基于DNA的连续优化算法  

Continuous optimization algorithm based on DNA

作　　者：朱越[1] 

机构地区：[1]南京师范大学计算机科学与技术学院,南京210097

出　　处：《计算机工程与应用》2011年第22期48-52,共5页Computer Engineering and Applications

基　　金：国家自然科学基金No.60673102~~

摘　　要：用遗传算法进行函数优化已得到了较好应用。与遗传算法相比,DNA的编码更具丰富性,译码更具多样性,基因级操作更灵活,且更容易用DNA计算机实现。鉴于基于DNA的函数优化研究还较少,提出了一种基于DNA的连续优化算法。该算法用表示DNA的基本元素符号进行碥码,用其对应的密码子表征变量参数,用DNA的复制、重组、变异和倒位等操作实现对解空间的搜索。在这些过程中,参考了精英保留策略和模拟退火算法等思想方法,采取了若干加快收敛、同时满足搜索多样性要求的措施,以使算法加快收敛且不易早熟和陷入局部最优。计算机仿真实验表明该算法具有收敛快,精度高等特点,效果令人满意。The application of genetic algorithm for continuous optimization field has been proved to be effective.Compared with traditional genetic algorithm,DNA algorithm enjoys more advantages,such as more abundant encoding,more diversified decoding,more convenient gene operation and easier realization on DNA computers.So a continuous optimization algorithm based on DNA is proposed.In this algorithm,notations of DNA base pair are used for DNA encoding;translated DNA codens represent function variables;reproduction,crossover,mutation and inversion operations are used to realize the search for solution space.Meanwhile,some other strategies,elitist model and simulated annealing for example,are used to obtain a higher convergence speed and a better diversity.As a result,the algorithm will have lower probability plunging into a local minimum or premature situation.The computer simulation experiments indicate that the algorithm is of high search efficiency and convergence speed.The result is satisfied.

关 键 词：函数优化 DNA计算 遗传算法 DNA编码 

分 类 号：TP301[自动化与计算机技术—计算机系统结构]