CN103345657A - Task scheduling method based on heredity and ant colony in cloud computing environment - Google Patents

Abstract

Provided in the invention is a task scheduling method based on heredity and ant colony in a cloud computing environment. The method comprises the following methods: S1, initializing population; S2, selecting individuals according to a wheel disc type selection strategy; S3, carrying out crossover operation on the individuals according to crossover probability and carrying out reversion mutation operation according to a mutation probability so as to generate a new colony; S4, updating the new generated colony; S5, determining whether a dynamic fusion condition is met; S6, initializing ant pheromone by using an optimal solution found by heredity; S7, calculating probabilities of moving to next nodes by all ants and moving all the ants to the next nodes according to the probabilities; S8, enabling M ants to travelling N resource nodes and carrying out pheromone updating on an optimal ant cycle; S9, carrying out pheromone updating on all paths; and S10, determining whether an ant end condition is met and outputting an optimal solution. According to the invention, respective advantages of a genetic algorithm and an ant colony algorithm are drawn and respective defects are overcome; and on the basis of dynamic fusion of the two algorithms, time and efficiency of exact solution solving are both considered.

Description

Task scheduling method based on heredity and ant colony in cloud computing environment

Technical Field

The invention relates to the technical field of task scheduling algorithms in a cloud computing environment, in particular to a task scheduling method based on dynamic fusion of a genetic algorithm and an ant colony algorithm in the cloud computing environment.

Background

Cloud computing is a product of development and fusion of traditional computer technologies and network technologies, such as grid computing, distributed computing, network storage, virtualization, load balancing and the like. The system aims to integrate a plurality of relatively low-cost computing entities into a perfect system with strong computing power through a network, and distributes the strong computing power to end users by means of advanced business modes such as SaaS, PaaS, IaaS, MSP and the like. A core idea of cloud computing is to reduce the processing burden of a user terminal by continuously improving the processing capacity of the cloud, finally simplify the user terminal into a simple input and output device, and enjoy the powerful computing processing capacity of the cloud as required.

Cloud computing emphasizes resource sharing, heterogeneity and dynamic collaboration, which brings convenience to users and also puts higher requirements on task scheduling technology. The cloud computing task scheduling refers to a process of resource adjustment among different resource users in a specific cloud environment according to a certain resource use rule. Most of the conventional task scheduling strategies are to attempt to schedule tasks for applications inside a virtual machine by combining a certain scheduling strategy through a scheduling technology on the virtual machine level (that is, in a virtualization technology, multiple operating systems can be run simultaneously, and each operating system has multiple programs running therein, and each operating system runs on one virtual CPU or virtual machine).

The task amount faced by cloud computing is quite huge, and the problems of task scheduling and resource allocation are key points and difficulties faced by cloud computing efficiency. Much research work has been done at home and abroad on the task scheduling problem in the cloud computing environment. The document, namely a task scheduling algorithm based on an improved genetic algorithm in a cloud environment (Lijiafeng, Penghai computer application 2011, 31 (1): 184-. In the document cloud environment task scheduling research based on the improved ant colony algorithm (Wangyonggui Han Lian computer measurement and control 2011, 19 (5): 1203-1211), the solution time and the optimization precision are good. However, after the genetic algorithm is executed for a period of time, the convergence speed of the optimal solution is slow, and the accurate solution is low in efficiency; the ant colony algorithm lacks pheromones in the initial running stage, and further improvement of efficiency is limited.

In view of the above, it is necessary to provide a task scheduling method based on inheritance and ant colony in a cloud computing environment to solve the above problems.

Disclosure of Invention

The invention aims to provide a task scheduling method based on dynamic fusion of a genetic algorithm and an ant colony algorithm in a cloud computing environment. The method comprises the steps of generating a task scheduling initial solution by utilizing the rapid, random and global search capability of a genetic algorithm in the early stage of the algorithm, simultaneously initializing the solution to pheromone distribution, then seeking optimal task scheduling by utilizing the advantages of a positive feedback mechanism and high-efficiency convergence of an ant colony algorithm in the later stage, recording the evolutionary rate of a filial generation group in the iterative process of the genetic algorithm in the solving process, terminating the genetic algorithm and switching to the ant colony algorithm after continuous iteration within a preset iterative frequency range and the evolutionary theory of the filial generation is still lower than the given minimum evolutionary rate, thereby ensuring the optimal time of the fusion algorithm. Therefore, the algorithm draws the advantages of the genetic algorithm and the ant colony algorithm, overcomes the defects of the genetic algorithm and the ant colony algorithm, and dynamically fuses the genetic algorithm and the ant colony algorithm, so that the time and the efficiency of accurate solution are both considered.

The invention discloses a task scheduling method based on heredity and ant colony in a cloud computing environment, which comprises the following steps:

s1, initializing a population, and randomly placing M individuals of the population on N task nodes;

s2, selecting individuals X, Y according to a wheel disc type selection strategy;

s3, carrying out cross operation on the individuals X, Y according to the cross probability, and carrying out reverse mutation operation according to the mutation probability to generate a new population;

s4, updating the generated new cluster;

s5, judging whether the dynamic fusion condition is met, if yes, executing a step S6, and if not, returning to execute a step S2;

s6, initializing ant pheromones by using the genetically found optimal solution, and randomly placing ant numbers, namely task numbers M, on N resource nodes;

s7, calculating the probability of each ant moving to the next node, and moving each ant to the next node according to the probability;

s8, M ants traverse N resource nodes, and the optimal ant circle is used for pheromone updating;

s9, performing pheromone updating on all paths;

and S10, judging whether the ant end condition is met, if so, outputting an optimal solution, otherwise, returning to the step S7.

As a further improvement of the present invention, the calculation formulas of the cross probability and the mutation probability in step S3 are respectively:

$P_c=\left\{\begin{array}{cc}{P}_{c1}-\frac{(P_{c1}-P_{c2})(f^{\&prime;}-f_{\mathrm{avg}})}{f_{\max }-f_{\mathrm{avg}}},& f^{\&prime;}\&GreaterEqual;f_{\mathrm{avg}}\\ P_{c1},& f^{\&prime;}<f_{\mathrm{avg}}\end{array}\right.;$

$P_m=\left\{\begin{array}{cc}{P}_{m1}-\frac{(P_{m1}-P_{m2})(f_{\max }-f)}{f_{\max }-f_{\mathrm{avg}}},& f\&GreaterEqual;f_{\mathrm{avg}}\\ P_{m1},& f<f_{\mathrm{avg}}\end{array}\right.;$

wherein p is_cAnd p_mRespectively representing the cross probability and the mutation probability, f_maxIs the maximum fitness value of the population, f_avgIs the average fitness value of each generation population, f is the fitness value of the variant individual, and f' is the greater fitness value of the two individuals to be crossed.

As a further improvement of the present invention, the step S3 specifically includes:

generating a random number r;

if r is less than or equal to the mutation probability p_mPerforming a mutation operation on the selected two individuals and inserting the resulting two offspring into the new population;

if r is greater than the mutation probability p_mBut less than or equal to the probability of variation p_mAnd cross probability p_cPerforming a crossover operation on the selected two individuals and inserting the resulting two offspring into the new population;

otherwise, the breeding operation is carried out on the two selected individuals, and the two individuals are inserted into the new population unchanged.

As a further improvement of the present invention, before the step S1, the method further includes: calculating the adaptive value of each individual in the population; the step S4 is followed by: and calculating the adaptive value of each individual in the new population.

As a further improvement of the present invention, the step S6 of "initializing the ant pheromone with the genetically found optimal solution" specifically includes:

and selecting individuals with strong adaptability according to the individual adaptive values in the new colony as an optimal solution set for initializing the pheromone value of the ant colony algorithm.

As a further improvement of the present invention, the calculation formula of "the probability of each ant moving to the next node" in step S7 is as follows:

wherein,

is the probability of each ant moving to the next node, τ_j(t) pheromone concentration of the resource allocated by task i at time t; eta_jRepresenting an inherent property of the resource; α and β represent the importance of the resource pheromone and the resource intrinsic property, respectively.

As a further improvement of the present invention, the equation of "path pheromone update" in step S9 is:

${\mathrm{\&tau;}}_{\mathrm{ij}}(t+1)=\mathrm{\&rho;}.{\mathrm{\&tau;}}_{\mathrm{ij}}\left(t\right)+(1-\mathrm{\&rho;})\mathrm{\&Sigma;\&Delta;}{\mathrm{\&tau;}}_{\mathrm{ij}}^k\left(t\right);$

wherein the intensity of the pheromone track of the path (i, j) at the time t is tau_ijNumber of unit length trace pheromones ant k leaves on way (i, j)

The persistence of the track is rho and satisfies that rho is more than or equal to 0<1。

As a further improvement of the present invention, the method for determining that the dynamic fusion condition is satisfied in step S5 includes:

within the preset iteration number range, after continuous iteration, the evolutionary theory of the filial generation is still lower than the preset minimum evolutionary rate.

The invention has the beneficial effects that: the algorithm after dynamic fusion draws the advantages of heredity and ant colony, overcomes the defects of the heredity and ant colony, is applied to task scheduling in a cloud computing environment, and is superior to a single algorithm in the aspects of time efficiency, accurate solution efficiency and the like of solution.

Drawings

Fig. 1 is a schematic flow chart of a task scheduling method based on inheritance and ant colony in a cloud computing environment according to the present invention.

Detailed Description

The present invention will be described in detail below with reference to embodiments shown in the drawings. These embodiments are not intended to limit the present invention, and structural, methodological, or functional changes made by those skilled in the art according to these embodiments are included in the scope of the present invention.

Referring to fig. 1, the task scheduling method based on heredity and ant colony in a cloud computing environment according to the present invention includes the following steps:

s1, initializing a population, and randomly placing M individuals of the population on N task nodes;

s2, selecting individuals X, Y according to a wheel disc type selection strategy;

s3, carrying out cross operation on the individuals X, Y according to the cross probability, and carrying out reverse mutation operation according to the mutation probability to generate a new population;

s4, updating the generated new cluster;

s5, judging whether the dynamic fusion condition is met, if yes, executing a step S6, and if not, returning to execute a step S2;

s6, initializing ant pheromones by using the genetically found optimal solution, and randomly placing ant numbers, namely task numbers M, on N resource nodes;

s7, calculating the probability of each ant moving to the next node, and moving each ant to the next node according to the probability;

s8, M ants traverse N resource nodes, and the optimal ant circle is used for pheromone updating; s9, performing pheromone updating on all paths;

and S10, judging whether the ant end condition is met, if so, outputting an optimal solution, otherwise, returning to the step S7.

In step S3, the calculation formulas of the cross probability and the mutation probability are respectively:

$P_c=\left\{\begin{array}{cc}{P}_{c1}-\frac{(P_{c1}-P_{c2})(f^{\&prime;}-f_{\mathrm{avg}})}{f_{\max }-f_{\mathrm{avg}}},& f^{\&prime;}\&GreaterEqual;f_{\mathrm{avg}}\\ P_{c1},& f^{\&prime;}<f_{\mathrm{avg}}\end{array}\right.;$

$P_m=\left\{\begin{array}{cc}{P}_{m1}-\frac{(P_{m1}-P_{m2})\left(f_{\max }-f\right)}{f_{\max }-f_{\mathrm{avg}}},& f\&GreaterEqual;f_{\mathrm{avg}}\\ P_{m1},& f<f_{\mathrm{avg}}\end{array}\right.;$

p_cand p_mRespectively representing the cross probability and the mutation probability, f_maxIs the maximum fitness value of the population, f_avgIs the average fitness value of each generation population, f is the fitness value of the variant individual, and f' is the greater fitness value of the two individuals to be crossed.

Further, step S3 specifically includes:

generating a random number r;

if r is less than or equal to the mutation probability p_mPerforming a mutation operation on the selected two individuals and inserting the resulting two offspring into the new population;

if r is greater than the mutation probability p_mBut less than or equal to the probability of variation p_mAnd cross probability p_cPerforming a crossover operation on the selected two individuals and inserting the resulting two offspring into the new population;

otherwise, the breeding operation is carried out on the two selected individuals, and the two individuals are inserted into the new population unchanged.

Further, step S1 is preceded by: calculating the adaptive value of each individual in the population; step S4 is followed by: and calculating the adaptive value of each individual in the new population.

The "initializing ant pheromones with the genetically found optimal solution" in step S6 is specifically:

and selecting individuals with strong adaptability according to the individual adaptive values in the new colony as an optimal solution set for initializing the pheromone value of the ant colony algorithm. Preferably, the first 10% of individuals with strong adaptability can be selected.

Further, the calculation formula of "the probability of each ant moving to the next node" in step S7 is:

wherein,is the probability of each ant moving to the next node, τ_j(t) pheromone concentration of the resource allocated by task i at time t; eta_jRepresenting an inherent property of the resource; α and β represent the importance of the resource pheromone and the resource intrinsic property, respectively.

The equation of "path pheromone update" in step S9 is:

${\mathrm{\&tau;}}_{\mathrm{ij}}(t+1)=\mathrm{\&rho;}.{\mathrm{\&tau;}}_{\mathrm{ij}}\left(t\right)+(1-\mathrm{\&rho;})\mathrm{\&Sigma;\&Delta;}{\mathrm{\&tau;}}_{\mathrm{ij}}^k\left(t\right);$

wherein the intensity of the pheromone track of the path (i, j) at the time t is tau_ijNumber of trace elements per unit length left by ant k on way (i, j)

The persistence of the track is rho and satisfies that rho is more than or equal to 0<1。

In the present invention, an iteration number range and a minimum evolutionary rate are preset, and the method of determining that the dynamic fusion condition is satisfied in step S5 is:

within the preset iteration number range, after continuous iteration, the evolutionary theory of the filial generation is still lower than the preset minimum evolutionary rate.

The idea of dynamic fusion of the invention comprises:

1) the method comprises the steps of generating a task scheduling initial solution by utilizing the rapid, random and global search capability of a genetic algorithm in the early stage, meanwhile, initializing the solution into pheromone distribution, and seeking optimal task scheduling by utilizing the advantages of an ant colony algorithm positive feedback mechanism and high-efficiency convergence in the later stage;

2) and recording the evolutionary rate of the filial generation population in the iterative process of the genetic algorithm, and terminating the genetic algorithm and switching to the ant colony algorithm after continuous iteration within a preset iteration number range and the evolutionary theory of the filial generation is still lower than a given minimum evolutionary rate, thereby ensuring the optimal time of the fusion algorithm.

Setting an initial value of the pheromone of the ant colony algorithm, and setting the initial value of the pheromone on each path as a maximum value tau_maxObtaining a certain path pheromone by using a genetic algorithmThe initial value of the initial pheromone is set as:

${\mathrm{\&tau;}}_{\mathrm{ij}}^S={\mathrm{\&tau;}}_{\mathrm{ij}}^C+{\mathrm{\&tau;}}_{\mathrm{ij}}^G---\left(1\right)$

wherein,

based on a pheromone constant set to solve the problem,

is the pheromone value converted from the solving result of the genetic algorithm.

And (3) converting the solving result of the genetic algorithm into the ant colony algorithm pheromone: the invention selects 10% of individuals with the best fitness value in the population when the genetic algorithm is terminated as an optimal solution set of heredity. By task m_iIs allocated to resource n_jThe above probabilities are part of the ant colony algorithm initial pheromone. The calculation formula is as follows:

$\begin{array}{cc}{\mathrm{\&tau;}}_{\mathrm{ij}}^G=\frac{x}{y}\&times;{\mathrm{\&eta;}}_j& m_i\&Element;M---\left(2\right)\end{array}$

wherein x optimizes task m in solution set_iIs allocated to resource n_jThe total number of times, y, is the number of solutions in the optimized solution set, η_jIs a resource n_jThe intrinsic properties of (a).

The optimal fusion time of the genetic algorithm and the ant colony algorithm is determined by the end condition of the genetic algorithm, so that the minimum and maximum genetic iteration times, the minimum evolutionary rate and the continuous iteration times need to be set. The present invention preferably sets the genetic maximum minimum number of iterations to G_max=50 and G_min=15, minimum evolution rate and number of successive iterations G_minratio=3% and G_die=3。

Is provided with L_kThe length of the path taken by the kth ant in the cycle

Wherein Q is a constant if eta_ijFor the visibility of the side path, it is generally taken as 1/d_ijWhere d is_ijIs the length of the path (i, j), the relative importance of path visibility beta (beta ≧ 0), the relative importance of path trajectory alpha (alpha ≧ 0), U is the set of feasible vertices, the transition probability of ant k at time t is

Then

The following were used:

the ant circle model is a better algorithm for global optimization, and the pheromone track intensity of a path (i, j) at the time t is set to be tau_ijNumber of unit length trace pheromones ant k leaves on way (i, j)

Persistence of the track ρ (0 ≦ ρ)<1) Then the update equation of the track strength is:

${\mathrm{\&tau;}}_{\mathrm{ij}}(t+1)=\mathrm{\&rho;}.{\mathrm{\&tau;}}_{\mathrm{ij}}\left(t\right)+(1-\mathrm{\&rho;})\mathrm{\&Sigma;\&Delta;}{\mathrm{\&tau;}}_{\mathrm{ij}}^k\left(t\right)---\left(5\right)$

in the cloud environment, the pheromone needs to be represented by correlation coefficients of computing, communication capacity and the like of resources, and heuristic information in the ant colony algorithm needs to be represented by attributes inherent to the resources. Therefore, the probability of calculating the next task of the existing resources in the cloud computing environment is as follows:

in the formula, τ_j(t) pheromone concentration of the resource allocated by task i at time t; eta_jRepresenting an inherent property of the resource; α and β represent the importance of the resource pheromone and the resource intrinsic property, respectively.

Chromosome coding in genetic algorithms: and adopting an indirect resource-task coding mode, wherein the number of the subtasks is the length of the chromosome, and the value of each gene in the chromosome is the serial number of the resource acquired by the subtask corresponding to the position.

Generation of an initial population: distributing M tasks to N resources in a cloud environment, taking the cluster scale as S, and randomly generating S chromosomes by the system, wherein the length of the S chromosomes is N, and the gene value is randomly selected from [1, M ].

Selecting an operator: and adopting a wheel disc type selection strategy which is most widely applied in genetic algorithms.

Crossover operators and mutation operators: the cross operation determines the global search capability of the genetic algorithm, and is a main method for generating new individuals, while the mutation operation can improve the local search capability of the genetic algorithm, keep the diversity of the population and prevent the premature phenomenon. The cross probability function and the variant probability function are respectively:

$P_c=\left\{\begin{array}{cc}{P}_{c1}-\frac{(P_{c1}-P_{c2})(f^{\&prime;}-f_{\mathrm{avg}})}{f_{\max }-f_{\mathrm{avg}}},& f^{\&prime;}\&GreaterEqual;f_{\mathrm{avg}}\\ P_{c1},& f^{\&prime;}<f_{\mathrm{avg}}\end{array}\right.---\left(7\right)$

$P_m=\left\{\begin{array}{cc}{P}_{m1}-\frac{(P_{m1}-P_{m2})\left(f_{\max }-f\right)}{f_{\max }-f_{\mathrm{avg}}},& f\&GreaterEqual;f_{\mathrm{avg}}\\ P_{m1},& f<f_{\mathrm{avg}}\end{array}\right.---\left(8\right)$

in the above two formulae, p_cAnd p_mRespectively representing the cross probability and the mutation probability, f_maxIs the maximum fitness value of the population, f_avgIs the average fitness value of each generation population, f is the fitness value of the variant individual, and f' is the greater fitness value of the two individuals to be crossed.

One embodiment of the present invention comprises the steps of:

1. initializing a population P (0), and randomly placing M individuals of the population on N task nodes;

2. calculating the adaptive value of the individual in P (0);

3. repeatedly executing the following operations until a genetic algorithm end condition is met:

a) determining the selection probability pi of each individual in P (g) according to the individual adaptive value and the wheel disc type selection strategy;

b)For(k=0；k<N；k=k+2)

{

i. selecting two parents within P (g) according to the probability pi;

if (r < = Pm), performing mutation operations on the selected 2 parents, and inserting the resulting 2 offspring into the new population P (g + 1);

else if (r < = Pm + Pc), performing a crossover operation, and inserting the resulting 2 offspring into the new population P (g + 1);

else, performing a reproduction operation, inserting 2 parents into the new population P (g +1) unchanged;

}

c) calculating the fitness value of the individual in P (g +1), g = g + 1;

4. selecting the first 10% individuals with strong adaptability from P (g) as an optimization solution set for initializing the pheromone value of the ant colony algorithm;

5. for each optimal solution in the optimal solution set, initializing a pheromone value according to a conversion strategy from the solution of the genetic algorithm to the pheromone value;

6. randomly placing N ants on N task nodes, wherein each ant carries corresponding information;

7. calculating the probability of each ant in the optimizing process by utilizing the ant colony moving type (6);

8. updating the pheromone concentration of the resources distributed by the tasks according to the formula (5);

9. if the target is met, outputting a final optimization solution, otherwise, continuously executing 7;

according to the technical scheme, the task scheduling method based on the heredity and the ant colony dynamically fuses the genetic algorithm and the ant colony algorithm in the cloud computing environment, determines the optimal fusion time of the two fusion and converts the genetic solving result into the initialization of the ant colony algorithm pheromone. Meanwhile, the dynamic fusion algorithm is applied to task scheduling of cloud computing, so that task scheduling and resource allocation are more reasonable.

Compared with the prior art, the algorithm after dynamic fusion draws the advantages of heredity and ant colony, overcomes the defects of the heredity and the ant colony, is applied to task scheduling in a cloud computing environment, and is superior to a single algorithm in the aspects of solving time efficiency, accurate solution efficiency and the like.

It should be understood that although the present description refers to embodiments, not every embodiment contains only a single technical solution, and such description is for clarity only, and those skilled in the art should make the description as a whole, and the technical solutions in the embodiments can also be combined appropriately to form other embodiments understood by those skilled in the art.

The above-listed detailed description is only a specific description of a possible embodiment of the present invention, and they are not intended to limit the scope of the present invention, and equivalent embodiments or modifications made without departing from the technical spirit of the present invention should be included in the scope of the present invention.

Claims (8)

1. A task scheduling method based on heredity and ant colony in a cloud computing environment is characterized by comprising the following steps:

s1, initializing a population, and randomly placing M individuals of the population on N task nodes;

s2, selecting individuals X, Y according to a wheel disc type selection strategy;

s3, carrying out cross operation on the individuals X, Y according to the cross probability, and carrying out reverse mutation operation according to the mutation probability to generate a new population;

s4, updating the generated new cluster;

s5, judging whether the dynamic fusion condition is met, if yes, executing a step S6, and if not, returning to execute a step S2;

s6, initializing ant pheromones by using the genetically found optimal solution, and randomly placing ant numbers, namely task numbers M, on N resource nodes;

s7, calculating the probability of each ant moving to the next node, and moving each ant to the next node according to the probability;

s8, M ants traverse N resource nodes, and the optimal ant circle is used for pheromone updating;

s9, performing pheromone updating on all paths;

and S10, judging whether the ant end condition is met, if so, outputting an optimal solution, otherwise, returning to the step S7.

2. The method according to claim 1, wherein the calculation formulas of the cross probability and the mutation probability in step S3 are respectively:

$P_c=\left\{\begin{array}{cc}{P}_{c1}-\frac{(P_{c1}-P_{c2})(f^{\&prime;}-f_{\mathrm{avg}})}{f_{\max }-f_{\mathrm{avg}}},& f^{\&prime;}\&GreaterEqual;f_{\mathrm{avg}}\\ P_{c1},& f^{\&prime;}<f_{\mathrm{avg}}\end{array}\right.;$

$P_m=\left\{\begin{array}{cc}{P}_{m1}-\frac{(P_{m1}-P_{m2})(f_{\max }-f)}{f_{\max }-f_{\mathrm{avg}}},& f\&GreaterEqual;f_{\mathrm{avg}}\\ P_{m1},& f<f_{\mathrm{avg}}\end{array};\right.$

wherein p is_cAnd p_mRespectively representing the cross probability and the mutation probability, f_maxTo the maximum fitness value of the population，f_avgIs the average fitness value of each generation population, f is the fitness value of the variant individual, and f' is the greater fitness value of the two individuals to be crossed.

3. The method according to claim 2, wherein the step S3 specifically includes:

generating a random number r;

if r is less than or equal to the mutation probability p_mPerforming a mutation operation on the selected two individuals and inserting the resulting two offspring into the new population;

if r is greater than the mutation probability p_mBut less than or equal to the probability of variation p_mAnd cross probability p_cPerforming a crossover operation on the selected two individuals and inserting the resulting two offspring into the new population;

otherwise, the breeding operation is carried out on the two selected individuals, and the two individuals are inserted into the new population unchanged.

4. The method according to claim 1, wherein the step S1 is preceded by: calculating the adaptive value of each individual in the population; the step S4 is followed by: and calculating the adaptive value of each individual in the new population.

5. The method as claimed in claim 4, wherein the step S6 of initializing ant pheromones with the genetically found optimal solution is specifically as follows:

and selecting individuals with strong adaptability according to the individual adaptive values in the new colony as an optimal solution set for initializing the pheromone value of the ant colony algorithm.

6. The method as claimed in claim 1, wherein the calculation formula of the "probability of each ant moving to the next node" in the step S7 is as follows:

wherein,

is the probability of each ant moving to the next node, τ_j(t) pheromone concentration of the resource allocated by task i at time t; eta_jRepresenting an inherent property of the resource; α and β represent the importance of the resource pheromone and the resource intrinsic property, respectively.

7. The method according to claim 1, wherein the equation of "path pheromone update" in step S9 is:

${\mathrm{\&tau;}}_{\mathrm{ij}}(t+1)=\mathrm{\&rho;}.{\mathrm{\&tau;}}_{\mathrm{ij}}\left(t\right)+(1-\mathrm{\&rho;})\mathrm{\&Sigma;\&Delta;}{\mathrm{\&tau;}}_{\mathrm{ij}}^k\left(t\right);$

wherein the intensity of the pheromone track of the path (i, j) at the time t is tau_ijNumber of unit length trace pheromones ant k leaves on way (i, j)The persistence of the track is rho and satisfies that rho is more than or equal to 0<1。

8. The method according to claim 1, wherein the determination method of "dynamic fusion condition is satisfied" in step S5 is:

within the preset iteration number range, after continuous iteration, the evolutionary theory of the filial generation is still lower than the preset minimum evolutionary rate.

Images