CN116306919A - Large-Scale Multi-objective Combinatorial Optimization Method Based on Problem Recombination and Its Application - Google Patents

Abstract

The invention relates to a large-scale multi-objective combination optimization method based on problem recombination and application thereof, wherein the method comprises the following steps: randomly initializing a population; based on a decision variable clustering technology, dividing decision variables into convergence variables and diversity variables; adopting directional cross variation facing convergence, and combining a convergence environment selection mechanism to select an optimal individual as a new parent population; when the population encounters selection pressure, problem reconstruction is carried out, and large-scale multi-objective optimization is converted into single-objective optimization; adopting diversity-oriented directional cross variation and combining a diversity environment selection mechanism to select a population as a new parent population; and after the termination condition is met, outputting the parent population as an optimal solution set of the optimization target. Compared with the prior art, the complementary search strategy and the problem reconstruction strategy provided by the invention can solve the problem of large-scale multi-objective optimization, respectively process the problems of convergence and diversity in different optimization stages, and avoid the situation of sinking into local optimum.

Description

Large-Scale Multi-objective Combinatorial Optimization Method Based on Problem Recombination and Its Application

technical field

The invention relates to the field of multi-objective evolutionary calculation, in particular to a large-scale multi-objective combined evolutionary optimization method based on problem reorganization.

Background technique

A variety of multi-objective evolutionary algorithms have emerged in the past two decades, including Pareto-based multi-objective evolutionary algorithms (Multi Objective Evolutionary Algorithms, MOEAs), decomposition-based MOEAs and index-based MOEAs. Although most of the existing multi-objective evolutionary algorithms show good performance in solving multi-objective problems with a small number of decision variables, when solving multi-objective optimization problems (MOPs) with hundreds or even thousands of decision variables, Namely large-scale MOPs (LSMOPs), their performance drops dramatically. As the number of decision variables increases linearly, the volume (and thus complexity) of the search space will grow exponentially, causing the algorithm to converge prematurely to local optima or to very large regions. In recent years, academia and industry have successively proposed a variety of multi-objective evolutionary algorithm frameworks based on co-evolution framework, decision variable analysis and problem transformation to optimize LSMOPs, but there are still many problems to be solved, mainly in The following aspects:

LSMOEAs based on the co-evolutionary framework need to spend a lot of time analyzing decision variables to complete the grouping of decision variables. In addition, due to improper grouping, when there is a correlation between the sub-problems, it needs to be optimized repeatedly in sequence, and the performance of the algorithm will also be severely degraded. It is worth noting that the assumption of separability between decision variables is not always true. Therefore, the algorithm has limitations and is not suitable for solving large-scale MOPs where various decision variables interact.

Although LSMOEAs based on decision variable analysis reduce the scale of the problem to a certain extent through the classification of decision variables, there are fewer categories (convergent variables, diversity variables and mixed variables), and the sub-problems decomposed may still be large. In terms of scale, the overall search efficiency of the algorithm needs to be improved.

LSMOEAs based on problem transformation need to find a problem transformation function that guarantees as little information loss as possible after the original problem is transformed into a new problem. However, it is very difficult to find a perfect problem transformation function, and it is even impossible in particularly complex problems. In addition, since one weight value corresponds to a set of decision variables, the search of the decision space is incomplete, and the quality of the final solution obtained needs to be improved.

Chinese patent application CN114819040A discloses a dual-population co-evolution method based on dual search, which cannot handle multi-objective optimization problems with as many as 500 or even 1000 or more decision variables.

Contents of the invention

The purpose of the present invention is to provide a large-scale multi-objective combined optimization method based on problem recombination in order to overcome the defect that it is difficult to achieve a balance between convergence and diversity in the optimization process of the prior art.

The purpose of the present invention can be achieved through the following technical solutions:

As a first aspect of the present invention, there is provided a large-scale multi-objective combined optimization method based on problem reorganization, the method includes a convergence optimization stage and a diversity optimization stage, and the specific steps include:

Randomly initialize the population P;

Based on decision variable clustering technology, decision variables are divided into convergent variables and diversity variables;

Convergent optimization phase:

In the process of evolution, the convergence-oriented directional crossover mutation is adopted, combined with the convergence environment selection mechanism, to select the optimal individual as the parent population P';

When the population encounters selection pressure, it reconstructs the problem, converts large-scale multi-objective optimization into single-objective optimization, and reselects the optimal individual as the parent population P″;

Diversity optimization phase:

In the evolution process, the diversity-oriented directional cross-mutation is adopted, combined with the diversity environment selection mechanism, and the population is selected as the new parent population;

After satisfying the termination condition, the parent population is output as the optimal solution set of the optimization target.

Further, the specific evolution steps of the convergence optimization stage are as follows:

Using convergence-based binary competition selection, using crossover and mutation operators, select individuals from the current population to generate N offspring Q;

Calculate the convergence degree of each individual in the union of the current population P and the offspring Q, and select the best N individuals from the union as the new parent population P′;

Repeat the above process until the termination condition is met;

Carry out problem reconstruction, and convert large-scale multi-objective optimization into single-objective optimization through two-way weight vector association;

Producing offspring population A through differential evolution;

The environment selection mechanism based on the convergence strategy selects the best N individuals from the union of the offspring population A and the weight vector population Q′ as the new parent population P″.

Further, the convergent directional crossover operation is performed in units of mutually independent convergent variable groups;

The convergent directional mutation selects a variable among the convergent variables to mutate.

Further, the calculation formula of the degree of convergence C _d of the solution x is as follows:

Among them, M represents the number of targets.

Further, the problem reorganization strategy is specifically: through two-way weight vector association, the original large-scale multi-objective optimization problem is re-expressed as a single-objective optimization with relatively small weight variables, and a group of candidate solutions with good convergence and uniform distribution are used To guide the algorithm to search in the direction of the optimal set.

Further, the specific evolution steps of the problem restructuring strategy are as follows:

Weight vector association, select r solutions from the current population as a reference solution set, each reference solution is associated with two direction vectors V _l and V _u and two weight variables λ _r1 and λ _r2 ;

Construct subproblems from direction vectors and weight variables:

Z′(Λ)={z ₁₁ (λ ₁₁ ), z ₁₂ (λ ₁₂ ), . . . , z _r1 (λ _r1 )z _r2 (λ _r2 )}

Among them, Λ={λ ₁₁ ,λ ₁₂ ,...,λ _r1 ,λ _r2 } is the reconstructed decision space;

The target space is reconstructed. When the sub-problem is reconstructed, the optimization of the decision vector x in the original decision space is transformed into the optimization of the weight vector Λ in the reconstructed decision space. The new optimization problem is expressed as:

maximize G(Λ)=H(Z′(Λ))

Among them, H can be any performance index.

Further, the specific steps of constructing sub-problems according to the direction vector and weight variables include:

Compute the two direction vectors for each reference solution:

V _l =s ₁ -o

V _u =ts ₁

Among them, s ₁ ={x ₁ ,...,x _d } is the reference solution, and t and o are the upper and lower boundary points of X respectively;

Compute two weight vectors for each reference solution:

Among them, λ11 and λ12 are two weight variables, l _max =||to||;

The expression of the constructor problem is:

Further, the specific evolution steps of the diversity optimization stage are as follows:

Select individuals from the current population P″ using binary competition selection based on crowding distance;

Use the individual selected above to only cross and mutate the diversity-related variables to generate N offspring Q";

According to the Pareto advantage and crowding distance, select half of the population from the union of the current population P" and the offspring Q" as the new parent population P"';

Repeat the above process until the termination condition is satisfied, and output the parent population P"' as the optimal solution set of the optimization goal.

Further, the directional crossover of the diversity performs the crossover operation in units of all diversity variables;

The directional variation of the diversity selects a variable among the diversity variables to mutate.

As a second aspect of the present invention, an application of the large-scale multi-objective combination optimization method based on problem reorganization as described above is provided, and the application scenarios of the method include resource scheduling and integration optimization of large-scale data centers; specifically Application steps include:

Taking the indicators including energy consumption, resource consumption, data communication flow and task completion delay as the problem optimization goal, and taking the remaining available computing capacity of the server as the constraint condition, a multi-objective constrained optimization model for large-scale data center resource scheduling and integration problems is established;

The number of virtual machines to be scheduled is the chromosome length, and the number of virtual machines is the gene position to realize the coding of the population individual;

Initialize the population P based on the encoding method, and perform the iterative optimization of the population according to the large-scale multi-objective combination optimization method based on problem recombination as described above, until the termination condition of the algorithm is met, and the output includes energy consumption, resource consumption, and data communication flow. Multiple conflicting indicators with the task completion delay are the Pareto optimal solution set of the optimization objective.

Compared with the prior art, the present invention has the following beneficial effects:

1) The large-scale multi-objective optimization method provided by the present invention, based on the problem reconstruction strategy, can convert the large-scale multi-objective optimization problem into the optimization of weight vectors, thereby solving large-scale problems.

2) Different from the existing large-scale multi-objective evolutionary algorithm that maintains the convergence and diversity of the population during an iterative optimization process of the population, the complementary search strategy proposed by the present invention handles the convergence and diversity problems separately in different optimization stages . In the first stage, the population quickly approaches the Pareto front, ignoring the diversity of the population. After the population converges, the second stage adopts the decision variable clustering method to emphasize the diversity of the population, so as to avoid falling into the local optimal situation.

3) Different from the existing multi-objective evolutionary algorithm that uses methods such as crowding degree in environment selection, the present invention proposes the concept of convergence degree, which is used to quantitatively judge the convergence of candidate solutions, so as to select the solution with better convergence .

4) The diversity evolution of the present invention can adopt other methods of multi-objective optimization algorithms, so that different instantiation algorithms can be designed for different types of multi-objective optimization problems, and the scalability is strong.

Description of drawings

Fig. 1 is a schematic flow chart of the large-scale multi-objective combinatorial evolution method based on problem reorganization of the present invention;

Fig. 2 is the approximate Pareto front schematic diagram of instantiation algorithm of the present invention on test function LSMOF2;

Fig. 3 is the approximate Pareto front schematic diagram of instantiation algorithm of the present invention on test function ZDT1;

Fig. 4 is an exemplary application flowchart of the large-scale multi-objective combination optimization method provided by the present invention.

Detailed ways

The present invention will be described in detail below in conjunction with the accompanying drawings and specific embodiments. This embodiment is carried out on the premise of the technical solution of the present invention, and detailed implementation and specific operation process are given, but the protection scope of the present invention is not limited to the following embodiments.

Example 1

The multi-objective optimization method proposed by the present invention handles convergence and diversity problems in different optimization stages through complementary search strategies. In the first stage, the population quickly approaches the Pareto front, ignoring the diversity of the population. When selection pressure is encountered, the Pareto optimal set (PS) is directly tracked through problem reformulation. The algorithm first obtains a set of reference directions in the decision space and associates them with a set of weight variables for locating the PS, and then transforms the original large-scale multi-objective optimization problem into a low-dimensional single-objective optimization problem. In the second stage, the optimal solutions are evenly distributed on the approximate Pareto optimal frontier using the multi-objective evolutionary algorithm. In the first stage, the population quickly approaches the Pareto front, ignoring the diversity of the population. After the population converges, the second stage adopts the decision variable clustering method to emphasize the diversity of the population, so as to avoid falling into the local optimal situation.

As shown in Figure 1, a large-scale multi-objective combinatorial optimization method based on problem recombination is divided into two phases to optimize the population in this framework: the convergence-oriented stage (Convergence-oriented Stage, CS) and the diversity-oriented stage (Diversity-oriented stage). Stage, DS). The convergence optimization stage (CS) first adopts the binary competition selection based on the degree of convergence, and uses crossover and mutation operators to select individuals from the current population P to generate N offspring Q, where N is the population size. Then, the degree of convergence of the union of P and Q is determined by formula (1). Finally, select the best N individuals from the union as the new parental population P ^′ , and when the population encounters selection pressure, use the method of problem reconstruction to maintain a set of weight vectors, and convert large-scale multi-objective optimization into For single-objective optimization, the offspring population A is produced through differential evolution, and the optimal N individuals are collectively selected from the offspring population A and the weight vector population Q′ based on the environment selection mechanism of the convergence strategy as the new parent population P ″, to further improve the convergence of the population; the diversity optimization stage (DS) adopts the binary competition selection based on the crowding distance to select individuals from the current population P″, and then use the above-mentioned selected individuals to only crossover and mutate the diversity-related variables , to generate offspring Q″. Finally, according to Pareto advantage and crowding distance, select half of the population from the union of parent P″ and offspring Q″ as the new parent population. The evolution steps of the specific framework are as follows:

Randomly initialize the population P, the population size is N;

Based on the decision variable clustering technique, the decision variables are divided into two categories (convergence variables and diversity variables).

Using convergence-based binary competition selection, using crossover and mutation operators, select individuals from the current population P to generate N offspring Q, where N is the population size;

Calculate the degree of convergence of each individual of P and Q, and select the best N individuals from the union as the new parental population P′; the calculation formula for the degree of convergence Cd of solving x is:

Among them, M represents the number of targets. Convergence is defined as the sum of the target values of x. For minimizing MOP, the smaller the value of Cd, the better the convergence of solution x.

After satisfying the termination condition, the problem-based reconstruction strategy is used to select r solutions from the current population P ^' as the reference solution set. Then use Equation 2 to calculate the two direction vectors for each reference solution.

Vl _l =s ₁ -o

V _{u u} =ts ₁ (2)

Wherein, s ₁ ={x ₁ ,…,x _d } is the reference solution, and o and t are the upper and lower boundary points of X respectively.

Compute the two weight vectors for each reference solution using Equation 3.

Wherein, λ ₁₁ and λ ₁₂ are two weight variables, l _max =||to||.

A set of reference solutions of size r is chosen, and once each reference solution is associated with two direction vectors and two weight variables, a total of 2r subproblems can be constructed.

The subproblems are Z′(∧)={z ₁₁ (λ ₁₁ ), z ₁₂ (λ ₁₂ ),…, z _r1 (λ _r1 )z _r2 (λ _r2 )}, where ∧={λ ₁₁ , λ ₁₂ , ..., λ _r1 , λ _r2 } is the reconstructed decision space.

Once the subproblems are reconstructed, the optimization of the decision vector x in the original decision space is transformed into the optimization of the weight vector. Accordingly, the target space can be reduced, and the new optimization problem can be reformulated as

maximize G(∧)＝H(Z′(∧)) (5)

Among them, H can be any performance index.

The offspring population A is produced through differential evolution, and the optimal N individuals are selected from the union of A and Q′ as the new parent population P″ based on the environment selection mechanism of the convergence strategy.

After satisfying the termination condition, enter DS, first, select individuals from the current population P″ by using binary competition selection based on crowding distance, and then use the above-mentioned selected individuals to crossover and mutate only the diversity-related variables to generate N offspring Q ". Finally, according to the Pareto advantage and crowding distance, select half of the population from the union of P″ and Q″ as the new parental population P″′, and output the parental population P″′ as the optimal optimization target after the termination condition is met. Excellent solution set.

1) CS adopts convergence-oriented directional crossover mutation in the evolution process, and combines the convergence environment selection mechanism to maintain the convergence of the population;

2) DS adopts diversity-oriented directional cross-mutation in the evolution process, and combines the diversity environment selection mechanism to maintain population diversity.

As a preference, the aforementioned convergent and diverse crossover and mutation operators are respectively implemented through convergence calculation and decision variable analysis, and the specific steps are as follows:

1) Clustering of decision variables: This method uses k-means clustering to divide variables into two categories (convergence variables and diversity variables).

2) Classification of convergent variables: the calculation formula (1) for the degree of convergence Cd of the solution x.

3) Convergent crossover is performed in units of independent convergent variable groups; diversity crossover is performed in units of all diversity variables;

4) Convergent mutation selects variables from convergent variables to mutate; diversity mutation selects variables from diverse variables to mutate.

The problem reconstruction strategy used by CS in the evolution process is as follows:

Problem Reframing Strategy: This strategy is used to address the selection pressure caused by large-scale decision variables. It is specifically divided into three steps: the first step, weight vector association, selects r solutions from the current population P as a reference solution set, and each reference solution is associated with two direction vectors and two weight variables. The second step is to construct sub-problems, according to formulas 4 and 5 to construct sub-problems Z′(∧)={z ₁₁ (λ ₁₁ ), z ₁₂ (λ ₁₂ ),…, z _r1 (λ _r1 ) z _r2 (λ _r2 )}, where ∧={λ ₁₁ ,λ ₁₂ ,…,λ _r1 ,λ _r2 } is the reconstructed decision space. The third step is the reconstruction of the target space. Once the subproblems are reconstructed, the optimization of the decision vector x in the original decision space is transformed into the optimization of the weight vector ∧ in the reconstructed decision space. The new optimization problem can be formulated as Equation 5.

In the stage of diversity evolution, other multi-objective optimization algorithms can be used, so that different instantiation algorithms can be designed for different types of multi-objective optimization problems, and the scalability is strong.

Table 1 shows the mean values of IGD obtained by the two multi-objective evolutionary algorithms on nine large-scale test cases. In order to reduce the impact of random errors on the calculation results, each algorithm in this embodiment is independently run 30 times in each calculation example, and the average value of the IGD index obtained by each algorithm on each calculation example is calculated. The IGD indicator measures the distance between the real Pareto front and the approximate Pareto obtained by the algorithm. Generally, the smaller the IGD index value, the better the convergence and diversity of the algorithm.

Table 1

Problem m D. The algorithm instantiated by this framework LSMOF LSMOP1 2 1000 2.6390e-1(1.63e-2)+ 6.2856e-1 (2.77e-2) LSMOP2 2 1000 1.1624e-2(3.74e-4)+ 1.8858e-2 (3.47e-4) LSMOP3 2 1000 1.5680e+0(1.61e-3)+ 1.5729e+0(4.09e-4) LSMOP4 2 1000 2.5755e-2(7.90e-4)+ 3.8112e-2 (1.95e-3) LSMOP5 2 1000 6.8558e-1(4.40e-2)+ 7.4209e-1 (3.39e-16) LSMOP6 2 1000 5.6957e-1 (1.62e-1)- 3.1236e-1 (4.03e-4) LSMOP7 2 1000 1.5087e+0(4.30e-4)- 1.5083e+0(7.70e-4) LSMOP8 2 1000 5.3281e-1(2.17e-1)+ 7.4209e-1 (3.39e-16) LSMOP9 2 1000 5.1013e-1(4.32e-3)+ 8.0686e-1 (8.94e-4)

Fig. 2 is that the method of the present invention obtains on 2-target, the approximate Pareto front on the LSMOP2 benchmark function of 1000 decision variables, and Fig. 3 is that the method of the present invention obtains on 2-target, the ZDT1 benchmark function of 1000 decision variables Approximate Pareto front.

Example 2

As a second aspect of the present invention, an application of the method described in the above embodiments is provided, and a typical application scenario of the method is resource scheduling and integration optimization of a large-scale data center. The specific application process is shown in Figure 4:

Step 1: Firstly, using energy consumption, resource consumption, data communication flow, task completion delay and other indicators as the optimization objectives of the problem, and taking the remaining available computing capacity of the server as the constraint condition, establish a multi-objective problem for large-scale data center resource scheduling and integration Constrained optimization model;

Step 2: Then, take the number of virtual machines to be scheduled as the chromosome length, and the virtual machine number as the gene position to realize the coding of the population individual;

Step 3: Finally, initialize the population P based on the encoding method, and perform iterative optimization of the population according to the large-scale multi-objective combination optimization method based on problem recombination shown in Figure 1, until the termination condition of the algorithm is met, and the output is in terms of energy consumption and resource consumption , data communication flow, task completion delay and other conflicting indicators as the optimal Pareto solution set for the optimization goal.

The preferred specific embodiments of the present invention have been described in detail above. It should be understood that those skilled in the art can make many modifications and changes according to the concept of the present invention without creative efforts. Therefore, all technical solutions that can be obtained by those skilled in the art based on the concept of the present invention through logical analysis, reasoning or limited experiments on the basis of the prior art shall be within the scope of protection defined by the claims.

Claims (10)

1. The large-scale multi-objective combination optimization method based on the problem recombination is characterized by comprising a convergence optimization stage and a diversity optimization stage, and comprises the following specific steps of:

randomly initializing a population P;

based on a decision variable clustering technology, dividing decision variables into convergence variables and diversity variables;

convergence optimization stage:

in the evolution process, adopting directional cross variation facing convergence, and combining with a convergence environment selection mechanism, selecting an optimal individual as a parent population P';

when the population encounters selection pressure, problem reconstruction is carried out, large-scale multi-objective optimization is converted into single-objective optimization, and the optimal individuals are reselected as parent populations P';

diversity optimization stage:

in the evolution process, adopting diversity-oriented directional cross variation, and combining a diversity environment selection mechanism to select a population as a new parent population;

and after the termination condition is met, outputting the parent population as an optimal solution set of the optimization target.

2. The problem recombination-based large-scale multi-objective combinatorial optimization method of claim 1, wherein the specific evolution steps of the convergence optimization stage are as follows:

using binary competition selection based on convergence, and selecting individuals from the current population to generate N sub-generations Q by using crossover and mutation operators;

calculating the convergence degree of each individual in the current population P and the offspring Q, and selecting the optimal N individuals from the union as a new parent population P';

repeating the above process until the termination condition is satisfied;

performing problem reconstruction, and converting large-scale multi-objective optimization into single-objective optimization through bi-directional weight vector association;

producing a progeny population a by differential evolution;

the convergence policy based environmental selection mechanism selects the optimal N individuals from the union of the offspring population a and the weight vector population Q' as the new parent population P ".

3. The method for large-scale multi-objective combinatorial optimization based on problem recombination according to claim 2, wherein,

the directional crossing of the convergence performs crossing operation by taking independent convergence variable groups as units;

the directional variation of the convergence selects a variable from the convergence variables to vary.

4. The problem recombination-based massive multi-objective combinatorial optimization method according to claim 2, wherein the convergence degree C of solution x _d The calculation formula is as follows:

where M represents the target number.

5. The large-scale multi-objective combination optimization method based on problem recombination according to claim 1, wherein the problem recombination strategy is specifically: through bi-directional weight vector association, the original large-scale multi-objective optimization problem is re-expressed into single-objective optimization with relatively smaller weight variables, and a group of candidate solutions with good convergence and uniform distribution are used for guiding the algorithm to search the direction of the optimal set.

6. The method for large-scale multi-objective combinatorial optimization based on problem recombination according to claim 5, wherein the specific evolution steps of the problem recombination strategy are as follows:

weight vector correlation, selecting r solutions from the current population as a reference solution set, each reference solution being associated with two direction vectors V _l And V _u And two weight variables lambda _r1 And lambda (lambda) _r2 Associating;

constructing a sub-problem according to the direction vector and the weight variable:

Z′(∧)＝(z ₁₁ (λ ₁₁ )，z ₁₂ (λ ₁₂ )，...，z _r1 (λ _r1 )z _r2 (λ _r2 )}

wherein, lambada= { lambda ₁₁ ，λ ₁₂ ，...，λ _r1 ，λ _r2 A reconstructed decision space;

target space reconstruction, when the sub-problem is reconstructed, the optimization of the decision vector x in the original decision space is converted into the optimization of the weight vector Λ in the reconstructed decision space, and the new optimization problem is expressed as follows:

maximize G(Λ)＝H(Z′(Λ))

where H may be any performance indicator.

7. The method for massive multi-objective combinatorial optimization based on problem recombination according to claim 6, wherein the specific step of constructing the sub-problem according to the direction vector and the weight variable comprises:

two direction vectors for each reference solution are calculated:

V _l ＝s ₁ -o

V _u ＝t-s ₁

wherein s is ₁ ＝{x ₁ ，...，x _d The } is a reference solution, and t and o are upper and lower boundary points of X respectively;

two weight vectors for each reference solution are calculated:

where λ11 and λ12 are two weight variables, l _max ＝||t-o||；

The expression of the constructor problem is:

8. the problem recombination-based large-scale multi-objective combinatorial optimization method of claim 1, wherein the specific evolution steps of the diversity optimization stage are as follows:

selecting individuals from the current population P' using a binary competition selection based on crowding distances;

utilizing the selected individuals to only cross and mutate the diversity related variables to generate N offspring Q';

selecting half of the current population P ' and offspring Q ' from the union set as a new parent population P ' according to the pareto dominance and crowding distance;

repeating the above process until the termination condition is satisfied, and outputting the parent population P' as the optimal solution set of the optimization target.

9. The method for large-scale multi-objective combinatorial optimization based on problem recombination according to claim 8, wherein,

the directional crossing of the diversity performs crossing operation by taking all diversity variables as units;

the directed variation of diversity selects variables among the diversity variables for variation.

10. Application of the problem recombination-based large-scale multi-objective combination optimization method according to any of claims 1-9, characterized in that the application scenario of the method comprises resource scheduling and integration optimization of a large-scale data center; the specific application steps comprise:

taking indexes including energy consumption, resource loss, data communication flow and task completion time delay as problem optimization targets, taking the residual available computing capacity of a server as constraint conditions, and establishing a multi-target constraint optimization model of a large-scale data center resource scheduling and integration problem;

taking the number of virtual machines to be scheduled as the length of a chromosome, and taking the number of the virtual machines as the gene position, so as to realize the coding of population individuals;

initializing a population P based on a coding mode, performing iterative optimization of the population according to the large-scale multi-objective combination optimization method based on the problem recombination as claimed in any one of claims 1-9 until the termination condition of an algorithm is met, and outputting a Pareto optimal solution set with a plurality of conflict indexes including energy consumption, resource loss, data communication flow and task completion time delay as optimization objectives.

Images