一种新的全局优化算法——统计归纳算法

在多极值问题的优化领域 ,主要有模拟退火算法(SA) ,遗传算法 (GA) ,人工神经网络 (ANN)算法 ,它们都是基于对自然现象模仿的算法。该文从更基本的优化思想出发 ,基于概率论提出了一种新的全局优化算法——统计归纳算法 (SIA)。在一些标准测试函数以及“货郎担问题”(TSP)上的计算结果表明 ,该算法在智能性 (所需的函数计算次数 )和解的全局性方面都远远好于 SA和 GA。在中国 144个城市的 TSP问题实例中 ,它甚至很快就找到了比参考计算中给出的“目前已知的最优路径”更短的路径。从这一算法思想的角度 ,阐述了 SA和 GA为何对全局优化问题有效 ,以及SA和 GA各自固有的不足之处

A new global optimization algorithm--Statistic inductive algorithm

Global optimization uses many algorithms such as the simulated annealing algorithm (SA), the genetic algorithm (GA), and artificial neural networks (ANN). All these algorithms are based on the imitation of naturel phenomena. This paper presents a new global optimization algorithm, the statistic inductive algorithm (SIA) based on a more fundamental optimization concept. Calculated results using standard tests and a 'travelling salesman problem' show that SIA has much better qualities than SA and GA. In addition to these results, the fundamental optimization concept in this paper is used to analyze how SA and GA affect the global search and the inherent defects in SA and GA.