基于变维分形理论的卡尔曼滤波实时跟踪预测模型在股票价格预测中的应用  

Application of a Real-time Trace and Prediction Model of Kalman Filter Based on Variable Dimension Fractal Theory in the Prediction of Stock Price

作　　者：彭继兵[1] 唐春艳[1] 

机构地区：[1]成都理工大学信息管理学院,成都610059

出　　处：《计算机工程与应用》2005年第13期218-220,223,共4页Computer Engineering and Applications

摘　　要：基于卡尔曼滤波的动态、实时性以及股票市场的分形特性,论文首创利用变维分形理论来建立关于股票市场的卡尔曼滤波状态方程和观测方程,提出了一种新的基于变维分形理论的卡尔曼滤波实时跟踪预测模型和算法。实例仿真结果分析表明,论文提出的算法具有可靠、计算简便、快速等特点,模型预测精度较高,并可实现实时跟踪预测,具有一定的理论价值和实用价值。Based on the dynamic and real-time features of Kalman filter and the fractal feature of stock market,the state equation and observation equation of Kalman filter about stock market are established with variable dimension fractal theory for the first time in this article,and a new real-time trace and prediction model and algorithm of Kalman filter that based on variable dimension fractal theory is posed.The result of the instance simulation indicates that the algorithm posed in this article is reliable,simple and rapid,and the model has high prediction precision,which can realize real-time trace and prediction and has definite value of both theory and practice.

关 键 词：变维分形 卡尔曼滤波 实时跟踪 预测 股票价格 

分 类 号：F830.9[经济管理—金融学]