CN110533152A - A kind of method for scheduling task based on improvement fireworks algorithm under CMP architecture - Google Patents
A kind of method for scheduling task based on improvement fireworks algorithm under CMP architecture Download PDFInfo
- Publication number
- CN110533152A CN110533152A CN201910738722.XA CN201910738722A CN110533152A CN 110533152 A CN110533152 A CN 110533152A CN 201910738722 A CN201910738722 A CN 201910738722A CN 110533152 A CN110533152 A CN 110533152A
- Authority
- CN
- China
- Prior art keywords
- fireworks
- fitness value
- fitness
- algorithm
- cmp
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
- 238000000034 method Methods 0.000 title claims abstract description 20
- 238000004880 explosion Methods 0.000 claims abstract description 9
- 230000035772 mutation Effects 0.000 claims abstract description 7
- 230000001174 ascending effect Effects 0.000 claims abstract description 6
- 239000002360 explosive Substances 0.000 claims abstract description 6
- 238000012163 sequencing technique Methods 0.000 claims description 4
- 235000019504 cigarettes Nutrition 0.000 claims 1
- 230000002068 genetic effect Effects 0.000 description 4
- 238000005457 optimization Methods 0.000 description 3
- 238000013507 mapping Methods 0.000 description 2
- 230000000717 retained effect Effects 0.000 description 2
- 238000012216 screening Methods 0.000 description 2
- 230000006978 adaptation Effects 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 230000009172 bursting Effects 0.000 description 1
- 238000006073 displacement reaction Methods 0.000 description 1
- 235000013399 edible fruits Nutrition 0.000 description 1
- 238000013178 mathematical model Methods 0.000 description 1
- 230000005855 radiation Effects 0.000 description 1
- 230000000452 restraining effect Effects 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
- G06F9/48—Program initiating; Program switching, e.g. by interrupt
- G06F9/4806—Task transfer initiation or dispatching
- G06F9/4843—Task transfer initiation or dispatching by program, e.g. task dispatcher, supervisor, operating system
- G06F9/4881—Scheduling strategies for dispatcher, e.g. round robin, multi-level priority queues
-
- 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]
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Software Systems (AREA)
- Physics & Mathematics (AREA)
- General Engineering & Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Artificial Intelligence (AREA)
- Life Sciences & Earth Sciences (AREA)
- Health & Medical Sciences (AREA)
- Biomedical Technology (AREA)
- Biophysics (AREA)
- Computational Linguistics (AREA)
- Data Mining & Analysis (AREA)
- Evolutionary Computation (AREA)
- General Health & Medical Sciences (AREA)
- Molecular Biology (AREA)
- Computing Systems (AREA)
- Mathematical Physics (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The invention belongs to mathematical technique fields, and in particular to a kind of method for scheduling task based on improvement fireworks algorithm under CMP architecture of execution efficiency can be improved.This method includes the following steps: N number of task schedule sequence in CMP Mission Scheduling being mapped as N number of fireworks, initialize;Calculate the fitness value, burst radius and explosive spark number of each fireworks;Carry out explosion and Gaussian mutation;The fireworks that will exceed feasible zone map back in feasible zone;Fireworks are subjected to ascending sort according to fitness value;The smallest fireworks of fitness value are selected as next-generation fireworks.Improved fireworks algorithm is used in CMP Mission Scheduling by the present invention, so that it is most short to have executed the time used in whole tasks, effectively raises the execution efficiency of task under CMP architecture.
Description
Technical field
The invention belongs to mathematical technique fields, and in particular to one kind that execution efficiency can be improved is based on improving fireworks algorithm
Method for scheduling task under CMP architecture.
Background technique
Dependence between by interior internuclear hardware resource sharing and task is influenced, and has the CMP task schedule of dependence to be easy
Contention for resources is generated, so that multi-core processor throughput declines, concurrency, high efficiency cannot be played effectively.
CMP Mission Scheduling is the complete combinatorial optimization problem of a kind of NP.Currently, on Mission Scheduling using compared with
More combinatorial optimization algorithms is genetic algorithm (Genetic Algorithm, GA), but genetic algorithm is easy " precocity ", and by
In its complicated cross and variation operation, the time complexity and execution efficiency of algorithm are easy to be influenced by task quantity, cause to appoint
Business scheduling result is not able to satisfy actual demand.
Fireworks algorithm (Fireworks Algorithm, FWA) is that the famous Tan Ying professor in China was put forward for the first time in 2010
A kind of Swarm Intelligence Algorithm, be a kind of mathematical model taken out according to fireworks explosion phenomenon in nature, by introduce with
Machine factor and selection strategy form a kind of parallel explosive way of search, and the complete of complicated optimum problem optimal solution can be solved by becoming
Office's probabilistic search method.Fireworks algorithm starts iteration, successively utilizes explosion operator, mutation operator, mapping ruler and selection strategy,
Until reaching termination condition, that is, meet the required precision of problem or reaching maximal function assessment number.
The realization of fireworks algorithm includes following several steps:
1) some fireworks are randomly generated in specific solution space, each fireworks represents a solution of solution space.
2) fitness value of each fireworks is calculated according to fitness function, and spark is generated according to fitness value.Spark
Number be to be calculated based on the thought of the immune concentration in immunology, i.e., the better fireworks of fitness value produce pyrophoric number
Mesh is more.
3) it according to the fireworks attribute in reality and the actual conditions of combination search problem, is generated in the radiation space of fireworks
Spark.The size of the explosion amplitude of some fireworks determines that fitness value is bigger by fitness value of the fireworks on function, explosion
Amplitude is bigger, and vice versa.Each spark represents a solution in solution space.In order to guarantee the diversity of population, need pair
Fireworks are suitably made a variation, such as Gaussian mutation.
4) optimal solution for calculating population, judges whether fitness value tends to be steady or whether reach maximum number of iterations, such as
Fruit is to stop search, and otherwise continues iteration.
Fireworks algorithm is there is also convergence rate is slow, the shortcomings that being easily trapped into locally optimal solution, in order to optimize these disadvantages,
Present invention fitness value from the aspect of selection strategy selects next-generation fireworks, enhances the search capability and convergence speed of algorithm
Degree.
Summary of the invention
The purpose of the present invention is to provide a kind of method for scheduling task based on improvement fireworks algorithm under CMP architecture.
The object of the present invention is achieved like this:
A kind of method for scheduling task based on improvement fireworks algorithm under CMP architecture, includes the following steps:
Step 1: N number of task schedule sequence in CMP Mission Scheduling is mapped as N number of fireworks, it is random after initialization
Generate the position vector of N number of fireworks;
Step 2: calculating the fitness value, burst radius and explosive spark number of each fireworks;
Step 3: carrying out explosion and Gaussian mutation;
Step 4: the fireworks that will exceed feasible zone map back in feasible zone;
Step 5: fireworks are subjected to ascending sort according to fitness value;
Step 6: the smallest fireworks of fitness value are selected as next-generation fireworks;
Step 7: in remaining N-1 fireworks, the difference of two fireworks fitness is then considered as search capability phase less than 2 if it exists
Seemingly, the lesser fireworks of fitness are retained in the two, the biggish fireworks of fitness are given up, and remaining fireworks sort according to fitness value
Table uniformly chooses next-generation fireworks;
Step 8: if the fitness value of next-generation fireworks restrains or reaches maximum number of iterations, exporting fitness value minimum
Fireworks as optimal task schedule sequence;Otherwise return step 2 carries out next iteration.
The beneficial effects of the present invention are: the selection strategy of basic fireworks algorithm is improved, fitness value is passed through
Next-generation fireworks are selected, makes to avoid the similar spark of selection search capability, selects the fitness value range of fireworks wider, avoid falling into
Enter locally optimal solution.Improved fireworks algorithm is used in CMP Mission Scheduling, so that having executed the time used in whole tasks
It is most short, effectively raise the execution efficiency of task under CMP architecture.
Detailed description of the invention
Fig. 1 is the flow chart of selection strategy in the present invention;
Fig. 2 is the method for scheduling task flow chart under the CMP architecture based on improved fireworks algorithm;
Fig. 3 is three classes algorithmic statement result figure;
Fig. 4 is three classes algorithm the convergence speed figure.
Specific embodiment
The present invention is described further with reference to the accompanying drawing.
The selection strategy of basic fireworks algorithm is using the mode of roulette, and this mode randomness is big, Bu Nengbao
Card leaves more high-quality fireworks, to reduce the precision and convergence rate of algorithm.In order to reduce this randomness, can pass through
A kind of selection strategy considering fitness value selects next-generation fireworks.And CMP Mission Scheduling is incited somebody to action according to a certain algorithm
In task schedule to different kernels, and it is most short to have executed the time used in whole tasks.Now improved fireworks algorithm is used
In CMP Mission Scheduling, steps are as follows for the execution of the CMP task scheduling strategy:
Input quantity: N number of task schedule sequence in CMP Mission Scheduling
(1) fireworks number N is set;
(2) it initializes;
(3) number of iterations adds 1;
(4) fitness value of each fireworks is assessed, and calculates the burst radius and explosive spark number of each fireworks;
(5) explosive spark is generated, displacement operation is carried out to every dimension of fireworks according to bursting strength and explosion amplitude;
(6) generate Gaussian mutation spark, using Gaussian Profile to any number of dimensions of any one fireworks in population into
Row Gaussian mutation;
(7) exceed the fireworks of feasible zone range using mapping policy processing;
(8) fireworks carry out ascending sort according to fitness value;
(9) the best fireworks of fitness value or spark are remained into the next generation, according to the present invention the choosing of described selection strategy
Remaining N-1 fireworks is selected, specific embodiment is as follows: if the difference of two fireworks fitness retains fitness in the two less than 2
Lesser fireworks, the biggish fireworks of fitness are given up, i.e. the similar fireworks of search capability only retain one;
(10) next-generation fireworks are uniformly chosen according to fitness value sequencing table;
(11) judge whether fitness value tends to be steady or whether reach maximum number of iterations, it is defeated if then iteration terminates
Optimal task schedule sequence out, algorithm stop;It is no to then follow the steps (3).
Table 1 is the parameter setting of Genetic Algorithms for comparison, basic fireworks algorithm and SIFWA:
The setting of table 1GA, FWA and SIFWA algorithm parameter
In terms of computational accuracy, as shown in figure 3, inventive algorithm SIFWA computational accuracy is greatly improved.It is restraining
In terms of speed, as shown in figure 4, SIFWA convergence speed of the algorithm is substantially better than FWA and GA algorithm.
Detailed elaboration is carried out to implementation process of the invention in conjunction with attached drawing, those skilled in the art is appreciated that above-mentioned mistake
Journey is example, and therefore, protection scope of the present invention should be determined by the content of appended claims.
The present invention relates to a kind of fireworks algorithms of optimum choice strategy, and propose one based on improved fireworks algorithm
Method for scheduling task under kind CMP architecture, is detailed below the present invention.
A kind of fireworks algorithm of optimum choice strategy, and proposed under a kind of CMP architecture based on improved fireworks algorithm
Method for scheduling task, fitness value the smallest is selected as next-generation fireworks with knowing from experience being determined property, and to remaining fireworks root
It is selected according to fitness value.
Fireworks are subjected to ascending sort according to fitness value, first select the smallest fireworks of fitness value.In remaining fireworks
In, the similar fireworks of search capability only retain one, and specific implementation method is as follows: if the difference of two fireworks fitness less than 2,
It is similar to be considered as search capability, the lesser fireworks of fitness are retained in the two, the biggish fireworks of fitness are given up.By screening it
Afterwards, a certain number of fireworks are uniformly chosen as next-generation fireworks according to fitness value sequencing table.
It is an object of the invention to improve the selection strategy of fireworks algorithm, the similar spark of selection search capability, choosing are avoided
The fitness value range for selecting fireworks is wider.And based on improved fireworks algorithm propose it is a kind of solution have rely on CMP task schedule
The method of problem optimal scheduling sequence.
The selection strategy of basic fireworks algorithm is using the mode of roulette, and this mode randomness is big, Bu Nengbao
Card leaves more high-quality fireworks, to reduce the precision and convergence rate of algorithm.In order to reduce this randomness, can pass through
A kind of selection strategy considering fitness value selects next-generation fireworks.This selection strategy looks after similar fitness and different adaptations
The fireworks of degree reduce randomness, keep the selection of next-generation fireworks more reasonable.
Mode after optimization: fireworks are subjected to ascending sort according to fitness value.Fireworks algorithm passes through fitness value
Size evaluates the superiority and inferiority of fireworks or spark, and high-quality fireworks fitness value is small;Ropy fireworks fitness value is big.But it is straight
Connect the preceding fireworks of selected and sorted be it is unreasonable because fitness value it is close fireworks search capability it is similar, this often by
Have similar search capability because of being limited by explosion amplitude in the spark that same fireworks generate, can not most probably jump
Local optimum out can also seriously undermine the global exploring ability of fireworks.So the smallest fireworks of fitness value should first be selected, so
Heel row deconditioning angle value differs the fireworks within 2.After screening, a fixed number is uniformly chosen according to fitness value sequencing table
The fireworks of amount are as next-generation fireworks.Multiplicity of the poor fireworks of a part of fitness value to guarantee next-generation fireworks population
Property, to maintain the global exploring ability of fireworks.
In conclusion the present invention provides a kind of fireworks algorithm SIFWA for improving selection strategy, and calculated based on improved fireworks
Method proposes the method for scheduling task under a kind of CMP architecture.
Basic fireworks algorithm, which is used, selects next-generation fireworks based on the selection strategy of the roulette rule of distance.This choosing
Select that tactful randomness is big, it cannot be guaranteed that more high-quality fireworks are left, to reduce the precision and convergence rate of algorithm.For
This problem, the present invention improve the selection strategy of fireworks algorithm, and next-generation fireworks are screened according to fitness value, are avoided
The spark for selecting search capability similar and inferior.The present invention can effectively promote fireworks algorithm and find the ability of optimal solution and search
The efficiency of rope.
Now improved fireworks algorithm is used in CMP Mission Scheduling, according to improved fireworks algorithm by task schedule
Onto different kernels, so that having executed, the time used in whole tasks is most short, and the present invention can effectively improve CMP architecture lower
The execution efficiency of business.
Claims (1)
1. a kind of method for scheduling task based on improvement fireworks algorithm under CMP architecture, which comprises the steps of:
Step 1: N number of task schedule sequence in CMP Mission Scheduling being mapped as N number of fireworks, generates N after initialization at random
The position vector of a fireworks;
Step 2: calculating the fitness value, burst radius and explosive spark number of each fireworks;
Step 3: carrying out explosion and Gaussian mutation;
Step 4: the fireworks that will exceed feasible zone map back in feasible zone;
Step 5: fireworks are subjected to ascending sort according to fitness value;
Step 6: the smallest fireworks of fitness value are selected as next-generation fireworks;
Step 7: in remaining N-1 fireworks, the difference of two fireworks fitness is less than 2 if it exists, then it is similar to be considered as search capability, and two
Retain the lesser fireworks of fitness in person, the biggish fireworks of fitness are given up, and remaining fireworks are equal according to fitness value sequencing table
The next-generation fireworks of even selection;
Step 8: if the fitness value of next-generation fireworks restrains or reaches maximum number of iterations, exporting the smallest cigarette of fitness value
Flower is used as optimal task schedule sequence;Otherwise return step 2 carries out next iteration.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201910738722.XA CN110533152A (en) | 2019-08-12 | 2019-08-12 | A kind of method for scheduling task based on improvement fireworks algorithm under CMP architecture |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201910738722.XA CN110533152A (en) | 2019-08-12 | 2019-08-12 | A kind of method for scheduling task based on improvement fireworks algorithm under CMP architecture |
Publications (1)
Publication Number | Publication Date |
---|---|
CN110533152A true CN110533152A (en) | 2019-12-03 |
Family
ID=68662849
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201910738722.XA Pending CN110533152A (en) | 2019-08-12 | 2019-08-12 | A kind of method for scheduling task based on improvement fireworks algorithm under CMP architecture |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN110533152A (en) |
Cited By (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN111124633A (en) * | 2019-12-06 | 2020-05-08 | 上海交通大学 | Earth-moon relay satellite task scheduling method considering storage limit of detector |
CN112257297A (en) * | 2020-11-27 | 2021-01-22 | 西南交通大学 | Welding shop comprehensive scheduling method based on improved firework algorithm |
CN113297785A (en) * | 2021-04-27 | 2021-08-24 | 河南工业大学 | Medical material emergency dispatching optimization method based on firework optimization algorithm |
CN113505975A (en) * | 2021-06-18 | 2021-10-15 | 宁波沙塔信息技术有限公司 | Order insertion and scheduling method based on genetic algorithm and firework algorithm |
CN113505974A (en) * | 2021-06-18 | 2021-10-15 | 宁波沙塔信息技术有限公司 | Multi-target scheduling method based on firework algorithm and genetic algorithm |
Citations (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN107704319A (en) * | 2017-10-18 | 2018-02-16 | 哈尔滨工程大学 | Improve the CMP method for scheduling task of fireworks algorithm |
CN108197696A (en) * | 2018-01-31 | 2018-06-22 | 湖北工业大学 | A kind of network navy account recognition methods and system |
CN108280209A (en) * | 2018-01-31 | 2018-07-13 | 湖北工业大学 | A kind of image search method and system based on fireworks algorithm |
CN109508818A (en) * | 2018-10-23 | 2019-03-22 | 保定正德电力技术有限公司 | A kind of online NOx prediction technique based on LSSVM |
-
2019
- 2019-08-12 CN CN201910738722.XA patent/CN110533152A/en active Pending
Patent Citations (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN107704319A (en) * | 2017-10-18 | 2018-02-16 | 哈尔滨工程大学 | Improve the CMP method for scheduling task of fireworks algorithm |
CN108197696A (en) * | 2018-01-31 | 2018-06-22 | 湖北工业大学 | A kind of network navy account recognition methods and system |
CN108280209A (en) * | 2018-01-31 | 2018-07-13 | 湖北工业大学 | A kind of image search method and system based on fireworks algorithm |
CN109508818A (en) * | 2018-10-23 | 2019-03-22 | 保定正德电力技术有限公司 | A kind of online NOx prediction technique based on LSSVM |
Cited By (9)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN111124633A (en) * | 2019-12-06 | 2020-05-08 | 上海交通大学 | Earth-moon relay satellite task scheduling method considering storage limit of detector |
CN111124633B (en) * | 2019-12-06 | 2023-05-30 | 上海交通大学 | Ground-month relay satellite task scheduling method considering detector storage limit |
CN112257297A (en) * | 2020-11-27 | 2021-01-22 | 西南交通大学 | Welding shop comprehensive scheduling method based on improved firework algorithm |
CN112257297B (en) * | 2020-11-27 | 2021-06-25 | 西南交通大学 | Welding shop comprehensive scheduling method based on improved firework algorithm |
CN113297785A (en) * | 2021-04-27 | 2021-08-24 | 河南工业大学 | Medical material emergency dispatching optimization method based on firework optimization algorithm |
CN113505975A (en) * | 2021-06-18 | 2021-10-15 | 宁波沙塔信息技术有限公司 | Order insertion and scheduling method based on genetic algorithm and firework algorithm |
CN113505974A (en) * | 2021-06-18 | 2021-10-15 | 宁波沙塔信息技术有限公司 | Multi-target scheduling method based on firework algorithm and genetic algorithm |
CN113505975B (en) * | 2021-06-18 | 2024-04-09 | 宁波沙塔信息技术有限公司 | Plug sheet scheduling method based on genetic algorithm and firework algorithm |
CN113505974B (en) * | 2021-06-18 | 2024-04-09 | 宁波沙塔信息技术有限公司 | Multi-target scheduling method based on firework algorithm and genetic algorithm |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN110533152A (en) | A kind of method for scheduling task based on improvement fireworks algorithm under CMP architecture | |
Lagoudakis et al. | Algorithm Selection using Reinforcement Learning. | |
CN104866904B (en) | A kind of BP neural network parallel method of the genetic algorithm optimization based on spark | |
CN104035816A (en) | Cloud computing task scheduling method based on improved NSGA-II | |
CN106845642B (en) | A kind of adaptive multi-target evolution method of belt restraining cloud workflow schedule | |
Cruz-Cortés et al. | Handling constraints in global optimization using an artificial immune system | |
CN110134007A (en) | Multiple no-manned plane cooperates with target assignment method | |
Chang et al. | A new approach to weapon-target assignment in cooperative air combat | |
US10425343B2 (en) | Packet classification | |
Cabrera et al. | Micro-MOPSO: A multi-objective particle swarm optimizer that uses a very small population size | |
CN109948828A (en) | Working procedure compositor method and device | |
CN109145449A (en) | A kind of level of factory thermal power plant load distribution method | |
Chowdhury et al. | Improvements to single-objective constrained predator–prey evolutionary optimization algorithm | |
CN107329826A (en) | A kind of heuristic fusion resource dynamic dispatching algorithm based on Cloudsim platforms | |
Zhao et al. | Improved backtracking search algorithm based on population control factor and optimal learning strategy | |
CN109635034A (en) | Training data method for resampling, device, storage medium and electronic equipment | |
CN108133240A (en) | A kind of multi-tag sorting technique and system based on fireworks algorithm | |
CN109858617A (en) | A kind of fireworks calculation method optimizing initialization mode | |
CN108921207A (en) | A kind of hyper parameter determines method, device and equipment | |
CN105589896B (en) | Data digging method and device | |
Nakano et al. | An asynchronous P system with branch and bound for solving the knapsack problem | |
CN102254040A (en) | SVM (Support Vector Machine)-based Web partitioning method | |
CN106445960A (en) | Data clustering method and device | |
CN108197696A (en) | A kind of network navy account recognition methods and system | |
CN109542623A (en) | The Optimization deployment method of virtual system in the server |
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 | ||
RJ01 | Rejection of invention patent application after publication |
Application publication date: 20191203 |
|
RJ01 | Rejection of invention patent application after publication |