基于数据流的织造设备三分量数据清洗算法

Three-Component Data Cleaning Algorithm for Weaving Equipment Based on Data Flow

织造车间的信息化以数据采集为基础,但采集过程易产生脏数据,为保证数据被准确采集,文章研究了织造车间的数据采集与清洗算法。首先,针对设备多样、数据并发高的特性,设计了分频采集方案和服务器均衡负载方案,以及织造设备数据流处理有向网。其次,针对织造车间的数据特点,将数据分为常分量、增分量和状态分量,并结合箱线图、滑动时间窗研究了三分量清洗算法对各分量数据的清洗。最后,通过实验证明采集方案和数据清洗方法能保证数据采集的实时性、有效性及准确性。

The informatization of weaving workshops relies on data collection,but this process often leads to the generation of dirty data.In order to ensure accurate data collection,this paper proposes to study data collection and cleaning algorithms for weaving workshops.Firstly,considering the diverse equipment and high data concurrency,a split-frequency acquisition scheme and a server load balancing scheme are designed,as well as a directional network for data flow processing of weaving equipment.Secondly,based on the characteristics of weaving workshop data,the data is categorized into constant components,incremental components,and status components,and the three-component cleaning algorithm is studied by combining with box plots and sliding time window to clean each component data.Finally,experimental results demonstrate that the collection scheme and data cleaning method can ensure the real-time,effective,and accurate data collection.