基于遗传算法和RBF网络的番茄生长模型辨识  

Model building of growing process of tomato based on RBF neural network and genetic algorithm

作　　者：张娟[1] 陈杰[1] 王珊珊[1] 

机构地区：[1]北京理工大学自动控制系,北京100081

出　　处：《控制与决策》2005年第6期682-685,共4页Control and Decision

基　　金："十五"国家重点科技攻关项目(2001BA503B01).

摘　　要：采用神经网络和遗传算法,对温室栽培番茄生长过程中主要器官——茎的生长过程进行了建模.温室番茄的生长过程具有控制变量多、生长过程复杂等特点.采用基于径向基函数(RBF)神经网络的辨识方法建立了温室栽培番茄生长的模型,以温室中番茄的实测数据为训练和预测样本,采用遗传算法进行训练.仿真结果表明,该方法较其他方法更适合于温室番茄生长过程的建模.The modeling of the growing status of the cauline of tomato in growing process is discussed, by using NN and genetic algorithm. The model of the growing process of tomato is constructed based on RBF neural network. The real data of the tomato planted in greenhouse are used as the sample data and the genetic algorithm is used for getting the parameters. Simulation results show that the model has better performance than those based on other method.

关 键 词：径向基函数网络 遗传算法 建模 

分 类 号：TP273[自动化与计算机技术—检测技术与自动化装置]