SOM-BP神经网络在水稻钵育栽植机关键部件设计中的应用初探

Application on SOM-BP neural network in critical mechanism design of transplanter for rice growing in cupulate tray

水稻钵育栽植机的设计是实现水稻钵育栽培技术的关键环节,区别于传统的水稻栽植机其关键部件的设计方法还不成熟,在理论研究方面还不完善。为探索新的设计方法,论述了神经网络算法对农业机械设计的适用性,以自组织特征映射网络SOM和误差反向传播网络BP为理论基础,将SOM-BP集成神经网络模型应用于水稻钵育栽植机关键部件设计领域,建立神经网络模型,通过集成网络训练,得到设计结果。验证了SOM-BP神经网络在农业机械机设计中的正确性和精确性。

The design of transplanter for rice growing in cupulate tray is the critical link in rice growing in cupulate tray technology, the design method is not mature, for exploring a new design method, having discussed neural network＇s serviceability in agriculture machinery mechanism design with the theoretical basis of self-organizing feature map（SOM）and the back-propagation （BP） network, the integrated SOM-BP nerve network model is applied in the field of transplanter＇s critical mechanism design for rice growing, also building the neural network＇s model. Through training integration network, a design result has been obtained, which verified SOM-BP neural network＇s correctness and accuracy in designing for agriculture machinery mechanism optimization.