Unlock instant, AI-driven research and patent intelligence for your innovation.

Subspace elimination type random search optimization method

A technology of random search and optimization method, applied in the fields of instruments, artificial life, computing, etc., can solve problems such as poor real-time performance and large amount of calculation

Inactive Publication Date: 2018-11-23
JILIN UNIV
View PDF0 Cites 2 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0003] These swarm intelligence methods have a large amount of calculation, are more likely to fall into local optimal solutions, and have poor real-time performance.

Method used

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
View more

Image

Smart Image Click on the blue labels to locate them in the text.
Viewing Examples
Smart Image
  • Subspace elimination type random search optimization method
  • Subspace elimination type random search optimization method
  • Subspace elimination type random search optimization method

Examples

Experimental program
Comparison scheme
Effect test

Embodiment

[0165] Embodiment: The characteristic of this algorithm is illustrated by the application of typical functions to the optimization algorithm test

[0166] In the Matlab environment, the typical functions of the three optimization algorithms are tested, and the search speed of the optimal value of this algorithm is compared with that of the particle swarm optimization algorithm alone.

example 1

[0167] Example 1: Optimization of Booth function

[0168] 1) Objective function

[0169] The Booth function is a function of two arguments, the formula:

[0170] f(x,y)=-(((x-10)+2(y-10)-7) 2 +(2(x-10)+(y-10)-5) 2 )

[0171] The range of the function argument is x, y∈[0, 20], and the corresponding function image is drawn according to this range, such as image 3 , the maximum value of the function is 0, and the precision is greater than -e -8 stop searching.

[0172] 2) Simple use of standard particle swarm optimization algorithm

[0173] After adjusting the total number of particles of the standard particle swarm algorithm, the particle swarm is 490, and the number of iterations is 200. The search to achieve the specified accuracy is the fastest, and the average time of running 30 times is 1.37s.

[0174] 3) use the algorithm of the present invention to optimize

[0175] (1) Blocking of decision variable space

[0176] The number of segments on the X axis is 5, the n...

example 2

[0179] Example 2: Optimization of Schaffer function

[0180] 1) Objective function

[0181] The Schaffer function is a function of two independent variables, the formula:

[0182] f(x,y)=-((x-10) 2 +(y-10) 2 ) 0.25 ((sin(50((x-10) 2 +(y-10) 2 ) 0.1 ) 0.2 )+1.0)

[0183] The range of the function argument is x, y∈[0, 20], and the corresponding function image is drawn according to this range, such as Figure 4 , the maximum value of the function is 0, and the search stops when the specified precision is >-0.01.

[0184] 2) Simple use of standard particle swarm optimization algorithm

[0185]After adjusting the total number of particles of the standard particle swarm algorithm, the particle swarm is 360, and the number of iterations is 86. The search to achieve the specified accuracy is the fastest, and the average time of running 30 times is 0.50s.

[0186] 3) use the algorithm of the present invention to optimize

[0187] (1) Blocking of decision variable space

[...

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to View More

PUM

No PUM Login to View More

Abstract

The invention relates to a sub-space elimination type random search optimization method, and belongs to the field of random optimization. The method comprises the following steps of: dividing a decision variable space into sub-space blocks; preliminarily searching local optimum values in the sub-spaces via a certain search method; carrying out prioritizing according to a result of the preliminarysearch; leaving the potential sub-space blocks and eliminating non-potential sub-blocks according to searched local optimization solutions; and continuously searching the potential sub-blocks, and carrying out continuous elimination so as to finally search an optimum solution. In the search, the parts which do not possibly have optimum points are eliminated as soon as possible, and more search resources are put into the areas which possibly have optimum threshold values, so that benefit is brought to improve the search efficiency which comprises speed and correctness. The method has the advantages of being high in search efficiency, effectively reducing the calculation complexity, improving the timeliness and carrying out flexible selection between optimization indexes and the calculationcomplexity through parameter adjustment.

Description

technical field [0001] The invention relates to the field of stochastic optimization, in particular to a fast search method for an optimal solution oriented to optimization of a single objective function in a limited decision variable space, in particular to a subspace elimination random search optimization method. It is suitable for the decision variable space composed of digital variables, such as the fast solution of optimization problems in systems such as digital image processing and computer vision. The method can be widely used in the fields of image processing, object tracking, single-objective optimization of discrete systems in intelligent systems, and the like. Background technique [0002] Finding the optimal solution of a certain function (fitness function) is a typical problem of optimization, and it is also a research field with strong application. With the development of computer-based technology methods, computer-based numerical calculation methods and opti...

Claims

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to View More

Application Information

Patent Timeline
no application Login to View More
Patent Type & Authority Applications(China)
IPC IPC(8): G06N3/00
CPCG06N3/006
Inventor 秦俊黄玥王楚婷陈海鹏申铉京秦贵和
Owner JILIN UNIV