摘要
在系统结构信息缺乏的情况下,要确立离散动态系统的动态方程组非常困难。利用遗传规划(GP)融合系统结构估计与参数估计的特点,阐明GP在离散动态系统建模中的具体运用,并提出一种利用误差估计进行模型评价的方法。用Lotka-Volterra公式演化的数据对该建模方法进行了实证研究,结果表明遗传规划在离散动态系统建模中是有效的。
In the case of lacking of system structure information,it is difficult to determine the dynamic equation group in the discrete dynamic system.The application of genetic programming(GP) in constructing discrete dynamic system mode is put forwarded by using the characteristics of genetic programming(GP) blending system structure estimation and parameter estimation.Also,this paper suggests a mode evaluation method by using a kind of error estimation.The data calculated by Lotka-Volterra formula are used to test the model establishing method,whose results indicate that genetic programming is effective in model estabishment of discrete dynamic system.
出处
《西安理工大学学报》
CAS
2004年第4期374-378,共5页
Journal of Xi'an University of Technology
基金
陕西省自然科学基金资助项目(2001K04-G15)。
关键词
离散动态系统
遗传规划
复杂系统
discrete dynamic system
genetic programming
complex system
