基于大数据的传感器异常诊断的智能算法研究
|
摘要
预拌混凝土行业在自动化生产过程中,因计量传感器会发生信号偏移等异常情况,可能导致计量偏差而影响预拌混凝土质量的稳定性,故对计量传感器进行异常诊断并做好异常预警,对于安全高效生产具有重要的指导意义。为实现计量传感器状态的实时监测与异常诊断,本文基于大数据技术,以历史数据加实时数据驱动的自学习方法,提出了一种傅里叶变换与移动平均-求和-自回归(ARIMA)相混合的智能化模型,实现异常诊断。为预防大数据预处理过程中将异常数据误清洗而屏蔽异常信号识别,本文采取傅里叶变换法对实时数据进行预处理,使时间序列平滑化,同时保留异常征兆,然后建立ARIMA模型完成数据的时间序列分析,并实现计量传感器异常预测。该异常诊断智能算法已应用于实际的搅拌站配料系统称重设备的异常诊断中,其方法的可行性得到验证。
In industrial automation production process, equipment anomalies may lead to a chain reaction,causing huge economic losses of the enterprise, and even cause casualties serious social impact. Thus, the abnormal diagnosis of equipment and abnormal warning has important guiding significance to the safe and efficient production. In order to carry out real-time monitoring and abnormal diagnosis of equipment status, various equipment abnormal diagnosis technology emerged. Based on the statistical learning method driven by signal, this paper presents an abnormal diagnosis method based on signal smoothing and ARIMA model. After the signal is smoothed, the ARIMA model is established for time series analysis. the signal prediction is finally made based on the ARIMA model, and the abnormal diagnosis of equipment is determined when the predicted value exceed the abnormal threshold value.In this paper, the abnormal diagnosis method is applied to the abnormal diagnosis of weighing equipment in the batching system of commercial concrete mixing station, vertifying the reliability of the method.
In industrial automation production process, equipment anomalies may lead to a chain reaction,causing huge economic losses of the enterprise, and even cause casualties serious social impact. Thus, the abnormal diagnosis of equipment and abnormal warning has important guiding significance to the safe and efficient production. In order to carry out real-time monitoring and abnormal diagnosis of equipment status, various equipment abnormal diagnosis technology emerged. Based on the statistical learning method driven by signal, this paper presents an abnormal diagnosis method based on signal smoothing and ARIMA model. After the signal is smoothed, the ARIMA model is established for time series analysis. the signal prediction is finally made based on the ARIMA model, and the abnormal diagnosis of equipment is determined when the predicted value exceed the abnormal threshold value.In this paper, the abnormal diagnosis method is applied to the abnormal diagnosis of weighing equipment in the batching system of commercial concrete mixing station, vertifying the reliability of the method.
引文
[1]范雪艳,高超.浅谈混凝土搅拌站的整体设计及生产工艺流程[J].科技信息,2010,33:507-506.
[2]程旭,李圆圆.水泥混凝土搅拌设备配料称量系统控制模式的改进[J].工程机械与维修,2016(05):83-85.
[3]潘玉松.基于主元分析的传感器故障检测与诊断[D].华北电力大学,2006(14):12.
[4]张娅玲,陈伟民,章鹏,胡顺仁,黄晓微,郑伟.传感器故障诊断技术概述[J].传感器与微系统,2009,28(01):4-6.
[5]吴浩中,戴小文,王开文.摆式列车用传感器的故障检测与识别[J].测控技术,2001(06):19-21.
[6]赵志刚,赵伟.基于动态不确定度理论的多传感器系统传感器失效检测方法[J].传感技术学报,2006(06):2723-2726.
[7]ISERMANN R.Model-based fault-detection and diagnosis Status and applications[J].Annual Reviews in Control,2005,29(1):71-85.
[8]陈溪南,钟伟,吉旭,等.基于贝叶斯网络的预拌混凝土质量可靠度评价模型[J].西南科技大学学报, 2018,33(1):34-39.
[9]刘希亮,陈桂明,李方溪,张倩.基于多传感器信息融合的贝叶斯网络故障诊断方法研究及应用[J].机械科学与技术,2013,32(01):91-95.
[10]房方,魏乐.传感器故障的神经网络信息融合诊断方法[J].传感技术学报,2000(04):272-276.
[11]张成军,阴妍,鲍久圣,纪洋洋.多源信息融合故障诊断方法研究进展[J].河北科技大学学报,2014,35(03):213-221.
[12]于洋,王栋,吉旭.混凝土企业基于知识管理的质量管控模型分析[J].混凝土, 2011(3):90-92.
[13]崔伟杰.以ARIMA模型预测全国生产安全事故[J].安全, 2017,38(6):27-28.
[14]林桂粉,罗立民.基于三维Savitzky-Golay滤波的蒙特卡罗剂量分布去噪[J].数据采集与处理, 2006, 21(4):439-443.
[15]杜艺,刘文国,葛帅.基于傅里叶变换的条带噪声去除方法研究[J].测绘与空间地理信息, 2014(8):165-167.
[2]程旭,李圆圆.水泥混凝土搅拌设备配料称量系统控制模式的改进[J].工程机械与维修,2016(05):83-85.
[3]潘玉松.基于主元分析的传感器故障检测与诊断[D].华北电力大学,2006(14):12.
[4]张娅玲,陈伟民,章鹏,胡顺仁,黄晓微,郑伟.传感器故障诊断技术概述[J].传感器与微系统,2009,28(01):4-6.
[5]吴浩中,戴小文,王开文.摆式列车用传感器的故障检测与识别[J].测控技术,2001(06):19-21.
[6]赵志刚,赵伟.基于动态不确定度理论的多传感器系统传感器失效检测方法[J].传感技术学报,2006(06):2723-2726.
[7]ISERMANN R.Model-based fault-detection and diagnosis Status and applications[J].Annual Reviews in Control,2005,29(1):71-85.
[8]陈溪南,钟伟,吉旭,等.基于贝叶斯网络的预拌混凝土质量可靠度评价模型[J].西南科技大学学报, 2018,33(1):34-39.
[9]刘希亮,陈桂明,李方溪,张倩.基于多传感器信息融合的贝叶斯网络故障诊断方法研究及应用[J].机械科学与技术,2013,32(01):91-95.
[10]房方,魏乐.传感器故障的神经网络信息融合诊断方法[J].传感技术学报,2000(04):272-276.
[11]张成军,阴妍,鲍久圣,纪洋洋.多源信息融合故障诊断方法研究进展[J].河北科技大学学报,2014,35(03):213-221.
[12]于洋,王栋,吉旭.混凝土企业基于知识管理的质量管控模型分析[J].混凝土, 2011(3):90-92.
[13]崔伟杰.以ARIMA模型预测全国生产安全事故[J].安全, 2017,38(6):27-28.
[14]林桂粉,罗立民.基于三维Savitzky-Golay滤波的蒙特卡罗剂量分布去噪[J].数据采集与处理, 2006, 21(4):439-443.
[15]杜艺,刘文国,葛帅.基于傅里叶变换的条带噪声去除方法研究[J].测绘与空间地理信息, 2014(8):165-167.