基于多源异类信息的化学羽流追踪研究  

Study on chemical plume tracking based on multi-source and heterogeneous information

作　　者：周杰[1] 袁杰[1] 卫晓军 湾玥 ZHOU Jie;YUAN Jie;WEI Xiaojun;WAN Yue(Xinjiang University,Urumqi 830000,China)

机构地区：[1]新疆大学,新疆乌鲁木齐830000

出　　处：《现代电子技术》2018年第15期152-156,共5页Modern Electronics Technique

基　　金：新疆维吾尔自治区自然科学基金项目(2016D01C032)~~

摘　　要：为了在多种信息融合中提高化学羽流追踪效果,采用基于多源异类信息的化学羽流追踪方法。通过加权平均融合算法将浓度、红外避障及视觉检测传感器对化学羽流源模型采集的数据进行融合,计算出决策阈值;再通过建立的先验知识库和化学羽流源模型库的支持下,提出基于多源异类信息的化学羽流追踪策略,包括化学羽流搜寻、追踪及羽流源确认过程。通过实验验证,化学羽流源位置误差?e=0.216,实现了机器人对羽流源位置的精确定位。A chemical plume tracing method based on multi.source and heterogeneous information is adopted to improve the chemical plume tracking effect in a variety of information fusions.The concentration,infrared obstacle avoidance and visual detection sensor are used in the weighted average fusion algorithm to fuse the data collected by the chemical plume model,and calculate the decision threshold.On the basis of the established prior knowledge base and chemical plume source model,a chemical plume tracking strategy based on multi.source and heterogeneous information is proposed,including chemical plume search,tracking and plume source recognition.The experimental results show that the position errorΔe of chemical plume source is 0.216,which realizes the precise positioning of the robot for plume source position.

关 键 词：化学羽流 加权平均融合算法 知识库 多源异类信息 决策阈值 先验知识库 

分 类 号：TN964-34[电子电信—信号与信息处理] TP18[电子电信—信息与通信工程]