基于改进多传感器数据融合算法的温室环境检测研究  

Research on greenhouse environment detection based on improved multi⁃sensor data fusion algorithm

作　　者：宋坤 李雨婷 张钰颖 高佳乐 杨玉强 李依潼 SONG Kun;LI Yuting;ZHANG Yuying;GAO Jiale;YANG Yuqiang;LI Yitong(College of Electronic and Information Engineering,Guangdong Ocean University,Zhanjiang 524088,China)

机构地区：[1]广东海洋大学电子与信息工程学院,广东湛江524088

出　　处：《现代电子技术》2023年第20期178-182,共5页Modern Electronics Technique

基　　金：广东省普通高校重点领域专项(新一代信息技术)(2021ZDZX1015)。

摘　　要：采用传统方法进行温室环境参数检测,通常存在环境复杂、检测可靠性低、精度差等问题。为提高温室大棚多参数检测数据的准确性,提出一种改进的多传感器数据融合算法。首先利用箱线图算法剔除偏离大的传感器数据,得到最优数据集;其次使用支持度和置信距离理论构建新的支持矩阵,将剔除的异常数据用支持度最高值代替,提高参与融合的数据可靠性;然后利用改进的自适应加权算法对数据进行融合;最后经测试,对传感器数据融合算法和算术加权平均融合算法处理结果进行分析比较。实验结果表明,所提算法能够提高温室环境参数检测的精度,融合值的相对误差更低,稳健性较好。In the process of traditional greenhouse environmental parameter detection,problems such as complex environment,low detection reliability and poor accuracy are often encountered.In order to improve the accuracy of multi⁃parameter detection data in greenhouses,an improved multi⁃sensor data fusion algorithm is proposed.The boxplot algorithm is used to eliminate the sensor data with large deviation,so as to obtain the optimal data set.The support degree and confidence distance theory are used to construct a new support matrix to replace the excluded abnormal data with the highest value of support degree,so as to improve the reliability of the data participating in the fusion.The improved adaptive weighting algorithm is used to fuse the data.After testing,the processing results of the sensor data fusion algorithm and the arithmetic weighted average fusion algorithm are analyzed and compared.The experimental results show that the proposed algorithm can improve the accuracy of greenhouse environment parameter detection,and the relative error of the fusion value is lower and the robustness is better.

关 键 词：温室 参数检测 多传感器 数据融合 箱线图算法 自适应加权 支持度 

分 类 号：TN92-34[电子电信—通信与信息系统] TP212.9[电子电信—信息与通信工程]