GEP优化的多输出RBF网络作物生理参数建模

发布时间：2018-05-10 07:35

  本文选题：作物模型 + 基因表达式编程　； 参考：《安徽农业大学学报》2017年01期

【摘要】：针对常用的回归和神经网络作物建模方法存在的输出单一、参数优化困难和预测精度不足等问题,利用基因表达式编程优异的全局搜索能力和RBF神经网络多输出任意非线性函数逼近特点,设计了1种GEP优化的RBF多输出模型算法GEP-RBF。以水稻和番茄的5个关键环境因子为输入、以叶片CO_2交换率和蒸腾速率为输出,进行建模验证。结果显示,在预测的均方根误差指标上,GEP-RBF模型与GA-RBF和RBF相比,水稻的CO_2交换率和蒸腾速率分别降低了约28.4%、38.0%和89.9%、62.8%,番茄的CO_2交换率和蒸腾速率则分别降低了约56.9%、48.4%和75.3%、67.1%;在多输出结果的平衡性指标上,相比GA-RBF和RBF,GEP-RBF模型提高了约16.4%~77.4%。结果表明,GEP-RBF模型具有良好的预测精度和多输出平衡性,是一种有效的作物生长建模方法。
[Abstract]:In order to solve the problems of single output, difficulty in parameter optimization and low precision of prediction, the commonly used methods of regression and neural network for crop modeling are discussed. Taking advantage of the excellent global searching ability of gene expression programming and the characteristic of multi-output arbitrary nonlinear function approximation of RBF neural network, a GEP optimized RBF multi-output model algorithm GEP-RBF is designed. Five key environmental factors of rice and tomato were taken as input and leaf CO_2 exchange rate and transpiration rate as output. The results show that the GEP-RBF model is compared with GA-RBF and RBF on the root-mean-square error index. The CO_2 exchange rate and transpiration rate of rice decreased about 28.480% and 89.9%, respectively, while the CO_2 exchange rate and transpiration rate of tomato decreased about 56.9% and 75.3%, respectively, compared with GA-RBF and RBFGP-RBF model, the CO_2 exchange rate and transpiration rate of tomato decreased by about 16.4% and 77.4%, respectively. The results show that the GEP-RBF model has good prediction accuracy and multi-output balance. It is an effective modeling method for crop growth.
【作者单位】： 安徽农业大学信息与计算机学院;
【基金】：农业部国际科技合作项目(948计划,2015-Z44和2016-X34) 安徽省自然科学基金(1508085MF110) 安徽省科技攻关项目(1501031102)共同资助
【分类号】：S311;S126
，

本文编号：1868425