论文部分内容阅读
提出了一种基于试验设计方法学的响应曲面模型方法，用于建立算法的性能模型．将一种小群体自组织的遗传算法用于算法参数的优化．遗传算法的改进，使得结构更加合理简单，收敛速度明显加快．实验结果表明，所提出的方法能够随着场景条件的变化较好地调整算法参数，从而有效地提高了算法性能．
A reactive surface model method based on experimental design methodology is proposed to establish the performance model of the algorithm. A small group of self-organizing genetic algorithm is used to optimize the algorithm parameters. The improvement of genetic algorithm makes the structure more reasonable and simple, and the convergence speed is obviously accelerated. Experimental results show that the proposed method can adjust the algorithm parameters well with the changes of scene conditions, which improves the performance of the algorithm effectively.