基于矩阵编码的CGP电路进化设计  

Research on Circuit Evolutionary Design Based on Matrix Coding CGP

作　　者：王晓霞 宋学君[1] 郭振兴 李充存 盖彦荣[1] WANG Xiaoxia;SONG Xuejun;GUO Zhenxing;LI Chongcun;GE Yanrong(College of Physics Science and Information Engineering,Hebei Normal University,Hebei Shijiazhuang 050024,China)

机构地区：[1]河北师范大学物理科学与信息工程学院,河北石家庄050024

出　　处：《河北师范大学学报（自然科学版）》2018年第4期312-317,共6页Journal of Hebei Normal University：Natural Science

基　　金：河北省自然科学基金和重点基础研究专项(F2018205178);河北师范大学研究生创新资助项目(CXZZSS2017050)

摘　　要：利用笛卡尔遗传规划进行电路进化设计时,常用的一维向量编码易出现早熟、收敛速度慢等现象,针对该问题,采用基于笛卡尔遗传规划的二维矩阵编码方案对一维向量编码进行改进.研究了连接度与收敛速度、激活节点之间,可编程逻辑单元数与激活节点、收敛速度之间的关系,结果表明,连接度越大,收敛速度快,激活节点数越少;可编程逻辑单元数适当时,进化代数少,但二维矩阵编码收敛速度更快.Using Cartesian genetic programming for circuit evolutionary design,common one-dimensional vector coding tends to appear premature,slow convergence and so on.To solve this problem,two-dimensional matrix coding scheme based on Cartesian genetic programming is used to improve one-dimensional vector coding.The relationship among the levels-back,the speed of convergence and the activation of nodes,the relationship among the number of programmable logic cells,convergence speed and the activation of nodes are studied.The results prove that the grenter the levels-back,the faster convergence speed and the smaller the mumber of active nodes is.The number of programmable logic cells is appropriate,the number of evolutionary generations is small,but the fewer activation of nodes.

关 键 词：演化硬件 电路进化设计 笛卡尔遗传规划 矩阵编码 

分 类 号：TN79[电子电信—电路与系统]