一种基于文法演化自动拟合非线性数据的蜂群算法

Grammatical evolution based nonlinear data automatic fitting bee colony algorithm

将文法演化方法引入蜂群算法,基于上下文文法定义多种数学函数,提出了自动拟合非线性数据的蜂群算法BCGE,并给出了通过基因截断、基因增补及利用领域知识加速BCGE的方法。基于文法演化的BCGE比基于文法规划的其他算法更为简洁,且存储基因型所需的空间也远比其他算法存储语法树所需的空间少。通过五个测试函数的实验表明,BCGE能够有效地拟合非常复杂的非线性数据。

This paper introduced grammatical evolution into bee colony algorithm, and proposed a new nonlinear data automa-tic fitting bee colony algorithm BCGE, which was based on multiple functions defined in context free grammar, and presented speedup method for BCGE by using gene truncation and supplement and utilizing domain knowledge. This grammatical evolution based BCGE was much more concise than the genetic planning based algorithms, and it needed far less memory for storing genotypes than the genetic planning based algorithms do for storing syntax trees. The experimental results on 5 test functions indicate that BCGE can effectively fit very complicated nonlinear data.