论文部分内容阅读
建立了一种近似估计下一环境进化种群和问题的Pareto最优解集的核分布估计方法,当问题环境发生改变时,算法利用以前不同环境搜索到的有用解信息对下一环境进化种群及Pareto最优解集进行近似估计,极大地提高了算法的搜索效率。在对进化算子的合理设计基础上提出了一种核分布估计的动态多目标优化进化算法。通过对4个常用标准测试函数所作的数据仿真实验表明:提出的算法是十分有效的.
A kernel estimation method for estimating the Pareto optimal solution set of an evolutionary population and its approximate problem is established. When the problem environment changes, the algorithm makes use of the available solution information searched in different environments to evaluate the evolutionary population of the next environment, Pareto optimal solution set for approximate estimation, which greatly improves the search efficiency of the algorithm. Based on the rational design of evolution operators, a dynamic multi-objective optimization evolutionary algorithm for kernel distribution estimation is proposed. The simulation results of four commonly used standard test functions show that the proposed algorithm is very effective.