工业加热电炉高耗电异常DCS数据主动采集技术研究

Study on Active DCS Data Acquisition Technology of High Power Consumption Anomaly in Industrial Heating Electric Furnace

工业加热电炉在运行过程中容易出现高耗电现象,且异常用电数据产生的频率较高、数据量较大,使得对异常数据的采集效率降低。为此,提出工业加热电炉高耗电异常分布式控制系统数据主动采集技术研究。利用局部平均值计算工业加热缺失DCS数据,采用拉格朗日插值法对其插补,对插补后的数据标准化处理。构建用电不平衡特征矩阵,采用局部离群因子检测算法,提取出用电异常数据特征集,构建基于长短期记忆网络(long short-term memory,简称LSTM)的分位数回归模型,实现工业电炉异常DCS数据的主动采集。实验结果表明,所提方法有效地提高了采集效率和采集精度,采集平均用时仅为029ms,其准确率-召回率在98%以上。

The phenomenon of high power consumption is easy to occur in the operation of industrial heating electric furnace,and the abnor-mal power data is generated with high frequency and large amount of data,which reduces the collection efficiency of abnormal data.Therefore,the active data acquisition technology of DCS(Distributed Control System)for high power consumption anomaly of industrial heating electric furnace is proposed.The local mean value is used to calculate the missing DCS data of industrial heating,the Lagrange interpolation method is used to interpolate it,and the data after interpolation is standardized.The power unbalance feature matrix was constructed,the Local Outlier Factor(LOF)algorithm was used to extract the feature set of power consumption anomaly data,and the quantile regression model based on Long Short-Term Memory network(LSTM)was constructed.To realize the active collection of abnormal DCS data of industrial electric fur-nace.The experimental results show that the proposed method can effectively improve the acquisition efficiency and accuracy,the average ac-quisition time is only 0.29ms,and the accuracy rate-recall rate is over 98%.