题目-数据采集与处理.DOCVIP

国家自然科学基金；西北工业大学翱翔之星计划项目 基于标准差的自适应激素调节遗传算法 刘奕晨1，王 毅1，牛奕龙2，樊养余1，郝重阳1 （1. 西北工业大学电子信息学院，陕西 西安 710072 2. 西北工业大学航海学院，陕西 西安 710072） 摘 要：内分泌调节系统对体内各种激素的调控，具备较好的自适应性和稳定性。基于生物内分泌系统的激素调节原理，提出了一种新的自适应遗传算法。该算法以内分泌激素调节的Hill函数下降形式为基础，设计了自适应交叉算子和自适应变异算子，使交叉率和变异率在遗传算法迭代过程中，能够根据函数适应度值的标准差进行自适应调节，使得整个进化过程中将种群多样性维持在合理水平，从而保证算法的正常进化。4种测试函数及三维人脑图像分割的实验结果显示，提出的自适应遗传算法可较好地保持种群多样性并克服早熟现象，性能优于其他三种自适应遗传算法及传统遗传算法。 关键词：遗传算法；人工内分泌系统；激素调节；自适应；三维图像分割 中图分类号：TP18 TP391.9 TP301.6 文献标志码：A An Adaptive Genetic Algorithm Based on Hormone Regulation LIU Yichen 1，WANG Yi1，NIU Yilong2，FAN Yangyu1，HAO Chongyang1 (1. School of Electronics and Information, Northwestern Polytechnical University, Xi’an 710072, China 2. School of Marine Engineering, Northwestern Polytechnical University, Xi’an 710072, China) Abstract: The neuroendocrine system regulates all hormones in the body with better self-adaptation and stability. Based on the principle of hormone modulation in endocrine system, an improved adaptive genetic algorithm was proposed. An adaptive crossover operator and an adaptive mutation operator based on the downward form of Hill function were designed in this algorithm. The crossover rate and mutation rate were made self-regulated according to the standard deviation of fitness value in each generation. And the diversity was maintained at a reasonable level in the whole process of evolution. In order to ensure the normal evolution of genetic algorithm. Experimental results of 4 test functions and 3D brain image segmentation showed the improved genetic algorithm can not only maintain the diversity of the population effectively, but also overcome the premature problem, and performances of this algorithm are better than those of the other three adaptive genetic algorithms and traditional genetic algorithm. Keywords: genetic algorithm(GA); artificial endocrine system; hormone modulation; self-adaptation; 3D image segmentation 0 引 言 遗传算法（Genetic Algorithm，GA）是人工智能的一个重要分支，但实际应用时存在许多不足，如在进化