注册 登录  
 加关注
   显示下一条  |  关闭
温馨提示!由于新浪微博认证机制调整,您的新浪微博帐号绑定已过期,请重新绑定!立即重新绑定新浪微博》  |  关闭

洪荒世界_江波的科幻天地

关于科学,还有未来的想象

 
 
 

日志

 
 

有趣的实验,使用机器人模拟行为遗传  

2007-02-28 08:56:48|  分类: 默认分类 |  标签: |举报 |字号 订阅

  下载LOFTER 我的照片书  |
 
科学试验总是很有趣,比坐着空想出来的东西还要有趣。
全文如下:

Robot swarms 'evolve' effectivecommunication

Robots that artificially evolve ways to communicate with oneanother have been demonstrated by Swiss researchers. Theexperiments suggest that simulated evolution could be a useful toolfor those designing of swarms of robots.

Roboticists Dario Floreano, Sara Mitri, and Stéphane Magnenatat the Swiss Federal Institute of Technology in Lausannecollaborated with biologist Laurent Keller from the University ofLausanne.

They first evolved colonies of robots in software then testeddifferent strategies on real bots, called s-bots. Both simulatedand real robots were set loose in an arena containing two types ofobjects – one classified as "food" and another designated "poison"– both lit up red.

Each bot had a built-in attraction to food and aversion topoison. They also have a randomly-generated set of parameters,dubbed "genomes" that define the way they move, process sensoryinformation, and how they flash their own blue lights.

"They start with completely random behaviour," Keller explains."All they can do is discriminate food from poison." The robots cansee both food or poison from a distance of several metres but canonly tell them apart when almost touching.

Unnatural selection

The simulated bots were allowed to explore their surroundings.The genomes of the bots that found food and avoided poison mostefficiently were recombined, mimicking biological naturalselection. Their "genomes" were combined and randomised in a waydesigned to mimick mating and mutation and used this to create thenext generation robot.

In each experiment, five hundred generations were evolved thisway, under different selective pressures. "Under some conditions,sophisticated communication evolved," says Keller. "We saw coloniesthat used their lights to signal when they found food and othersthat used signals to communicate they had found poison."

Cooperative communication evolved when selective success wasjudged at the group level – when many robots displayed efficientbehaviour – or when the genomes of the robots were most similar –like biological relatives.

Real s-bots with genomes evolved through simulations displayedthe same behaviour. This video shows the robots in action. The sames-bots have previously been used to test different cooperationstrategies (see Robot swarm works together to shift heavy objects).

Genes that promote sharing spread when an individual's survivalis tied to that of the whole group – like in bees or ants. Theyalso do well in populations of close relatives, since helping kinhelps the same genes reach the next generation.

Misleading behaviour

In contrast, in simulations with bots not closely related andwhen group performance was less important, some robots evolvedmisleading behaviour. Some even produced signals that lured otherbots away far from food.

Further experiments involving real robots will be used toinvestigate ways that evolution could be used as a practical designtool. Keller also plans to test what happens when evolved andun-evolved bots mix.

"These are significant findings for the biologically-inspiredrobotics community," says Noel Sharkey, who is also evolving robotbehaviour at Sheffield University, UK. Learning how to evolverobots instead of writing their behaviour from scratch couldultimately lead to more sophisticated behaviour, he says.

However, artificial evolution has yet to mimic all the powersshown in biology. Sharkey points out that his robots already startwith a means of communication. "In real biology no suchcommunication channel exists initially and has to be created ratherthan discovered," he notes.

  评论这张
 
阅读(73)| 评论(0)
推荐 转载

历史上的今天

评论

<#--最新日志,群博日志--> <#--推荐日志--> <#--引用记录--> <#--博主推荐--> <#--随机阅读--> <#--首页推荐--> <#--历史上的今天--> <#--被推荐日志--> <#--上一篇,下一篇--> <#-- 热度 --> <#-- 网易新闻广告 --> <#--右边模块结构--> <#--评论模块结构--> <#--引用模块结构--> <#--博主发起的投票-->
 
 
 
 
 
 
 
 
 
 
 
 
 
 

页脚

网易公司版权所有 ©1997-2017