基于数据挖掘的设备状态知识获取方法研究

Device Status Knowledge Acquisition Method Based on Data Mining

数据挖掘是知识发现过程的一个重要步骤,在大数据特征的工程问题中有着广阔的应用前景。通过分析设备的实时状态数据,建立一套基于时间序列的设备知识获取模型以及模型度量,运用相似性预测算法以及聚类分析中的k-means算法对设备实时数据进行挖掘,从而获取设备实时数据所蕴含的知识,为后续通过知识推理进行设备故障预警奠定了基础。最后通过对某数控机床实时采集数据的分析实验,验证了模型和算法获取状态知识的有效性,并提出了改进的方法。

Data mining is an important step in the process of knowledge discovery and has broad application prospects in large data characteristics engineering problems.By analyzing real-time status data of equipment,a set of equipment knowledge acquisition model and model measurement based on time series is established.The real-time data of equipment is explored by using similarity prediction algorithm and K-means algorithm in clustering analysis to obtain knowledge of equipment real-time data,which lays the fundation for equipment failure warning through knowledge reasoning.Finally,through the analysis experiment of realtime collection data from NC machine tool,the validity of the model and algorithm to obtain the status knowledge is verified and the improved method is put forward.