基于HO算法的NURBS拟合过程中的权重优化

Weight Optimization in NURBS Fitting Process Based on HO Algorithm

摘要: 非均匀有理B样条(NURBS)在复杂形状表达上有很强的适应性，高精度的曲线拟合可以有效提高其在多领域的应用效果。针对NURBS曲线逼近拟合，结合河马优化算法(Hippopotamus optimization, HO)，提出一种针对NURBS曲线逼近拟合中的权重优化方法。首先，将权重引入HO算法中作为种群成员；然后，结合型值点与拟合曲线位置关系建立拟合目标函数；最后，通过HO算法迭代机制，寻找最佳权重，应用最佳权重再次进行NURBS曲线拟合。实验数据表明，通过HO算法对NURBS曲线逼近拟合中的权重进行优化，有效提高了曲线拟合精度。

Abstract: NURBS (Non-Uniform Rational B-Splines) have strong adaptability in expressing complex shapes. High-precision curve fitting can effectively improve their application effects in various fields. Aiming at the NURBS curve approximation fitting, combined with the HO (Hippopotamus Optimization) algorithm, a weight optimization algorithm for NURBS curve approximation fitting is proposed. Firstly, the weights are introduced into the HO algorithm as population members; then, the fitting objective function is established based on the positional relationship between the control points and the fitting curve; finally, through the iterative mechanism in the HO algorithm, the best individual member, that is, the best weight, is found, and the NURBS curve fitting is performed again with the best weight. Experimental data show that optimizing the weights in the NURBS curve approximation fitting process through the HO algorithm effectively improves the curve fitting accuracy.