th service; The target resource requirement for the i-th service; is the adjustment factor; is the Lagrange multiplier; The maximum total amount of allocatable resources.
[0093] S2046: Determine a resource allocation strategy based on total required resources.
[0094] For example, the resource allocation strategy is detailed configuration information such as the frequency allocation of the central processing unit, the space allocation of the memory bar, and the I / O allocation of the disk for each service within a corresponding time period.
[0095] S205: Determine a target operating mode from a plurality of preset operating modes according to the predicted load.
[0096] In this embodiment, the predicted load corresponds to the amount of resources required by the server; accordingly, step S205 specifically includes:
[0097] S2051: Acquire multiple preset working modes of the server, and construct an energy consumption model for each preset working mode to obtain an energy consumption value for each preset working mode.
[0098] In this embodiment, the plurality of preset operating modes include a high-performance operating mode, a standard operating mode, an energy-saving operating mode, and other preset operating modes.
[0099] In this embodiment, the calculation formula of the energy consumption model corresponding to each preset working mode includes:
[0100]
[0101] Where, is the energy consumption of any preset working mode M; is the mode coefficient; is the exponential coefficient; Basic correction item.
[0102] The quantitative index of the preset working mode M may be determined, and a mapping relationship between the preset working mode and the quantitative index may be established to calculate the quantitative value of the preset working mode M.
[0103] in, The actual energy consumption data under different preset working modes is measured experimentally, and the relationship curve between energy consumption and preset working mode is fitted.
[0104] S2052: Determine the matching degree between the required resource amount and each energy consumption value.
[0105] In this embodiment, the calculation formula for determining the matching degree between each target required resource amount and each energy consumption value includes:
[0106]
[0107] Where, The matching degree between resources and energy consumption for any preset working mode; The amount of resources required for the server; is the energy consumption of any preset working mode M; is the adjustment factor.
[0108] S2053: Determine a target operating mode from a plurality of preset operating modes according to the matching degrees.
[0109] Specifically, a target matching degree with the minimum matching degree value is determined from each matching degree, and a target operating mode is determined from a plurality of preset operating modes according to the target matching degree.
[0110] The target working mode can also be called the preferred working mode M opt .
[0111] S206: Update the resource configuration strategy according to the target working mode to obtain an updated resource configuration strategy.
[0112] S207: Generate an energy consumption optimization control strategy for the server according to the updated resource allocation strategy and target working mode.
[0113] Specifically, the updated resource allocation strategy and target working mode are integrated to generate an energy consumption optimization control strategy for the server.
[0114] S208: Perform energy consumption optimization control on the server according to the energy consumption optimization control strategy.
[0115] In summary, the energy consumption optimization control method provided in this embodiment obtains the operating data of multiple devices in the server; packages the operating data of each device to obtain an operating data set; inputs the operating data set into a pre-trained load prediction model for processing to obtain the predicted load of the server; performs resource configuration analysis and processing based on the predicted load to obtain a resource configuration strategy; determines a target working mode from multiple preset working modes based on the predicted load; updates the resource configuration strategy based on the target working mode to obtain an updated resource configuration strategy; generates an energy consumption optimization control strategy for the server based on the updated resource configuration strategy and the target working mode; performs energy consumption optimization control on the server based on the energy consumption optimization control strategy, and obtains the predicted load of the server by packaging each operating data into an operating data set and processing the operating data set through a trained load prediction model, so that the predicted load is more accurate, which is beneficial to the subsequent precise energy consumption optimization control of the server.
[0116] In addition, the energy consumption optimization control method provided in this embodiment obtains the operating data of multiple devices in the server through a monitoring tool based on the Simple Network Management Protocol, making it more efficient to obtain the operating data of each device and ensuring the accuracy and consistency of each operating data.
[0117] In addition, the energy consumption optimization control method provided in this embodiment obtains an operation data set by cleaning, removing outliers, reducing noise, formatting, sorting and packaging the operation data of each device, thereby improving the data quality of each operation data.
[0118] In addition, the energy consumption optimization control method provided in this embodiment can effectively capture the long-term dependencies in time series data by training the long short-term memory network through various key features, thereby improving the prediction accuracy of future load change trends and enhancing robustness.
[0119] In addition, the energy consumption optimization control method provided in this embodiment can reasonably allocate resources according to the importance of services and resource consumption by determining the required resources of the server and the priority of each service, ensuring that critical services obtain sufficient resource support during high-load periods, while avoiding excessive resource occupation by non-critical services. The refined resource management method not only improves the overall performance and service quality of the server, but also effectively reduces operating costs and realizes the maximum benefit of resource utilization.
[0120] In addition, the energy consumption optimization control method provided in this embodiment determines the target matching degree with the smallest matching degree value from each matching degree, and is used to determine the working mode with the smallest energy consumption as the preferred mode. It can significantly reduce energy consumption while meeting the server performance requirements, thereby achieving the goal of energy conservation and emission reduction, and meeting the requirements of green operation of modern data centers.
[0121] Figure 3 Schematic diagram of the energy consumption optimization control method provided in this embodiment Figure 2 In the embodiment of the present application, Figure 2 Based on the embodiment provided, a detailed description is given of the specific implementation method for updating the resource configuration strategy according to the target working mode in step S206 to obtain the updated resource configuration strategy. Figure 3 As shown, the method includes:
[0122] S301: Determine an operation adjustment coefficient of the server according to a target operating mode.
[0123] In this embodiment, the calculation formula for determining the operation adjustment coefficient of each device according to the target working mode includes:
[0124]
[0125] Where, is the server operation adjustment coefficient, It is the preferred working mode, which is equivalent to the target working mode; is the frequency adjustment factor of the CPU; is the maximum frequency of the CPU; Adjustment factor for memory bank; The maximum available memory capacity of the memory module.
[0126] S302: Adjust the operating status of each device according to the operating adjustment coefficient to obtain the operating status parameters of each device.
[0127] Specifically, the operating status of each device is adjusted according to the operating adjustment coefficient, and the operating status parameters of each device are collected using the implementation monitoring system.
[0128] S303: Evaluate the operating status parameters of each device to obtain a status evaluation value of the server.
[0129] In this embodiment, the operating status parameters of each device are evaluated to obtain the status evaluation value of the server, and the calculation formula includes:
[0130]
[0131] Where, is the status evaluation value of the server from time 0 to time t; is the actual utilization of the CPU; The actual usage of the memory bank; The actual I / O data of the disk; is the memory impact factor; is the disk impact factor.
[0132] The value of the disk impact factor can be determined by analyzing disk performance indicators, evaluating the impact of disk performance on server status, and determining a value range of the disk impact factor.
[0133] In addition, it is clear 、 and The original dimension of each parameter is converted into the same dimension range by using the minimum-maximum normalization method.
[0134] S304: Generate server operation status feedback information according to the status evaluation value.
[0135] In this embodiment, the operating status feedback information can be expressed as V(S(t)).
[0136] S305: Evaluate the operation status feedback information to obtain an evaluation result.
[0137] Specifically, a comprehensive evaluation of the overall health status and resource utilization efficiency is performed on the operation status feedback information to obtain an evaluation result.
[0138] S306: Optimize the trained prediction load model according to the operation status feedback information and the evaluation result to obtain an optimized prediction load model.
[0139] S307: Update the resource configuration strategy according to the optimized forecast load model to obtain an updated resource configuration strategy.
[0140] In this embodiment, the operation corresponding to the update may specifically be optimization, recalculation or other updating operations.
[0141] Exemplarily, the resource allocation strategy is recalculated based on the optimized forecast load model to obtain an updated resource allocation strategy.
[0142] In summary, the energy consumption optimization control method provided in this embodiment determines the operation adjustment coefficient of the server according to the target working mode; adjusts the operation status of each device according to the operation adjustment coefficient to obtain the operation status parameters of each device; evaluates the operation status parameters of each device to obtain the status evaluation value of the server; generates the operation status feedback information of the server according to the status evaluation value; evaluates the operation status feedback information to obtain the evaluation result; optimizes the trained prediction load model according to the operation status feedback information and the evaluation result to obtain the optimized prediction load model; updates the resource allocation strategy according to the optimized prediction load model to obtain the updated resource allocation strategy, so that the updated resource allocation strategy is more in line with the current resource demand, which is beneficial to the accurate energy consumption optimization control of the server.
[0143] In addition, the energy consumption optimization control method provided in this embodiment can timely discover potential problems and respond quickly through dynamic adjustment of each device in the server and collection and analysis of the operating status parameters of each device, ensuring that the server is always in the optimal operating state. The continuous monitoring and feedback mechanism not only improves the stability and reliability of the server, but also provides data support for subsequent resource optimization, which helps to continuously improve and enhance server performance.
[0144] Through the description of the above implementation methods, those skilled in the art can clearly understand that the method according to the above embodiment can be implemented by means of software plus the necessary general hardware platform, and of course it can also be implemented by hardware, but in many cases the former is a better implementation method.
[0145] Figure 4 This is a schematic diagram of the structure of the energy consumption optimization control device provided in the embodiment of the present application. Figure 4 As shown, an embodiment of the present application also provides an energy consumption optimization control device, including: a first acquisition module 401, a first packaging module 402, an input module 403, an analysis module 404, a determination module 405, an update module 406, a generation module 407 and a control module 408.
[0146] The first acquisition module 401 is used to acquire the operating data of multiple devices in the server;
[0147] The first packaging module 402 is used to package the operation data of each device to obtain an operation data set;
[0148] Input module 403, used to input the operating data set into the pre-trained load prediction model for processing to obtain the predicted load of the server;
[0149] Analysis module 404, configured to perform resource allocation analysis and processing based on the predicted load to obtain a resource allocation strategy;
[0150] A determination module 405 is configured to determine a target operating mode from a plurality of preset operating modes according to the predicted load;
[0151] An updating module 406 is configured to update the resource configuration strategy according to the target working mode to obtain an updated resource configuration strategy;
[0152] A generation module 407 is used to generate an energy consumption optimization control strategy for the server based on the updated resource allocation strategy and target working mode;
[0153] The control module 408 is configured to perform energy consumption optimization control on the server according to the energy consumption optimization control strategy.
[0154] In a possible implementation, the packaging module 402 specifically includes:
[0155] A cleaning unit is used to clean the operating data of each device to obtain the cleaned operating data;
[0156] a clearing unit, configured to perform an outlier clearing process on each cleaned operation data to obtain each cleaned operation data;
[0157] a noise reduction unit, configured to perform noise reduction processing on each cleared operation data to obtain each processed operation data;
[0158] an assimilation unit, configured to perform format assimilation processing on the processed operation data to obtain operation data in the same format;
[0159] A sorting unit is used to sort the operation data of the same format in chronological order to obtain an operation data sequence;
[0160] The packaging unit is used to package the running data sequence into a running data set.
[0161] In a possible implementation, the apparatus further includes:
[0162] The second acquisition module is used to obtain historical operation data of multiple devices in the server;
[0163] The second packaging module is used to package the historical operation data to obtain a historical operation data set;
[0164] The extraction module is used to extract features from the historical operation data set to obtain one or more key features;
[0165] The training module is used to train the load prediction model according to various key features to obtain a pre-trained load prediction model.
[0166] In one possible implementation, the load preset model is constructed using a machine learning algorithm; accordingly, the training module specifically includes:
[0167] The training unit is used to input each key feature into the machine learning algorithm for training to obtain the corresponding hidden layer state data;
[0168] The determination unit is used to determine the pre-trained load prediction model according to the hidden layer state data.
[0169] In a possible implementation, the analysis module 404 specifically includes:
[0170] A first determining unit, configured to determine a required amount of resources of the server according to the predicted load;
[0171] A second determining unit, configured to determine one or more corresponding services according to the running data set;
[0172] A third determining unit is used to determine the target required resource amount of each service according to the required resource amount;
[0173] An evaluation unit is used to evaluate the priority of corresponding services according to the resource requirements of each target and obtain the priority of each service;
[0174] A fourth determining unit, configured to determine the total required resources for each service according to each target required resource amount and the priority of each service;
[0175] The fifth determining unit is configured to determine a resource allocation strategy according to the total required resources.
[0176] In a possible implementation, the updating module 406 specifically includes:
[0177] a determination unit, configured to determine an operation adjustment coefficient of the server according to a target operation mode;
[0178] An adjustment unit, configured to adjust the operating state of each device according to an operating adjustment coefficient to obtain operating state parameters of each device;
[0179] An evaluation unit, configured to evaluate the operating status parameters of each device and obtain a status evaluation value of the server;
[0180] A generating unit, configured to generate server operation status feedback information according to the status evaluation value;
[0181] An evaluation unit, used to evaluate the operation status feedback information and obtain an evaluation result;
[0182] An optimization unit is used to optimize the trained prediction load model according to the operation status feedback information and evaluation results to obtain an optimized prediction load model;
[0183] The updating unit is used to update the resource configuration strategy according to the optimized forecast load model to obtain an updated resource configuration strategy.
[0184] In one possible implementation, the predicted load corresponds to the amount of resources required by the server; accordingly, the determination module 405 specifically includes:
[0185] A construction unit, configured to obtain a plurality of preset working modes of the server, and construct an energy consumption model for each preset working mode to obtain an energy consumption value for each preset working mode;
[0186] A first determining unit is used to determine the matching degree between the required resource amount and each energy consumption value;
[0187] The second determining unit is configured to determine a target operating mode from a plurality of preset operating modes according to the matching degrees.
[0188] For the description of the features in the embodiment corresponding to the energy consumption optimization control device, please refer to the relevant description of the embodiment corresponding to the energy consumption optimization control method, and no further details will be given here.
[0189] Figure 5 This is a schematic diagram of the server structure provided for this application. Figure 5 As shown, the server provided in this embodiment includes: at least one processor 501 and a memory 502. Optionally, the server also includes a communication component 503. The processor 501, the memory 502 and the communication component 503 are connected via a bus.
[0190] During the specific implementation process, at least one processor 501 executes the computer-executable instructions stored in the memory 502, so that the at least one processor 501 executes the above-mentioned energy consumption optimization control method embodiment.
[0191] The specific implementation process of the processor 501 can be found in the above method embodiment. Its implementation principle and technical effects are similar and will not be repeated here in this embodiment.
[0192] In the above embodiments, it should be understood that the processor may be a central processing unit (CPU), other general-purpose processors, digital signal processors (DSP), application-specific integrated circuits (ASIC), etc. A general-purpose processor may be a microprocessor or any conventional processor. The steps of the method disclosed in the application may be directly executed by a hardware processor or by a combination of hardware and software modules within the processor.
[0193] The memory may include random access memory (RAM) and may also include non-volatile memory (NVM), such as at least one disk storage.
[0194] A bus can be an Industry Standard Architecture (ISA) bus, a Peripheral Component Interconnect (PCI) bus, or an Extended Industry Standard Architecture (EISA) bus. Buses can be categorized as address buses, data buses, and control buses. For ease of illustration, the buses in the drawings of this application are not limited to just one bus or just one type of bus.
[0195] An embodiment of the present application further provides a computer-readable storage medium, in which a computer program is stored, wherein the computer program is configured to execute the steps of any of the above-mentioned energy consumption optimization control method embodiments when running.
[0196] In an exemplary embodiment, the computer-readable storage medium may include, but is not limited to, various media that can store computer programs, such as a USB flash drive, a read-only memory (ROM), a random access memory (RAM), a mobile hard disk, a magnetic disk, or an optical disk.
[0197] An embodiment of the present application further provides a computer program product, which includes a computer program. When the computer program is executed by a processor, the steps in any of the above energy consumption optimization control method embodiments are implemented.
[0198] An embodiment of the present application also provides another computer program product, including a non-volatile computer-readable storage medium, which stores a computer program. When the computer program is executed by a processor, it implements the steps in any of the above-mentioned energy consumption optimization control method embodiments.
[0199] Professionals may further appreciate that the units and algorithm steps of each example described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, computer software, or a combination of the two. In order to clearly illustrate the interchangeability of hardware and software, the above description has generally described the components and steps of each example according to their functions. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professionals and technicians may use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of this application.
[0200] The above is a detailed introduction to the energy consumption optimization control method, device, server and storage medium provided by the present application. This article uses specific examples to illustrate the principles and implementation methods of the present application. The description of the above embodiments is only used to help understand the method of the present application and its core idea. It should be pointed out that for ordinary technicians in this technical field, without departing from the principles of the present application, several improvements and modifications can be made to the present application, and these improvements and modifications also fall within the scope of protection of the claims of the present application.
Claims
1. An energy consumption optimization control method, characterized in that: include: Get the operating data of multiple devices in the server; The operation data of each device is packaged and processed to obtain the operation data set; Inputting the running data set into a pre-trained load prediction model for processing to obtain a predicted load of the server; Performing resource allocation analysis and processing according to the predicted load to obtain a resource allocation strategy; determining a target operating mode from a plurality of preset operating modes according to the predicted load; Updating the resource allocation strategy according to the target working mode to obtain an updated resource allocation strategy; generating an energy consumption optimization control strategy for the server according to the updated resource allocation strategy and the target operating mode; According to the energy consumption optimization control strategy, energy consumption optimization control is performed on the server.
2. The energy consumption optimization control method according to claim 1, characterized in that: The operation data of each device is packaged and processed to obtain an operation data set, including: Cleaning the operation data of each device to obtain cleaned operation data; performing outlier removal processing on the cleaned operation data to obtain cleaned operation data; performing noise reduction processing on the cleared operation data to obtain processed operation data; Performing format assimilation processing on the processed operation data to obtain operation data of the same format; Sorting the operation data in the same format in chronological order to obtain an operation data sequence; The running data sequence is packaged into a running data set.
3. The energy consumption optimization control method according to claim 1, characterized in that: The training process of the load prediction model includes: Get historical operation data of multiple devices in the server; Packaging and processing each historical operation data to obtain a historical operation data set; Performing feature extraction on the historical operation data set to obtain one or more key features; The load prediction model is trained according to the key features to obtain a pre-trained load prediction model.
4. The energy consumption optimization control method according to claim 3, characterized in that: The load preset model is constructed using a machine learning algorithm. Accordingly, the load prediction model is trained according to the key features to obtain a pre-trained load prediction model, including: Inputting each key feature into the machine learning algorithm for training to obtain corresponding hidden layer state data; The pre-trained load prediction model is determined according to the hidden layer state data.
5. The energy consumption optimization control method according to any one of claims 1 to 4, characterized in that: The performing resource configuration analysis and processing according to the predicted load to obtain a resource configuration strategy includes: determining a required amount of resources for the server according to the predicted load; Determining one or more corresponding services according to the operating data set; Determine the target required resource amount for each service based on the required resource amount; Evaluate the priority of the corresponding services according to the resource requirements of each target to obtain the priority of each service; Determine the total required resources for each service based on the target required resource amounts and the priority of each service; A resource allocation strategy is determined based on the total required resources.
6. The energy consumption optimization control method according to claim 1, characterized in that: The updating of the resource configuration strategy according to the target working mode to obtain an updated resource configuration strategy includes: determining an operation adjustment coefficient of the server according to the target operating mode; Adjust the operating status of each device according to the operation adjustment coefficient to obtain the operating status parameters of each device; Evaluate the operating status parameters of each device to obtain a status evaluation value of the server; generating operating status feedback information of the server according to the status evaluation value; Evaluating the operating status feedback information to obtain an evaluation result; Optimizing the trained prediction load model according to the operating status feedback information and the evaluation result to obtain an optimized prediction load model; The resource configuration strategy is updated according to the optimized forecast load model to obtain an updated resource configuration strategy.
7. The energy consumption optimization control method according to claim 5, characterized in that: wherein the predicted load corresponds to the required amount of resources of the server; Accordingly, determining a target operating mode from a plurality of preset operating modes according to the predicted load includes: Acquire multiple preset working modes of the server, and construct an energy consumption model for each preset working mode to obtain an energy consumption value for each preset working mode; Determining the degree of matching between the required resource amounts and the energy consumption values; According to each matching degree, a target operating mode is determined from the plurality of preset operating modes.
8. An energy consumption optimization control device, characterized in that: include: A first acquisition module is used to acquire operation data of multiple devices in the server; The first packaging module is used to package the operation data of each device to obtain an operation data set; An input module, configured to input the operating data set into a pre-trained load prediction model for processing to obtain a predicted load of the server; An analysis module, configured to perform resource allocation analysis and processing based on the predicted load to obtain a resource allocation strategy; a determination module, configured to determine a target operating mode from a plurality of preset operating modes according to the predicted load; An updating module, configured to update the resource configuration strategy according to the target working mode to obtain an updated resource configuration strategy; A generating module, configured to generate an energy consumption optimization control strategy for the server according to the updated resource allocation strategy and the target operating mode; The control module is used to perform energy consumption optimization control on the server according to the energy consumption optimization control strategy.
9. A server, characterized in that: include: memory for storing computer programs; A processor, configured to implement the steps of the energy consumption optimization control method according to any one of claims 1 to 7 when executing the computer program.
10. A computer-readable storage medium, characterized in that The computer-readable storage medium stores a computer program, wherein when the computer program is executed by a processor, the steps of the energy consumption optimization control method according to any one of claims 1 to 7 are implemented.