一种优于混沌优化的对分插值逼近算法

A Bisection-Interpolation Approach Algorithm Superion to Chaotic Optimization Algorithm

提出一种新的求解函数最优值的算法——对分插值逼近算法。该算法产生均匀分布于[a,b]区间的稠密点集,理论证明了该点集可以无限逼近[a,b]区间内的任何实数,且以概率1收敛于任何待优化函数的全局最优值。与混沌优化算法进行了比较,以一维、二维变量的仿真为例,结果表明,该算法在寻优过程中优于混沌优化算法。

A new method 'bisectioninterpolation approach' algorithm to find the optimal value of a function is advanced in this paper. This new method produces dense pointset having a uniform distribution . It is proved theoretically in this paper that the value calculated by this method can infinitely approach any real value in , and can converge to the global optimal value of any function with total probability. Comparisons between chaotic optimization method and this new method are made,and simulations (based on one dimension and two dimensions) in the end show that this algorithm is superior to the chaotic optimization method.