中文摘要：

复合事件探测是RFID事件处理的核心，本文对其中的非自发事件探测技术进行了深入研究，提出一种后继事件驱动的非自发事件探测方法SD—EventD：SD—EventD将查询树中的非自发操作上提，与其父节点操作进行语义融合，直至最近的双目运算父节点；仅为双目运算符设置单队列，并基于单队列的后继事件驱动进行双目运算的语义探测；针对融合后的双目运算，给出了基于操作符语义的语义探测方法，即出入队规则．本文详细描述了SD．EventD的处理策略并给出了算法描述．实验表明SD—EventD有效减少了内存使用量的同时降低了算法的处理时间．

英文摘要：

Composite evem detection is the core of RFID event processing, in this paper, non-spontaneous event detection technologies are studied deeply, and this paper proposes a detection approach of the successive event-driven non-spontaneous event which is called SD-EventD ： SD-EventD pulls up the non-spontaneous operators of query trees and makes semantic fusion with their parent nodes until to the nearest binary operators ; SD-EventD just sets a single queue for each binary operator, and detects the semanteme of binary op- erators based on the single queue＇s successive event drive; In allusion to the fusion binary operation, SD-EventD gives out the seman- tic detection approaches which are based on the semanteme of operators, namely the rules of enqueue and dequeue. This paper de- scribes the processing strategy of SD-EventD in detail and proposes the algorithm description. Experiment shows that SD-EventD de- creases memory usage, at the same time reduces processing time of the algorithm.