Bio-inspired decision-making and control: From honeybees and neurons to network design
Bio-inspired decision-making and control: From honeybees and neurons to network design
复制标题
DOI：
10.23919/acc.2017.7963250
复制
发表时间：
2017-05
期刊：
影响因子：
--
通讯作者：
Vaibhav Srivastava;Naomi Ehrich Leonard
中科院分区：
文献类型：
--
作者：
Vaibhav Srivastava;Naomi Ehrich Leonard
我们提出了非线性确定性模型和线性随机模型之间的替代品，连接不同的蜜蜂和神经元的生物群体的决策。使用这些模型，我们解释了生物群体，分散控制和有限的传感和通信，选择最高质量的替代品，抛硬币几乎相等的替代品，最佳平衡速度和准确性，保持鲁棒性面对不确定性，并利用异质性。这些显着的行为的动机，我们提出了一个通用的基于代理的模型的设计和控制网络动态与生物群体中观察到的有利功能。
We present nonlinear deterministic models and linear stochastic models of decision-making between alternatives that connect biological groups as diverse as honeybees and neurons. Using these models we explain how biological groups, with decentralized control and limited sensing and communication, select the highest quality alternative, flip a coin for nearly equal alternatives, optimally balance speed and accuracy, maintain robustness in the face of uncertainty, and leverage heterogeneity. Motivated by these remarkable behaviors, we present a generalizable agent-based model for the design and control of network dynamics with the advantageous features observed in the biological groups.