分别用BP神经网络和RBF神经网络实现近红外光谱的汽油辛烷值预测，两种方法用相对误差以及决定系数进行性能评价，最后绘制图仿真曲线。有较为详细的注释，并有仿真数据-Respectively with BP neural network and RBF neural network Near Infrared Spectroscopy Gasoline Octane prediction, two methods relative error and the coefficient of determination for performance　uation, the final sketch simulation curve. There are more detailed comments, and a simulation data