本文研究了多重分形的统计物理方法,以上证综合指数长达四年的一分钟高频数据为研究

对象,计算了实际交易数据的多重分形谱及其特征参数,并确定了权重因子的取值范围。结果表

明奇异指数和相应的谱函数作为多重分形谱的重要参数,一定程度上反映了股指本身的变化范围

和高低价位出现频率的变化,然而谱函数可以预测股市趋势的断言在沪市并不成立。

关键词:统计物理 高频数据 多重分形谱 奇异指数-This paper studies the multifractal statistical physics methods, the Shanghai Composite Index more than four years of high frequency data for the study of one minute

Object, calculated multifractal spectrum and characteristic parameters of the actual transaction data, and determine the weighting factor ranges. Results Table

Ming singularity index and the corresponding spectral function as a multi-fractal spectrum of important parameters, to some extent, reflect the range of variation of the index itself

And the level of price changes in the frequency of appearance, but the spectral function can predict stock market trends assertion does not hold in the Shanghai stock market.

Keywords: Statistical Physics　high frequency data　multifractal spectrum　singularity exponent