摘要
本文应用基于形式文法的方法,结合计算机科学中新兴有效的进化计算优化方法,进行了文法进化算法的有效性探索。通过运用巴科斯描述文法的四元组,强调了BNF产生式的设计,充分利用遗传操作来优化求解问题。并以有机化合物的分解(以离解能适应值进行适应度评价)为例子进行了算法设计,通过适应度函数择优,复制、交换和突变操作获得了实验结果。经过对结果的分析比较证明了算法的有效性。可以说是文法进化算法的成功探索。
Grammatical Evolution algorithm is a method based on the Form Grammar and Evolutionary Algorithms. In this study, some related aspects were discussed. First, a quatemion of form grammar is defined based on the BNF （Backus-Naur Form） through an example. Second, program is developed by using this form grammar based on the evolutionary algorithm. As an efficient idea, the operators such as crossover and mutation of genetic algorithm are also added as the rule of survival of the fittest to help choosing the fittest individuals. Finally, the decomposition of the organic compound （based on the energies） was adopted as an example to proof the effective under experiments with different condition. Comparing with the results, Grammatical Evolution is a powerful method, especially in the operation of genetic algorithm. In a word, Grammatical Evolution is a useful and effective method to optimize and evolve automatic programming. This research is also a beginning of applying Grammatical Evolution to the system of chemical reaction.
出处
《计算机与应用化学》
CAS
CSCD
北大核心
2006年第6期511-514,共4页
Computers and Applied Chemistry
基金
云南省教育科学研究基金资助
关键词
形式文法
进化算法
文法算法
form grammar, evolutionary algorithm, grammatical evolution
