CN117331706B - Calculation force optimization method and system in electric power data maintenance - Google Patents
Calculation force optimization method and system in electric power data maintenance Download PDFInfo
- Publication number
- CN117331706B CN117331706B CN202311631800.9A CN202311631800A CN117331706B CN 117331706 B CN117331706 B CN 117331706B CN 202311631800 A CN202311631800 A CN 202311631800A CN 117331706 B CN117331706 B CN 117331706B
- Authority
- CN
- China
- Prior art keywords
- power
- task
- calculation
- computing
- public
- 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.)
- Active
Links
- 238000004364 calculation method Methods 0.000 title claims abstract description 296
- 238000005457 optimization Methods 0.000 title claims abstract description 51
- 238000012423 maintenance Methods 0.000 title claims abstract description 44
- 238000000034 method Methods 0.000 title claims abstract description 43
- 238000012545 processing Methods 0.000 claims abstract description 61
- 238000012795 verification Methods 0.000 claims abstract description 27
- 238000005516 engineering process Methods 0.000 claims description 16
- 238000004458 analytical method Methods 0.000 claims description 12
- 238000012163 sequencing technique Methods 0.000 claims description 11
- 230000003321 amplification Effects 0.000 claims description 6
- 238000012937 correction Methods 0.000 claims description 6
- 238000003199 nucleic acid amplification method Methods 0.000 claims description 6
- 238000004088 simulation Methods 0.000 claims description 6
- 238000007619 statistical method Methods 0.000 claims description 4
- 238000009826 distribution Methods 0.000 abstract description 6
- 230000000694 effects Effects 0.000 abstract description 4
- 230000006872 improvement Effects 0.000 abstract description 3
- 230000008569 process Effects 0.000 description 4
- 238000012986 modification Methods 0.000 description 3
- 230000004048 modification Effects 0.000 description 3
- 238000000605 extraction 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
- 238000004422 calculation algorithm Methods 0.000 description 1
- 238000010276 construction Methods 0.000 description 1
- 238000013461 design Methods 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 238000003745 diagnosis Methods 0.000 description 1
- 238000010586 diagram Methods 0.000 description 1
- 238000004519 manufacturing process Methods 0.000 description 1
- 238000004904 shortening Methods 0.000 description 1
- 238000006467 substitution reaction Methods 0.000 description 1
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F9/00—Arrangements for program control, e.g. control units
- G06F9/06—Arrangements for program control, e.g. control units using stored programs, i.e. using an internal store of processing equipment to receive or retain programs
- G06F9/46—Multiprogramming arrangements
- G06F9/50—Allocation of resources, e.g. of the central processing unit [CPU]
- G06F9/5061—Partitioning or combining of resources
- G06F9/5072—Grid computing
-
- 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
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- Software Systems (AREA)
- Theoretical Computer Science (AREA)
- Mathematical Physics (AREA)
- General Engineering & Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Supply And Distribution Of Alternating Current (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The disclosure provides a calculation force optimization method and a calculation force optimization system in electric power data maintenance, and relates to the technical field of calculation force optimization, wherein the method comprises the following steps: acquiring task processing records; extracting private calculation force use records, public calculation force use records and task processing delay records of all task nodes; obtaining a plurality of sequences of computing power demands; obtaining a plurality of optimized private computing resources; obtaining an optimized public computing power resource; acquiring a plurality of real-time computing tasks, and performing calculation force constraint verification on the plurality of real-time computing tasks and a plurality of optimized private calculation force resources; and generating a public power scheduling instruction, performing power scheduling on the optimized public power resource, and performing execution of a plurality of real-time computing tasks. According to the power data maintenance method and device, the technical problem that the power data maintenance quality is poor due to the fact that the power distribution efficiency is low in the power data maintenance in the prior art can be solved, the power optimization improvement goal is achieved, and the technical effect of improving the power data maintenance quality is achieved.
Description
Technical Field
The disclosure relates to the technical field of calculation force optimization, in particular to a calculation force optimization method and a calculation force optimization system in electric power data maintenance.
Background
Along with the continuous development of productivity level, a fused digital technology is gradually generated, the digital level of the whole link and the whole production and operation process of the power grid is improved, and the traditional power grid is converted into digital and intelligent phenomena. However, the existing intelligent methods for power data maintenance such as state monitoring, fault monitoring and fault diagnosis all need to be calculated through calculation resources, and thus the calculation pressure of a certain node is too high. There is a need for a method to solve the above problems.
In summary, in the prior art, the technical problem of poor quality of power data maintenance is caused by low efficiency of computing power distribution in power data maintenance.
Disclosure of Invention
The disclosure provides a calculation force optimization method and a calculation force optimization system in power data maintenance, which are used for solving the technical problem that the quality of the power data maintenance is poor due to lower calculation force distribution efficiency in the power data maintenance in the prior art.
According to a first aspect of the present disclosure, there is provided a method of computing power optimization in power data maintenance, comprising: acquiring a plurality of task nodes of a target power maintenance system, and acquiring task processing records of the plurality of task nodes in a preset time period; extracting private calculation power use records, public calculation power use records and task processing delay records of all task nodes under a plurality of time nodes based on the task processing records; performing calculation force demand trend analysis on the plurality of task nodes based on the private calculation force use record, the public calculation force use record and the task processing delay record, and obtaining a plurality of calculation force demand sequences based on a digital twin technology; calculating a plurality of frequently-required computing forces of the plurality of task nodes according to the plurality of computing force demand sequences, and optimizing and adjusting private computing force resources of the plurality of task nodes to obtain a plurality of optimized private computing force resources; calculating and acquiring total calculation force demands of a plurality of task nodes under a plurality of time nodes according to the plurality of calculation force demand sequences, calculating and acquiring the maximum total calculation force demands, and optimizing public calculation force resources to acquire optimized public calculation force resources; acquiring a plurality of real-time computing tasks of the task nodes, and performing calculation force constraint verification on the real-time computing tasks and the optimized private calculation force resources; and generating a public power scheduling instruction based on the power constraint verification result, performing power scheduling on the optimized public power resource through the public power scheduling instruction, and executing a plurality of real-time computing tasks.
According to a second aspect of the present disclosure, there is provided a computing power optimization system in power data maintenance, comprising: the task processing record acquisition module is used for acquiring a plurality of task nodes of the target power maintenance system and acquiring task processing records of the plurality of task nodes in a preset time period; the private calculation force use record obtaining module is used for extracting private calculation force use records, public calculation force use records and task processing delay records of all task nodes under a plurality of time nodes based on the task processing records; the calculation force demand sequence obtaining module is used for carrying out calculation force demand trend analysis on the plurality of task nodes based on the private calculation force use record, the public calculation force use record and the task processing delay record and obtaining a plurality of calculation force demand sequences based on a digital twin technology; the optimizing private computing power resource obtaining module is used for calculating a plurality of frequently required computing power of the plurality of task nodes according to the plurality of computing power demand sequences, and optimizing and adjusting the private computing power resources of the plurality of task nodes to obtain a plurality of optimizing private computing power resources; the optimization public computing power resource obtaining module is used for calculating and obtaining total computing power demands of a plurality of task nodes under a plurality of time nodes according to the plurality of computing power demand sequences, calculating and obtaining maximum total computing power demands, and optimizing public computing power resources to obtain the optimization public computing power resources; the computing force constraint checking module is used for acquiring a plurality of real-time computing tasks of the task nodes and performing computing force constraint checking on the real-time computing tasks and the optimizing private computing force resources; and the real-time calculation task execution module is used for generating a public calculation power scheduling instruction based on the calculation power constraint verification result, and carrying out calculation power scheduling on the optimized public calculation power resource through the public calculation power scheduling instruction so as to execute a plurality of real-time calculation tasks.
One or more technical solutions provided in the present disclosure have at least the following technical effects or advantages: according to the method, a plurality of task nodes of a target power maintenance system are acquired, and task processing records of the plurality of task nodes in a preset time period are acquired; extracting private calculation power use records, public calculation power use records and task processing delay records of all task nodes under a plurality of time nodes based on the task processing records; performing calculation force demand trend analysis on the plurality of task nodes based on the private calculation force use record, the public calculation force use record and the task processing delay record, and obtaining a plurality of calculation force demand sequences based on a digital twin technology; calculating a plurality of frequently-required computing forces of the plurality of task nodes according to the plurality of computing force demand sequences, and optimizing and adjusting private computing force resources of the plurality of task nodes to obtain a plurality of optimized private computing force resources; calculating and acquiring total calculation force demands of a plurality of task nodes under a plurality of time nodes according to the plurality of calculation force demand sequences, calculating and acquiring the maximum total calculation force demands, and optimizing public calculation force resources to acquire optimized public calculation force resources; acquiring a plurality of real-time computing tasks of the task nodes, and performing calculation force constraint verification on the real-time computing tasks and the optimized private calculation force resources; a public power dispatching instruction is generated based on a power constraint verification result, power dispatching is carried out on the public power resource optimized through the public power dispatching instruction, and execution of a plurality of real-time computing tasks is carried out, so that the technical problem that the power data maintenance quality is poor due to low power distribution efficiency in the power data maintenance in the prior art is solved, the power optimization improvement target is realized, and the technical effect of improving the power data maintenance quality is achieved.
It should be understood that the description of this section is not intended to identify key or critical features of the embodiments of the disclosure, nor is it intended to be used to limit the scope of the disclosure. Other features of the present disclosure will become apparent from the following specification.
Drawings
For a clearer description of the present disclosure or of the prior art, the drawings used in the description of the embodiments or of the prior art will be briefly described, it being obvious that the drawings in the description below are only exemplary and that other drawings may be obtained, without inventive effort, by a person skilled in the art, from the provided drawings.
Fig. 1 is a schematic flow chart of a computing power optimization method in power data maintenance according to an embodiment of the disclosure;
FIG. 2 is a flow chart of obtaining a plurality of calculation force demand sequences in a calculation force optimization method in power data maintenance according to an embodiment of the disclosure;
fig. 3 is a schematic structural diagram of a computing power optimization system in power data maintenance according to an embodiment of the present disclosure.
Reference numerals illustrate: the system comprises a task processing record obtaining module 11, a private calculation force use record obtaining module 12, a calculation force demand sequence obtaining module 13, an optimizing private calculation force resource obtaining module 14, an optimizing public calculation force resource obtaining module 15, a calculation force constraint checking module 16 and a real-time calculation task executing module 17.
Detailed Description
Exemplary embodiments of the present disclosure are described below in conjunction with the accompanying drawings, which include various details of the embodiments of the present disclosure to facilitate understanding, and should be considered as merely exemplary. Accordingly, one of ordinary skill in the art will recognize that various changes and modifications of the embodiments described herein can be made without departing from the scope and spirit of the present disclosure. Also, descriptions of well-known functions and constructions are omitted in the following description for clarity and conciseness.
Example 1
An embodiment of the present disclosure provides a method for optimizing power in power data maintenance, and is described with reference to fig. 1, where the method includes:
the method provided by the embodiment of the disclosure comprises the following steps:
acquiring a plurality of task nodes of a target power maintenance system, and acquiring task processing records of the plurality of task nodes in a preset time period;
specifically, the target power maintenance system is a system to be subjected to power data calculation optimization. Further, a plurality of task nodes of the target power maintenance system are obtained. The task nodes are nodes for performing power task processing through calculation. Further, task processing records of a plurality of task nodes in a preset time period are acquired and acquired. The preset time period is historical processing task time. For example, the preset time period is a history of one month or one week or the like.
Extracting private calculation power use records, public calculation power use records and task processing delay records of all task nodes under a plurality of time nodes based on the task processing records;
specifically, a private calculation power use record, a public calculation power use record and a task processing delay record of each task node under a plurality of time nodes are extracted from the task processing records. The private computing power is the computing power of a certain task node and can only be used by the current task node. The common computing power is the computing power that can be used by any of the task nodes. The task processing delay is the time for compensating the private calculation power or calling the public calculation power to process the task when the private calculation power or the public calculation power of a certain task node is insufficient. Further, the plurality of time nodes are historical plurality of task processing time records, and the obtained private calculation force use records, public calculation force use records and task processing delay records are historical use or calling records.
Performing calculation force demand trend analysis on the plurality of task nodes based on the private calculation force use record, the public calculation force use record and the task processing delay record, and obtaining a plurality of calculation force demand sequences based on a digital twin technology;
Specifically, a private calculation force demand sequence of each task node in the plurality of task nodes under the plurality of time nodes is obtained from a private calculation force use record, and a public calculation force borrowing sequence is obtained according to the public calculation force use record. And superposing the private calculation force demand sequence and the public calculation force borrowing sequence to obtain a calculation force use sequence. Further, a time-delay task sequence is obtained according to the task processing time-delay record. And (3) carrying out simulation on the time-delay task sequence by a digital twin technology to obtain a calculation force demand sequence which is needed besides the calculation force use sequence when the task is normally executed and is used as a calculation force demand sequence to be expanded. And carrying out calculation force superposition on the calculation force demand sequence to be expanded and the calculation force use sequence to obtain a plurality of calculation force demand sequences.
Calculating a plurality of frequently-required computing forces of the plurality of task nodes according to the plurality of computing force demand sequences, and optimizing and adjusting private computing force resources of the plurality of task nodes to obtain a plurality of optimized private computing force resources;
specifically, the plurality of calculation force demand sequences corresponding to the plurality of task nodes are ordered according to the occurrence frequency from more to less, and the calculation force demand with the highest occurrence frequency in the plurality of calculation force demand sequences is obtained and used as a plurality of frequent demand calculation forces. Comparing the frequently required computing power with the private computing power resources of the corresponding task nodes, and performing difference value calculation on the frequently required computing power and the private computing power resources to obtain a plurality of compensating computing power resources of the task nodes. And adding and calculating the private computing power resources and the compensation computing power resources corresponding to the task nodes to obtain a plurality of optimized private computing power resources.
Calculating and acquiring total calculation force demands of a plurality of task nodes under a plurality of time nodes according to the plurality of calculation force demand sequences, calculating and acquiring the maximum total calculation force demands, and optimizing public calculation force resources to acquire optimized public calculation force resources;
specifically, according to the multiple calculation force demand sequences, adding calculation is performed on the total calculation force demand corresponding to each time node of each task node, and the maximum total calculation force demand is obtained through statistics. The computational power resources required by different task nodes of the same time node may be different, for example, at a first time node, the first task node needs more computational power, at a second time node, the third task node needs more computational power, the first task node needs less computational power, the optimization of the public computational power resources is performed, and the computational power distribution is performed on the task nodes with more computational power requirements preferentially at the current time node.
Acquiring a plurality of real-time computing tasks of the task nodes, and performing calculation force constraint verification on the real-time computing tasks and the optimized private calculation force resources;
specifically, the optimized private computing power resource is used as the self-used computing power resource of the corresponding task node. And acquiring a plurality of real-time computing tasks of a plurality of task nodes, and performing calculation force constraint verification on the plurality of real-time computing tasks and a plurality of optimized private calculation force resources to obtain constraint verification results. When the self-used computing power resources of each task node meet the size of the real-time task, the current task node can process, and the verification is passed without the scheduling support of the public computing power resources. When the verification fails, scheduling support of the common computational resources is required.
And generating a public power scheduling instruction based on the power constraint verification result, performing power scheduling on the optimized public power resource through the public power scheduling instruction, and executing a plurality of real-time computing tasks.
Specifically, when the self-used computing power resources of each task node do not meet the size of the real-time task, the scheduling support of the public computing power resources is needed, and the computing power constraint verification result does not pass. Thereby generating a public power dispatching instruction. And extracting the public computing resources meeting the scheduling computing power demand identification, namely extracting the quantity of public computing power needed by the task nodes from the public computing power resources distributed in a preset distance range of the scheduling task nodes according to the extraction quantity of the destination and the delivery destination. And executing the real-time computing task of the scheduling task node by the scheduled public computing power resource and the private computing power resource of the scheduling task node.
The technical problem that in the prior art, the power data maintenance quality is poor due to low power distribution efficiency in the power data maintenance can be solved, the goal of improving power optimization is achieved, and the technical effect of improving the power data maintenance quality is achieved.
The method provided by the embodiment of the disclosure further comprises the following steps:
carrying out statistical analysis on the private calculation force use records to obtain a first private calculation force demand sequence of a first task node in the plurality of task nodes;
carrying out public calculation power use analysis on the first task node based on the public calculation power use record to obtain a first public calculation power borrowing sequence;
superposing the first private calculation force demand sequence and the first public calculation force borrowing sequence to obtain a first node calculation force use sequence;
based on a digital twin technology, compensating and correcting the calculation force demand of the first node calculation force using sequence according to the task processing delay record to obtain a first calculation force demand sequence;
and constructing a plurality of calculation force demand sequences according to the first calculation force demand sequence.
As shown in fig. 2, specifically, the extracted private calculation force usage record is subjected to statistical analysis, and the statistical number of the private calculation force usage record of each of the plurality of task nodes is obtained. Further, one task node of the plurality of task nodes is randomly acquired and used as a first task node, and a first private calculation power demand sequence of the first task node is acquired. The first private power demand sequence is a sequence set obtained by combining a plurality of private power demands of the first task node under a plurality of time nodes.
Further, the public power use records under a plurality of time nodes matched with the first task node are extracted according to the public power use records, and the public power borrowing sequences are obtained through combination.
Further, the plurality of private power demands in the first private power demand sequence and the plurality of public power borrowing in the first public power borrowing sequence correspond to each other in terms of time nodes under the condition of the same task node. And adding the private calculation force demands and the public calculation force borrowing corresponding to the private calculation force demands to obtain a plurality of use calculation forces, and further obtaining a first node calculation force use sequence under a plurality of time nodes.
Further, a first time delay task sequence under a plurality of time nodes corresponding to the first task node is extracted according to the task processing time delay record. And carrying out simulation on the first delay task sequence through a digital twin technology to obtain a first calculation force demand sequence to be expanded, wherein the calculation force demand sequence is required by the current task node and corresponds to the first delay record except for the calculation force use sequence of the first node when the task is normally executed. And performing superposition compensation of the calculation force demand on the first node calculation force using sequence by using the first calculation force demand sequence to be expanded to obtain a first calculation force demand sequence.
Further, a plurality of power demand sequences of the plurality of task nodes are acquired in a manner that the first power demand sequence is acquired.
And carrying out calculation force demand trend analysis on a plurality of task nodes based on the private calculation force use record, the public calculation force use record and the task processing delay record, and obtaining a plurality of calculation force demand sequences based on a digital twin technology, thereby improving the accuracy degree of obtaining the plurality of calculation force demand sequences.
The method provided by the embodiment of the disclosure further comprises the following steps:
extracting a first delay task sequence of the first task node according to the task processing delay record;
acquiring a preset cloud calculator, and constructing a digital twin calculator based on the preset cloud calculator;
based on the digital twin calculator, performing simulation calculation on the first time delay task sequence to obtain a task normal execution demand calculation force sequence as a first to-be-calculated force demand sequence;
and carrying out compensation correction of the calculation force demand on the first node calculation force using sequence by using the first calculation force demand sequence to be expanded to obtain the first calculation force demand sequence.
Specifically, a first time delay task sequence under a plurality of time nodes corresponding to a first task node is extracted according to the task processing time delay record.
Further, the predetermined cloud calculator is a calculator for performing task processing on the target power system. And extracting a preset cloud calculator, and constructing a digital twin calculator for the preset cloud calculator through a digital twin technology for simulating a processing system of the cloud calculator.
Further, the digital twin calculator is used for carrying out simulation calculation and emulation on the first time delay task sequence, so that when the task is normally executed, a public calculation force or a private calculation force sequence which is required by the current task node is removed from the first node calculation force use sequence and is used as a first calculation force demand sequence to be expanded.
Further, the first node calculation force using sequence is overlapped with the first calculation force to-be-expanded calculation force demand sequence to complete compensation correction, and the first calculation force demand sequence is obtained. The first calculation force demand sequence is a calculation force sequence which can normally execute tasks by the first task node.
And correcting and acquiring the first calculation force demand sequence through the first time delay task sequence so as to improve the accuracy of acquiring the first calculation force demand sequence.
The method provided by the embodiment of the disclosure further comprises the following steps:
extracting the computing power demand with highest occurrence frequency in the computing power demand sequences as the frequently required computing power;
Comparing the frequent demand computing power with the private computing power resources of the task nodes to obtain a plurality of compensation computing power resources;
and optimizing and adjusting the private computing power resources of the task nodes by using the compensation computing power resources to obtain the optimized private computing power resources.
Specifically, the plurality of calculation force demand sequences corresponding to the plurality of task nodes are ordered according to the occurrence frequency from more to less, and the calculation force demand with the highest occurrence frequency in the plurality of calculation force demand sequences is obtained and used as a plurality of frequent demand calculation forces. Wherein the frequent demand computing force may be a plurality of computing force demand sequences of a plurality of task nodes. For example, the frequency of occurrence of the calculation force demand sequence of the first task node and the calculation force demand sequence of the second task node is highest and is 5 times, and the frequent demand calculation force is 2.
Further, comparing each frequently-required computing power with the private computing power resources of the corresponding task nodes, and performing difference calculation on the frequently-required computing power and the private computing power resources to obtain the compensation computing power resources of each task node, so as to obtain a plurality of compensation computing power resources of a plurality of task nodes.
Further, the plurality of compensation computing power resources are optimized and adjusted to the corresponding private computing power resources of each task node, namely, each private computing power resource and the corresponding compensation computing power resource are added and calculated, and a plurality of optimized private computing power resources are obtained.
And calculating a plurality of frequently-required computing forces of a plurality of task nodes according to the computing force demand sequences, and optimizing and adjusting private computing force resources of the plurality of task nodes to obtain a plurality of optimized private computing force resources so as to improve computing force optimization accuracy.
The method provided by the embodiment of the disclosure further comprises the following steps:
acquiring peak task nodes with maximum calculation force demands under a plurality of time nodes, and counting occurrence frequencies of all task nodes in the acquired peak task nodes;
performing calculation power tolerance amplification according to the maximum total calculation power requirement to obtain an optimization target;
and according to the sequence of the occurrence frequencies from large to small, different numbers of public computing power resources are distributed in a preset distance range of each task node according to the optimization target, and the distributed public computing power resources are obtained and used as the optimized public computing power resources.
Specifically, the common calculation power demands in the plurality of time nodes are processed in sequence according to the sequence from more to less, the time node with the largest common calculation power demand of the plurality of task nodes in the plurality of time nodes is obtained and used as the peak task node, and the occurrence frequency of each task node in the peak task node is obtained in a statistics mode. For example, in the first time node and the second time node, the total computational power demands of all task nodes are ordered, such that the total computational power demands of the first time node are greater than the total computational power demands of the second time node.
Further, the tolerance amplification of public calculation force is carried out according to the maximum total calculation force demand, the maximum total calculation force demand is subtracted by a plurality of optimized private calculation force resources, and tolerance processing is carried out on the obtained difference value, so that an optimization target is obtained. For example, the tolerance treatment is to increase the obtained difference by 20% to obtain 120% of the fair optimization goal.
Further, the task nodes are extracted according to the order of the occurrence frequencies from large to small. And sequentially distributing different amounts of public computing resources in a preset distance range of each task node according to the sequence from high frequency to low frequency of each occurrence according to the public computing optimization target, and obtaining distributed public computing resources as optimized public computing resources. The preset distance range is set by a person skilled in the art in a self-defining way according to actual conditions. For example, the preset distance range is within the distance between each task node and the adjacent task node.
And optimizing the public computing power resources to obtain the optimized public computing power resources so as to improve the efficiency and accuracy of obtaining computing power optimization.
The method provided by the embodiment of the disclosure further comprises the following steps:
if the calculation force constraint verification result is that verification is not passed, generating a public calculation force dispatching instruction, wherein the public calculation force dispatching instruction is provided with a dispatching calculation force demand identifier and a dispatching task node identifier;
And extracting public computing power resources meeting the scheduling computing power demand identification from public computing power resources distributed in a preset distance range of the scheduling task node based on the scheduling computing power demand identification and the scheduling task node identification, and executing real-time computing tasks of the scheduling task node together with private computing power resources of the scheduling task node.
Specifically, when the self-used computing power resources of each task node do not meet the size of the real-time task, the current task node cannot process, scheduling support of the public computing power resources is needed, and the computing power constraint verification result is that verification is not passed. Thereby generating a public power dispatching instruction. The public power scheduling instruction is provided with a scheduling power demand identifier and a scheduling task node identifier. The scheduling power calculation requirement is identified as the power calculation scheduling quantity, and the scheduling task node is identified as the task node. For example, the dispatch task node is identified as the destination and the dispatch algorithm force demand is identified as the pick-up number to the destination.
Further, based on the scheduled computing power demand identifier and the scheduled task node identifier, namely according to the extraction quantity of the destination and the delivery destination, extracting the public computing power resources meeting the scheduled computing power demand identifier, namely the quantity of the public computing power required by the extracted task node, from the public computing power resources distributed in the preset distance range of the scheduled task node. And executing the real-time computing task of the scheduling task node by the scheduled public computing power resource and the private computing power resource of the scheduling task node.
And generating a public power scheduling instruction based on the power constraint verification result, performing power scheduling on the public power resource optimized through the public power scheduling instruction, and performing execution of a plurality of real-time computing tasks so as to improve task processing efficiency.
The method provided by the embodiment of the disclosure further comprises the following steps:
if the total amount of public computing power resources distributed in the preset distance range of the dispatching task node does not meet the dispatching computing power requirement identification;
the method comprises the steps of obtaining public computing power resource scheduling information of other task nodes except the scheduling task node;
calculating the distance between other task nodes and the scheduling task node, and sequencing the other task nodes according to the sequence from small distance to large distance to obtain a sequencing result;
and scheduling the public computing resources of other task nodes by combining the sequencing result and the public computing resource scheduling information of other task nodes with the aim that the scheduling resources meet the scheduling computing requirement identification.
Specifically, if the total amount of the public computing power resources distributed in the preset distance range of the dispatching task node does not meet the dispatching computing power demand identification, the fact that the public computing power resources cannot be dispatched is indicated.
Further, the public computing resource scheduling information of a plurality of other task nodes except the scheduling task node is obtained, namely, the public computing resource in a far range is obtained in a similar way of shortening and finding.
Further, distances between a plurality of other task nodes and the scheduling task nodes which do not meet the scheduling calculation force requirement identification are calculated, and the plurality of other task nodes are ordered according to the order of the distances from small to large, so that an ordering result is obtained. For example, the distance acquisition may be an acquisition of euclidean distance.
Further, the first other task nodes are sequentially extracted and sequenced according to the sequencing result, the public computing power resource scheduling information of the other task nodes is scheduled to the scheduling task nodes until the scheduling resources of the scheduling task nodes meet the scheduling computing power requirement identification.
And scheduling the public computing power resources of other task nodes in sequence with the aim of scheduling resource meeting the scheduling computing power demand mark, so that the task processing efficiency is improved.
Example two
Based on the same inventive concept as the calculation force optimization method in the maintenance of electric power data in the foregoing embodiment, the disclosure will be described with reference to fig. 3, and the disclosure further provides a calculation force optimization system in the maintenance of electric power data, where the system includes:
the task processing record obtaining module 11 is used for obtaining a plurality of task nodes of the target power maintenance system and collecting and obtaining task processing records of the plurality of task nodes in a preset time period;
A private calculation force use record obtaining module 12, where the private calculation force use record obtaining module 12 is configured to extract a private calculation force use record, a public calculation force use record, and a task processing delay record of each task node under a plurality of time nodes based on the task processing record;
the calculation force demand sequence obtaining module 13, wherein the calculation force demand sequence obtaining module 13 is configured to perform calculation force demand trend analysis on the plurality of task nodes based on the private calculation force usage record, the public calculation force usage record and the task processing delay record, and obtain a plurality of calculation force demand sequences based on a digital twin technology;
an optimized private computing power resource obtaining module 14, where the optimized private computing power resource obtaining module 14 is configured to calculate a plurality of frequently-required computing powers of the plurality of task nodes according to the plurality of computing power demand sequences, and perform optimization adjustment on private computing power resources of the plurality of task nodes to obtain a plurality of optimized private computing power resources;
the optimizing public power resource obtaining module 15 is configured to calculate and obtain total power demands of a plurality of task nodes under a plurality of time nodes according to the plurality of power demand sequences, calculate and obtain a maximum total power demand, and optimize public power resources to obtain optimizing public power resources;
The computing force constraint checking module 16, wherein the computing force constraint checking module 16 is used for acquiring a plurality of real-time computing tasks of the task nodes and performing computing force constraint checking on the plurality of real-time computing tasks and the plurality of optimized private computing force resources;
the real-time computing task execution module 17 is used for generating a public power scheduling instruction based on the power constraint verification result, and performing power scheduling on the optimized public power resource through the public power scheduling instruction to execute a plurality of real-time computing tasks.
Further, the system further comprises:
the first private calculation force demand sequence obtaining module is used for carrying out statistical analysis on the private calculation force use records to obtain a first private calculation force demand sequence of a first task node in the plurality of task nodes;
the first public power borrowing and adjusting sequence obtaining module is used for carrying out public power use analysis on the first task node based on the public power use record to obtain a first public power borrowing and adjusting sequence;
the first node calculation force use sequence obtaining module is used for superposing the first private calculation force demand sequence and the first public calculation force borrowing sequence to obtain a first node calculation force use sequence;
The first calculation force demand sequence obtaining module is used for carrying out compensation correction on calculation force demands on the first node calculation force using sequence according to the task processing delay record based on a digital twin technology to obtain a first calculation force demand sequence;
a plurality of calculation force demand sequence obtaining modules for constructing the plurality of calculation force demand sequences according to the first calculation force demand sequence.
Further, the system further comprises:
the first delay task sequence acquisition module is used for extracting a first delay task sequence of the first task node according to the task processing delay record;
the digital twin calculator building module is used for acquiring a preset cloud calculator and building the digital twin calculator based on the preset cloud calculator;
the first to-be-amplified force demand sequence obtaining module is used for carrying out simulation calculation on the first time delay task sequence based on the digital twin calculator to obtain a task normal execution demand force sequence as a first to-be-amplified force demand sequence;
And the first calculation force demand sequence obtaining module is used for carrying out compensation correction on the calculation force demand of the first node calculation force using sequence by using the first calculation force demand sequence to be expanded to obtain the first calculation force demand sequence.
Further, the system further comprises:
a plurality of frequent need computing force obtaining modules for extracting computing force demands, which occur most frequently in the plurality of computing force demand sequences, as the plurality of frequent need computing forces;
the compensation computing power resource obtaining modules are used for comparing the frequently-required computing power with the private computing power resources of the task nodes to obtain a plurality of compensation computing power resources;
the plurality of optimized private computing power resource obtaining modules are used for optimizing and adjusting the private computing power resources of the plurality of task nodes by the plurality of compensating computing power resources to obtain the plurality of optimized private computing power resources.
Further, the system further comprises:
the system comprises an occurrence frequency acquisition module, a calculation power calculation module and a calculation power calculation module, wherein the occurrence frequency acquisition module is used for acquiring peak task nodes with the largest calculation power demands under a plurality of time nodes and counting the occurrence frequency of each task node in the acquired peak task nodes;
The optimization target obtaining module is used for carrying out calculation power tolerance amplification according to the maximum total calculation power requirement to obtain an optimization target;
the optimization public computing resource obtaining module is used for distributing different numbers of public computing resources in a preset distance range of each task node according to the optimization target according to the sequence from big to small of each occurrence frequency, and obtaining distributed public computing resources as the optimization public computing resources.
Further, the system further comprises:
the public power dispatching instruction obtaining module is used for generating a public power dispatching instruction if the power constraint verification result is that verification is not passed, wherein the public power dispatching instruction is provided with a dispatching power demand identifier and a dispatching task node identifier;
the real-time computing task executing module is used for extracting public computing power resources meeting the dispatching computing power demand identification from public computing power resources distributed in a preset distance range of the dispatching task node based on the dispatching computing power demand identification and the dispatching task node identification, and executing the real-time computing task of the dispatching task node together with the private computing power resources of the dispatching task node.
Further, the system further comprises:
the scheduling computing power demand identification judgment module is used for judging whether the total amount of public computing power resources distributed in a preset distance range of the scheduling task node does not meet the scheduling computing power demand identification or not;
the public power resource scheduling information acquisition module is used for acquiring public power resource scheduling information of other task nodes except the scheduling task node;
the sequencing result obtaining module is used for calculating the distance between other task nodes and the scheduling task node, sequencing the other task nodes according to the sequence from the small distance to the large distance, and obtaining a sequencing result;
and the public power resource scheduling module is used for scheduling the public power resources of other task nodes by combining the sequencing result and public power resource scheduling information of other task nodes and taking the scheduling resource meeting the scheduling power demand identification as a target.
The specific example of the calculation optimization method in the maintenance of electric power data in the first embodiment is also applicable to the calculation optimization system in the maintenance of electric power data in the present embodiment, and the detailed description of the calculation optimization method in the maintenance of electric power data in the present embodiment is clearly known to those skilled in the art, so that the details of the calculation optimization system in the maintenance of electric power data in the present embodiment are not described herein for brevity. The device disclosed in the embodiment corresponds to the method disclosed in the embodiment, so that the description is simpler, and the relevant points refer to the description of the method.
It should be appreciated that various forms of the flows shown above may be used to reorder, add, or delete steps. For example, the steps recited in the present disclosure may be performed in parallel or sequentially or in a different order, provided that the desired results of the technical solutions of the present disclosure are achieved, and are not limited herein.
The above detailed description should not be taken as limiting the scope of the present disclosure. It will be apparent to those skilled in the art that various modifications, combinations, and alternatives are possible, depending on design requirements and other factors. Any modifications, equivalent substitutions and improvements made within the spirit and principles of the present disclosure are intended to be included within the scope of the present disclosure.
Claims (6)
1. A method of computing power optimization in power data maintenance, the method comprising:
acquiring a plurality of task nodes of a target power maintenance system, and acquiring task processing records of the plurality of task nodes in a preset time period;
extracting private calculation power use records, public calculation power use records and task processing delay records of all task nodes under a plurality of time nodes based on the task processing records;
Performing calculation force demand trend analysis on the plurality of task nodes based on the private calculation force use record, the public calculation force use record and the task processing delay record, and obtaining a plurality of calculation force demand sequences based on a digital twin technology;
calculating a plurality of frequently-required computing forces of the plurality of task nodes according to the plurality of computing force demand sequences, and optimizing and adjusting private computing force resources of the plurality of task nodes to obtain a plurality of optimized private computing force resources;
calculating and acquiring total calculation force demands of a plurality of task nodes under a plurality of time nodes according to the plurality of calculation force demand sequences, calculating and acquiring the maximum total calculation force demands, and optimizing public calculation force resources to acquire optimized public calculation force resources;
acquiring a plurality of real-time computing tasks of the task nodes, and performing calculation force constraint verification on the real-time computing tasks and the optimized private calculation force resources;
generating a public power scheduling instruction based on a power constraint verification result, performing power scheduling on the optimized public power resource through the public power scheduling instruction, and executing a plurality of real-time computing tasks;
the calculating force demand trend analysis is performed on the plurality of task nodes based on the private calculating force use record, the public calculating force use record and the task processing delay record, and a plurality of calculating force demand sequences are obtained based on a digital twin technology, and the calculating force demand sequence comprises the following steps:
Carrying out statistical analysis on the private calculation force use records to obtain a first private calculation force demand sequence of a first task node in the plurality of task nodes;
carrying out public calculation power use analysis on the first task node based on the public calculation power use record to obtain a first public calculation power borrowing sequence;
superposing the first private calculation force demand sequence and the first public calculation force borrowing sequence to obtain a first node calculation force use sequence;
based on a digital twin technology, compensating and correcting the calculation force demand of the first node calculation force using sequence according to the task processing delay record to obtain a first calculation force demand sequence;
constructing the plurality of computing force demand sequences according to the first computing force demand sequence;
generating a public power scheduling instruction based on the power constraint verification result, performing power scheduling on the optimized public power resource through the public power scheduling instruction, and performing execution of a plurality of real-time computing tasks, wherein the method comprises the following steps:
if the calculation force constraint verification result is that verification is not passed, generating a public calculation force dispatching instruction, wherein the public calculation force dispatching instruction is provided with a dispatching calculation force demand identifier and a dispatching task node identifier;
And extracting public computing power resources meeting the scheduling computing power demand identification from public computing power resources distributed in a preset distance range of the scheduling task node based on the scheduling computing power demand identification and the scheduling task node identification, and executing real-time computing tasks of the scheduling task node together with private computing power resources of the scheduling task node.
2. The method of claim 1, wherein the performing compensation correction of the computing power demand for the first node computing power usage sequence based on the task processing delay record based on a digital twin technique to obtain a first computing power demand sequence comprises:
extracting a first delay task sequence of the first task node according to the task processing delay record;
acquiring a preset cloud calculator, and constructing a digital twin calculator based on the preset cloud calculator;
based on the digital twin calculator, performing simulation calculation on the first time delay task sequence to obtain a task normal execution demand calculation force sequence as a first to-be-calculated force demand sequence;
and carrying out compensation correction of the calculation force demand on the first node calculation force using sequence by using the first calculation force demand sequence to be expanded to obtain the first calculation force demand sequence.
3. The method of claim 1, wherein the computing the plurality of frequently demanded computing forces for the plurality of task nodes according to the plurality of computing force demand sequences and performing optimization adjustment on the private computing force resources for the plurality of task nodes to obtain a plurality of optimized private computing force resources comprises:
extracting the computing power demand with highest occurrence frequency in the computing power demand sequences as the frequently required computing power;
comparing the frequent demand computing power with the private computing power resources of the task nodes to obtain a plurality of compensation computing power resources;
and optimizing and adjusting the private computing power resources of the task nodes by using the compensation computing power resources to obtain the optimized private computing power resources.
4. The method of claim 1, wherein calculating according to the plurality of calculation force demand sequences to obtain total calculation force demands of a plurality of task nodes under a plurality of time nodes, calculating to obtain maximum total calculation force demands, and optimizing a public calculation force resource to obtain an optimized public calculation force resource, comprises:
acquiring peak task nodes with maximum calculation force demands under a plurality of time nodes, and counting occurrence frequencies of all task nodes in the acquired peak task nodes;
Performing calculation power forgiving amplification according to the maximum total calculation power demand to obtain an optimization target, wherein the calculation power forgiving amplification refers to the forgiving amplification of public calculation power according to the maximum total calculation power demand, subtracting a plurality of optimized private calculation power resources from the maximum total calculation power demand, and performing forgiving processing on the obtained difference value to obtain the optimization target, wherein the forgiving processing is to increase the obtained difference value by 20% and obtain an optimization target of 120% of public calculation power;
and according to the sequence of the occurrence frequencies from large to small, distributing different numbers of public computing resources in a preset distance range of each task node according to the optimization target, and obtaining distributed public computing resources as the optimized public computing resources.
5. The method of claim 1, wherein the method further comprises:
if the total amount of public computing power resources distributed in the preset distance range of the dispatching task node does not meet the dispatching computing power requirement identification;
the method comprises the steps of obtaining public computing power resource scheduling information of other task nodes except the scheduling task node;
calculating the distance between other task nodes and the scheduling task node, and sequencing the other task nodes according to the sequence from small distance to large distance to obtain a sequencing result;
And scheduling the public computing resources of other task nodes by combining the sequencing result and the public computing resource scheduling information of other task nodes with the aim that the scheduling resources meet the scheduling computing requirement identification.
6. A power optimization system in power data maintenance for implementing a power optimization method in power data maintenance according to any one of claims 1-5, the system comprising:
the task processing record acquisition module is used for acquiring a plurality of task nodes of the target power maintenance system and acquiring task processing records of the plurality of task nodes in a preset time period;
the private calculation force use record obtaining module is used for extracting private calculation force use records, public calculation force use records and task processing delay records of all task nodes under a plurality of time nodes based on the task processing records;
the calculation force demand sequence obtaining module is used for carrying out calculation force demand trend analysis on the plurality of task nodes based on the private calculation force use record, the public calculation force use record and the task processing delay record and obtaining a plurality of calculation force demand sequences based on a digital twin technology;
The optimizing private computing power resource obtaining module is used for calculating a plurality of frequently required computing power of the plurality of task nodes according to the plurality of computing power demand sequences, and optimizing and adjusting the private computing power resources of the plurality of task nodes to obtain a plurality of optimizing private computing power resources;
the optimization public computing power resource obtaining module is used for calculating and obtaining total computing power demands of a plurality of task nodes under a plurality of time nodes according to the plurality of computing power demand sequences, calculating and obtaining maximum total computing power demands, and optimizing public computing power resources to obtain the optimization public computing power resources;
the computing force constraint checking module is used for acquiring a plurality of real-time computing tasks of the task nodes and performing computing force constraint checking on the real-time computing tasks and the optimizing private computing force resources;
and the real-time calculation task execution module is used for generating a public calculation power scheduling instruction based on the calculation power constraint verification result, and carrying out calculation power scheduling on the optimized public calculation power resource through the public calculation power scheduling instruction so as to execute a plurality of real-time calculation tasks.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202311631800.9A CN117331706B (en) | 2023-12-01 | 2023-12-01 | Calculation force optimization method and system in electric power data maintenance |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202311631800.9A CN117331706B (en) | 2023-12-01 | 2023-12-01 | Calculation force optimization method and system in electric power data maintenance |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN117331706A CN117331706A (en) | 2024-01-02 |
| CN117331706B true CN117331706B (en) | 2024-02-13 |
Family
ID=89283428
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202311631800.9A Active CN117331706B (en) | 2023-12-01 | 2023-12-01 | Calculation force optimization method and system in electric power data maintenance |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN117331706B (en) |
Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104102544A (en) * | 2014-06-30 | 2014-10-15 | 武汉理工大学 | Multi QoS (quality of service)-constrained parallel task scheduling cost optimizing method under mixed cloud environment |
| CN107274053A (en) * | 2017-05-03 | 2017-10-20 | 浙江工商大学 | The wisdom logistics data method for digging dispatched based on mixed cloud |
| CN110308967A (en) * | 2019-06-06 | 2019-10-08 | 东南大学 | A kind of workflow cost based on mixed cloud-delay optimization method for allocating tasks |
| WO2023005702A1 (en) * | 2021-07-28 | 2023-02-02 | 腾讯科技（深圳）有限公司 | Data processing method and apparatus based on edge computing, and device and storage medium |
| CN116306324A (en) * | 2023-05-25 | 2023-06-23 | 安世亚太科技股份有限公司 | Distributed resource scheduling method based on multiple agents |
| CN117032962A (en) * | 2023-08-07 | 2023-11-10 | 中国电信股份有限公司技术创新中心 | Distribution method and device of computing power resources, storage medium and electronic equipment |
Family Cites Families (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US9948514B2 (en) * | 2014-06-30 | 2018-04-17 | Microsoft Technology Licensing, Llc | Opportunistically connecting private computational resources to external services |
-
2023
- 2023-12-01 CN CN202311631800.9A patent/CN117331706B/en active Active
Patent Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104102544A (en) * | 2014-06-30 | 2014-10-15 | 武汉理工大学 | Multi QoS (quality of service)-constrained parallel task scheduling cost optimizing method under mixed cloud environment |
| CN107274053A (en) * | 2017-05-03 | 2017-10-20 | 浙江工商大学 | The wisdom logistics data method for digging dispatched based on mixed cloud |
| CN110308967A (en) * | 2019-06-06 | 2019-10-08 | 东南大学 | A kind of workflow cost based on mixed cloud-delay optimization method for allocating tasks |
| WO2023005702A1 (en) * | 2021-07-28 | 2023-02-02 | 腾讯科技（深圳）有限公司 | Data processing method and apparatus based on edge computing, and device and storage medium |
| CN116306324A (en) * | 2023-05-25 | 2023-06-23 | 安世亚太科技股份有限公司 | Distributed resource scheduling method based on multiple agents |
| CN117032962A (en) * | 2023-08-07 | 2023-11-10 | 中国电信股份有限公司技术创新中心 | Distribution method and device of computing power resources, storage medium and electronic equipment |
Non-Patent Citations (1)
| Title |
|---|
| 基于二次聚类的多目标混合云任务调度算法;李建丽;丁丁;李涛;;浙江大学学报(工学版)(06);全文 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN117331706A (en) | 2024-01-02 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN110555785B (en) | Monthly plan safety and stability checking method and system | |
| CN110704231A (en) | Fault processing method and device | |
| CN100580594C (en) | Machine group scheduling method based on process capacity characteristic model | |
| CN108876140A (en) | A kind of dispatching method, device, server and the medium of power communication maintenance task | |
| CN104217978A (en) | Semiconductor lot handling system and method | |
| CN112446526A (en) | Production scheduling system and method | |
| CN112333342B (en) | Intelligent voice calling method, device, equipment and storage medium | |
| CN112565422B (en) | Method, system and storage medium for identifying fault data of power internet of things | |
| CN117331706B (en) | Calculation force optimization method and system in electric power data maintenance | |
| CN111177128A (en) | Batch processing method and system for big metering data based on improved outlier detection algorithm | |
| CN107908731A (en) | Based on PSD BPA Guangxi Power Grids barrier load data batch modification method | |
| CN117235062A (en) | Service system data modeling method based on data center | |
| CN104915250B (en) | It is a kind of to realize the method for making MapReduce data localization in the industry | |
| CN104850923A (en) | Semiconductor production simulation system | |
| CN115329995A (en) | Optimization method and system for state maintenance decision of power system | |
| CN109359870A (en) | A kind of distribution network failure recovery scheme comprehensive estimation method based on selection elimination approach | |
| CN114676931A (en) | Electric quantity prediction system based on data relay technology | |
| CN106843101A (en) | Data analysis processing method and device | |
| CN109522590B (en) | Engine blade frequency ordering method | |
| CN107403027A (en) | A kind of method for establishing meter and the failure rate of electrical equipment correction model of maintenance influence | |
| CN109492913B (en) | Modular risk prediction method and device for power distribution terminal and storable medium | |
| CN109242308B (en) | Power distribution network fault recovery scheme interval evaluation method considering load uncertainty | |
| CN114330527B (en) | Building ammeter distribution identification method, system, device and storage medium | |
| CN116566859B (en) | Network anomaly detection method and system for switch | |
| CN109697571A (en) | Product quality analysis method, apparatus and storage medium |
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 | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |