基团贡献人工神经网络集成法估算有机物物性研究进展

Research progress of group contribution-artificial neural network-integration means for estimating physical propertiess of organic compounds

介绍了基团贡献人工神经网络集成法的原理,综述了其在各种物性估算中的应用(如有机物的临界参数、常压凝固点、闪点、比容积及偏心因子等),并对该法与单一的基团贡献法、人工神经网络法进行了简单比较。最后对基团贡献人工神经网络集成法的发展方向进行了展望。

The principle of group contribution-artificial neural network-integration means was introduced, the some applications of aforesaid means were summarized in estimating various physical properties （such as critical parameters of organic compounds, freezing point at atmospheric pressure, flash point, specific volume, and acentric factor）. And the aforesaid means, single group contribution means, and single artificial neural network means were simply compared. Finally, the development direction of group contribution-artificial neural network-integration means was expected.