摘要
介绍了一种基于Dignet ANN无监督学习聚类算法和自适应模糊控制算法的智能火灾探测算法模型.详细阐述了算法模型的思想和实现,给出了环境模式阈值自适应的方法和简单的多类型火灾探测器探测数据融合的方法,较好地解决了环境阈值的自适应问题.在实验室条件下利用欧洲标准火对算法进行了检测,结果表明该智能算法可以有效地对火灾进行探测.
An algorithm for intelligent fire detection was proposed, which is based on the Dignet ANN and fuzzy algorithm. The idea and implementation of the detection algorithm were introduced in detail. Method for self-adaptation to environment thresholds and for primary data fusion of multi-type fire detectors were given. The algorithm was successfully tested with European standard experimental fire in laboratory.
基金
中国科学技术大学研究生创新基金(KD2007083)资助