关键词: 数据同化, 动态重构, 时序预测, 产能状态, 钢铁企业, 数据处理

Abstract: Current methods of production capacity state time-series data preprocessing were difficult to ensure accurate data update timely and distortion-free production capacity sate description. To deal with these problems, a novel method supporting data dynamic reconfiguration for production capacity state time-series prediction was presented. Handling methods of production capacity sate data grooming, data cleaning, and data mapping procedures were studied and proposed. Data assimilation model of production capacity sate was set up based on dynamic data reconfiguration, rapid retrieval method of assimilation data was presented based on process feature attributes matching. Application studies and example analyses were used to validate the actual handling process of production capacity state time-series data in certain steel enterprise.

Key words: data assimilation, dynamic reconfiguration, time-series prediction, production capacity state, steel enterprise