遗传算法中适应度函数的改进

Improvement of fitness function in genetic algorithm

遗传算法引导搜索的主要依据就是个体的适应度值,因此适应度函数的设计显得尤为重要。本文兼顾保持种群的多样性和算法的收敛性,提出了一种基于指数变换的、指数系数可随进化代数动态调整的非线性适应度函数。以两个典型的测试函数为例,在相同的遗传操作和参数下,分别采用本文提出的适应度函数、线性拉伸变换及一般的指数变换适应度函数进行优化计算,计算结果表明采用提出的新适应度函数能极大地提高算法的优化精度、收敛速度和收敛概率。

The primary basis of genetic algorithm guiding the search is the individual fitness value, so the design of fitness function is particularly important. To keep the diversity of population and the convergence of algorithm,it proposed a non-linear fitness function which based on index transformation. The index coefficient in it can adapt to evolutionary process of algorithm. Under the same genetic operators and the same parameters ,calculating with the proposed fitness function,linear scaling transformation fitness function of Goldberg and general index transformation fitness function respectively, The simulation results of two typical testing functions show that the proposed fitness function can greatly improve the accuracy of optimization algorithms ,the convergence speed and the probability of convergenee.