MATLAB神经网络分析上海光源光束线的运行数据

摘要

The operation data of the beamline in SSRF is non-linear, complex and irregular. The data analysis tool selects the Matlab. The connection between the Matlab and archive database are introduced, neural network algorithm is selected for data analysis of the historic running data of the beamline. A temperature data from the monochromator of the BL13W beamline during one month is applied and the analysis results indicate the pre-alarm information of the monitored equipment is the same as the assumption of the classification before.%上海光源光束线运行状态的记录数据具有非线性、复杂和无规律的特点，利用神经网络可有效地分析和处理这些数据。选用Matlab进行数据提取和分析，介绍了Matlab与archive数据库间的数据连接，并用神经网络模式识别对光束线设备运行的状态参量进行数据训练，给出了BL13W光束线站单色器中某一温度运行参量的分类预警及其运行数据分析处理的结果。