中南带杉木林乔木层碳储量生长与成熟模拟

Simulating for carbon growth and maturity of Chinese fir forest layer in central south China region

以中南带杉木林乔木层为研究对象,用人工神经网络表达碳储量与林龄之间的非线性关系,构建了BP人工神经网络模型.以乔木层各组分的碳储量年龄序列对所建模型进行训练和分析,得适宜的模型结构为1∶1∶4,总体拟合准确度为97.46%.以年均碳储量最大值对应的年龄定义碳储量成熟龄,基于所建模型,应用MATLAB编程语言编制了计算碳储量生长与成熟的计算机模拟程序,模拟的生长曲线符合森林生长规律,计算的碳储量成熟龄符合研究对象实际,表明所编计算机模拟程序有效.

In the paper, the Chinese fir forest layer in central south China region was taken as the re- search object. The artificial neural network was used to express nonlinear relationship between carbon and age. A BP artificial neural network model was constructed. The constructed model was trained and analyzed with the carbon age sequence of each composition in the tree layer so that the optimal model structure is 1 ： 1 ： 4, and the overall fitting accuracy is 97.46%. The age corresponding with the largest annual average carbon was used to define the carbon mature age. Based on the model, a computer sim- ulation program was compiled with the MATLAB language for calculating carbon growth and maturity. The simulating growth curve is in accordance with forest growth rule, and the calculating carbon mature age conforms to the practice of the research object. The results show that the computer simulation pro- gram is effective.