基于动态时间错位的多元批次轨迹同步化

Synchronization of Multivariate Trajectories Based on Dynamic Time Warping

动态时间错位 (Dynamictimewarping)是一种灵活的、定常的模式匹配方案 ,它是基于动态规划理论提出的 ,能使模式内的相似特征合理匹配 ,它已在语音识别领域得到了成功的应用。笔者将动态时间错位 (DTW )的理论运用于分析和处理间歇反应过程中批次轨迹不同步的问题。在间歇反应过程中 ,由于批次与批次之间受物理特性和约束的影响 ,批次轨迹常具有持续时间不同步的特点。如果要用统计的方法分析和比较两个批次轨迹的数据特征 ,必须使两个批次的持续时间长度保持一致。动态时间错位 (DTW )理论可适时转换、扩张或压缩两个批次轨迹的局部模式特征 。

Dynamic time warping (DTW) is a flexible pattern matching method. Based on dynamic programming, this technique can reconcile similar characteristics of two trajectories, and has been used in the area of speech recognition successfully. This paper discusses the application of dynamic time warping (DTW) to the analysis and disposal of unsynchronized trajectories of batch processes. In batch process, due to the presence of batch batch disturbances and existence of physical constraints, batch processes are often characterized by unsynchronized trajectories. To compare these batch histories and apply statistical analysis, one needs to reconcile the timing difference among these histories first. Dynamic time warping (DTW) has the ability to synchronize two trajectories by appropriately translating, expanding, and contracting localized segments within both trajectories to achieve a minimum distance between the trajectories and optimal path, then synchronize two trajectories.