Recruitment-Based Robotic Colony Allocation
Recruitment-Based Robotic Colony Allocation
复制标题
基于招募的机器人群体分配
DOI：
10.1007/978-3-030-05816-6_6
复制
发表时间：
2018
期刊：
影响因子：
--
通讯作者：
J. Adams
中科院分区：
文献类型：
--
作者：
Chloë Fleming;J. Adams
用于分布式现场操作的机器人群体需要基于时变任务要求分配资产的智能算法。农业应用，包括作物授粉，涉及动态的时间表是未知的先验，因此，一个有效的殖民地必须表征任务完成和感知的变化作出反应。昆虫群体依靠个体的观察来分配他们的劳动力任务，例如在最丰富的花丛中觅食。生物启发的策略，收集意见，同时提供现场覆盖制定和集成到一个殖民地招聘模型。策略的覆盖结果和效率进行了评估，不同的任务时间表，殖民地的位置，和多个殖民地，产生50-100%的授粉在所有的独立变量。将单个菌落分裂成多个小菌落提高了覆盖率和效率。
Robotic colonies for distributed field operations require intelligent algorithms for allocating assets based on time-varying task requirements. Agricultural applications, including crop pollination, involve dynamic schedules that are unknown a priori; thus, an effective colony must characterize task completion and respond to perceived changes. Insect colonies rely on individuals’ observations to allocate their workforces to tasks, such as foraging in the richest flower patches. Biologically-inspired strategies for gathering observations while providing field coverage were formulated and integrated into a colony recruitment model. The strategies’ coverage results and efficiency were evaluated with varying task schedules, colony placements, and multiple colonies, yielding 50–100% pollination across all independent variables. Splitting a single colony into multiple small colonies improved coverage and efficiency.
DOI：
10.1098/rspb.2013.2440
发表时间：
2014-01-22
期刊：
Proceedings. Biological sciences
影响因子：
--
作者：
Klatt BK;Holzschuh A;Westphal C;Clough Y;Smit I;Pawelzik E;Tscharntke T
通讯作者：
Tscharntke T