模糊聚类分析在火灾早期探测系统中的应用

Fuzzy clustering analysis in early fire detection

针对基于阈值的传统火灾探测算法在外界干扰下易频繁误报或漏报,以及响应时间过长等缺点,提出了采用模糊聚类分析的方法,对火灾早期的过程特征参数进行模糊聚类划分。实验证明,该方法不但可以准确、可靠地进行火灾探测,还可以提前火灾报警时间。

Due to the weakness of traditional algorithm of firedetectors detect the fire by parameters of a single moment, such as frequent misreporting or failed-report under the interference of environment, long response to the fire parameters. Describes a method of fuzzy clustering to identify the pattern of the fire process parameters. The experiment results indicate this approach can alarm the fire earlier and make it more exactly.