中文摘要：

针对基于物联网的焊机监测系统中异常情况检测和预警的需要,提出一种基于多种群萤火虫的突发检测算法;通过不同种群萤火虫的协同工作实现滑动窗口大小的优化选择与配置,提高突发检测模型的处理速度与检测性能。仿真实验结果表明,基于多种群萤火虫的突发检测算法在突发概率或者最大滑动窗口大小相同的情况下,处理时间少于传统的突发检测算法,并具有更高的准确率和召回率。

英文摘要：

For the need of detecting and early warning of abnormal conditions in an Internet of things-based welding machine monitoring system,this paper presented a multi-population firefly based burst detection algorithm for reducing processing time and improving detection performance through optimizing the sizes of slide windows.The simulations experimental results demonstrate that the burst detection algorithm consumes less processing time in the same conditions of burst probability or maximum slide window size,and also achieves higher precision and recall rate than traditional algorithms.