引用本文:任子武,熊蓉,褚健.混合量子差分进化算法及应用[J].控制理论与应用,2011,28(10):1349~1355.[点击复制]
REN Zi-wu,XIONG Rong,CHU Jian.Hybrid quantum differential evolutionary algorithm and its applications[J].Control Theory and Technology,2011,28(10):1349~1355.[点击复制]
混合量子差分进化算法及应用
Hybrid quantum differential evolutionary algorithm and its applications
摘要点击 2754  全文点击 1574  投稿时间：2010-03-11  修订日期：2010-11-30
查看全文  查看/发表评论  下载PDF阅读器
DOI编号  10.7641/j.issn.1000-8152.2011.10.CCTA100235
  2011,28(10):1349-1355
中文关键词  量子进化算法  差分进化  和声搜索  量子非门
英文关键词  quantum-inspired evolutionary algorithm  differential evolution  harmony search  quantum non-gate
基金项目  国家“863”计划重点资助项目(2008AA042602); 国家自然科学基金资助项目(61075078).
作者单位E-mail
任子武* 苏州大学 机器人与微系统研究中心 zwren@iipc.zju.edu.cn 
熊蓉 浙江大学 智能系统与控制研究所  
褚健 浙江大学 智能系统与控制研究所  
中文摘要
      量子进化算法基于量子旋转门更新量子比特状态影响了算法搜索性能. 提出一种差分进化(DE)与和声搜索(HS)相结合更新量子比特状态的混合量子差分进化算法(HQDE). 该方法采用实数量子角形式编码染色体, 设计一种由差分进化计算更新量子位状态的量子差分进化算法(QDE)和一种由和声搜索更新量子位状态的量子和声搜索(QHS), 并相互机制融合, 采用两种不同进化策略共同作用产生种群新量子个体以克服常规算法中早熟及收敛速度慢等缺陷; 在此基础上, 算法还引入量子非门算子对当前最劣个体以一定概率选中的量子比特位进行变异操作增强算法跳出局部最优解能力. 理论分析证明该算法收敛于全局最优解. 0/1背包问题及旅行商问题实例测试结果验证了该方法有效性.
英文摘要
      Standard quantum-inspired evolutionary algorithm uses quantum gate to update the state of Q-bits, which deteriorates its optimization performance. A novel hybrid quantum-inspired evolutionary algorithm(HQDE) based on a hybrid of quantum differential evolutionary algorithm(QDE) and quantum harmony search(QHS) is presented. The HQDE adopts real-valued quantum angle to express the Q-bits of chromosome, and the new quantum population is produced through two approaches, i.e. QDE strategy and QHS strategy. Therein QDE strategy uses differential evolution to update the state of Q-bits, and QHS strategy employs harmony search to update the state of Q-bits. In addition, to avoid the disadvantage of easily getting in the local optimum, the HQDE performs quantum non-gate operation to transform the selected Q-bits of the current worst chromosome with a specified probability. Theoretical analysis proves that HQDE converges to the global optimum. The experimental results in solving 0-1 knapsack problem and 14 cities traveling salesman problem(TSP) demonstrate its effectiveness.