【作者】 王影；

【导师】 孟庆浩；

【作者基本信息】 天津大学 ， 控制科学与工程， 2017， 硕士

【摘要】 有毒、有害气体的泄漏严重威胁着人们的生命和财产安全,因此开展泄漏源定位研究显得非常重要。气味(或气体)泄漏源定位研究主要包括烟羽发现、烟羽追踪和气味源确认三个部分。以往的工作中有关气味源确认的研究很少,且均基于地面机器人或二维传感器网络(采集的数据为二维平面上的气味信息)。针对现实环境下的空中泄漏源(例如架空的输气管道破损),在三维环境中对气味源确认问题进行研究是非常必要的。本文将质量散度和质量通量概念引入气味源确认问题,重点解决如何利用三维多传感器节点采集的气味浓度和风速信息判断某可疑位置是否为气味源,主要开展了如下工作:首先,提出了基于质量散度的气味源确认方法。该方法源于质量守恒定律,当某位置散度大于零时,表明该位置为源;当散度等于零时,表明该位置仅是流动过程中的一点;当散度小于零时,表明该位置为汇。本文采用泰勒级数展开式法求解散度的近似值,设计了多组数值仿真,验证了该方法的可行性,并讨论了传感器节点数目和计算误差之间的关系。最后给出整个三维空间中散度的分布情况再次验证了算法的正确性。然后,设计了基于质量通量的气味源确认方法。该方法采用四个传感器节点组成正四面体,计算该正四面体的质量通量。文中提出正四面体的质量通量等于组成该正四面体的各表面的质量通量之和。利用四个顶点的浓度和风速信息通过插值、网格划分等方法获取更多位置的信息,以此求解各表面的质量通量。在质量通量算法的仿真中,同样设置了多组风速条件。仿真结果给出了当闭合曲面内包含和不包含气味源时,正四面体不同尺度时的质量通量。最后,将两种算法对比分析,散度方法判断气味源的位置准确性高,但由于计算的是传感器节点中心点的散度,如果多个传感器节点的中心点恰好没有遍历到气味源所在的点,可能会造成漏警。而通量方法改进了散度算法的缺点,将气味源的位置确定在一个很小的范围内,而且算法的可行性很强。更多还原

【Abstract】 The leakage of poisonous and harmful gas is a serious threat to people’s life and property,so it is of great significance to carry out the research of leakage source location.Odor(or gas)source location research mainly includes three parts: plume finding,plume tracking and odor source identification.In previous work,there are few studies about the declaration of odor sources,and most of the studies are based on the ground robots or sensor nets which make use of odor information in the two-dimension space.For the actual environment of the odor source(such as the damage of overhead gas pipeline),it is necessary to study the declaration of the odor source in the three-dimensional environment.This thesis introduces the concept of mass divergence and mass flux in the problems of odor source declaration.The key point is how to use the three-dimensional sensor network to collect the odor concentration and wind speed information in order to determine whether a suspicious location is the odor source.The main work is as follows:Firstly,an odor source declaration algorithm based on mass divergence is proposed.The algorithm derives from the law of mass conservation.When the divergence is greater than zero,it indicates the position is a source,while on the opposite the negative one represents a sink.And when the divergence is equal to zero,it means the position is only a point in the process of flow.Taylor series expansion method is used to solve the approximation of divergence,and several numerical simulations are designed to verify the feasibility of the method.The relationship between the number of sensor nodes and the calculation error has been discussed.Finally,the distributions of divergence are given in the whole three-dimensional space,with which the method proposed before can also be proved.Then,a mass flux algorithm is designed for odor source declaration.The method uses four sensor nodes to form a regular tetrahedron,and calculates the mass flux of the tetrahedron.It illustrates that the mass flux of a regular tetrahedron is equal to the sum of the mass fluxes of the surfaces forming the regular tetrahedron.The mass flux of each surface is solved by interpolation,mesh partition and so on.In the simulation of the mass flux algorithm,multiple kinds of wind speed conditions are set.Considering whether the enclosed curve surface includes an odor source or not,different results are derived given different scales.Finally,compared with another algorithm,the divergence method can locate the position of the odor source more accurately.However,if the center point of multiple sensor nodes does not exactly traverse the point where the odor source is located,it may cause a false alarm.The shortcoming can be further improved by the flux method,in which the position of the odor source is determined in a small range,and the algorithm is of great practicability.更多还原