番茄形态结构模型参数的多目标拟合估算方法研究  

Estimation method for model parameters of tomato morphological architecture by multi-targets plant fitting

作　　者：董乔雪[1] 王一鸣[1] Jean Francois Barczi 贺超兴[3] 

机构地区：[1]中国农业大学(东校区)信息与电气工程学院,北京100083 [2]中国科学院自动化研究所中法联合实验室LIAMA,北京100080 [3]中国农业科学院蔬菜花卉研究所,北京100081

出　　处：《农业工程学报》2006年第2期1-5,共5页Transactions of the Chinese Society of Agricultural Engineering

基　　金：国家863子课题(2002AA241221)"温室番茄生长建模与智能管理研究及软件开发"的资助项目;国家863课题"农林植物生长模型与数字化设计技术(2003AA209020)"的资助项目;中-法国际合作研究项目

摘　　要：对番茄生长模型参数进行准确估算是模型得以验证和面向应用的前提条件。基于番茄拓扑结构,采用非线性最小二乘优化方法,通过对植株形态数据的多目标拟合,实现模型参数的最优估计和校准。非线性最小二乘法的权值根据目标数据自动调整,使得不同量纲和不同数值范围的观测数据可以在同一水平上进行,提高了计算的精度和稳定性;另一方面用差商近似代替导数加快了算法的收敛速度。通过应用拟合优度检验,得出模型输出对观测值的拟合程度较好,说明模型合理,参数可靠。实践表明该估算方法是对模型进行验证的一个行之有效的方法。Accurate estimation of model parameters gives a better prediction basis for model validation and application. In tomato model, non-linear least square method （NLSM） was applied to calibrate growth parameters by fitting multi-stages tomato morphology data based on the description of their topological structure. The weight values in NLSM were adjusted automatically according to target data, which allowed the measured data with different dimensions or ranges to be treated at the same level, and thus the accuracy and stability of algorithm were improved accordingly. Derivative was replaced by the difference quotient in NLSM which further improved the algorithm speed. The check results of the optimum grade about the data fitting indicated that model output fitted well the trend of observed data, which verified that estimated parameters were credible, and model was valid on a certain scale. The practice proved that the multi-target fitting method is an effective and powerful tool applied to the crop model parameter calibration.

关 键 词：番茄 模型参数 估算 多目标拟合 非线性最小二乘方法 

分 类 号：S601[农业科学—园艺学] S641.2