摘要
针对维数计算速度慢、耗时多、成本高这一缺陷,本文从算法和程序结构两方面对此做了改进和优化,使CPU耗时仅为原来的百分之一左右。在数据后处理方面,本文提出了加权最小二乘法,使维数计算更精确,更适合实际。
In order to cover the shortage of speeding slowly and taking too more time in the calculation of attractor dimension from time series by the Grassberger and Procaccia algorithm , this paper privides several improvements in the program of structural optimization and design. And the weighted least squares method is presented for gaining more accurate results of correlation dimensions.
关键词
吸引子
维数计算
最小二乘法
chaos
correlation dimension
weighted least squares method
