量子遗传算法的改进与研究

发布时间：2018-01-18 02:26

本文关键词：量子遗传算法的改进与研究　出处：《渤海大学》2017年硕士论文　论文类型：学位论文

【摘要】：量子遗传算法(Quantum Genetic Algorithm,简称QGA),它来源于量子计算和遗传算法的有机整合,具有寻求最优解能力突出、计算速率较快和整体计算规模较小的优点。学术界已经证实对于复杂问题的求解,用这种计算方法能够有效的提高精度和运算效率。但是,量子遗传算法在复杂函数优化问题上仍存在迭代次数多、收敛速度慢、较易陷入局部最优解的不足。为此本文基于量子遗传算法的不足作出了改进。具体工作如下:(1)深入研究了量子计算、遗传算法和量子遗传算法基本原理,分析归纳出它们的优缺点。(2)为提高算法的收敛速度和精确度,提出了一种基于小生境改进的量子遗传算法,将小生境协同进化策略引入量子遗传算法的种群初始化过程中对算法进行改进,在更新过程中采用动态调整量子旋转门策略和量子非门变异操作。并通过测试函数(Schaffer函数和DeJong函数)对算法进行验证。(3)由于小生境改进的量子遗传算法加入了量子非门变异操作,使得算法可能丢失优秀的信息,导致算法陷入局部极值。因此,利用Hadamard门变异操作代替了原有的量子非门变异操作,算法在增加种群多样性的同时还避免了在更新过程中丢失优秀信息,提高算法的全局搜索能力。(4)为验证改进算法的有效性和可行性,通过经典测试函数对算法进行优化验证,并将改进算法应用到了实际医疗数据分析。
[Abstract]:Quantum Genetic algorithm (QGAA) is derived from the organic integration of quantum computation and genetic algorithm. It has the advantages of searching for the best solution, fast computing rate and small scale of overall calculation. The academic circles have proved that the solution of complex problems. This method can effectively improve the accuracy and computational efficiency. However, quantum genetic algorithm still has many iterations and slow convergence rate in complex function optimization problems. It is easy to fall into the deficiency of local optimal solution. For this reason, this paper makes an improvement based on the deficiency of quantum genetic algorithm. The specific work is as follows: 1) the basic principles of quantum computation, genetic algorithm and quantum genetic algorithm are studied in depth. In order to improve the convergence speed and accuracy of the algorithm, an improved quantum genetic algorithm based on niche is proposed. The niche coevolution strategy is introduced into the population initialization of quantum genetic algorithm (QGA) to improve the algorithm. In the course of updating, the dynamic adjusting quantum revolving gate strategy and quantum non-gate mutation operation are adopted. The algorithm is verified by testing the function of Schaffer function and DeJong function. Because of niche improved quantum genetic algorithm, quantum non-gate mutation operation is added. The algorithm may lose excellent information, resulting in the algorithm falling into local extremum. Therefore, the original quantum non-gate mutation operation is replaced by the Hadamard gate mutation operation. The algorithm not only increases population diversity, but also avoids the loss of excellent information in the updating process, and improves the global search ability of the algorithm. 4) to verify the effectiveness and feasibility of the improved algorithm. The algorithm is optimized and verified by classical test function, and the improved algorithm is applied to actual medical data analysis.
【学位授予单位】：渤海大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：TP18

【相似文献】