基于信息熵的自适应遗传算法

On the adaptive genetic algorithms based on Shannon entropy

现有自适应遗传算法陷入局部极小点后很难跳出，本文提出一种改进算法，用信息熵来估计系统分散度，使变异率随系统的分散度而变化。试验结果验证了该方法不仅速度快，而且几乎不陷入局部极小点。

It is difficult for AGA(Adaptive Genetic Algorithm) proposed by Srinivas M and Patnaik L M to escape from local optimum. An improved approach is proposed. By utilizing Shannon entropy to evaluate the diversity of solutions in population, the probabilities of crossover and mutation are adjusted based on the diversity of solutions, Through simulation examples, the algorithm is proved to converge to the global optimum quickly, and hardly gets stuck at a local optimum.