【摘要】 基因表达式编程(GEP)是一种基于基因组和表现型组的新型遗传算法,本文提出了一种基于GEP的非线性系统辨识算法。针对非线性系统的特点,本文基于GEP建模作了如下工作:在对象的结构和参数未知的情况下,首先进行参数恒定时的系统辨识,然后进行运行期间非线性环节参数发生变化时的辨识。实验结果表明该算法能够更加迅速的得出非线性部分直观近似的数学表达式,比用遗传编程(GP)优越两个数量级以上,是一种有效可行的算法。更多还原

【Abstract】 In this paper,GEP,a genotype/phenotype genetic algorithm , is applied to recognition of non- linear system. Based on the features of non- linear system, the main contribution of this paper includes: for objects with unknown structure and parameters, the recognition with constant parameters is accomplished first. then the system recognition during operation period is realized with varying non- linear element parameters. Experiments show that visual and approximate mathematical expressions of non- linear sections are all obtained rapidly by GEP. Furthermore, GEP outperforms GP more than two magnitudes and is a effective and feasible algorithms.更多还原