CN117762632A - Calculation management method based on calculation operation system - Google Patents
Calculation management method based on calculation operation system Download PDFInfo
- Publication number
- CN117762632A CN117762632A CN202311800788.XA CN202311800788A CN117762632A CN 117762632 A CN117762632 A CN 117762632A CN 202311800788 A CN202311800788 A CN 202311800788A CN 117762632 A CN117762632 A CN 117762632A
- Authority
- CN
- China
- Prior art keywords
- task
- computing
- node
- parallel
- abnormal
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
- 238000007726 management method Methods 0.000 title claims abstract description 40
- 238000004364 calculation method Methods 0.000 title claims description 22
- 238000012545 processing Methods 0.000 claims abstract description 143
- 238000013507 mapping Methods 0.000 claims abstract description 93
- 238000004088 simulation Methods 0.000 claims abstract description 51
- 238000005457 optimization Methods 0.000 claims abstract description 17
- 230000002159 abnormal effect Effects 0.000 claims description 83
- 238000000034 method Methods 0.000 claims description 23
- 238000004458 analytical method Methods 0.000 claims description 9
- 230000000694 effects Effects 0.000 abstract description 5
- 238000013468 resource allocation Methods 0.000 abstract description 4
- 238000012986 modification Methods 0.000 description 4
- 230000004048 modification Effects 0.000 description 4
- 238000013500 data storage Methods 0.000 description 2
- 238000012544 monitoring process Methods 0.000 description 2
- RYAUSSKQMZRMAI-YESZJQIVSA-N (S)-fenpropimorph Chemical compound C([C@@H](C)CC=1C=CC(=CC=1)C(C)(C)C)N1C[C@H](C)O[C@H](C)C1 RYAUSSKQMZRMAI-YESZJQIVSA-N 0.000 description 1
- 230000005540 biological transmission Effects 0.000 description 1
- 238000013075 data extraction Methods 0.000 description 1
- 238000010586 diagram Methods 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
- 238000012163 sequencing technique Methods 0.000 description 1
- 238000012549 training Methods 0.000 description 1
- 239000002699 waste material Substances 0.000 description 1
Classifications
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02D—CLIMATE CHANGE MITIGATION TECHNOLOGIES IN INFORMATION AND COMMUNICATION TECHNOLOGIES [ICT], I.E. INFORMATION AND COMMUNICATION TECHNOLOGIES AIMING AT THE REDUCTION OF THEIR OWN ENERGY USE
- Y02D10/00—Energy efficient computing, e.g. low power processors, power management or thermal management
Landscapes
- Supply And Distribution Of Alternating Current (AREA)
Abstract
The invention discloses a power management method based on a power operation system, which relates to the technical field of data processing, and comprises the following steps: acquiring multi-level computing resource deployment information of a target computing power operating system, and establishing a multi-level computing resource deployment network; and carrying out configuration of the multi-stage computing nodes and storage of an original database on the preset processing tasks, carrying out task execution simulation on the multi-stage computing nodes, obtaining independent task processing precision, parallel task processing precision, independent task time delay and parallel task time delay, carrying out optimization resetting of computing power computing resources on the multi-stage computing nodes, obtaining updated node-task mapping results, and carrying out computing power management. The invention solves the technical problem that the current computational management method cannot meet the computational processing precision requirement along with the increase of the complexity of computational processing tasks in the prior art, and achieves the technical effect of optimizing the computational resource allocation through the split nodes and improving the computational processing precision.
Description
Technical Field
The invention relates to the technical field of data processing, in particular to a power management method based on a power operation system.
Background
The power management is a process of reasonably configuring, monitoring, controlling and optimizing the computing resources, and the main aims of the power management are to realize resource sharing, save cost, improve efficiency and guarantee safety. Through reasonable resource allocation and sharing, the computing resources can be utilized to the maximum extent, and waste is avoided. Meanwhile, by effectively configuring and monitoring the resources, the investment of enterprises in hardware and software can be reduced, and the cost is further saved. In addition, the calculation efficiency of enterprises can be improved through calculation management, the business process is accelerated, and the competitiveness of the enterprises is improved. However, at present, the complexity of management tasks is higher and higher, and the current management method cannot meet the accuracy requirement of the management of the calculation.
Disclosure of Invention
The application provides a power management method based on a power operation system, which is used for solving the technical problem that the current power management method cannot meet the power processing precision requirement along with the increase of the complexity of power processing tasks in the prior art.
In a first aspect of the present application, there is provided a method of managing a computing force based on a computing force operating system, the method comprising: acquiring multi-level computing resource deployment information of a target computing power operating system, and establishing a multi-level computing resource deployment network, wherein the multi-level computing resources comprise cloud computing resources, edge computing resources and terminal computing resources; acquiring a preset processing task of the target computing power operating system and an original database of a preset intelligent model, wherein the preset processing task has a user privacy grade identifier; performing configuration of multi-stage computing nodes and storage of an original database on the preset processing task based on the user privacy grade identification and the multi-stage computing resource deployment network to obtain a node-task mapping result; establishing a multi-stage computing node digital twin model based on the node-task mapping result, and performing task independent execution simulation and task parallel execution simulation on the multi-stage computing node to acquire independent task processing precision, parallel task processing precision, independent task time delay and parallel task time delay; performing optimization reset of computing power computing resources on the multi-stage computing nodes according to the independent task processing precision, the parallel task processing precision, the independent task time delay and the parallel task time delay to obtain updated node-task mapping results; and performing calculation power management according to the updated node-task mapping result.
In a second aspect of the present application, there is provided a computing management system based on a computing power operating system, the system comprising: the system comprises a multistage computing resource deployment network establishment module, a terminal computing resource and a target computing power operation system, wherein the multistage computing resource deployment network establishment module is used for acquiring multistage computing resource deployment information of the target computing power operation system and establishing a multistage computing resource deployment network, and the multistage computing resources comprise cloud computing resources, edge computing resources and terminal computing resources; the preset processing task acquisition module is used for acquiring a preset processing task of the target computing power operating system and an original database of a preset intelligent model, and the preset processing task has a user privacy grade identifier; the node-task mapping result acquisition module is used for carrying out configuration of multi-stage computing nodes and storage of an original database on the preset processing task based on the user privacy grade identification and the multi-stage computing resource deployment network to obtain a node-task mapping result; the task execution simulation module is used for establishing a multi-stage computing node digital twin model based on the node-task mapping result, performing task independent execution simulation and task parallel execution simulation on the multi-stage computing node, and acquiring independent task processing precision, parallel task processing precision, independent task time delay and parallel task time delay; the updating node-task mapping result acquisition module is used for carrying out optimization reset on computing power computing resources of the multi-stage computing nodes according to the independent task processing precision, the parallel task processing precision, the independent task time delay and the parallel task time delay to acquire an updating node-task mapping result; and the calculation management module is used for carrying out calculation management according to the updated node-task mapping result.
One or more technical solutions provided in the present application have at least the following technical effects or advantages:
the utility model provides a power management method based on a power operation system, which relates to the technical field of data processing, and aims at solving the technical problem that the current power management method cannot meet the power processing precision requirement along with the increase of the complexity of the power processing task in the prior art, and realizes the technical effects of optimizing power resource allocation and improving the power processing precision through the nodes by performing the power resource allocation optimization through the nodes.
Drawings
In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings required for the description of the embodiments will be briefly described below, and it is apparent that the drawings in the following description are only some embodiments of the present invention, and other drawings may be obtained according to these drawings without inventive effort for a person skilled in the art.
Fig. 1 is a schematic flow chart of a power management method based on a power operation system according to an embodiment of the present application;
FIG. 2 is a schematic flow chart of obtaining a node-task mapping result in a power management method based on a power operation system according to an embodiment of the present application;
FIG. 3 is a flowchart of obtaining an updated node-task mapping result in a power management method based on a power operation system according to an embodiment of the present application;
fig. 4 is a schematic structural diagram of a power management system based on a power operation system according to an embodiment of the present application.
Reference numerals illustrate: the system comprises a multistage computing resource deployment network establishment module 11, a preset processing task acquisition module 12, a node-task mapping result acquisition module 13, a task execution simulation module 14, an update node-task mapping result acquisition module 15 and a power management module 16.
Detailed Description
The application provides a power management method based on a power operation system, which is used for solving the technical problem that the current power management method cannot meet the power processing precision requirement along with the increase of the complexity of power processing tasks in the prior art.
The technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the drawings in the embodiments of the present application. It will be apparent that the described embodiments are only some, but not all, of the embodiments of the present application. All other embodiments, which can be made by one of ordinary skill in the art based on the embodiments herein without making any inventive effort, are intended to be within the scope of the present application.
It should be noted that, the terms "first," "second," and the like in the description of the present application and the above figures are used for distinguishing between similar objects and not necessarily for describing a particular sequential or chronological order. It is to be understood that the data so used may be interchanged where appropriate such that embodiments of the present application described herein may be implemented in sequences other than those illustrated or otherwise described herein. Furthermore, the terms "comprises," "comprising," and "having," and any variations thereof, are intended to cover a non-exclusive inclusion, such that a process, method, system, article, or server that comprises a list of steps or elements is not necessarily limited to those steps or elements expressly listed but may include other steps or modules not expressly listed or inherent to such process, method, article, or apparatus.
Example 1
As shown in fig. 1, the present application provides a method for managing power based on a power operation system, the method comprising:
p10: acquiring multi-level computing resource deployment information of a target computing power operating system, and establishing a multi-level computing resource deployment network, wherein the multi-level computing resources comprise cloud computing resources, edge computing resources and terminal computing resources;
optionally, the computing power operating system is an operating system for computing power distribution for various computing network clients and computing network services, where the computing power operating system includes cloud computing resources, edge computing resources and terminal computing resources, where the cloud computing resources refer to computing resources of a cloud center, such as a central server and a database, the edge computing resources are computing power devices used for performing edge computing, the edge computing is a method for moving computing tasks from a cloud computing device to a network edge, which can reduce delay of data in a transmission process, improve response speed and reduce burden of a data center, the terminal computing resources refer to computing resources owned by various terminal devices, such as intelligent devices and various sensors, further, the computing resource deployment information refers to arrangement information of the computing resources in the system, such as connection relations of the computing resources, and deployment positions, numbers and configurations of the computing resources, such as servers, databases, storage devices and chips. Furthermore, the multistage computing resource deployment network is established according to the multistage computing resource deployment information, so that the distribution condition, connection relation and the like of the multistage computing resources can be clearly reflected.
P20: acquiring a preset processing task of the target computing power operating system and an original database of a preset intelligent model, wherein the preset processing task has a user privacy grade identifier;
in one possible embodiment of the present application, the target computing power operating system includes a plurality of preset computing power processing tasks, that is, preset processing tasks, and a plurality of intelligent models for performing computing power task processing, that is, preset intelligent models, and an original database of the preset processing tasks and the preset intelligent models is collected, where the original database is a database for performing training data extraction of the preset intelligent models and task data storage. And, the preset processing task has a user privacy level identification, that is, different preset processing tasks have different user security requirement levels.
P30: performing configuration of multi-stage computing nodes and storage of an original database on the preset processing task based on the user privacy grade identification and the multi-stage computing resource deployment network to obtain a node-task mapping result;
further, as shown in fig. 2, step P30 in the embodiment of the present application further includes:
p31: the multi-stage computing nodes comprise cloud computing nodes, edge computing nodes and terminal computing nodes, and the privacy protection grade relation of the multi-stage computing nodes is that the terminal computing nodes are larger than the cloud computing nodes and larger than the edge computing nodes;
p32: performing node-task matching mapping according to the privacy protection level relation and the user privacy level mark to obtain an initial mapping result;
p33: and carrying out operation index balance analysis on the initial mapping result and optimizing the initial mapping result to obtain the node-task mapping result.
Specifically, the configuration of the multi-stage computing nodes is performed for the preset processing task by referring to the user privacy grade identifier and the residual computing power distribution condition in the multi-stage computing resource deployment network, and the storage of an original database is performed on the corresponding computing nodes, so that a node-task mapping result, namely the corresponding relationship between the computing nodes and the preset processing task, is obtained. The multi-stage computing nodes comprise cloud computing nodes, edge computing nodes and terminal computing nodes, task processing can be carried out through a cloud center, edge computing equipment and terminal equipment respectively, and the privacy protection class relationship of the multi-stage computing nodes is that the terminal computing nodes are larger than the cloud computing nodes and larger than the edge computing nodes, that is to say, the data security of the cloud computing nodes, the edge computing nodes and the terminal computing nodes is gradually increased.
Further, according to the privacy protection level relation of each type of computing node and the privacy level identification of the user, performing node matching for each preset processing task to obtain an initial mapping result, and further, performing operation index balance analysis on each computing node in each mapping result aiming at the initial mapping result, namely analyzing whether the residual computing power of each computing node can meet the task requirement, and performing adjustment and optimization on the initial mapping result according to the analysis result, such as replacing the node with more residual computing power to obtain the node-task mapping result, thereby improving the accuracy of computing power distribution.
Further, step P33 in the embodiment of the present application further includes:
p33-1: acquiring an operation capability index and a storage space of the multistage computing node;
p33-2: acquiring a starting demand operation index and a starting demand data space of the preset intelligent model of the preset processing task;
p33-3: and performing task allocation on the initial mapping result by using the operation capability index and the storage space, the starting requirement operation index and the starting requirement data space to obtain the node-task mapping result.
The computing capability index, that is, the residual computing power, and the idle computing power, for example, the floating point computing number or the integer computing number, are respectively collected, and meanwhile, the intelligent computing requirements of the preset intelligent model corresponding to the preset processing task are obtained, including the starting requirement computing index and the starting requirement data space, that is, the numerical value and the data storage space size of each computing index required by the normal operation of the model, further, according to the computing capability index and the storage space of each computing node, the matching and the judgment are performed with the starting requirement computing index and the starting requirement data space of the model, the initial mapping result that the node computing capability index and the storage space do not meet the model computing requirement is adjusted and optimized, and the computing node meeting the requirement is selected, so that the node-task mapping result is obtained.
P40: establishing a multi-stage computing node digital twin model based on the node-task mapping result, and performing task independent execution simulation and task parallel execution simulation on the multi-stage computing node to acquire independent task processing precision, parallel task processing precision, independent task time delay and parallel task time delay;
further, step P40 of the embodiment of the present application further includes:
p41: performing association of cloud computing nodes, edge computing nodes and terminal computing nodes based on the node-task mapping result, acquiring a plurality of computing node chains, and acquiring a multi-stage task set of a first computing node chain;
p42: randomly combining the tasks in the multi-stage task set to obtain M groups of parallel simulation tasks, wherein M is an integer greater than 0,n is the total number of tasks in the multi-stage task set, i is the number of combined tasks, and C is the sign of the combined number;
p43: performing multiple simulation on any independent task in the multi-stage task set through the multi-stage computing node digital twin model, and recording the independent task processing precision and the independent task time delay;
p44: and performing multiple-time task parallel simulation on the M groups of parallel simulation tasks through the multi-stage computing node digital twin model, and recording the parallel task processing precision, independent task time delay and parallel task time delay.
It should be understood that, based on the node-task mapping result, a digital twin model of the multi-stage computing node is established through a digital twin technology, further, the preset processing task is started by the terminal and executed by the terminal, and the cloud or edge supply computing resource is used for processing the task and then transmitting the task processing result back to the terminal, so that a connection relationship exists between the multi-stage computing nodes, for example, the task of the edge node is started by the terminal node, and the edge node returns to the terminal node after processing. Therefore, the association of the cloud computing node, the edge computing node and the terminal computing node is performed based on the node-task mapping result, a plurality of computing node chains can be obtained, and a multi-stage task set of the first computing node chain is obtained, wherein the multi-stage task set comprises a plurality of independent tasks. Further, a plurality of independent tasks in the multi-stage task set are randomly combined to obtain M groups anda line simulation task, wherein M is an integer greater than 0,n is the total number of tasks in the multi-stage task set, i is the number of combined tasks, i.e. the number of parallel tasks, and C is the combined number symbol.
Further, the multi-stage computing node digital twin model is used for performing multiple simulation on any independent task in the multi-stage task set, recording the processing precision and the task time delay of the independent task, and similarly, the multi-stage computing node digital twin model performs multiple task parallel simulation on the M groups of parallel simulation tasks, and recording the parallel task processing precision, the independent task time delay and the parallel task time delay.
Further, the task processing precision is an index describing the accuracy and consistency of task processing, and the method for obtaining the task processing precision comprises the following steps:
p45: obtaining a calibration execution result of any task in the multi-stage task set;
p46: obtaining multiple simulation results and comparing the simulation results with the calibration execution results in accuracy to obtain accuracy indexes;
p47: consistency comparison is carried out on the multiple simulation results, and consistency indexes are obtained;
p48: and carrying out weighted average calculation on the accuracy index and the consistency index to obtain the task processing precision.
Optionally, setting a calibration execution result of any task in the multi-stage task set, that is, a standard execution result, including an accuracy threshold value of calculation force distribution, and the like, comparing the multiple simulation results with the calibration execution result, judging an accuracy standard rate to obtain an accuracy index, further, comparing consistency of results, that is, similarity, aiming at multiple simulation results of the same task to obtain a consistency index, and carrying out weighted average calculation on the accuracy index and the consistency index according to processing result requirements to obtain the task processing precision.
P50: performing optimization reset of computing power computing resources on the multi-stage computing nodes according to the independent task processing precision, the parallel task processing precision, the independent task time delay and the parallel task time delay to obtain updated node-task mapping results;
further, as shown in fig. 3, step P50 in the embodiment of the present application further includes:
p51: extracting a first abnormal task if the independent task processing precision is smaller than the preset independent processing precision and/or the independent task time delay is larger than the preset independent processing time delay;
p52: if the parallel task processing precision is smaller than the preset parallel processing precision and/or the parallel task time delay is larger than the preset parallel processing time delay, extracting a first abnormal parallel task combination;
p53: performing task independent execution analysis on tasks in the first abnormal parallel task combination, and extracting a second abnormal task;
p54: computing node optimization is carried out based on the user privacy grade identifiers of the first abnormal task and the second abnormal task, an abnormal optimizing node is obtained, and abnormal mapping association between the abnormal optimizing node and the first abnormal task and the second abnormal task is established;
p55: updating the node-task mapping result with the abnormal mapping association.
Illustratively, the computing power computing resources of the multi-stage computing nodes are optimized and reset according to the independent task processing precision, the parallel task processing precision, the independent task time delay and the parallel task time delay. Specifically, if the processing precision of any independent task is smaller than the preset independent processing precision and/or the independent task time delay is larger than the preset independent processing time delay, the processing precision and the processing time delay of the independent task are at least one of which does not meet the preset requirement, and the independent task is taken as the first abnormal task.
Further, if the processing precision of any parallel task is smaller than the preset parallel processing precision and/or the parallel task time delay is larger than the preset parallel processing time delay, the parallel task is used as a first abnormal parallel task combination, further, tasks in the first abnormal parallel task combination are independently executed and analyzed, and the task with the minimum independent task processing precision and the maximum independent task time delay is extracted as a second abnormal task.
Further, computing node optimizing is performed based on the user privacy grade identifiers of the first abnormal task and the second abnormal task, a node capable of meeting the processing time delay and the processing precision of the abnormal task is reselected to be used as an abnormal optimizing node, abnormal mapping association between the abnormal optimizing node and the first abnormal task and the abnormal mapping association between the abnormal optimizing node and the second abnormal task are established, updating and replacing are performed on the node-task mapping result according to the abnormal mapping association, and an updated node-task mapping result is obtained, so that privacy protection performance of user data is improved.
Further, step P54 in the embodiment of the present application further includes:
p54-1: performing serialization processing of a calculation node chain based on the parallel task processing precision and the parallel task time delay to obtain an initial optimizing sequence;
p54-2: based on the user privacy level identifiers of the first abnormal task and the second abnormal task, taking the balanced user privacy level as optimizing constraint to obtain optimizing space;
p54-3: the optimizing space comprises original computing nodes of which the privacy protection level in the initial optimizing sequence is more than or equal to that of the first abnormal task and the second abnormal task;
p54-4: and in the optimizing space, performing node parallel optimization of the first abnormal task and the second abnormal task through a multi-stage computing node digital twin model, and obtaining the abnormal optimizing node.
Optionally, according to the processing precision of the parallel task and the size of the delay of the parallel task, sequencing the corresponding computing node chains from top to bottom to obtain an initial optimizing sequence, aiming at the user privacy levels of the first abnormal task and the second abnormal task, taking the balanced user privacy level as optimizing constraint, wherein the balanced user privacy level refers to the privacy protection level of the computing node with the privacy level larger than that of the abnormal task, extracting the original computing nodes with the privacy protection level larger than or equal to that of the first abnormal task and the second abnormal task in the initial optimizing sequence, constructing an optimizing space, further, traversing and selecting any original computing node, adding the first abnormal task and the second abnormal task into the original task of the node, and executing parallel simulation through a multi-stage computing node digital twin model until the processing delay and the processing precision are met, thereby obtaining the abnormal optimizing node.
P60: and performing calculation power management according to the updated node-task mapping result.
Specifically, referring to the updated node-task mapping result, the computing power management of the user computing power distribution task is performed, and the computing power processing precision and the privacy protection performance of the user data are improved.
In summary, the embodiments of the present application have at least the following technical effects:
according to the method, a multistage computing resource deployment network is established by collecting multistage computing resource deployment information of a target computing power operating system, configuration of multistage computing nodes and storage of an original database are carried out on preset processing tasks, task execution simulation is carried out, optimization and reset of computing power computing resources are carried out on the multistage computing nodes according to simulation results, updated node-task mapping results are obtained, and computing power management is carried out.
The technical effect of optimizing the distribution of the computing power resources through the sub-nodes and improving the processing precision of the computing power is achieved.
Example two
Based on the same inventive concept as the power management method based on the power operation system in the foregoing embodiments, as shown in fig. 4, the present application provides a power management system based on the power operation system, and the embodiments of the system and the method in the embodiments of the present application are based on the same inventive concept. Wherein the system comprises:
the multi-stage computing resource deployment network establishment module 11 is used for acquiring multi-stage computing resource deployment information of the target computing power operating system and establishing a multi-stage computing resource deployment network, wherein the multi-stage computing resources comprise cloud computing resources, edge computing resources and terminal computing resources;
the preset processing task acquisition module 12 is used for acquiring a preset processing task of the target computing power operating system and an original database of a preset intelligent model, and the preset processing task has a user privacy grade identifier;
the node-task mapping result obtaining module 13 is configured to perform configuration of a multi-stage computing node and storage of an original database on the preset processing task based on the user privacy level identifier and the multi-stage computing resource deployment network, so as to obtain a node-task mapping result;
the task execution simulation module 14 is configured to establish a multi-stage computing node digital twin model based on the node-task mapping result, and perform task independent execution simulation and task parallel execution simulation on the multi-stage computing node, so as to obtain independent task processing precision, parallel task processing precision, independent task time delay and parallel task time delay;
the update node-task mapping result obtaining module 15 is configured to perform optimization reset of computing power computing resources on the multi-level computing node according to the independent task processing precision, the parallel task processing precision, the independent task time delay and the parallel task time delay to obtain an update node-task mapping result;
the computing management module 16 is configured to perform computing management according to the updated node-task mapping result by the computing management module 16.
Further, the node-task mapping result obtaining module 13 is further configured to perform the following steps:
the multi-stage computing nodes comprise cloud computing nodes, edge computing nodes and terminal computing nodes, and the privacy protection grade relation of the multi-stage computing nodes is that the terminal computing nodes are larger than the cloud computing nodes and larger than the edge computing nodes;
performing node-task matching mapping according to the privacy protection level relation and the user privacy level mark to obtain an initial mapping result;
and carrying out operation index balance analysis on the initial mapping result and optimizing the initial mapping result to obtain the node-task mapping result.
Further, the node-task mapping result obtaining module 13 is further configured to perform the following steps:
acquiring an operation capability index and a storage space of the multistage computing node;
acquiring a starting demand operation index and a starting demand data space of the preset intelligent model of the preset processing task;
and performing task allocation on the initial mapping result by using the operation capability index and the storage space, the starting requirement operation index and the starting requirement data space to obtain the node-task mapping result.
Further, the task execution simulation module 14 is further configured to perform the following steps:
performing association of cloud computing nodes, edge computing nodes and terminal computing nodes based on the node-task mapping result, acquiring a plurality of computing node chains, and acquiring a multi-stage task set of a first computing node chain;
randomly combining the tasks in the multi-stage task set to obtain M groups of parallel simulation tasks, wherein M is an integer greater than 0,n is the total number of tasks in the multi-stage task set, i is the number of combined tasks, and C is the sign of the combined number;
performing multiple simulation on any independent task in the multi-stage task set through the multi-stage computing node digital twin model, and recording the independent task processing precision and the independent task time delay;
and performing multiple-time task parallel simulation on the M groups of parallel simulation tasks through the multi-stage computing node digital twin model, and recording the parallel task processing precision, independent task time delay and parallel task time delay.
Further, the task execution simulation module 14 is further configured to perform the following steps:
obtaining a calibration execution result of any task in the multi-stage task set;
obtaining multiple simulation results and comparing the simulation results with the calibration execution results in accuracy to obtain accuracy indexes;
consistency comparison is carried out on the multiple simulation results, and consistency indexes are obtained;
and carrying out weighted average calculation on the accuracy index and the consistency index to obtain the task processing precision.
Further, the update node-task mapping result obtaining module 15 is further configured to perform the following steps:
extracting a first abnormal task if the independent task processing precision is smaller than the preset independent processing precision and/or the independent task time delay is larger than the preset independent processing time delay;
if the parallel task processing precision is smaller than the preset parallel processing precision and/or the parallel task time delay is larger than the preset parallel processing time delay, extracting a first abnormal parallel task combination;
performing task independent execution analysis on tasks in the first abnormal parallel task combination, and extracting a second abnormal task;
computing node optimization is carried out based on the user privacy grade identifiers of the first abnormal task and the second abnormal task, an abnormal optimizing node is obtained, and abnormal mapping association between the abnormal optimizing node and the first abnormal task and the second abnormal task is established;
updating the node-task mapping result with the abnormal mapping association.
Further, the update node-task mapping result obtaining module 15 is further configured to perform the following steps:
performing serialization processing of a calculation node chain based on the parallel task processing precision and the parallel task time delay to obtain an initial optimizing sequence;
based on the user privacy level identifiers of the first abnormal task and the second abnormal task, taking the balanced user privacy level as optimizing constraint to obtain optimizing space;
the optimizing space comprises original computing nodes of which the privacy protection level in the initial optimizing sequence is more than or equal to that of the first abnormal task and the second abnormal task;
and in the optimizing space, performing node parallel optimization of the first abnormal task and the second abnormal task through a multi-stage computing node digital twin model, and obtaining the abnormal optimizing node.
It should be noted that the sequence of the embodiments of the present application is merely for description, and does not represent the advantages and disadvantages of the embodiments. And the foregoing description has been directed to specific embodiments of this specification. In addition, the processes depicted in the accompanying figures do not necessarily require the particular order shown, or sequential order, to achieve desirable results. In some embodiments, multitasking and parallel processing are also possible or may be advantageous.
The foregoing description of the preferred embodiments of the present application is not intended to limit the invention to the particular embodiments of the present application, but to limit the scope of the invention to the particular embodiments of the present application.
The specification and drawings are merely exemplary of the application and are to be regarded as covering any and all modifications, variations, combinations, or equivalents that are within the scope of the application. It will be apparent to those skilled in the art that various modifications and variations can be made in the present application without departing from the scope of the application. Thus, if such modifications and variations of the present application fall within the scope of the present application and the equivalents thereof, the present application is intended to cover such modifications and variations.
Claims (8)
1. A method of power management based on a power operation system, comprising:
acquiring multi-level computing resource deployment information of a target computing power operating system, and establishing a multi-level computing resource deployment network, wherein the multi-level computing resources comprise cloud computing resources, edge computing resources and terminal computing resources;
acquiring a preset processing task of the target computing power operating system and an original database of a preset intelligent model, wherein the preset processing task has a user privacy grade identifier;
performing configuration of multi-stage computing nodes and storage of an original database on the preset processing task based on the user privacy grade identification and the multi-stage computing resource deployment network to obtain a node-task mapping result;
establishing a multi-stage computing node digital twin model based on the node-task mapping result, and performing task independent execution simulation and task parallel execution simulation on the multi-stage computing node to acquire independent task processing precision, parallel task processing precision, independent task time delay and parallel task time delay;
performing optimization reset of computing power computing resources on the multi-stage computing nodes according to the independent task processing precision, the parallel task processing precision, the independent task time delay and the parallel task time delay to obtain updated node-task mapping results;
and performing calculation power management according to the updated node-task mapping result.
2. The method of claim 1, wherein the obtaining the node-task mapping result comprises:
the multi-stage computing nodes comprise cloud computing nodes, edge computing nodes and terminal computing nodes, and the privacy protection grade relation of the multi-stage computing nodes is that the terminal computing nodes are larger than the cloud computing nodes and larger than the edge computing nodes;
performing node-task matching mapping according to the privacy protection level relation and the user privacy level mark to obtain an initial mapping result;
and carrying out operation index balance analysis on the initial mapping result and optimizing the initial mapping result to obtain the node-task mapping result.
3. The method of claim 2, wherein said performing the operation index balance analysis of the multi-stage computing node on the initial mapping result, performing the optimization of the initial mapping result, comprises:
acquiring an operation capability index and a storage space of the multistage computing node;
acquiring a starting demand operation index and a starting demand data space of the preset intelligent model of the preset processing task;
and performing task allocation on the initial mapping result by using the operation capability index and the storage space, the starting requirement operation index and the starting requirement data space to obtain the node-task mapping result.
4. The method of claim 1, wherein performing task independent execution simulation and task parallel execution simulation on the multi-level computing node comprises:
performing association of cloud computing nodes, edge computing nodes and terminal computing nodes based on the node-task mapping result, acquiring a plurality of computing node chains, and acquiring a multi-stage task set of a first computing node chain;
randomly combining tasks in the multi-stage task set to obtain M groups of parallel simulation tasks, wherein M is an integer greater than 0, and M=C n i N is the total number of tasks in the multi-stage task set, i is the number of combined tasks, and C is the sign of the combined number;
performing multiple simulation on any independent task in the multi-stage task set through the multi-stage computing node digital twin model, and recording the independent task processing precision and the independent task time delay;
and performing multiple-time task parallel simulation on the M groups of parallel simulation tasks through the multi-stage computing node digital twin model, and recording the parallel task processing precision, independent task time delay and parallel task time delay.
5. The method of claim 4, wherein the task processing accuracy is an index describing task processing accuracy and consistency, and wherein the method of obtaining the task processing accuracy comprises:
obtaining a calibration execution result of any task in the multi-stage task set;
obtaining multiple simulation results and comparing the simulation results with the calibration execution results in accuracy to obtain accuracy indexes;
consistency comparison is carried out on the multiple simulation results, and consistency indexes are obtained;
and carrying out weighted average calculation on the accuracy index and the consistency index to obtain the task processing precision.
6. The method of claim 1, wherein the obtaining updated node-task mapping results comprises:
extracting a first abnormal task if the independent task processing precision is smaller than the preset independent processing precision and/or the independent task time delay is larger than the preset independent processing time delay;
if the parallel task processing precision is smaller than the preset parallel processing precision and/or the parallel task time delay is larger than the preset parallel processing time delay, extracting a first abnormal parallel task combination;
performing task independent execution analysis on tasks in the first abnormal parallel task combination, and extracting a second abnormal task;
computing node optimization is carried out based on the user privacy grade identifiers of the first abnormal task and the second abnormal task, an abnormal optimizing node is obtained, and abnormal mapping association between the abnormal optimizing node and the first abnormal task and the second abnormal task is established;
updating the node-task mapping result with the abnormal mapping association.
7. The method of claim 6, wherein the computing node optimizing based on the user privacy level identification of the first abnormal task and the second abnormal task, obtaining an abnormal optimizing node, comprises:
performing serialization processing of a calculation node chain based on the parallel task processing precision and the parallel task time delay to obtain an initial optimizing sequence;
based on the user privacy level identifiers of the first abnormal task and the second abnormal task, taking the balanced user privacy level as optimizing constraint to obtain optimizing space;
the optimizing space comprises original computing nodes of which the privacy protection level in the initial optimizing sequence is more than or equal to that of the first abnormal task and the second abnormal task;
and in the optimizing space, performing node parallel optimization of the first abnormal task and the second abnormal task through a multi-stage computing node digital twin model, and obtaining the abnormal optimizing node.
8. A computing management system based on a computing operating system, the system comprising:
the system comprises a multistage computing resource deployment network establishment module, a terminal computing resource and a target computing power operation system, wherein the multistage computing resource deployment network establishment module is used for acquiring multistage computing resource deployment information of the target computing power operation system and establishing a multistage computing resource deployment network, and the multistage computing resources comprise cloud computing resources, edge computing resources and terminal computing resources;
the preset processing task acquisition module is used for acquiring a preset processing task of the target computing power operating system and an original database of a preset intelligent model, and the preset processing task has a user privacy grade identifier;
the node-task mapping result acquisition module is used for carrying out configuration of multi-stage computing nodes and storage of an original database on the preset processing task based on the user privacy grade identification and the multi-stage computing resource deployment network to obtain a node-task mapping result;
the task execution simulation module is used for establishing a multi-stage computing node digital twin model based on the node-task mapping result, performing task independent execution simulation and task parallel execution simulation on the multi-stage computing node, and acquiring independent task processing precision, parallel task processing precision, independent task time delay and parallel task time delay;
the updating node-task mapping result acquisition module is used for carrying out optimization reset on computing power computing resources of the multi-stage computing nodes according to the independent task processing precision, the parallel task processing precision, the independent task time delay and the parallel task time delay to acquire an updating node-task mapping result;
and the calculation management module is used for carrying out calculation management according to the updated node-task mapping result.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202311800788.XA CN117762632A (en) | 2023-12-26 | 2023-12-26 | Calculation management method based on calculation operation system |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202311800788.XA CN117762632A (en) | 2023-12-26 | 2023-12-26 | Calculation management method based on calculation operation system |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN117762632A true CN117762632A (en) | 2024-03-26 |
Family
ID=90321817
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202311800788.XA Pending CN117762632A (en) | 2023-12-26 | 2023-12-26 | Calculation management method based on calculation operation system |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN117762632A (en) |
-
2023
- 2023-12-26 CN CN202311800788.XA patent/CN117762632A/en active Pending
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Ozfatura et al. | Gradient coding with clustering and multi-message communication | |
| CN103618652A (en) | Audit and depth analysis system and audit and depth analysis method of business data | |
| CN111596924B (en) | Micro-service dividing method and device | |
| CN114841374A (en) | Method for optimizing transverse federated gradient spanning tree based on stochastic greedy algorithm | |
| CN115733754A (en) | Resource management system based on cloud native middle platform technology and elastic construction method thereof | |
| CN111324429B (en) | Micro-service combination scheduling method based on multi-generation ancestry reference distance | |
| CN100493001C (en) | Automatic clustering method for multi-particle size network under G bit flow rate | |
| CN115510249A (en) | Knowledge graph construction method and device, electronic equipment and storage medium | |
| CN108268655A (en) | A kind of government subsidy service push method and system based on big data | |
| CN117762632A (en) | Calculation management method based on calculation operation system | |
| CN115883392B (en) | Data perception method and device of computing power network, electronic equipment and storage medium | |
| CN107609194B (en) | Cloud computing-oriented storage method for time redundancy power load data | |
| CN116166440A (en) | FLINK stream batch integrated data processing method based on dynamic management | |
| CN113656187B (en) | Public security big data computing power service system based on 5G | |
| CN104507150A (en) | Method for clustering virtual resources in baseband pooling | |
| CN115544697A (en) | Broadband node position planning method, device, equipment and storage medium | |
| CN115150152A (en) | Method for rapidly reasoning actual authority of network user based on authority dependency graph reduction | |
| Sha et al. | Optimizing data placement of mapreduce on ceph-based framework under load-balancing constraint | |
| CN111813542A (en) | Load balancing method and device for parallel processing of large-scale graph analysis tasks | |
| CN114356712A (en) | Data processing method, device, equipment, readable storage medium and program product | |
| Cheng et al. | GSLAC: A general scalable and low-overhead alert correlation method | |
| CN115905167B (en) | Intelligent data storage method and system capable of rapidly migrating data | |
| CN106888199B (en) | Role-driven demand response secure access method in smart grid | |
| CN109284278A (en) | Calculating logic moving method and terminal device based on data analysis technique | |
| CN116680625B (en) | Cloud edge end cooperation-based distribution network multi-scene matching data processing method and system |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination |