首页> 外文期刊>Artificial life and robotics >Genetic algorithms produce individual robotic rat pup behaviors that match Norway rat pup behaviors at multiple scales
【24h】

Genetic algorithms produce individual robotic rat pup behaviors that match Norway rat pup behaviors at multiple scales

机译:遗传算法可产生与多个规模的挪威大白鼠行为相匹配的单个机器人大白鼠行为

获取原文
获取原文并翻译 | 示例
           

摘要

We designed input/output cognitive architectures for individual robotic rat pups using genetic algorithms with the aim of achieving insight into Norway rat pup behavior. Our cognitive architectures were evolved using only metrics of Norway rat pup macro-behavior (e.g., percent of time spent in corners, along walls, and center of an arena during animal experiments). Robotic rat pups not only quantitatively matched Norway rat pups at the macro-level (by design), but additionally qualitatively matched Norway rat pup behavior at the micro-behavior scales (corner- snooping, punting). The complexity of the resulting deterministic input-output controllers may lend support to previous studies that show random-like control codes (possibly emerging from complex underlying input-output interactions, or brain-initiated spontaneous behavior) can produce apparently realistic rat pup behavior below a certain age.
机译:我们使用遗传算法设计了个体机器人幼崽的输入/输出认知架构,旨在深入了解挪威幼崽的行为。我们的认知架构仅使用挪威大鼠幼崽的宏观行为来衡量(例如,在动物实验中花费在角落,墙壁和竞技场中心的时间百分比)。机器人老鼠幼崽不仅在宏观水平上与挪威老鼠幼崽在数量上匹配(通过设计),而且在微观行为尺度上(角落侦探,打孔)在质量上与挪威老鼠幼崽的行为相匹配。结果确定性输入输出控制器的复杂性可能会为以前的研究提供支持,这些研究表明类似随机的控制代码(可能来自复杂的基础输入输出交互或大脑启动的自发行为)可以在以下条件下产生明显逼真的大鼠幼崽行为。一定年龄。

著录项

相似文献

  • 外文文献
  • 中文文献
  • 专利
获取原文

客服邮箱:kefu@zhangqiaokeyan.com

京公网安备:11010802029741号 ICP备案号:京ICP备15016152号-6 六维联合信息科技 (北京) 有限公司©版权所有
  • 客服微信

  • 服务号