论文部分内容阅读
在建瓯设置 40块毛竹林标准地 ,分别测定了毛竹单株各部分干重与能量 ,建立了各部分生物量模型 ,并在此基础上 ,运用人工神经网络方法对毛竹林各组分能量进行估测 .结果表明 :毛竹林各组分秆、枝叶和地下部分的平均能量依次为 4.2 32 2 5× 10 8、9.2 2 95× 10 7和 1.76 43× 10 8kJhm-2 ,所占的比例分别为 6 1.32 %、13.11%和 2 5 .5 7% ;所建立的毛竹林秆、枝叶和地下部分各组分能量人工神经网络模型的平均模拟精度分别为 87.88%、82 .99%和 82 .95 % ,平均预测精度分别为 87.13%、81.32 %和 81.86 % .从而为揭示毛竹林生产潜力提供科学依据 .图 4表 7参 19
In Jian’ou, 40 bamboo forests were set up in a standard way. The dry weight and energy of each part of the plant were measured, and the biomass models of each part were established. On the basis of this, the artificial neural network The results showed that the average energy of stems, branches and underground parts of the bamboo forest was 4.2 32 2 5 × 10 8, 9.22 × 95 × 10 7 and 1.76 43 × 10 8 kJhm-2, respectively Respectively, 6 1.32%, 13.11% and 25.57% respectively. The average simulated accuracy of the energy artificial neural network models of the components of the stems, leaves and underground parts of the established bamboo forests were 87.88%, 82.99% and 82 .95% and the average prediction accuracy was 87.13%, 81.32% and 81.86%, respectively, which provided a scientific basis for revealing the potential of bamboo forest production.