摘要
针对量子进化算法求解二进制编码问题比较有效,而求解多进制编码问题则比较困难的情况,本文提出了一种多进制概率角复合位编码量子进化算法。该算法将量子进化算法中量子位的概率幅表示法转化为复合位的概率角表示法,采用随机观测方法得到观测个体,采用概率角增减对个体进行更新,该算法适用于采用任意进制编码的问题。实验表明,与量子进化算法和传统遗传算法相比,多进制概率角复合位编码量子进化算法在适用范围、搜索能力和运算速度上具有较明显优势。
Considering that the Quantum-inspired evolutionary algorithm （QEA） is useful to binary coding problems, but useless to military coded problems, a novel evolutionary algorithm called Multinary Compound States of Probability Angle Coded Quantum-Inspired Evolutionary Algorithm （MQEA） is proposed in this paper. In the algorithm, the probability angle is used to represent the individual, instead of the probability amplitude which is used in QEA. The random observation method is used to obtain the observing individual. In the individual＇s updating, the probability angles of the individual increase or decrease according to the observing individual＇ s fitness compared with the current best solution ＇ s fitness. MQEA can be used in the problems with multinary coding. The experimental results show that MQEA has apparent superiority to QEA and the Conventional Genetic Algorithm （CGA） in applieation area, searching ability and computing time.
出处
《模式识别与人工智能》
EI
CSCD
北大核心
2005年第6期657-663,共7页
Pattern Recognition and Artificial Intelligence
关键词
进化算法
量子进化算法
多进制概率角复合位编码量子进化算法
背包问题
Evolutionary Algorithm, Quantum-Inspired Evolutionary Algorithm, Multinary Compound States of Probability Angle Coded Quantum-Inspired Evolutionary Algorithm, Knapsack Problem
