论文部分内容阅读
A transition identification method considering dynamic information is proposed in this paper.First,a definition for the transition process is presented here to make a brief description about the characteristics of transition.Then,mutual information(MI)algorithm is revisited to analyze correlation between two data blocks.Because of the existence of dynamic behavior in industrial processes,dynamic mutual information(DMI)is first proposed in this paper to consider the dynamic relationship between historical samples and current samples.The differences in statistical characteristics between transition and stable modes can be well revealed through DMI algorithm.Extensive testing is carried out both on numerical example and Tennessee Eastman(TE)benchmark process to validate the effectiveness of the proposed method.