基于ELM实现对聚酯纤维工业数据质量的提高

Improvement of Data Quality of Polyester Fiber Industry Based on ELM

聚酯纤维工业现场的传感器在采集数据和传输数据过程中,由于故障或速率不一致会导致数据存在缺失,从而影响数据质量。根据聚酯纤维工业数据流的特点,提出对数据质量进行判定的规则,并通过定义准确性、完整性与一致性的度量方法,基于ELM算法对纺丝过程和聚合过程中的缺失数据进行填补,实现数据质量的提高。

During the process of collecting data and transmitting data,the sensors in the polyester fiber industry field will lack data due to faults or in⁃consistent rates,which will affect the data quality.According to the characteristics of the polyester fiber industry data stream,rules for judg⁃ing the quality of data are proposed,and by defining measurement methods for accuracy,completeness,and consistency,filling missing da⁃ta in the spinning process and aggregation process based on the ELM algorithm to achieve the improvement of data quality.