基于遗传算法GA解甘蔗收割顺序最优化问题的研究-中国生态农业学报.PDF

2009 11 17 6 Chinese Journal of Eco-Agriculture, Nov. 2009, 17(6): 1273−1277 DOI: 10. 3724/SP.J.1011.2009.01273 (GA)* 1 2 3 孙丽娅 上野正实 大岭政朗 (1. 215104; 2. 9010213; 3. 8850091) 本研究以已建立的栽培方法、收割时期预测甘蔗单产和含糖量的数学模型为基础, 提出将寻求最佳 收割顺序、收割期归结为最优化组合问题, 并探讨了遗传算法(GA) 的解题适用性, 同时进行了几种基本设定的 栽培方法推移、栽培方法构成项量、行列式特性的定性分析, 为求最佳的甘蔗收割顺序、收割期、不同栽培 方法的面积等提供解析方法。为不使栽培方法的构成比例发生变化, 与通常GA 算法采用交叉在不同个体间互 相交替进行不同, 开发了适用本研究的GA 的自己交叉具体解析法。将收割第1 年期作为过渡状态, 第2 年期 为安定状态, 将两期目的函数的合并作为一个适应度进行评价。 甘蔗 收割顺序 收割期 遗传算法(GA) 自己交叉 最优化 : S5047; S566.1 : A : 1671-3990(2009)06-1273-05 A genetic algorithm methodology for optimized harvesting sequence of sugarcane 1 2 3 SUN Li-Ya , Masami UENO , Masaaki OMINE (1. Suzhou Vocational University, Suzhou 215104, China; 2. University of the Ryukyus, Okinawa 9010213, Japan; 3. National Agricultural Research for Kyushu Okinawa Region, Miyazaki 8850091, Japan) Abstract New mathematical models capable of predicting sugarcane sugar content based on planting method and harvest time were used to develop a methodology for optimizing harvesting sequence, and harvesting time of sugarcane. The applicability of GA was discussed too. The methodology based on Genetic Algorithm (GA) was used to qualitatively analyze the determinant variables of planting mode to obtain optimized combination harvest sequence, harvest time and area of different planting methods under several setted planting mode transitions. Self-crossing, as opposed to conventional GA crossing methods, was used to retain the stru