摘要
针对非均匀有理B样条(NURBS)表示的空间曲线,利用进化算法对复杂空间曲线形状进行误差计算,阐述了测量点到理论曲线最短距离的求解算法和步骤,构建了误差评定的目标函数。采用实数编码的遗传算法以及微粒群算法对测量点与理论曲线的距离进行计算,并与BFGS方法进行比较。实验结果表明该方法能够快速获得误差评定结果。
Aiming at the Non-Uniform Rationl B-Spline （NURBS） spatial curve, a evolution algorithm is proposed for evaluating free form curve profile. The process of computing the distance between measured points and NURBS design curve is presented in detail. Genetic algorithm （GA） and particle swarm optimization （PSO） are taken for computing the error and compared with BFGS method. Experiments show the proposed method can rapidly acquire a fine result.
出处
《工程图学学报》
CSCD
北大核心
2008年第3期93-98,共6页
Journal of Engineering Graphics
关键词
计算机应用
误差评定
遗传算法
微粒群算法
空间曲线
computer application
error evaluation
genetic algorithm
particle swarm optimization
spatial free form curve
