断电流法测阳极过电位的模拟及其遗传算法优化  

Simulation of current interrupt technique for measuring anodic overpotential of aluminum electrolysis and its genetic algorithm for optimization

作　　者：陈晓春[1] 杨建红[1] 陈建军[1] 谭会义[2] 刘业翔[1] 郑军 

机构地区：[1]中南大学冶金科学与工程系,长沙410083 [2]郑州信息工程学院应用数学系,郑州450002 [3]贵州铝厂,贵阳550058

出　　处：《中国有色金属学报》2000年第6期930-933,共4页The Chinese Journal of Nonferrous Metals

基　　金：国家自然科学基金资助项目! 5 970 10 0 5

摘　　要：利用断电流法测定铝电解阳极过电位 ,用拟牛顿法确定过电位衰减极化曲线方程中的各项参数 ,结果表明采用拟牛顿法往往使得曲线的模拟结果与实验结果有一定偏差 ,从而影响其预测功能。采用实数编码方法 ,把遗传算法应用于过电位衰减曲线方程中的参数辨识 ,通过在遗传进化过程中动态改变交叉率和变异率以增加其全局寻优的机会 。Anodic overpotential was measured by current interrupt technique, and the quasi Newton method was used to determine the coefficient of overpotential decay equation . The result indicated that a little difference exists between the fit data and the experiment data, which affects its predicting accuracy. So the genetic algorithm was introduced to recognize the parameters. Using real number coding, this algorithm was devised to change recombination rate and mutation rate dynamically, therefore more chance can be obtained to optimize the parameters overall situation. The simulated curve can be well agreed with the experiment result. [

关 键 词：电解 断电流法 遗传算法 炼铝 阳极过电位 

分 类 号：TF821.01[冶金工程—有色金属冶金]