乙醇-水溶液电化学浸渍体系的神经网络优化  

Process optimization for electrochemical impregnation in ethanol-water solution by neural network

作　　者：刘浩杰[1] 胡树清[1] 呙成[1] 

机构地区：[1]中国电子科技集团公司第十八研究所,天津300381

出　　处：《电源技术》2008年第6期383-385,共3页Chinese Journal of Power Sources

摘　　要：乙醇-水溶液体系电化学浸渍具有很高的实用价值。采用人工神经网络对实验数据进行建模。采用基于学习参数模糊自适应调整的误差反向传播算法提高网络训练速度。适当拓宽实验各因素的水平范围,经过不同因素、不同水平间的组合模拟,利用建立的神经网络模型预测出二元电化学浸渍体系的适当工艺条件。The electrochemical impregnation （ECI） in ethanol-water solution has high application value. By using the experimental data as sample set, an artificial neural network （ANN） was constructed, The training velocity of the model was improved by back-propagation algorithm based on fuzzy adaptive control of learning parameters. The range of levels of the factors were broadened moderately, with combinations of all widened factors at different levels as simulation input sample, the optimal conditions for ECI in ethanol-water solution can be obtained by trained ANN model.

关 键 词：电化学浸渍 神经网络 模糊控制 优化 

分 类 号：TM912[电气工程—电力电子与电力传动]