模拟退火法用于气体红外光谱数据的识别分类

Simulated Annealing Algorithm Used for Identification and Classfication of Gas Infrared Remote Sensing Data

本文介绍了一种新的分类方法,它克服了传统的线性局部搜索的缺点,引入了跳跃式搜索,使得算法能够跳出局部极优解的陷阱,从而找到全局最优解。文中详细地介绍了其原理,并利用模拟退火的跳跃式搜索方法进行算法训练。通过实例比较,说明本方法应用于大气遥感红外光谱数据的分类和识别时,是行之有效的。

A new classification technology is proposed in the paper. It abandons defects of traditional linear local search, introducing jumping search. It can jump out the trap of the local better solution and find the best solution. The article relates its principles detailedly. And using the jumping search of simulated annealing, we have done algorithm training. Experimental results show that this method is more effective than local search on identification and classfication of gas's infrared remote sensing data.