【摘要】 根据生物嗅觉信息获取的原理,构造了装有传感器阵列的仿生鼻鼻流道和控制装置,实现在嗅觉区域主动控制气体流量和气味分子浓度,提高了嗅觉灵敏度.针对人工嗅觉系统目前广泛采用的模式识别算法———人工神经网络BP算法存在的网络学习算法收敛缓慢的难题,使用更接近生物嗅觉的混沌优化神经网络算法,采用混沌算法对BP神经网络误差修正进行改进.对标准BP算法和混沌优化算法进行仿真比较,验证了混沌优化神经网络算法的优越性.最后用研究的仿生鼻对白酒进行了实验研究,最高识别率达100%.更多还原

【Abstract】 Based on the principle of biologic olfaction information,the bionic nasal cavity system equipped with sensors array and a control device is developed,which can control the flux of airflow and the concentration of odorant molecules conveniently,and meanwhile,enhances the delicacy of olfaction.As to the slowness of BP net training,the identification algorithm of ANN with chaos optimized is put forward in the paper,which is more similar to biologic olfaction.The error modification of ANN is ameliorated by algorithm with chaos optimized.The standard BP algorithm and BP algorithm chaos optimized are compared by simulation method,and the superiority of the later is testified.Finally,experimental study is carried through a bionic nose system,and the maximal identification rate achieves 100%.更多还原