自组织人工神经网络在氟化物非晶态形成的判别条件中的应用  

APPLICATION OF SELF-ORGANIZATION ARTIFICIAL NEURAL NETWORK TO DISCRIMINATION TO AMORPHOUS STATE OF TRINARY FLUORIDE

作　　者：蔡煜东[1] 许伟杰[1] 

机构地区：[1]中国科学院上海冶金研究所

出　　处：《计算机与应用化学》1995年第3期205-208,共4页Computers and Applied Chemistry

摘　　要：运用三维Ｔ．Ｋｏｈｏｎｅｎ自组织人工神经网络，利用化学键参数法，研究了三元系氟化物非晶态形成的判别条件，识别成功率达１００％。结果表明，该方法性能良好，可望成为研究化合物非晶态形成条件的有效的辅助手段。In this paper, the T. Kohonen's self-organization neural network of 3-dimension is presented to discriminate the amorphous state of trinary fluoride on the basis of chemical bond parameters. The sucessful rate reaches 100%. The results indicated that the performance of the neural network method are good, and therefore it might be referred as an effective assistant technique for discrimination to amorphous state of compounds.

关 键 词：氟化物 非晶态 形成条件 人工神经网络 

分 类 号：O613.41[理学—无机化学] O641.121[理学—化学]