机构地区: 中原工学院理学院
出 处: 《数学的实践与认识》 2009年第15期17-20,共4页
摘 要: Hurst指数是描述分数布朗运动的重要指标.利用R/S分析方法计算出我国专利申请量年增长率的Hurst指数稳定在0.62附近,分形维数稳定在1.38附近.表明我国专利申请量年增长率时间序列,服从分数布朗运动,专利申请量年增长率具有长期相关性.历史数据的影响力随时间间隔的长度增长而缩小,当间隔长度达到10年以上时,历史数据的影响力趋于稳定. The annual increasing rate time series of patent applications filed is independent random process. It does not obey Brownian motion, but fractal Brownian motion. We compute the Hurst exponent of the annual increasing rate time series of patent applications filed of China, by R/S analysis method. The result shows that the Hurst exponent of the annual increasing rate time series of patent applications filed of China has the level of 0. 62, and the fractal dimmension is 1. 38. It is conclude that the long range correlation does exist in the annual increasing rate time series of patents received of China. The long range correlation becomes lower and lower, and tend to be stable, when the length of data is longer than 10 years.