煤矿多传感器混沌时序数据融合预测  

Fusion prediction of mine multi-sensor chaotic time series data

作　　者：穆文瑜[1] 李茹[1,2] 阴志洲[1] 王齐[1] 张宝燕[1] 

机构地区：[1]山西大学计算机与信息技术学院,太原030006 [2]山西大学计算智能与中文信息处理教育部重点实验室,太原030006

出　　处：《计算机应用》2012年第6期1769-1773,共5页journal of Computer Applications

基　　金：科技部科技型中小企业技术创新基金资助项目(07c26211401103);山西省教育厅科技开发项目(20041201);太原市科技局专项(08121005);太原市科技局创新计划项目(2006);太原市科技局中小企业创新项目(0503037);太原市科技局中小企业创新服务平台建设专项(110263);山西省高等学校中青年拔尖人才基金资助项目(2007)

摘　　要：针对单传感器煤矿数据预测存在的片面性问题,提出将信息融合技术与相空间重构技术相结合的多传感器煤矿数据的预测模型,对井下多种传感器,包括瓦斯浓度、风速、温度传感器,进行融合预测。以多类传感器时序数据为研究对象,首先利用信息融合的方法分别对各类传感器数据依次进行数据层融合、特征层融合;然后采用关联积分方法对两级融合之后的传感器数据分别确定相重构的时间延迟和嵌入维数两个参数;最后结合多变量相空间重构技术,将各类传感器数据融合重构相空间,运用基于K-Means聚类的加权一阶局域法构建多传感器数据的预测模型。实验结果表明:对于特征层的融合,每15 min时间段内的数据经融合后可有效作为衡量这段时间内的特征,经过预测模型计算后,与时间段为5 min、10 min、20 min相比较误差达到最小仅0.003,较目前的最小误差值0.05大大下降,融合预测效果较好,可以较准确地预测15 min后的传感器数据,有充足时间进一步为井下的安全评估提供决策依据。Concerning the one-sidedness for the single sensor data mining prediction,a multi-sensor data mining prediction model of combining information fusion technology and phase-space reconstruction technology was proposed.A variety of underground sensors,including gas concentration,wind speed,temperature sensors,were forecasted in fusion.Taking multi-sensor time-series data as the subject,information fusion methods were firstly used to do data fusion and feature level fusion in sequence on various sensor data.Secondly,the correlation integral method was used to determine the phase reconstruction time delay and embedding dimension of two parameters.Finally,combining multivariate phase space reconstruction,the various types of sensor data fusion reconstructed phase space,the use of weighted one-rank local-region method based on the K-Means clustering was adopted to build a multi-sensor data prediction model.The results show that： for the feature level fusion,the data every 15 minutes period of time after fusion can be effective as a measure of the characteristics of this period,after the prediction model calculations,compared with the time period,5 minutes,10 minutes,20 minutes,the error is minimum 0.003,compared with the current minimum error value of 0.05,the error is greatly decreased.Therefore,the integration forecast is better,and it can more accurately predict the future after 15 minutes of sensor data.People have sufficient time to further provide safety assessment of underground basis for making decision.

关 键 词：多传感器 信息融合 相空间重构 加权一阶局域法 融合预测 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]