采用人工神经网络方法拟合杉木林分蓄积量与立地质量、林龄、林分密度之间的非线性关系 ,建立了杉木可变密度蓄积量收获预估 BP网络模型 ,并对所建立的模型进行精度检验。检验表明 ,所建模型预测精度高 ,可应用于森林经理调查、数据更新与经营决策优化以及编制可变密度收获表 The artificial neural network method was used to fit the non-linear relationship between stand volume and stand quality, stand age and stand density of Chinese fir plantation, and a BP neural network model was established for predicting variable density accumulation of Chinese fir. The model Accuracy test. The test shows that the model has high predictive accuracy and can be applied to the investigation of forest managers, data updating and optimization of business decisions, as well as the preparation of variable density harvest tables