Abstract: Bi-group evolutionary programming algorithm uses Gauss oscillating mutation operator to realize the large-scale and sightless exploration for solution space, which brings poor efficiency. Aiming at the problem, this paper proposes an improved bi-group evolutionary programming algorithm. The algorithm replaces Gauss mutation operator in the old bi-group evolutionary programming algorithm with new mutation operator. The new mutation operator is correlated with fitness function. Oriented search with high efficiency of solution space is realized. Simulation results show that the improved algorithm has higher performance than the old one.

Key words: evolutionary programming, bi-group, mutation operator

摘要： 双群进化规划算法采用高斯振荡变异算子对解空间进行大范围盲搜索，效率较低。针对该问题提出一种改进的双群进化规划算法。采用与适应度函数相关的变异算子替换原双群进化规划算法中的高斯振荡变异算子，实现对解空间的导向性高效搜索。仿真结果表明，改进算法性能高于原有算法。

关键词: 进化规划, 双群, 变异算子

CLC Number: 

• TP18