﻿ 更多范文-Ant colony algorithm-51Due留学教育

24小时客服

#### 关于我们

51Due提供Essay，Paper，Report，Assignment等学科作业的代写与辅导，同时涵盖Personal Statement，转学申请等留学文书代写。

51Due将让你达成学业目标
51Due将让你达成学业目标
51Due将让你达成学业目标
51Due将让你达成学业目标

# Ant colony algorithm

2018-09-29 来源: 51due教员组 类别: 更多范文

In nature, the ant group showed a highly structured organization, ants have shown far beyond the ability of a single individual, scientists of ants foraging behavior, to build their nests and task allocation ability of the study found that ants have the unique ability to control their own surroundings, such as Cross as early as 1989 S through the famous build experiment to determine the role of pheromone for ants foraging process guidance, and also lay a foundation for later model of ant colony algorithm.

Ant colony algorithm originates from the simulation of ant colony behavior in nature, but there is still a certain difference between it and the real ant individual. The purpose of abstracting the ant individual in nature is to describe the natural behavior of ant colony more conveniently, and at the same time throw out the factors irrelevant to the problem modeling. In this way, the individual ants can be regarded as some intelligent agents, and they can communicate and influence each other through certain mechanisms, and also complete the construction process of the simple solution of the problem together.

Ant colony algorithm is based on the abstraction of three - dimensional space in nature. At the same time, the space for ant foraging in nature is a continuous two-dimensional space, and most of the problem models in computer simulation belong to discrete events. Therefore, it is necessary to discretize the problem to a solution space composed of points. The feasibility of this abstract process lies in that although the ant is moving in the continuous plane, its movement process is composed of discrete points, so the abstract process of problem space only improves the granularity of discretization, which is not in conflict with the foraging mechanism of the ant itself. Graph structure is often used to describe the problem space in most application problems.

Foraging between ant colony in food and nest in the process of constructing a specific space, in this space there are a large number of ant colony inherent information, such information can guide the ant movement direction in this space, in solving the problem of optimization, artificial ants search path on the plane node corresponds to the process of understanding the structure of the process; In the process of artificial ant movement, the path between the planar nodes provides guidance information, which is equivalent to pheromone. The artificial ant determines the direction of the next step according to a certain probability according to the concentration of pheromone on the path, and so on, it reaches the target node after a certain period of time, so that the feasible solution can be obtained.

In nature, the ant pheromone volatilization is a continuous process, is the continuous function related to time, and in a computer need to processing of the process, generally in the ant colony algorithm to step after a certain time or complete an iteration time as fixed time parameters, pheromone accumulation will over time units according to certain proportion after attenuation, so by the discrete points between the pheromone model can fully simulate the mechanism of ants foraging in the nature.

Through the abstraction of various factors in the foraging process of ants, the collective behavior of ants can be highly simulated, but this simulation needs to go through certain iterations in the evolutionary process, which consumes a lot of time. However, in practical application, certain requirements are often put forward for the execution time of the algorithm. Therefore, when the ant selects the next step in the discrete point space, a random process is used as the induction, and an heuristic factor or an heuristic function is introduced. In the process of solving the ant colony algorithm, an initial guidance information is usually set according to the characteristics of the final solution. The convergence of the ant colony algorithm can be accelerated by using the heuristic function, and its advantages can be better demonstrated in solving practical engineering problems.

Based on individual ants, problem space, foraging path, pheromone model and stimulating factor of abstraction, we can set up a complete model of basic ant colony algorithm a problem space description of ant colony algorithm is usually done by graph structure, using ant colony algorithm of solving process is a self-organizing process, any individual ants are without the guidance of the global information.

Ant individuals are the basic units of ant colony algorithm. They are self-adaptive and can dynamically accumulate knowledge, which can be obtained through the communication between ant individuals or the perception of the surrounding environment on the walking path. The core of ant colony algorithm is the cooperation and distribution of individual. In the solving process, the selection of ant individual is usually guided by the random decision mechanism and mutual coordination mechanism.

Ants in the ant colony algorithm, can according to the experience knowledge and environmental information organization and constantly learning, to reconstruct the knowledge base of itself, so the ant colony algorithm to realize the natural evolution, have very strong self-educated ability, changes in the environment and the learning ability of the algorithm are a pair of interacting factors, the uncertainties of the environmental change and strengthen the unpredictability of the algorithm and result in the implementation process.

Since 1996, five years, ant colony algorithm has received great attention of academia, has entered a period of rapid development, in 1998, hosted by Dorigo M the first seminar of ant colony algorithm in Belgium, the meeting with the boom of ant colony algorithm research, 2000, Dorigo M et al., summarizes the achievements and applications of ant colony algorithm has been made, and published the first paper reviews studies of ant colony algorithm, which laid a foundation for subsequent research. Gutjahr WJ proved the convergence of the ant colony algorithm in 1999 and 2000, and simplified the ant colony algorithm to the process of walking on the directed graph. On this basis, the convergence of the graph search ant colony algorithm was analyzed theoretically.

The research on ant colony algorithm in China started at the end of last century. The earliest research on ant colony algorithm was carried out in 1997 by Dr. Zhang jihui, professor xu xin and professor at Northeastern University. After entering the 21st century, a large number of scholars have devoted themselves to the study of ant colony algorithm. Scholars at home and abroad have applied it to practical work in many fields.

In conclusion, the basic ant colony algorithm is briefly explained. Ant colony algorithm is introduced to the individual, the problem space, path, pheromone volatilization and a series of problem, ant colony algorithm is the result of people the biological ant colony foraging behavior observation and simulation, compared with the traditional optimization method, the ant colony algorithm in solving all kinds of optimization has the very strong adaptability and robustness, suitable for all kinds of combinatorial optimization problems. The ant colony algorithm has been widely used in many scientific research and manufacturing fields because of its simple parameter setting and easy implementation.

51due留学教育原创版权郑重声明：原创assignment代写范文源自编辑创作，未经官方许可，网站谢绝转载。对于侵权行为，未经同意的情况下，51Due有权追究法律责任。主要业务有assignment代写、essay代写、paper代写服务。

51due为留学生提供最好的assignment代写服务，亲们可以进入主页了解和获取更多assignment代写范文 提供美国作业代写服务，详情可以咨询我们的客服QQ：800020041。