Based upon this subsection function fitting,artificial neural network algorithm and other methods are proposed to raise fitting accuracy in order to reduce error to obtain correct test results.

• 比较研究了混凝土无损检测中的三种常用测强曲线的拟合精度,在此基础上,提出了分段函数拟合,人工神经网络算法等提高曲线拟合精度的一些方法和措施,以得到正确的测试结果。