【作者】 李丽；

【导师】 郑国强；

【作者基本信息】 河南科技大学 ， 检测技术与自动化装置， 2012， 硕士

【摘要】 无线传感器网络(Wireless sensor networks, WSN)是传感器节点随机部署在空间区域内，通过无线多跳的方式传输监测数据，最终传送到用户终端的网络系统。近年来无线传感器网络运用越来越广泛，但是由于网络的能量受限，因此如何权衡应用设计要求和最大程度的节省能量成为研究无线传感器网络的核心问题。论文首先介绍了基于分簇的无线传感器网络目标跟踪算法。为了达到能量高效利用的目的，基于分簇提出一种目标跟踪算法，该算法将目标的运动过程看作是高斯马尔可夫模型，根据目标历史轨迹，估计下一时刻目标的位置坐标信息和运动速度，然后基于估计结果优化选择分簇的簇头和簇成员，形成一个动态簇来实现目标跟踪。在前述研究算法基础上，为了节省能源和均衡网络能效，提出了一种改进的基于能量高效的无线传感器网络分簇算法。该算法通过建立网络能量优化模型来计算最优簇头数，并结合节点向心度、无线通信模型和数据流守恒等约束条件组成非线性规划方程来选择簇头，形成动态簇。最后对本文提出的两种算法进行了仿真验证。仿真结果表明，前者使目标的跟踪精度得到提高，基本上满足了实际应用要求；后者侧重于均衡网络能效，延长网络寿命，通过计算最优簇头数，加以划分小区域，进行簇内选举，与经典算法相比较达到了均衡网络能耗和延长寿命的效果。更多还原

【Abstract】 Wireless sensor network is a network system in which sensor nodes are randomlydeployed in the region of space. Data are transmitted by multi-hop and ultimately sentto the user terminal. Wireless sensor networks have found more and more applicationsand developments in a variety of fields. However, how to get better performance tosupport energy conservation and tracking accuracy over WSN remains to be anontrivial and challenging task.This paper first introduces a cluster-based target tracking algorithm in wirelesssensor networks. In order to achieve the purpose of energy efficient using, a targettracking algorithm is proposed based on sub-clusters. This paper formulates themovement of the target as a Gauss-Markov model and depending on the targethistorical trajectory, this algorithm estimates the target position coordinatesinformation and velocity of the next time, and then based on the estimated results andsub-clusters of cluster head and cluster members are selected based on the principle ofoptimization. Then a dynamic cluster is formulated to achieve the target tracking.Furthermore, in order to save the energy and balance the network load in wirelesssensor networks, cluster head election strategy is proposed based on an energyoptimization model. The algorithm through the establishment of the network energyoptimization model to build the clustering structure, and composed a nonlinearprogramming equation which combined with the optimal number of cluster head、thedistance to the sensor nodes’ center、wireless communication model and data flowconservation constraints. The simulation results show that: compared with otherenergy-efficient clustering, the clustering algorithm can better save energy and extendthe network lifetime.The above three algorithms have all been simulated on Matlab, and they worktogether to offer accurate and timely information for decision-making in targettracking application.更多还原