Abstract: Real-time massive data streams in the manufacturing IOT（Internet of Things） call for efficient event detection and processing methods. Efficient event detection means that more input events can be handled with less storage space per unit time. This paper presents a multi-pattern complex event detection method based on double-array trie-tree. The method reduces the detection and computation of redundancy in the query by constructing a multi-pattern matching automata model, and fully compresses the storage space by using the double-array trie-tree, thus improves the efficiency of the complex event processing. Simulation results show that compared with the traditional single-pattern complex event detection methods, the proposed scheme has less space and time consumption.

Key words: manufacturing IOT, complex event processing, multi-pattern matching, automata model, double-array trie-tree

摘要： 制造物联网中海量实时数据流急需高效的事件检测与处理方法，高效意味着单位时间内使用较小的存储空间处理更多的输入事件。提出一种基于双数组trie树的多模式复杂事件检测方法，通过构建多模式匹配自动机模型减少查询过程中冗余的检测和计算，并利用双数组trie树充分压缩存储空间，从而提高了复杂事件处理的效率。仿真实验表明，提出的方案相比传统的单模式复杂事件检测，具有较小的空间和时间消耗。

关键词: 制造物联网, 复杂事件处理, 多模式匹配, 自动机模型, 双数组trie树