Abstract: This paper uses back propagation neural network to fit the volume coefficient model of asphalt.If the model formed by fitting is used to measure the flow of asphalt,the result shows that it is better than that of the model of linear and nonlinear regression.And it makes the precision measurement and control of flow of the asphalt weight possible.For the insufficiency of table data,spline interpolation is used to make the table data more available.The effect is satisfactory.

Key words: ANN； measurement of the flow of asphalt； fitting； interpolation；