CN103793745A - Distributed particle swarm optimization method - Google Patents

Abstract

The invention discloses a distributed particle swarm optimization method. The distributed particle swarm optimization method comprises the following steps that a high-dimension particle swarm function is divided into n low-dimension particle swarm sub-functions according to the importance of dimensionality, where n is an integer which is larger than or equal to 2; different sub-problems can be processed at different client sides in a distributed mode through a cloud server terminal, each calculating unit processes the corresponding low-dimension particle swarm sub-function through a particular swarm integrated algorithm, and therefore the optimal value of each low-dimension particle swarm sub-function corresponding to the corresponding dimensionality is obtained; optimal values corresponding to the dimensionalities of all the low-dimension particle swarm sub-functions are integrated, and therefore the optimal value of the high-dimension particle swarm function; the high-dimension particle swarm function integrated optimal value is used as the basis, iteration is conducted on the high-dimension particle swarm function through the particle swarm integrated algorithm, and therefore a distributed particle swarm optimization value is generated. According to the distributed particle swarm optimization method, the high-dimension particle swarm function is divided into the n low-dimension particle swarm sub-functions, all the low-dimension particle swarm sub-functions are processed simultaneously, as a result, the complexity of distributed particle swarm optimization is simplified, running time is shortened, and the efficiency of distributed particle swarm optimization is improved.

Description

A kind of distributed particle group optimizing method

Technical field

The present invention relates to information data analysis technical field, relate in particular to a kind of distributed particle group optimizing method.

Background technology

Particle swarm optimization algorithm (Particle Swarm Optimization, PSO) be one representative in multi-objective optimization algorithm, it is to be subject to the inspiration of some biological phenomenon in biology subject and a kind of relatively simple heuristic optimization algorithm based on probabilistic search that proposes.The behavior that this algorithm simulation flock of birds is looked for food, makes colony reach optimum object by the mutual cooperation between individuality, is a kind of optimization method based on colony intelligence.

The advantage of particle swarm optimization algorithm is that computation model is easy to describe, the parameter that need to debug is fewer, realize simple operation speed fast, without centralized control constraint, strong robustness, but be easily absorbed in local optimum and lose globally optimal solution, also referred to as precocious phenomenon, because the individuality in most of colony all evolves near optimal location, whole middle group position cannot be further optimized, in the time carrying out higher-dimension particle group optimizing, optimize difficulty large, complex structure, and optimize overlong time.

Summary of the invention

The embodiment of the present invention provides a kind of distributed particle group optimizing method, in order to solve the problems such as in the distributed particle group optimizing process of higher-dimension, processing is complicated, optimization efficiency is low.

Distributed particle group optimizing method, comprises the following steps:

Higher-dimension population function is divided into n low-dimensional population subfunction, and wherein n is more than or equal to 2 integer;

Utilize population Integrated Algorithm respectively each low-dimensional population subfunction to be processed, obtain the optimal value of the corresponding each dimension of each low-dimensional population subfunction;

Optimal value to the corresponding each dimension of each low-dimensional population subfunction is integrated, and obtains higher-dimension population function and integrates optimal value;

Integrate optimal value as basis take higher-dimension population function, utilize population Integrated Algorithm to carry out iteration to described higher-dimension population function, generate distributed particle group optimizing value.

Preferably, higher-dimension population function is divided into n low-dimensional population subfunction by described step, and wherein n is more than or equal to 2 integer, comprising:

Significance level according to dimension to higher-dimension population function, is divided into n low-dimensional population subfunction by higher-dimension population function, wherein, n >=2, n is integer.

Preferably, higher-dimension population function is divided into n low-dimensional population subfunction by described step, wherein, n >=2, n is integer, comprising:

Low-dimensional population subfunction number

wherein, N is the number of dimensions of described higher-dimension population function.

Preferably, described step utilizes population Integrated Algorithm respectively each low-dimensional population subfunction to be processed, and obtains the optimal value of the corresponding each dimension of each low-dimensional population subfunction, comprising:

Significance level according to each low-dimensional population subfunction to distributed particle group optimizing, arranges described low-dimensional population subfunction according to significance level order from big to small;

Each low-dimensional population subfunction is carried out to iteration, and the iterations of described low-dimensional population subfunction is:

$P_i=\frac{C}{2(i-1)}$

Wherein, i >=2, are the order sequence number of low-dimensional population subfunction, and Pi is the iterations of i low-dimensional population subfunction, and C is the iterations of the first low-dimensional population subfunction.

Preferably, described step utilizes population Integrated Algorithm respectively each low-dimensional population subfunction to be processed, and obtains the optimal value of the corresponding each dimension of each low-dimensional population subfunction, comprising:

Initialization underlying parameter, utilizes described low-dimensional population subfunction to obtain the initial position of particle, initial velocity;

Utilize respectively primary importance update method and second place update method to carry out iteration to described low-dimensional population subfunction, obtain adaptive value corresponding to described primary importance update method, current optimal location and historical optimal location, adaptive value, current optimal location and historical optimal location that described second place update method is corresponding;

The adaptive value that adaptive value corresponding primary importance update method is corresponding with second place update method compares, using current optimal location corresponding population Integrated Algorithm larger adaptive value and historical optimal location as underlying parameter;

Judge whether to reach maximum iteration time, if, output optimal location, repeat if not described primary importance update method and the second place update method utilized respectively described low-dimensional population subfunction is carried out to iteration, obtain adaptive value corresponding to described primary importance update method, current optimal location and historical optimal location, adaptive value corresponding to described second place update method, current optimal location and historical optimal location are to the adaptive value corresponding primary importance update method adaptive value corresponding with second place update method compared, using current optimal location corresponding population Integrated Algorithm larger adaptive value and historical optimal location as underlying parameter.

Preferably, described primary importance update method adopts First Speed update method;

Described second place update method adopts second speed update method.

Preferably, described step is integrated optimal value as basis take higher-dimension population function, utilizes population Integrated Algorithm to carry out iteration to described higher-dimension population function, generates distributed particle group optimizing value, comprising:

Each low-dimensional problem is deployed in the asynchronous distributed arithmetic that carries out of different computing units, connect and realize the unified management of attribute (dimension) data by Web cloud network, integrate optimal value as basis take higher-dimension population function, utilize population Integrated Algorithm to carry out described higher-dimension population function

inferior iteration, generates distributed particle group optimizing value, and wherein, C is the iterations of the first low-dimensional population subfunction.

Preferably, utilize different particle cluster algorithms respectively each low-dimensional population subfunction to be processed, obtain the optimal value of the corresponding each dimension of each low-dimensional population subfunction, comprising:

Utilize population Integrated Algorithm each low-dimensional population subfunction to be processed respectively simultaneously, obtain the optimal value of the corresponding each dimension of each low-dimensional population subfunction.

Preferably, described step utilizes population Integrated Algorithm respectively each low-dimensional population subfunction to be processed, and obtains the optimal value of the corresponding each dimension of each low-dimensional population subfunction, comprising:

The optimal value of the corresponding each dimension of each low-dimensional population subfunction is uploaded to network;

Described step is integrated the optimal value of the corresponding each dimension of each low-dimensional population subfunction, and synthetic higher-dimension population optimized value, comprising:

Utilize network to integrate the optimal value of the corresponding each dimension of each low-dimensional population subfunction, synthetic higher-dimension population optimized value.

In the embodiment of the present invention, by higher-dimension population function being divided into n low-dimensional population subfunction, simplifying the complexity of distributed particle group optimizing, and each low-dimensional population subfunction is processed simultaneously, reduce working time, improve the efficiency of distributed particle group optimizing.

Other features and advantages of the present invention will be set forth in the following description, and, partly from instructions, become apparent, or understand by implementing the present invention.Object of the present invention and other advantages can be realized and be obtained by specifically noted structure in write instructions, claims and accompanying drawing.

Below by drawings and Examples, technical scheme of the present invention is described in further detail.

Accompanying drawing explanation

Accompanying drawing is used to provide a further understanding of the present invention, and forms a part for instructions, for explaining the present invention, is not construed as limiting the invention together with embodiments of the present invention.

In the accompanying drawings:

Fig. 1 is the process flow diagram of an embodiment of the distributed particle group optimizing method of the present invention;

Fig. 2 is the process flow diagram of distributed another embodiment of particle group optimizing method of the present invention;

Fig. 3 is the logical flow chart of distributed another embodiment of particle group optimizing method of the present invention.

Embodiment

Below in conjunction with accompanying drawing, the preferred embodiments of the present invention are described, should be appreciated that preferred embodiment described herein, only for description and interpretation the present invention, is not intended to limit the present invention.

Referring to Fig. 1, be the process flow diagram of an embodiment of the distributed particle group optimizing method of the present invention, the method comprises:

Step 101: higher-dimension population function is divided into n low-dimensional population subfunction, and wherein n is more than or equal to 2 integer.

Wherein, this step can be divided in module and carry out at the function of the calculation processing apparatus such as computing machine, computing unit is deployed in high in the clouds, adopt the registration of Web Service encapsulation mode to issue, each computing unit can asynchronously serve (Software-as-a-service) according to software independently, solve corresponding subproblem, divide and can divide the significance level of higher-dimension population function according to dimension for higher-dimension population function, low-dimensional population subfunction number can be set to wherein, N is the number of dimensions of described higher-dimension population function, those skilled in the art also can be as required or concrete condition set.

In the present embodiment, by higher-dimension population function being divided into n low-dimensional population subfunction, reduce calculating and the treatment capacity of particle group optimizing, reduced system computing complexity.

Step 102: utilize population Integrated Algorithm respectively each low-dimensional population subfunction to be processed, obtain the optimal value of the corresponding each dimension of each low-dimensional population subfunction.

Wherein, this step can be carried out in the first optimal value acquisition module of the calculation processing apparatus such as computing machine, can utilize population Integrated Algorithm each low-dimensional population subfunction to be processed respectively simultaneously, obtain the optimal value of the corresponding each dimension of each low-dimensional population subfunction, and the optimal value of the corresponding each dimension of each low-dimensional population subfunction is uploaded to cloud network and store.

Significance level according to each low-dimensional population subfunction to distributed particle group optimizing, arranges described low-dimensional population subfunction according to significance level order from big to small;

Each low-dimensional population subfunction is carried out to iteration, and the iterations of described low-dimensional population subfunction is:

$P_i=\frac{C}{2(i-1)}$

Wherein, i >=2, are the order sequence number of low-dimensional population subfunction, and Pi is the iterations of i low-dimensional population subfunction, and C is the iterations of the first low-dimensional population subfunction.

The weight of the first low-dimensional population subfunction is the highest, need the number of times of iteration maximum, thereby can guarantee that it obtains top quality solution and at utmost reduces the error effect to this source problem, the weight of n low-dimensional population subfunction is minimum, needs the least number of times of iteration.

In the present embodiment, by each low-dimensional population subfunction is processed simultaneously, improve the efficiency that population is optimized, optimized thereby realized real-time population.

Step 103: the optimal value to the corresponding each dimension of each low-dimensional population subfunction is integrated, obtains higher-dimension population function and integrates optimal value.

Wherein, this step can be carried out in the optimal value integrate module of the calculation processing apparatus such as computing machine, for each low-dimensional population subfunction in the enterprising row operation of different computing units, obtain the optimal value of the corresponding each dimension of each low-dimensional population subfunction, then can realize mutual information interaction and integrate by network, obtain higher-dimension population function and integrate optimal value, realize Fractal distribution formula and calculate, those skilled in the art can freely choose the optimal value of the corresponding each dimension of method of integration and integrate.

Step 104: integrate optimal value as basis take higher-dimension population function, utilize population Integrated Algorithm to carry out iteration to described higher-dimension population function, generate distributed particle group optimizing value.

Wherein, this step can be carried out in the distributed particle group optimizing value generation module of the calculation processing apparatus such as computing machine, for fitness between balance subproblem solution, coordinate each dimension, on the basis of higher-dimension population function integration optimal value, carry out global optimization calculating, can integrate optimal value as basis take higher-dimension population function, utilize population Integrated Algorithm to carry out described higher-dimension population function inferior iteration, generates distributed particle group optimizing value, and wherein, C is the iterations of the first low-dimensional population subfunction.

In the present embodiment, by higher-dimension population function being divided into n low-dimensional population subfunction, and each low-dimensional population subfunction is processed simultaneously, simplified the complexity of distributed particle group optimizing, reduce working time, improve the efficiency of distributed particle group optimizing.

Participate in Fig. 2, for the process flow diagram of distributed another embodiment of particle group optimizing method of the present invention, in the method, utilize population Integrated Algorithm respectively each low-dimensional population subfunction to be processed, the optimal value that obtains the corresponding each dimension of each low-dimensional population subfunction, comprising:

Step 201: initialization underlying parameter, utilizes described low-dimensional population subfunction to obtain the initial position of particle, initial velocity.

Step 202: utilize respectively primary importance update method and second place update method to carry out iteration to described low-dimensional population subfunction, obtain adaptive value corresponding to described primary importance update method, current optimal location and historical optimal location, adaptive value, current optimal location and historical optimal location that described second place update method is corresponding.

Wherein, in the time carrying out iteration first, can utilize respectively primary importance update method and second place update method to carry out iteration to described low-dimensional population subfunction according to initial position and initial velocity.Primary importance update method adopts First Speed update method; Described second place update method adopts second speed update method.

Step 203: the adaptive value that adaptive value corresponding primary importance update method is corresponding with second place update method compares, using current optimal location corresponding population Integrated Algorithm larger adaptive value and historical optimal location as underlying parameter.

Step 204: judge whether to reach maximum iteration time, if so, and output optimal location, repeating step 202 is to step 204 if not.

In the present embodiment, mutually coordinate to carry out particle group optimizing by utilizing two kinds of update methods, when locally optimal solution appears being absorbed in a kind of particle swarm optimization wherein, can jump out optimum solution by contrasting the result that another kind of method obtains, greatly reduce the possibility that is absorbed in locally optimal solution.

Referring to Fig. 3, for the logical flow chart of distributed another embodiment of particle group optimizing method of the present invention, utilize population Integrated Algorithm respectively each low-dimensional population subfunction to be processed, obtain the optimal value of the corresponding each dimension of each low-dimensional population subfunction, comprising:

Step 301: initialization underlying parameter, utilizes described low-dimensional population subfunction to obtain the initial position of particle, initial velocity.

Step 302: judge whether to reach maximum iteration time, enter if not step 303, if enter step 309.

Step 303: utilize primary importance update method to carry out iteration to described low-dimensional population subfunction, obtain adaptive value F1 corresponding to described primary importance update method, current optimal location and historical optimal location.

Step 304: utilize second place update method to carry out iteration to described low-dimensional population subfunction, obtain adaptive value F2 corresponding to described second place update method, current optimal location and historical optimal location.

Step 305: judge whether adaptive value F1 is greater than adaptive value F2.If so, enter step 306, if not, enter step 307,

Step 306: using current optimal location corresponding adaptive value F1 and historical optimal location as underlying parameter.

Step 307: using current optimal location corresponding adaptive value F2 and historical optimal location as underlying parameter.

Wherein, step 305-307 is both using current optimal location corresponding population Integrated Algorithm larger adaptive value and historical optimal location as underlying parameter.

Step 308: utilize underlying parameter, use respectively primary importance update method and second place update method described low-dimensional population subfunction to be carried out to iteration, repeating step 302 to 308.

Step 309: relatively two kinds of optimal values that method obtains, getting the preferably position coordinates of is optimal location, output optimal location.

In the present embodiment, by Fractal distribution higher-dimension population is optimized, adopt in point whole process that dimension is calculated than the not minimum saving of calculating of fractal model inferior computing time, in having reduced optimization complexity, improve optimization efficiency, and adopt two kinds of update methods to utilize population set computing to low-dimensional population function, when locally optimal solution appears being absorbed in a kind of particle swarm optimization wherein, can jump out optimum solution by contrasting the result that another kind of method obtains, can reduce the possibility that is absorbed in locally optimal solution.

The present embodiment is with TP _mO-1trial function is example explanation:

Trial function TP _mO-1: $F_1\left(x\right)=\frac{1}{N}\underset{i=1}{\overset{N}{\mathrm{\Σ}}}{x}_i^2,F_2\left(x\right)=\frac{1}{N}\underset{i=1}{\overset{N}{\mathrm{\Σ}}}{(x_i-2)}^2,$ Suppose N=60, according to

this 60 dimension problem is divided into 5 low-dimensional population subfunction: P ₁, P ₂, P ₃, P ₄, P ₅.

P ₁solve 1-12 dimension population subfunction, trial function becomes: $F_1\left(x\right)=\frac{1}{12}\underset{i=1}{\overset{12}{\mathrm{\Σ}}}{x}_i^2,F_2\left(x\right)=\frac{1}{12}\underset{i=1}{\overset{12}{\mathrm{\Σ}}}{(x_i-2)}^2$

P ₂solve 13-24 dimension population subfunction, trial function becomes: $F_1\left(x\right)=\frac{1}{12}\underset{i=13}{\overset{24}{\mathrm{\Σ}}}{x}_i^2,F_2\left(x\right)=\frac{1}{12}\underset{i=13}{\overset{24}{\mathrm{\Σ}}}{(x_i-2)}^2$

P ₃solve 25-36 dimension population subfunction, trial function becomes: $F_1\left(x\right)=\frac{1}{12}\underset{i=25}{\overset{36}{\mathrm{\Σ}}}{x}_i^2,F_2\left(x\right)=\frac{1}{12}\underset{i=25}{\overset{36}{\mathrm{\Σ}}}{(x_i-2)}^2$

P ₄solve 37-48 dimension population subfunction, trial function becomes:

P ₅solve 49-60 dimension population subfunction, trial function becomes: $F_1\left(x\right)=\frac{1}{12}\underset{i=49}{\overset{60}{\mathrm{\Σ}}}{x}_i^2,F_2\left(x\right)=\frac{1}{12}\underset{i=49}{\overset{60}{\mathrm{\Σ}}}{(x_i-2)}^2$

Calculate these 5 population subfunctions are calculated at different computing units, application population Integrated Algorithm, false code is as table 1

Table 1

1. speed more new formula be:

$\begin{array}{c}{V}_i^{k+1}={\mathrm{random}}_3\×V_i^k+c_1\×{\mathrm{random}}_1\×({\mathrm{pbest}}_i-X_i^k)+c_2\×{\mathrm{random}}_2\×\\ (1-\frac{k}{{\mathrm{iter}}_{\max }})(\mathrm{lbest}-X_i^k)+c_2\×{\mathrm{random}}_3\×\frac{k}{{\mathrm{iter}}_{\max }}\×(\mathrm{gbest}-X_i^k).\end{array}$

2. speed more new formula be:

$w=w_{\min }\×{\left(\frac{w_{\max }}{w_{\min }}\right)}^{1/(1+c_3\×\frac{k}{{\mathrm{iter}}_{\max }})}$

$\begin{array}{c}{V}_i^{k+1}=w\×V_i^k+c_1\×{\mathrm{random}}_1\×({\mathrm{pbest}}_i-X_i^k)\\ +c_2\×{\mathrm{random}}_2\×(\mathrm{gbest}-X_i^k)\end{array}$

3. position more new formula be:

$X_i^{k+1}=X_i^k+V_i^{k+1}$

Separate the position obtaining after computing, deposits the database of networking in, is designated as Pos ₁, Pos ₂, Pos ₃, Pos ₄, Pos ₅, 60 dimensions that form higher-dimension population subfunction by integration are separated Pos, in new arithmetic element initialization speed, again adopt population Integrated Algorithm to carry out a small amount of computing, equilibrium degree between each dimension of balance, and gained is result.

Those skilled in the art should understand, embodiments of the invention can be provided as method, system or computer program.Therefore, the present invention can adopt complete hardware implementation example, completely implement software example or the form in conjunction with the embodiment of software and hardware aspect.And the present invention can adopt the form at one or more upper computer programs of implementing of computer-usable storage medium (including but not limited to magnetic disk memory and optical memory etc.) that wherein include computer usable program code.

The present invention is with reference to describing according to process flow diagram and/or the block scheme of the method for the embodiment of the present invention, equipment (system) and computer program.Should understand can be by the flow process in each flow process in computer program instructions realization flow figure and/or block scheme and/or square frame and process flow diagram and/or block scheme and/or the combination of square frame.Can provide these computer program instructions to the processor of multi-purpose computer, special purpose computer, Embedded Processor or other programmable data processing device to produce a machine, the instruction that makes to carry out by the processor of computing machine or other programmable data processing device produces the device for realizing the function of specifying at flow process of process flow diagram or multiple flow process and/or square frame of block scheme or multiple square frame.

These computer program instructions also can be stored in energy vectoring computer or the computer-readable memory of other programmable data processing device with ad hoc fashion work, the instruction that makes to be stored in this computer-readable memory produces the manufacture that comprises command device, and this command device is realized the function of specifying in flow process of process flow diagram or multiple flow process and/or square frame of block scheme or multiple square frame.

These computer program instructions also can be loaded in computing machine or other programmable data processing device, make to carry out sequence of operations step to produce computer implemented processing on computing machine or other programmable devices, thereby the instruction of carrying out is provided for realizing the step of the function of specifying in flow process of process flow diagram or multiple flow process and/or square frame of block scheme or multiple square frame on computing machine or other programmable devices.

Obviously, those skilled in the art can carry out various changes and modification and not depart from the spirit and scope of the present invention the present invention.Like this, if within of the present invention these are revised and modification belongs to the scope of the claims in the present invention and equivalent technologies thereof, the present invention is also intended to comprise these changes and modification interior.

Claims (9)

1. a distributed particle group optimizing method, is characterized in that, comprises the following steps:

Higher-dimension population function is divided into n low-dimensional population subfunction, and wherein n is more than or equal to 2 integer;

Utilize population Integrated Algorithm respectively each low-dimensional population subfunction to be processed, obtain the optimal value of the corresponding each dimension of each low-dimensional population subfunction;

Optimal value to the corresponding each dimension of each low-dimensional population subfunction is integrated, and obtains higher-dimension population function and integrates optimal value;

Integrate optimal value as basis take higher-dimension population function, utilize population Integrated Algorithm to carry out iteration to described higher-dimension population function, generate distributed particle group optimizing value.

2. the method for claim 1, is characterized in that, higher-dimension population function is divided into n low-dimensional population subfunction by described step, and wherein n is more than or equal to 2 integer, comprising:

Significance level according to dimension to higher-dimension population function, is divided into n low-dimensional population subfunction by higher-dimension population function, wherein, n >=2, n is integer.

3. the method for claim 1, is characterized in that, higher-dimension population function is divided into n low-dimensional population subfunction by described step, wherein, n >=2, n is integer, comprising:

Low-dimensional population subfunction number

wherein, N is the number of dimensions of described higher-dimension population function.

4. method as claimed in claim 2, is characterized in that, described step utilizes population Integrated Algorithm respectively each low-dimensional population subfunction to be processed, and obtains the optimal value of the corresponding each dimension of each low-dimensional population subfunction, comprising:

Significance level according to each low-dimensional population subfunction to distributed particle group optimizing, arranges described low-dimensional population subfunction according to significance level order from big to small;

Each low-dimensional population subfunction is carried out to iteration, and the iterations of described low-dimensional population subfunction is: $P_i=\frac{C}{2(i-1)}$

Wherein, i >=2, are the order sequence number of low-dimensional population subfunction, and Pi is the iterations of i low-dimensional population subfunction, and C is the iterations of the first low-dimensional population subfunction.

5. method as claimed in claim 2, is characterized in that, described step utilizes population Integrated Algorithm respectively each low-dimensional population subfunction to be processed, and obtains the optimal value of the corresponding each dimension of each low-dimensional population subfunction, comprising:

Initialization underlying parameter, utilizes described low-dimensional population subfunction to obtain the initial position of particle, initial velocity;

Utilize respectively primary importance update method and second place update method to carry out iteration to described low-dimensional population subfunction, obtain adaptive value corresponding to described primary importance update method, current optimal location and historical optimal location, adaptive value, current optimal location and historical optimal location that described second place update method is corresponding;

The adaptive value that adaptive value corresponding primary importance update method is corresponding with second place update method compares, using current optimal location corresponding population Integrated Algorithm larger adaptive value and historical optimal location as underlying parameter;

Judge whether to reach maximum iteration time, if, output optimal location, repeat if not described primary importance update method and the second place update method utilized respectively described low-dimensional population subfunction is carried out to iteration, obtain adaptive value corresponding to described primary importance update method, current optimal location and historical optimal location, adaptive value corresponding to described second place update method, current optimal location and historical optimal location are to the adaptive value corresponding primary importance update method adaptive value corresponding with second place update method compared, using current optimal location corresponding population Integrated Algorithm larger adaptive value and historical optimal location as underlying parameter.

6. method as claimed in claim 5, is characterized in that,

Described primary importance update method adopts First Speed update method;

Described second place update method adopts second speed update method.

7. the method for claim 1, is characterized in that, described step is integrated optimal value as basis take higher-dimension population function, utilizes population Integrated Algorithm to carry out iteration to described higher-dimension population function, generates distributed particle group optimizing value, comprising:

Integrating optimal value take higher-dimension population function utilizes population Integrated Algorithm to carry out described higher-dimension population function as basis

inferior iteration, generates distributed particle group optimizing value, and wherein, C is the iterations of the first low-dimensional population subfunction.

8. the method for claim 1, is characterized in that, utilizes population Integrated Algorithm respectively each low-dimensional population subfunction to be processed, and obtains the optimal value of the corresponding each dimension of each low-dimensional population subfunction, comprising:

Utilize population Integrated Algorithm each low-dimensional population subfunction to be processed respectively simultaneously, obtain the optimal value of the corresponding each dimension of each low-dimensional population subfunction.

9. the method for claim 1, is characterized in that, described step utilizes population Integrated Algorithm respectively each low-dimensional population subfunction to be processed, and obtains the optimal value of the corresponding each dimension of each low-dimensional population subfunction, comprising:

The optimal value of the corresponding each dimension of each low-dimensional population subfunction is uploaded to network;

Described step is integrated the optimal value of the corresponding each dimension of each low-dimensional population subfunction, and synthetic higher-dimension population optimized value, comprising:

Utilize network to integrate the optimal value of the corresponding each dimension of each low-dimensional population subfunction, synthetic higher-dimension population optimized value.

Images