1种人工林监测传感器网络节点簇及关键节点合理布设方法的研究

STUDY OF ARTIFICIAL FORESTRY DETECTING TRANSDUCER NET NODE CLUSTER AND DISPOSAL OF KEY NODE MODEL

将数量化理论和无线传感器网络技术相结合对人工林生长过程进行监测,以便能够科学准确地获得人工林生长量信息。本文运用数量化理论Ⅰ、Ⅱ对林地按林木生长量径级进行分类预测建模研究,结果表明数量化理论Ⅰ建立的回归模型的重相关系数值为0.851,用所筛选的生长环境描述因子可以有效反映林木生长量的情况。数量化理论Ⅱ建立的判别模型林木验证判对率随海拔高度增加呈现双峰值现象。在传感器有效交互通讯距离范围内,以数量化理论Ⅱ建立的判别模型的判对率的两个峰值为分类边界,林地内林木生长量信息可以用3个节点簇覆盖采集,提出侧重于分类界面目标树木的传感器网络的布设方案,使无线传感器网络应用于人工林生长量监测过程中存在的问题得到有效解决。

Theory of Quantification and wireless sensor Networks are used in long-term artificial forest monitoring to obtain exact mass growth information of the forest.In present study,Theory of Quantification I and II were applied in classification and predict of diameter of forests.Results showed that correlation coefficient of Quantification Theory I modeling could draw 0.851 on artificial forest.While,Quantification Theory II modeling could show that testing coefficient curve is double peak performance for forests individually.In the effective zones of wireless sensors,we raised a new wireless sensor Networks fixing proposal which make for classification interface aimed forests to make the wireless sensor Networks could overcome some shortcuts in monitoring the forests mass growth indexes.