CN110781003A - Load balancing method for particle swarm fusion variation control - Google Patents

Load balancing method for particle swarm fusion variation control Download PDF

Info

Publication number
CN110781003A
CN110781003A CN201911019387.4A CN201911019387A CN110781003A CN 110781003 A CN110781003 A CN 110781003A CN 201911019387 A CN201911019387 A CN 201911019387A CN 110781003 A CN110781003 A CN 110781003A
Authority
CN
China
Prior art keywords
particle swarm
load balancing
chromosome
population
variation
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.)
Granted
Application number
CN201911019387.4A
Other languages
Chinese (zh)
Other versions
CN110781003B (en
Inventor
徐贇
付蔚
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Chongqing University of Post and Telecommunications
Original Assignee
Chongqing University of Post and Telecommunications
Priority date (The priority date 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 date listed.)
Filing date
Publication date
Application filed by Chongqing University of Post and Telecommunications filed Critical Chongqing University of Post and Telecommunications
Priority to CN201911019387.4A priority Critical patent/CN110781003B/en
Publication of CN110781003A publication Critical patent/CN110781003A/en
Application granted granted Critical
Publication of CN110781003B publication Critical patent/CN110781003B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Images

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]
    • G06F9/5083—Techniques for rebalancing the load in a distributed system
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
    • G06N3/00—Computing arrangements based on biological models
    • G06N3/004—Artificial life, i.e. computing arrangements simulating life
    • G06N3/006—Artificial life, i.e. computing arrangements simulating life based on simulated virtual individual or collective life forms, e.g. social simulations or particle swarm optimisation [PSO]
    • 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)
  • Theoretical Computer Science (AREA)
  • Physics & Mathematics (AREA)
  • Software Systems (AREA)
  • General Physics & Mathematics (AREA)
  • General Engineering & Computer Science (AREA)
  • Data Mining & Analysis (AREA)
  • Computational Linguistics (AREA)
  • Biophysics (AREA)
  • Evolutionary Computation (AREA)
  • General Health & Medical Sciences (AREA)
  • Molecular Biology (AREA)
  • Computing Systems (AREA)
  • Biomedical Technology (AREA)
  • Artificial Intelligence (AREA)
  • Mathematical Physics (AREA)
  • Life Sciences & Earth Sciences (AREA)
  • Health & Medical Sciences (AREA)
  • Management, Administration, Business Operations System, And Electronic Commerce (AREA)

Abstract

The invention relates to a load balancing method for particle swarm fusion variation control, and belongs to the field of computer load balancing. Including S1: generating an initial population; s2: selecting operation; s3: performing cross operation; s4: performing mutation operation; s5: carrying out iterative updating; this can be terminated by achieving one of the following conditions: (1) finding an optimal solution; (2) the maximum number of iterations is reached. S7: and after the termination condition is reached, stopping the flow, and obtaining the optimal scheme of the processing relation between the tasks and the nodes by the last chromosome which is left according to the logical inverse solution generated by the sequence. The invention combines the ideas of the two methods, adds a control function and is assisted by a proper fitness function, thereby not only avoiding the situation of falling into local optimum in the early stage, but also avoiding the problem of poor convergence precision in the later stage.

Description

Load balancing method for particle swarm fusion variation control
Technical Field
The invention belongs to the technical field of computer load balancing, and relates to a load balancing method for particle swarm fusion variation control.
Background
The appearance of computers is used for assisting people to process some complex and repeated things, the things are more and more complex with the time, under the requirement, people have to improve the processing capacity of the computers, namely the processing capacity of a CPU, but the data information amount is increased explosively when the internet era comes, network data communication is the most common way, the server is accessed at high concurrency with large data amount, so that the server cannot work effectively or even generates errors, the problem can not be solved effectively by improving the hardware operation capacity of the server alone at the time, the processing capacity of the computers is wasted, and network congestion is caused, the effective solution is to utilize a server cluster which is formed by integrating a plurality of servers, the problem of high concurrency access can be solved effectively when the cluster appears, tasks needing to be processed are distributed to different computers for processing, the development of technology is always a problem of problem occurrence-solution occurrence-re-solution, and is accompanied by a problem of task allocation, namely how to realize load balancing.
Under the condition, it is crucial to design a load balancing mechanism to achieve reasonable task allocation, namely to design a task scheduler, the core of which is a load balancing algorithm, and a good load balancing algorithm, so that hardware resources can be utilized to the maximum extent, network tasks are processed efficiently, interaction experience of users is improved, and compared with the level of replacing hardware alone, the method has higher cost performance, and can save resources.
So far, load balancing algorithms can be roughly divided into static and dynamic, the static algorithms allocate requests in a fixed manner, although the implementation is easy, the effect is not good in practice, because the considered conditions are too ideal and simple, such as dead board operation and different performance of servers in a cluster, the dynamic algorithms consider more than the static algorithms, and are more comprehensive, the current research situation is basically based on the improvement or combination of the traditional dynamic algorithms, the improvement point depends on which aspect an author focuses on, and of course, a perfect algorithm cannot exist in reality, and each improved algorithm is prone to different application scenarios and is the point that the author optimizes different scenarios.
Disclosure of Invention
In view of the above, an object of the present invention is to provide a load balancing method for controlling fusion and variation of particle clusters, which can maintain load balancing as much as possible when a server cluster processes tasks by a more effective means, aiming at the problem of poor processing effect of the server due to high concurrent access in the internet, and the method has two advantages that firstly, on the basis of not increasing additional cost, load balancing is realized in a software manner, and higher cost performance is achieved; secondly, hardware resources are utilized more effectively.
In order to achieve the purpose, the invention provides the following technical scheme:
a load balancing method for particle swarm fusion variation control comprises the following steps:
s1: generating an initial population: the population is a set of 'chromosomes', the chromosome sequence comprises two pieces of information, the tasks correspond to the nodes, and an initial population is generated by a random method;
s2: selecting operation: based on the theory of survival of the fittest, selecting chromosome individuals in the population through a fitness function F (i);
s3: and (3) cross operation: the crossing is a means of generating new individuals and enriching a gene library, and the individuals in the population are used as parents, and partial genes between the parents are exchanged according to a designed crossing mode, so that the new individuals are formed;
s4: mutation operation:
variation control is added into the particle swarm algorithm, so that the algorithm avoids excessively fast shrinkage in the early stage, the local optimum is achieved, the convergence precision is ensured, and a variation control function is introduced to control the variation rate of variation operation:
Figure BDA0002246703820000021
if the predetermined variation rate is t, the variation rate controlled by the variation function is: t is t *＝tf(d)；
Where d is the number of current iterations, d maxIs the total number of pre-iterations, d ∈ [0, d max]And α is a control coefficient, k ═ d max+δ) αδ is a relative d maxA small number in order to ensure that d is taken to d maxAt times f (d) is small but not zero.
S5: iterative updating
And (3) carrying out iterative updating and evolutionary logic according to a particle swarm algorithm and carrying out appropriate modification:
Figure BDA0002246703820000022
Figure BDA0002246703820000023
Figure BDA0002246703820000024
wherein
Figure BDA0002246703820000025
Indicating that the history of the individual is optimal,
Figure BDA0002246703820000026
it is shown that the overall history is optimal, represents the value of the d-th generation of the individual,
Figure BDA0002246703820000028
representing the cumulative optimal difference; the power of updating comes from the history of the individual and the history of the whole body;
s6: and (4) terminating:
terminating the process if one of the following conditions is reached: (1) finding an optimal solution; (2) the maximum iteration times are reached;
s7: and (3) decoding:
and after the termination condition is reached, stopping the flow, and obtaining the optimal scheme of the processing relation between the tasks and the nodes by the last chromosome which is left according to the logical inverse solution generated by the sequence.
Further, the random method in step S1 generates an initial population, which specifically includes: setting the number of tasks as M and the number of nodes as n, randomly generating M numbers between [1, n ], then arranging the M numbers, taking the number of the arranging methods as the population scale, if the number of the arranged combinations is larger than a set population scale number M, selecting M types, and simultaneously determining the iteration number.
Further, in step S2, the fitness function is designed to:
F(i)＝exp{[Rstime min-Rstime(i)]·σ}·
wherein i represents the chromosome number, Rstime (i) represents the time consumed by the project processing task corresponding to the ith chromosome, and Rstime minRepresents the minimum value of the time consumed by each chromosome corresponding scheme in the same time period, and sigma represents the load variance, so that the fitness function is the chromosomes which consume less time and are more balanced in load, and the fitness is higher and is easier to keep.
Further, step S4 specifically includes: introducing a variation control function to control the variation rate of the variation operation:
Figure BDA0002246703820000031
if the predetermined variation rate is t, the variation rate controlled by the variation function is: t is t *＝tf(d)；
Where d is the number of current iterations, d maxIs the total number of pre-iterations, d ∈ [0, d max]And α is a control coefficient, k ═ d max+δ) αδ is a relative d maxA small number in order to ensure that d is taken to d maxAt times f (d) is small but not zero.
The invention has the beneficial effects that: the ideas of the two methods are combined, a control function is added, and a proper fitness function is used for assisting, so that the situation that the early stage is trapped in local optimization is avoided, and the problem that the later stage convergence precision is poor is also avoided.
Additional advantages, objects, and features of the invention will be set forth in part in the description which follows and in part will become apparent to those having ordinary skill in the art upon examination of the following or may be learned from practice of the invention. The objectives and other advantages of the invention may be realized and attained by the means of the instrumentalities and combinations particularly pointed out hereinafter.
Drawings
For the purposes of promoting a better understanding of the objects, aspects and advantages of the invention, reference will now be made to the following detailed description taken in conjunction with the accompanying drawings in which:
fig. 1 is a flow chart of a load balancing method for particle swarm fusion variation control according to the present invention.
Detailed Description
The embodiments of the present invention are described below with reference to specific embodiments, and other advantages and effects of the present invention will be easily understood by those skilled in the art from the disclosure of the present specification. The invention is capable of other and different embodiments and of being practiced or of being carried out in various ways, and its several details are capable of modification in various respects, all without departing from the spirit and scope of the present invention. It should be noted that the drawings provided in the following embodiments are only for illustrating the basic idea of the present invention in a schematic way, and the features in the following embodiments and examples may be combined with each other without conflict.
Wherein the showings are for the purpose of illustrating the invention only and not for the purpose of limiting the same, and in which there is shown by way of illustration only and not in the drawings in which there is no intention to limit the invention thereto; to better illustrate the embodiments of the present invention, some parts of the drawings may be omitted, enlarged or reduced, and do not represent the size of an actual product; it will be understood by those skilled in the art that certain well-known structures in the drawings and descriptions thereof may be omitted.
The same or similar reference numerals in the drawings of the embodiments of the present invention correspond to the same or similar components; in the description of the present invention, it should be understood that if there is an orientation or positional relationship indicated by terms such as "upper", "lower", "left", "right", "front", "rear", etc., based on the orientation or positional relationship shown in the drawings, it is only for convenience of description and simplification of description, but it is not an indication or suggestion that the referred device or element must have a specific orientation, be constructed in a specific orientation, and be operated, and therefore, the terms describing the positional relationship in the drawings are only used for illustrative purposes, and are not to be construed as limiting the present invention, and the specific meaning of the terms may be understood by those skilled in the art according to specific situations.
1 particle swarm algorithm
It is optimized by the mutual transmission competition of particle information in the population, which can be expressed as: randomly initializing a particle swarm with the number of m in an N-dimensional space, and updating the speed and the position of the example through individual optimization and overall optimization:
Figure BDA0002246703820000041
Figure BDA0002246703820000042
2GA algorithm
A simple genetic algorithm refers to a rule of biological evolution, namely 'survival of a suitable person', firstly, a realistic problem is randomly coded into chromosomes to form a population, and the chromosomes are continuously evolved towards the optimal state through operations such as evaluation, intersection, variation and the like of a fitness function, so that the optimal solution of the problem is finally obtained.
3 fusion improvement
Firstly, the ideas of the two algorithms are fused together, and then a variation control function is added, so that the variation rate is controlled, enough individuals in the early stage can be ensured, the local optimum is avoided, the convergence precision in the later stage can be accelerated, and whether the individuals are reserved or not can be judged more reasonably by being assisted with a proper fitness function.
F(i)＝exp{[Rstime min-Rstime(i)]·σ}
Wherein i represents the chromosome number, Rstime (i) represents the time consumed by the project processing task corresponding to the ith chromosome, and Rstime minThe method includes that the minimum value of time consumed by each chromosome corresponding to a scheme in the same time period is represented, sigma represents load variance, theoretically, the capacity of each distributed task amount corresponding to each chromosome is called balance, the total cluster performance on the total task ratio is a fixed value, the performance or the remaining performance of each node distributed task ratio is a variable value, the balance variance is dynamic, the difference between the individual ratio and the total ratio can be measured, and a fitness function can be understood as a chromosome which consumes less time and is more balanced in load, and the chromosome is higher in fitness and easier to survive.
In more detail: suppose that the ith meterPerformance of the compute node is C iThe factors affecting performance include CPU, bandwidth, memory, disk, etc., so the total computing power of the cluster is:
Figure BDA0002246703820000043
assuming that the calculation amount of the jth task is t for m tasks to be processed jThen the total number of computations for m tasks is
Figure BDA0002246703820000044
To node p kThe number of tasks in (a) is calculated as
Figure BDA0002246703820000045
Node p kHas the properties of
Figure BDA0002246703820000051
Then there should be the following:
Figure BDA0002246703820000052
is actually fluctuating, if it is assumed:
Figure BDA0002246703820000053
wherein k is [1, n ]]And sigma is the variance of n mu.
Thus, as shown in fig. 1, the present invention provides a load balancing method for controlling fusion variation of particle swarm, which includes the following steps:
s1: generating an initial population: population, i.e. the set of "chromosomes", generally the chromosome sequence contains two pieces of information, the relationship between tasks and nodes, and there are several methods for generating the initial population, and the initial population can be generated by using the most general random method, and assuming that the number of tasks is m and the nodes are n, the chromosome sequence can be generated by:
randomly generating M numbers between [1, n ], and then arranging the M numbers, wherein the number of the arranging method can be used as the population scale, of course, the situation can also be adopted, the number of the arranged combinations is larger than the set population scale number M, at this time, M kinds are selected, and the iteration number is determined at the same time.
S2: selecting operation: based on the theory of survival of the fittest, the chromosome individuals in the population must be selected through the fitness function f (i), and the fitter is left behind.
S3: and (3) cross operation: crossing is a means for generating new individuals and enriching gene libraries, and the individuals in the population are used as parents, and partial genes between the parents are exchanged according to a designed crossing mode, so that the new individuals are formed.
S4: mutation operation:
the particle swarm algorithm of a single unit obtains the optimal solution through the competition of particles, but is easy to fall into the local optimal solution, the genetic algorithm of a single unit is optimized through iteration and variation, but the convergence precision of the algorithm is not good, if at the moment, proper variation control is added into the particle swarm algorithm, namely when the two algorithms are fused, the two algorithms are improved simultaneously, the algorithm is prevented from being shrunk too fast in the early stage, the local optimal solution is fallen into, and the convergence precision is also ensured, so that a variation control function is introduced to control the variation rate of the variation operation:
Figure BDA0002246703820000054
if the predetermined variation rate is t, the variation rate controlled by the variation function is: t is t *＝tf(d)；
Where d is the number of current iterations, d maxIs the total number of pre-iterations, d ∈ [0, d max]And α is a control coefficient, k ═ d max+δ) αδ is a relative d maxA small number in order to ensure that d is taken to d maxAt times f (d) is small but not zero.
Through the structure of the function, the method is obtained through analysis, the larger variation rate is kept in the early stage of the hybrid algorithm, the population keeps diversity as much as possible, the search surface is larger, the local optimization is avoided, the variation rate is reduced in the later stage of the hybrid algorithm iteration, and only fine adjustment is needed because the optimization is already approached at the moment, and meanwhile, the convergence precision is guaranteed.
S5: iterative updating
And (3) carrying out iterative updating and evolutionary logic according to a particle swarm algorithm and carrying out appropriate modification:
Figure BDA0002246703820000061
Figure BDA0002246703820000062
Figure BDA0002246703820000063
description of the drawings:
Figure BDA0002246703820000064
indicating that the history of the individual is optimal,
Figure BDA0002246703820000065
it is shown that the overall history is optimal,
Figure BDA0002246703820000066
represents the value of the d-th generation of the individual,
Figure BDA0002246703820000067
indicating the cumulative optimal difference. In other words, the updated power comes from two aspects: one is the history of the individual, and the other is the history of the whole.
S6: and (4) terminating:
this can be terminated by achieving one of the following conditions: (1) finding an optimal solution; (2) the maximum number of iterations is reached.
S7: and (3) decoding:
and after the termination condition is reached, stopping the flow, and obtaining the optimal scheme of the processing relation between the tasks and the nodes by the last chromosome which is left according to the logical inverse solution generated by the sequence.
Optionally, the fitness function design emphasizes two factors, i.e., the load of the cluster and the time consumed by the processing task, generally, the design requirement of the fitness function is non-negative, and the chromosome is more adaptive to the environment as the fitness value is larger, so we design the fitness function as:
F(i)＝exp{[Rstime min-Rstime(i)]·σ}
wherein i represents the chromosome number, Rstime (i) represents the protocol corresponding to the ith chromosome, the time consumed for processing the task, and Rstime minRepresents the minimum value of the time consumed by each chromosome corresponding scheme in the same time period, and sigma represents the load variance, so the fitness function can be understood as the chromosomes which are less time consuming and more balanced in load, and the higher the fitness, the easier the fitness is to keep.
Finally, the above embodiments are only intended to illustrate the technical solutions of the present invention and not to limit the present invention, and although the present invention has been described in detail with reference to the preferred embodiments, it will be understood by those skilled in the art that modifications or equivalent substitutions may be made on the technical solutions of the present invention without departing from the spirit and scope of the technical solutions, and all of them should be covered by the claims of the present invention.

Claims (4)

1. A load balancing method for particle swarm fusion variation control is characterized in that: the method comprises the following steps:
s1: generating an initial population: the population is a set of 'chromosomes', the chromosome sequence comprises two pieces of information, the tasks correspond to the nodes, and an initial population is generated by a random method;
s2: selecting operation: based on the theory of survival of the fittest, selecting chromosome individuals in the population through a fitness function F (i);
s3: and (3) cross operation: the crossing is a means of generating new individuals and enriching a gene library, and the individuals in the population are used as parents, and partial genes between the parents are exchanged according to a designed crossing mode, so that the new individuals are formed;
s4: mutation operation:
variation control is added into the particle swarm algorithm, so that the algorithm avoids too fast shrinkage in the early stage, falls into local optimum, and ensures convergence accuracy;
s5: iterative updating
And (3) carrying out iterative updating and evolutionary logic according to a particle swarm algorithm and carrying out appropriate modification:
Figure FDA0002246703810000012
wherein
Figure FDA0002246703810000013
Indicating that the history of the individual is optimal,
Figure FDA0002246703810000014
it is shown that the overall history is optimal,
Figure FDA0002246703810000015
represents the value of the d-th generation of the individual,
Figure FDA0002246703810000016
representing the cumulative optimal difference; the power of updating comes from the history of the individual and the history of the whole body;
s6: and (4) terminating:
terminating the process if one of the following conditions is reached: (1) finding an optimal solution; (2) the maximum iteration times are reached;
s7: and (3) decoding:
and after the termination condition is reached, stopping the flow, and obtaining the optimal scheme of the processing relation between the tasks and the nodes by the last chromosome which is left according to the logical inverse solution generated by the sequence.
2. The load balancing method for particle swarm fusion variation control according to claim 1, wherein: the random method in step S1 generates an initial population, which specifically includes: setting the number of tasks as M and the number of nodes as n, randomly generating M numbers between [1, n ], then arranging the M numbers, taking the number of the arranging methods as the population scale, if the number of the arranged combinations is larger than a set population scale number M, selecting M types, and simultaneously determining the iteration number.
3. The load balancing method for particle swarm fusion variation control as claimed in claim 1, wherein: in step S2, the fitness function is designed as:
F(i)＝exp{[Rstime min-Rstime(i)]·σ}·
wherein i represents the chromosome number, Rstime (i) represents the time consumed by the project processing task corresponding to the ith chromosome, and Rstime minRepresents the minimum value of the time consumed by each chromosome corresponding scheme in the same time period, and sigma represents the load variance, so that the fitness function is the chromosomes which consume less time and are more balanced in load, and the fitness is higher and is easier to keep.
4. The load balancing method for particle swarm fusion variation control as claimed in claim 1, wherein: in step S4, the method specifically includes: introducing a variation control function to control the variation rate of the variation operation:
Figure FDA0002246703810000021
if the predetermined variation rate is t, the variation rate controlled by the variation function is: t is t *＝tf(d)；
Where d is the number of current iterations, d maxIs the total number of pre-iterations, d ∈ [0, d max]And α is a control coefficient, k ═ d max+δ) αδ is a relative d maxA small number in order to ensure that d is taken to d maxAt times f (d) is small but not zero.
CN201911019387.4A 2019-10-24 2019-10-24 Load balancing method for particle swarm fusion variation control Active CN110781003B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN201911019387.4A CN110781003B (en) 2019-10-24 2019-10-24 Load balancing method for particle swarm fusion variation control

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN201911019387.4A CN110781003B (en) 2019-10-24 2019-10-24 Load balancing method for particle swarm fusion variation control

Publications (2)

Publication Number Publication Date
CN110781003A true CN110781003A (en) 2020-02-11
CN110781003B CN110781003B (en) 2023-04-07

Family

ID=69386428

Family Applications (1)

Application Number Title Priority Date Filing Date
CN201911019387.4A Active CN110781003B (en) 2019-10-24 2019-10-24 Load balancing method for particle swarm fusion variation control

Country Status (1)

Country Link
CN (1) CN110781003B (en)

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN116737394A (en) * 2023-08-14 2023-09-12 中海智（北京）科技有限公司 Dynamic adjustment security check centralized graph judging task allocation method

Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN105959234A (en) * 2016-06-30 2016-09-21 南京理工大学 Load balance resource optimization method under safety perceptive cloud radio access network
CN108900428A (en) * 2018-06-26 2018-11-27 南京邮电大学 Controller load-balancing method based on interchanger dynamic migration
CN110109753A (en) * 2019-04-25 2019-08-09 成都信息工程大学 Resource regulating method and system based on various dimensions constraint genetic algorithm
CN110225096A (en) * 2019-05-20 2019-09-10 电子科技大学 Server load balancing method based on genetic algorithm

Patent Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN105959234A (en) * 2016-06-30 2016-09-21 南京理工大学 Load balance resource optimization method under safety perceptive cloud radio access network
CN108900428A (en) * 2018-06-26 2018-11-27 南京邮电大学 Controller load-balancing method based on interchanger dynamic migration
CN110109753A (en) * 2019-04-25 2019-08-09 成都信息工程大学 Resource regulating method and system based on various dimensions constraint genetic algorithm
CN110225096A (en) * 2019-05-20 2019-09-10 电子科技大学 Server load balancing method based on genetic algorithm

Non-Patent Citations (2)

* Cited by examiner, † Cited by third party
Title
刘震宇;谢胜利;: "网络处理器负载均衡中的遗传算法研究" *
梅东晖;李红灵;: "基于多目标混合粒子群算法的虚拟机负载均衡研究" *

Cited By (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN116737394A (en) * 2023-08-14 2023-09-12 中海智（北京）科技有限公司 Dynamic adjustment security check centralized graph judging task allocation method
CN116737394B (en) * 2023-08-14 2023-10-27 中海智（北京）科技有限公司 Dynamic adjustment security check centralized graph judging task allocation method

Also Published As

Publication number Publication date
CN110781003B (en) 2023-04-07

Similar Documents

Publication Publication Date Title
Srichandan et al. Task scheduling for cloud computing using multi-objective hybrid bacteria foraging algorithm
Liu et al. Deadline‐constrained coevolutionary genetic algorithm for scientific workflow scheduling in cloud computing
Zhang et al. Network-aware virtual machine migration in an overcommitted cloud
Junaid et al. Modeling an optimized approach for load balancing in cloud
Lee et al. Designing a parallel evolutionary algorithm for inferring gene networks on the cloud computing environment
Supreeth et al. Hybrid genetic algorithm and modified-particle swarm optimization algorithm (GA-MPSO) for predicting scheduling virtual machines in educational cloud platforms
CN111966495B (en) Data processing method and device
CN110008023B (en) Cloud computing system budget constraint random task scheduling method based on genetic algorithm
CN114461386A (en) Task allocation method and task allocation device
Özbakır et al. Clustering performance comparison of new generation meta-heuristic algorithms
Zhou et al. Tools to support mesh adaptation on massively parallel computers
Huang et al. Semantic integration of sensor knowledge on artificial internet of things
CN110781003B (en) Load balancing method for particle swarm fusion variation control
Xie et al. Subpopulation particle swarm optimization with a hybrid mutation strategy
Liu et al. Multi-objective dynamic population shuffled frog-leaping biclustering of microarray data
CN112836794B (en) Method, device, equipment and storage medium for determining image neural architecture
Li et al. An improved binary quantum-behaved particle swarm optimization algorithm for knapsack problems
Nair et al. Optimal knockout strategies in genome-scale metabolic networks using particle swarm optimization
Eldesokey et al. Hybrid swarm optimization algorithm based on task scheduling in a cloud environment
CN110175172B (en) Extremely-large binary cluster parallel enumeration method based on sparse bipartite graph
CN115794354A (en) Cloud video transcoding task scheduling based on K-Means clustering and improved AOA algorithm
Li et al. IPSOMC: An improved particle swarm optimization and membrane computing based algorithm for cloud computing
WO2021232907A1 (en) Neural network model training apparatus and method, and related device
Sun et al. An evolutionary many-objective algorithm based on decomposition and hierarchical clustering selection
Du et al. Comparative study of distributed deep learning tools on supercomputers

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