论文部分内容阅读
针对室内通风环境下的气味源定位问题,提出了一种基于模拟退火策略的单机器人气味源定位算法.受流场的控制,室内气味源释放的烟羽除了具有蜿蜒和间歇特性外,还会在涡流区域形成局部浓度极大值.本文将气味源定位看作是一种动态函数寻优问题,使用模拟退火策略求取浓度分布函数的最优解,即气味源所在位置.算法不依赖风信息,从而可以减少流场波动造成的影响.同时通过研究气味浓度与气味源距离之间的关系,提出了一种与气味源距离呈近似线性关系的模拟退火目标函数.真实室内通风环境下的实验表明,使用本文提出的算法,机器人能够在8m×6m区域内跟踪烟羽并定位气味源,平均定位时间约为10 min,且在搜索过程中可以有效地跳出局部极大值.
In order to solve the problem of odor source localization in indoor ventilation environment, a single robot scent source localization algorithm based on simulated annealing strategy is proposed.Under the control of flow field, smoke plume released by indoor odor source has meandering and intermittent characteristics The local concentration maximum will be formed in the vortex area.This paper regards the odor source localization as a dynamic function optimization problem, and uses the simulated annealing strategy to get the optimal solution of the concentration distribution function, ie, the location of the odor source.The algorithm does not depend on Wind information, which can reduce the impact of flow field fluctuations.At the same time through the study of the relationship between odor concentration and the distance between the odor source, and the distance from the odor source is an approximate linear simulated annealing objective function. Real indoor ventilation environment Experiments show that using the algorithm proposed in this paper, the robot can track the plume and locate the odor source in an area of 8m × 6m. The average positioning time is about 10 min, and the local maxima can be effectively extracted during the search.