神经网络在SnO_2气敏元件浓度测量中的应用

Application of Neural Network in Concentration Measurement of SiO_2 Gas Sensor

气体传感器属于传感器技术领域,在传感器行业中占有重要的地位。然而气体传感器阵列的交叉敏感性严重影响气体传感器对混合气体的测量。基于Matlab平台的神经网络工具箱,分别构建BP神经网络和RBF(径向基)神经网络,对由涂敷不同敏感材料的声表面波振荡器组成的阵列在4种混合气体灵敏度响应数据进行定量识别研究,结果表明RBF神经网络在气体定量识别方面更具优势。

The gas sensor belongs to the area of the sensor technology, and takes very important part in sensor technology. However, cross-sensitivity of gas sensor affects the measure precision severely. Back propagation （BP） neural network and radial basis function （RBF） neural network have been built separately in Matlab neural network toolbox. And with the sensitivities based on array of SAW oscillators in various sensitive materials, the quantitative identification of gas sensitivity response in 4 mixed gas has been done. The results indicate that the quantity recognition of RBF neural network is better than BP neural network.