模拟自然树生长的竞争算法及在曲线拟合中的应用

A Competition Algorithm of Simulated Growing Natural Tree and Its Application in Curve Fitting

利用自然树生长过程中生长、凋落矛盾统一的原理,建立了自然树生长的竞争模型,提出了模拟自然树生长的竞争算法.为验证算法的合理性和有效性,将其应用于复杂曲线的拟合,与标准遗传算法进行对比,该算法具有运行速度快、内存占用率低、拟合精确度高.与经典的最小二乘法进行对比,该算法内存占有率低且具有抗噪特性.该算法为优化设计和计算提供了一种新的思路.

According to the principle of unity of opposite on natural tree growing and wilting,a competition model is developed. Then, a competition algorithm of simulated growing natural tree is presented. To verify the algorithm, it is applied to the curves fitting. Compared with the standard genetic algorithm,it tans faster with less memory and has higher precision; compared with the classical least square method, it requires less memory with noiseproof feature. This algorithm may provide a novel solution to the design and calcaxlation of optimization.