欢迎您！东篱公司
退出
-
申报数据库
-
立项数据库
-
成果数据库
-
项目获奖数据库
中文摘要：
为了提高电子鼻对混合气体的识别率，针对气体传感器阵列的交叉敏感特性，探讨了在电子鼻系统中基于独立分量分析（IcA）算法与BP神经网络相结合进行模式识别的可行性。并对4个气体传感器组成的电子鼻对4种气体？昆合物所测得的原始数据进行处理，结果表明：ICA算法对数据进行有效预分类，减少了样本之间的相关性，将生成的新样本作为BP网络的输入，使网络结构简化，在保证一定正确率的前提下，大大提高网络的学习速度。利用该方法可以提高电子鼻识别混合气体的准确率。
英文摘要：
Independent component analysis（ICA） and back propagation（BP） neural network are used for pattern recognition of Electronic-nose. Considered the cross sensitivity of gas sensor array for improving the identify ratio to electronic-nose. The by using a sensor array of four sensors is analyzed to measure data which is obtained mixture of CO, C2 H5 OH, CH4 and C4 H io-The results showed that ICA can give a good classification for the gases data and eliminate the data correlation. The processing data is inputed to BP network, network structure is simplified, and the convergence speed of the BP network is enhanced greatly. Further the recognition ratio is improved highly with ICA and BP network to electronic-nose system.
同期刊论文项目
同项目期刊论文
期刊信息
位置：