基于序优化的嵌套分区算法

Combined Nested Partitions Algorithm Based on Ordinal Optimization

介绍了嵌套分区算法(NP)的基本思想,从理论上分析了NP算法的计算效率,提出了提高其优化效率的途径。介绍了序优化算法(OO),并将序优化的思想引入嵌套分区算法的可行域确定和希望指数选取算子中,给出了复合嵌套分区算法(OONP)的寻优思路。OONP算法将序优化思想集成到嵌套分区算法框架下,把两者的指数收敛能力结合起来,具有科学性和可行性。通过优化算例验证表明,该算法大大提高了优化效率。

The main ideas of nested partitions （NP） method were introduced, its efficiency was theoretically analyzed and a way was proposed to improve the optimization efficiency of the algorithm. Then the ordinal optimization （OO） algorithm was introduced and the ideas of OO were incorporated into the determining of feasible region and the estimating of the promising index. The combined OONP algorithm was analyzed that drew on the two optimization methods. The OONP algorithm adopted the exponential convergence rate ability of OO algorithm and NP algorithm, so that it improved the efficiency and feasibility. Simulation results demonstrate the OONP algorithm.