采用诸如幂律、关联、标度、列维分布、分形布朗运动等概念对高频金融数据进行统计分析，建立能够精确描述价格变化的分布特征、记忆效应和高阶统计性质的随机过程。提出和发展基于经纪人竞争有限资源相互作用的金融市场微观型，建立可容纳市场更多真实特征的金融物理模型，研究各种模型的自适应复杂系.统行为并对其物理机制有更深刻的理解。.

With the support of this project,we develop money market models and test their accuracy and predictive power using available high-frequency financial data;present statistical analysis of these data using such concepts as power-law distributions,correlations,scaling,unpredictable time series,and random processes.The focus has been on different aspects of the analysed stochastic process, e.g.,the shape of the distribution of price changes,the temporal memory and the higher-order statistical properties. We try to develop the most satisfactory stochastic model describing all the features encountered in empirical analyses. We consider analogies and differences between price dynamics in a financial market and such physical processes as turbulence and ecological systems. In order to clarify the time correlation of a financial series, the "technical analysis" has been reconsidered and developed for detection of the presence of a higher-order correlation in price changes.We study adaptive complexity behavior in a variety of microscopic models in which agents are competing for limited resources,develop agent-based model of money market and seek to understand the physics of adaptive behavior of a few existing models,such as minority game, the genetic model and its variations.We propose and study a generalized version of the minority model in which agents may stay away from entering the game by imposing criteria on the accumulative success rate of the strategies.Using the generalized minority model,such factors as an inhomogeneous population of agents with different capabilities and different confident levels have been investigated. These models have also been studied within the context of the bar-attendance model proposed by W.B.Arthur.The topics studied by us are of relevance to a variety of subjects of current interests such as econophysics,evolutionary biology,social dilemmas,and resource allocation in a distributed network..