NONLINEAR SIGNAL-PROCESSING AND SYSTEM-IDENTIFICATION - APPLICATIONS TO TIME-SERIES FROM ELECTROCHEMICAL REACTIONS

被引：48

作者：

HUDSON, JL

KUBE, M

ADOMAITIS, RA

KEVREKIDIS, IG

LAPEDES, AS

FARBER, RM

机构：

[1] PRINCETON UNIV,DEPT CHEM ENGN,PRINCETON,NJ 08544

[2] UNIV CALIF LOS ALAMOS SCI LAB,DIV THEORET,LOS ALAMOS,NM 87544

来源：

CHEMICAL ENGINEERING SCIENCE | 1990年 / 45卷 / 08期

基金：

美国国家科学基金会;

关键词：

Bifurcations; Electrochemical Oscillations; Neural Networks; Nonlinear Dynamics; Signal Processing;

D O I：

10.1016/0009-2509(90)80079-T

中图分类号：

TQ [化学工业];

学科分类号：

0817 ;

摘要：

We show how nonlinear signal processing techniques can be used for extracting simple dynamic models from complex experimental time series. A neural network analysis is applied to measurements of current versus time from an experimental system where the electrodissolution of copper in a phosphoric acid solution takes place. We investigate transitions from steady to oscillatory behavior and from period-one to period-two oscillations. Such procedures can be used in the analysis of systems for which no adequate phenomenological models exist. © 1990.

引用

页码：2075 / 2081

页数：7

共 19 条

[1]

Aizerman M. A., 1964, AUTOMAT REM CONTR, V25, P917

[2] AN EXPERIMENTAL INVESTIGATION OF PERIODIC AND CHAOTIC ELECTROCHEMICAL OSCILLATIONS IN THE ANODIC-DISSOLUTION OF COPPER IN PHOSPHORIC-ACID [J].