西北林学院学报2012，Vol.27Issue(1)：158-162,5.DOI:10.3969/j.issn.1001-7461.2012.01.33
云冷杉天然林林分年龄预测——以金沟岭林场为例
Forecasting the Spruce-fir Natural Forest Stand Age in Jingouling Forest Farm
摘要
Abstract
The back propagation (BP) artificial neural network( ANN), the projection pursuit regression (PPR) (ANN) and the multiple stepwise regression anatomic (MSRA) models were introduced to predict the nonlinear relation between the natural stand age and the stand factors. The precision and stabilities of the models were testified. The results indicated that 3 kinds of models were applicable for the prediction of natural forest age. The predication average relative error was 0. 04 for model of BP ANN, 0. 06 for PPR ANN, and 0. 08 for the MSRA model. The stability of BP ANN model was poor, and the PPR model and MSRA model were stable. It was concluded that the PPR model was better than the other two models, which can be applied to predict the natural forest stand age.关键词
BP神经网络模型/PPR神经网络模型/多元逐步回归分析模型/林分年龄Key words
BP artificial neural network model/ PPR artificial neural network mode/ multiple stepwise regression anatomic models/stands age分类
农业科技引用本文复制引用
宁杨翠,郑小贤,刘东兰,孔令红,陈宝升..云冷杉天然林林分年龄预测——以金沟岭林场为例[J].西北林学院学报,2012,27(1):158-162,5.基金项目
林业分益性行业科研专项的我国典型森林类型健康经营关键技术研究(20100400203) （20100400203）
