人工神经网络在示波计时电位法中的应用  

Application of Artificial Neural Network in Oscillographic Chronopotentiometry

作　　者：郑建斌[1] 张军[1] 

机构地区：[1]西北大学电分析化学研究所,西安710069

出　　处：《分析测试学报》1999年第6期26-29,共4页Journal of Instrumental Analysis

基　　金：国家自然科学基金!(29775018);陕西省自然科学基金!(98H07);中国博士后基金

摘　　要：研究了误差反传神经网络和小波变换在示波计时电位测定中应用的可能性。对Ｐｂ２ ＋ 等无机离子测定的实验结果表明： 人工神经网络可较好地用于高次微分和经小波变换处理后的示波计时电位测定。 与经典的示波测定方法相比较， 该法基本上消除了人为误差的影响， 提高了分析速度和分析结果的可靠性。The application of error back propagation artificial neural network(BP-ANN) and wavelet transform in oscillographic chronopotentiometry(OCPY) was studied. Experimental results of the determination for inorganic ions such as Pb2+ showed that the Pb2+ concentration could be accurately computed by the BP-ANN in the OCPY based on high order differential and the dE/dt-t curve after wavelet transform. Compared with classic OCPY, the new method not only basically avoids the effect of man _ made error but also enhances analytical speed and reliability of the results.

关 键 词：人工神经网络 示波分析 示波计时电位法 OCPY 

分 类 号：O657.1[理学—分析化学]