论文部分内容阅读
多模态过程中新出现的模态过程短期内无法获得充足的建模数据,且传统统计控制方法无法有效地估计过程特性.鉴于此,提出一种基于历史模型数据相关特性建立初步模型的方法,充分利用已有多模态历史数据的相关特性,从历史数据中寻找与当前数据特征相似的数据进行补充,建立初始模型,并利用新积累的数据迭代初步模型,逐步实现准确描述过程特性的算法.通过在田纳西-伊斯曼过程中的大量仿真,表明了所提出方法的可行性和有效性.
Due to the shortcomings of the emerging modal process in multi-modal process, sufficient modeling data can not be obtained and traditional statistical control methods can not effectively estimate the process characteristics. In view of this, a method of establishing a preliminary model based on the characteristics of historical model data is proposed , Make full use of the relevant characteristics of the existing multi-modal historical data, find the data similar to the current data from the historical data to supplement, establish the initial model, and use the newly accumulated data to iterate the preliminary model and gradually realize the accurate description of the process characteristics Algorithm.The large number of simulations in the Tennessee-Eastman process show the feasibility and effectiveness of the proposed method.