基于量子遗传聚类算法的质量控制方法  

Quality Control Method Based on Quantum Genetic Clustering Algorithm

作　　者：王杰 王艳 Wang Jie;Wang Yan(School of Internet of Things Engineering,Jiangnan University,Wuxi 214122,China)

机构地区：[1]江南大学物联网技术应用教育部工程研究中心

出　　处：《系统仿真学报》2019年第12期2591-2599,共9页Journal of System Simulation

基　　金：国家自然科学基金(61973138)

摘　　要：提出一种基于量子遗传聚类算法的质量控制图识别方法。该方法分为质量特征提取和模式分类2个流程。将量子遗传算法和K-means算法相结合,基于一种量子旋转门旋转方向确定机制提出一种量子遗传聚类算法,采用实验仿真的方式验证了该算法的性能。运用所提量子遗传算聚类算法对质量数据进行聚类分析,基于此提出一种控制图特征描述方法。以该特征为输入,运用支持向量机识别所对应的质量控制图模式。所提方法得到了更好的聚类结果,且该控制图识别方法识别精度达到了98.63%。This paper proposes a method of quality control chart recognition based on Quantum Genetic Clustering Algorithm. This method is divided into two parts: quality feature extraction and pattern classification. By combining Quantum Genetic Algorithm(QGA) and K-means algorithm, a quantum genetic clustering algorithm based on a mechanism for determining the rotation direction of a quantum rotary gate is proposed, and its performance is verified by experimental simulation. Based on the clustering analysis of quality data using the quantum genetic algorithm proposed in this paper, a control chart feature description method is proposed. With this feature as input, Support Vector Machine is used to identify the corresponding quality control chart pattern.The proposed quantum genetic clustering algorithm obtains better clustering results,and the accuracy of the proposed control chart recognition method reaches 98.63%.

关 键 词：质量控制 遗传聚类 控制图 K-MEANS 量子遗传算法 

分 类 号：TP391.9[自动化与计算机技术—计算机应用技术]