基于神经网络的杨树苗木注射针头阻力预测模型研究  

Study on Resistance Prediction Model of Poplar Seedling Injection Needle Based on Neural Network

作　　者：孙炜 陆一枝 商庆清[1] SUN Wei;LU Yi-zhi;SHANG Qing-qing(Nanjing Forestry University,Nanjing Jiangsu 210037,China)

机构地区：[1]南京林业大学,江苏南京210037

出　　处：《林业机械与木工设备》2021年第10期32-35,共4页Forestry Machinery & Woodworking Equipment

摘　　要：进行杨树苗木注射时,针头所受阻力与针头粗细和进针量都有关系。针对这种关系,设计了一种3层BP神经网络模型进行拟合预测。并对隐含层不同节点数的预测结果进行对比分析,最后得出,当隐含层节点数为8个时,可以很好的拟合与预测针头所受力大小。对开发苗木注射针头具有一定的参考价值。When injecting poplar seedlings,the resistance of the needle is related to the thickness of the needle and the amount of the needle.Aiming at this relationship,a 3-layer BP neural network model is designed for fitting prediction.And the prediction results of different nodes in the hidden layer are compared and analyzed.Finally,it is concluded that when the number of nodes in the hidden layer is 8,the stress on the needle can be well fitted and predicted.This study has a certain reference value for the development of seedling injection needle.

关 键 词：苗木注射 BP神经网络 决定系数 隐含层节点数 

分 类 号：S776[农业科学—森林工程]