中南大学学报（自然科学版）Issue(1)：100-107,8.DOI:10.11817/j.issn.1672-7207.2016.01.015
基于改进小波神经网络的极谱法多金属离子浓度检测信号的在线解析
Online analysis on polarographic detection signal of multi-metal ion concentrations based on improved wavelet neural networks
摘要
Abstract
For solving the overlapping peaks problem in multi-component detection of zinc hydrometallurgical process, an online analysis method for polarographic detection signal of multi-metal ion concentrations was proposed based on the improved wavelet neural network. Firstly, the first derivative of polarographic signal was obtained through the discrete wavelet transform, and consequently, the correspond feature points were obtained as the input of wavelet neural network based on the original signal and the first derivative of polarographic signal. Secondly, an improved state transition algorithm was proposed to optimize the parameters of wavelet neural network (WNN), and then the optimized WNN was adopted to describe the relationship between those feature points and the multi-metal ion concentrations so that it could be used to analyze online the polarographic detection signal of multi-metal ion concentrations. The method was verified by the actual polarographic overlapping peaks signal of zinc and cobalt. The results show that the proposed method is superior to those of the conventional curve fitting and the BP neural network algorithm.关键词
极谱曲线/多金属离子浓度/小波神经网络(WNN)/状态转移算法(STA)Key words
polarographic curve/multi-metal ion concentrations/wavelet neural networks (WNN)/state transition algorithm (STA)分类
信息技术与安全科学引用本文复制引用
王雅琳,黄凯华,黄天红,周晓君,阳春华..基于改进小波神经网络的极谱法多金属离子浓度检测信号的在线解析[J].中南大学学报（自然科学版）,2016,(1):100-107,8.基金项目
国家自然科学基金资助项目(61273187)；国家科技支撑计划项目(2012BAF03B05)；教育部博士点基金(优先发展领域)资助项目(20110162130011)(Project(61273187) supported by the National Natural Science Foundation of China （61273187）
Project(2012BAF03B05) supported by the National Science & Technology Pillar Program （2012BAF03B05）
Project(20110162130011) supported by the Ph.D Programs Foundation of Ministry of Education of China) （20110162130011）
