信息熵动态变异概率RNA遗传算法

摘要

约束优化问题是科学研究和工程应用的热点和难点.受生物RNA分子遗传信息表达机制和信息熵概念的启发,本文提出了一种信息熵动态变异概率RNA遗传算法来求解这类问题,算法采用碱基序列的个体编码方式,并用RNA分子再编码和蛋白质折叠操作代替传统遗传算法的交叉操作,在变异概率的设置中,借鉴信息熵对系统有序程度度量的概念,根据当前种群个体每一位的碱基分布情况对变异概率进行自适应调整.测试函数的仿真结果表明所提出的算法具有收敛速度快、搜索精度高的特点.将该算法用于求解短期汽油调合调度问题,能得到比其他几种算法更高的调合利润.%The constrained optimization problem becomes a focus and difficulty in science and engineering filed. Inspired by the expression of bio-genetic information of RNA molecular and entropy concept, a RNA genetic algorithm with entropy based dynamic mutation probability (edmp-RGA) is proposed. The algorithm adopts nucleotide base encoding, and RNA recoding operation and protein folding operation are designed to replace the conventional crossover operation. In the algorithm, the values of mutation probability are decided by nucleotide base distribution of the current bits of population. The numerical experiments on four benchmark functions show the effectiveness of the proposed algorithm. The solution to the short-time gasoline blending scheduling problem shows that the proposed algorithm gains a higher profit.