线阵式红外轴温探头的数据预处理

Data Preprocessing of Linear Array Infrared Hotbox Probe

列车轴温监测系统中,线阵式红外探头提高了数据的可靠性,也成倍增加了数据量。为了适应大规模的数据处理,采用实时采集实时上传的方法,由上位机完成数据处理。数据处理前的预处理包括提取轴温数据、滤波去噪声。根据车轮传感器间的距离和时序提取出轴温信号,通过不规则十字滤波法对轴温信号滤波,结果表明,该方法可准确地从大规模数据中提取出轴温信号,滤波能平滑轴温信号,提高信号的抗干扰性。

In hotbox train monitoring systems,the reliability of data is improved by linear array infrared sensor,but the amount of data is doubled.To adapt large-scale data processing,the method of real-time collection and real-time upload is used,the data is processed by PC.Data preprocessing includes the extraction of hotbox data and filter noise.Hotbox data is extracted according to the distance between the wheel sensor and timing,then these data are filtered by irregular cross-filtering method.Results show that hotbox data can be extracted accurately from the large-scale data,and the filter can smooth hotbox signal and improve the signal interference.