基于事件驱动的制造过程数据融合处理方法研究

Research of Manufacturing Data Fusion Method Based on Event-driven Architecture

为了解决物联网车间现场多源采集系统存在的延迟、缺失、冗余或语义模糊等数据质量问题,引入事件概念和D-S证据理论,依据"符号模式匹配+不确定性计算"的推理模式,提出基于事件驱动(Event-Driven Architecture,EDA)的制造数据融合方法.该方法利用制造物联场景下的业务约束和时序约束,定制了复杂事件处理(Complex Event Process,CEP)匹配模式,结合证据理论,对业务对象的状态置位过程给予量化评价,解决可能发生的数据质量问题,从而为离散制造车间有效合理地解决制造过程数据的实时采集、融合处理等问题提供了新的思路.

The multi-source collection system can not fully describe the status of business objects in the workshop based on the Internet of Things, which leads to data quality issues such as feedback delays, datadeletions, redundancy, and fuzzy semantics. To solve these problems, the concept of event and D-S evidence theory are introduced, and a man- ufacturing data fusion method based on the Event-driven Ar chitectore （EDA） is proposed, which implements the ＂sym- bol pattern matching and uncertainty calculation＂ reasoning mode. The method customized complex event process （CEP） match mode by defining the business constraints and timing constraints based on Internet of manufacturing-related things, and gave quantitative evaluation on the status feedback of business objects on-site via combining the evidence theory, so as to providing a new way of thinking about effectively and rationally solving real-time acquisition of manufac- turing process data, fusion and other issues in the discrete manufacturing workshop.