Abstract: CO，NO₂ and CH₄ are three kinds of the most terrible gases in traffic environment.An electronic nose system is structured to detect them，which is composed of an optimal metal oxide semiconductor sensor array using an ameliorated BP network called a repercussive algorithm.The results of experiment show that the system can recognize poison gases’ concentration in traffic environment by a high precision.

摘要： 针对交通环境中毒害较大的CO、NO₂和CH₄三种气体，优选金属氧化物半导体气体传感器构成阵列，并对BP网络进行改进，采用有弹回的BP算法，构造了电子鼻系统。实验结果表明，系统能够以较高的精度检测出交通环境中三种有毒气体的浓度。