A New Filled Function Method with Two Parameters for Global Optimization
A New Filled Function Method with Two Parameters for Global Optimization
复制标题
DOI：
10.1007/s10957-013-0515-1
复制
发表时间：
2014-01
影响因子：
1.9
通讯作者：
Fei Wei;Yuping Wang;Hongwei Lin
中科院分区：
文献类型：
--
作者：
Fei Wei;Yuping Wang;Hongwei Lin
填充函数法是寻找多峰函数全局极小点的有效方法。由于指数项或对数项以及对参数的敏感性，传统的填充函数在数值上往往是不稳定的。本文提出了一种新的填充函数，它是连续可微的，对参数不敏感，不容易引起溢出。然后给出了一种新的局部搜索算法。在此基础上，提出了一种新的填充函数方法。仿真结果表明，该方法对初始点和参数的变化具有较好的数值稳定性。与一些已有算法的比较表明，该方法具有更高的效率和有效性。
The filled function method is an effective approach to find the global minimizer of multi-modal functions. The conventional filled functions are often numerically unstable due to the exponential or logarithmic term and the sensitivity to parameters. In this paper, a new filled function is proposed, which is continuously differentiable, not sensitive to parameters, and not easy to cause overflow. Then a new local search algorithm is given. Based on this, a new filled function method is proposed. The simulations indicate that the proposed method is numerically stable to the variations of the initial points and the parameters. The comparison with some existing algorithms shows that the proposed method is more efficient and effective.