【作者】 李晓波；

【导师】 赵作鹏；

【作者基本信息】 中国矿业大学 ， 软件工程技术， 2016， 硕士

【摘要】 我国煤矿开采区域的结构相对封闭且复杂,使得矿难事故频发,构建基于无线传感器网络的智能安全监测系统将有效地改善安全数据监测能力和安全预警能力,是煤矿企业安全生产和科技创新发展的着力点。传感器节点的部署是无线传感器网络工作的基础,其影响着整个网络的性能,且无线传感器网络存在严重的能量约束问题,因此本文主要对节点部署和路由协议进行了研究,重点对分簇路由算法和三维空间节点部署进行了分析。在传感器节点部署方面主要分析了节点感知模型和覆盖问题,并对井下节点部署模型进行了分析,针对煤矿井下巷道环境的特殊性,提出了适用于井下巷道的部署模型。根据空间镶嵌理论,选择了三棱柱填充单元作为巷道部署模型;为了进一步提高节点覆盖效果并减少成本,对原有模型进行优化,设计了三角锥部署模型。在理论上,利用计算几何知识分别对两种模型进行了覆盖分析,最后通过仿真实验对部署模型进行了分析,两种模型均优于长方体部署模型,且三角锥部署模型效果更好。在WSN路由协议方面,对一些经典的分簇路由协议进行了研究,并分析了井下分簇路由协议的不足之处,针对井下巷道的空间特点,设计了两种分簇路由协议:GIDR和UCRPP协议。本文设计路由协议时首要考虑的问题是提高能量效率,GIDR算法通过感知节点的剩余能量和转播因子进行分簇;在建立簇间路由时利用贪心算法实现簇首与汇聚节点间的最优多跳通信,贪心依据综合了剩余能量期望、转播因子和剩余能量的影响;在路由选择时引入了定向区域的概念。仿真实验表明,GIDR协议在能量开销和均衡性方面具有较好的性能,但是其簇首分布存在不足之处。结合GAF算法的虚拟单元格思想,UCRPP协议对巷道区域进行非均匀分区处理。在传输路径选择时,分析了使用贪心算法进行路由选择的不足之处,UCRPP协议对贪心算法进行优化,采用半贪心算法构造通信路径,提出两跳通信能耗代价评估函数作为贪心手段。对传统的半贪心算法进行了改进,使用了不等概率的轮盘赌法竞选下一跳节点,使得剩余能量较大的节点有更大的机会竞选成功。仿真实验表明,UCRPP协议在能量消耗、网络生命周期、网络成本等方面优于其他协议,因此UCRPP协议更适用于井下巷道空间的数据监测。更多还原

【Abstract】 Due to the relatively closed and complex structure of coal mining region, the mine accidents occurred frequently in our country. Construction of WSN(wireless sensor network) based intelligent security monitoring system will effectively improve the ability of monitoring and early warning capabilities, making it be the focal point of coal mine enterprise safety production and technological innovation development. The deployment of sensor nodes is the basis of the work of wireless sensor network, which affect the performance of the whole network, and energy limited is a serious problem in WSN. Therefore, node deployment and routing protocols are studied in this paper, and it mainly focus on clustering routing algorithm and three-dimensional space nodes deployment.In the sensor nodes deployment, the node sensing model, coverage problem and deployment in mine are analyzed. According to the particularity of coal mine tunnel environment, the deployment model is put forward. The triangular prism filling unit is chosen as the model of tunnel deployment based on the spatial tessellations theory. In order to further improve the node coverage quality and reduce the cost of the deployment model, triangular pyramid model is designed based on the triangular prism model. Theoretically, the coverage problems of the two deployment models are analyzed based on the computational geometry. Finally, simulation experiments are carried out on the deployment models, the simulation results show that the two models are better than the rectangular deployment model, and the triangular pyramid model is better than triangular prism model. But the distribution of cluster heads is insufficient.In WSN routing protocols, some classical clustering routing protocols are studied, and the disadvantages of clustering routing protocols in mine are found. According to the characteristics of mine tunnel, two kinds of clustering routing protocols, GIDR protocol and UCRPP protocol, are designed. In this paper, improving energy efficiency is the first consideration for the design of routing protocols. Clustering is carried out through residual energy of nodes and broadcast factor in GIDR, and multi-hop communication can be achieved in optimal path between cluster heads and sink node by using greedy algorithm. The effect of residual energy expectation, retransmission factor and residual energy are combined with the greedy basis. The concept of directional area is introduced in the routing selection. The simulation results show that GIDR has better performance in energy consumption and energy balance, which is suitable for the tunnel safety data monitoring.According to the virtual cell idea of GAF algorithm, the mine tunnel is divided into non-uniform area in UCRPP. In the transmission path selection, the deficiency of the greedy algorithm is found. To optimize the greedy algorithm, UCRPP protocol uses semi-greedy algorithm to construct communication path. The two hop communication energy consumption cost evaluation function is proposed as the greedy method. Using the unequal probability of the roulette method to select the next hop node can improve the traditional semi greedy algorithm, which make the node with high residual energy easier to be selected as the next hop node. Simulation results show that the UCRPP protocol is superior to other protocols in terms of energy consumption, network lifetime, network cost and so on. Therefore, the UCRPP protocol is more suitable for the data monitoring of mine tunnel space.更多还原