基于神经网络的NURBS曲线插补改进研究

The improvement research of NURBS curve interpolation based on neural network

本文介绍了NURBS曲线的插补特点,针对复杂曲线在NURBS曲线插补中求导误差大、计算复杂等问题,应用BP神经网络进行NURBS曲线插补方法的优化。然后通过实例验证,表明了使用该方法可以满足所要求的插补精度,而且由于插补前先进行了参数u的预估,插补速率得以提高。

The characteristics of NURBS curve interpolation were introduced.Because the derivation error and the computational complexity for complex curves with NURBS curve is large,so that NURBS curve interpolation will be optimized by the application of BP neural network.Then a example was used to demonstrate that the method could meet the required accuracy.The results showed that the interpolation rate was improved because the parameter u had been estimated before interpolation.