论文部分内容阅读
人工神经网络是人工智能技术的一个分支。本文将人工神经网络技术用于化合物的正常沸点、临界温度和沸点汽化热的定量预测,取得了较好的结果。估算精度好于或相当于文献方法。通过对三种基础物性的试验,表明人工神经网络技术可以用于定量预测,从而为这种技术开辟了一个新的应用领域。
Artificial neural network is a branch of artificial intelligence technology. In this paper, the artificial neural network technology is applied to the quantitative prediction of the normal boiling point, the critical temperature and the boiling heat of vaporization of the compound, and the better results are obtained. The estimation accuracy is better than or equal to the literature method. Experiments on three basic physical properties show that artificial neural network technology can be used for quantitative prediction, thus opening up a new application field for this technology.