Woodland Quality Evaluation with Artificial Neural Network

摘要：

A growth equation of the stand topping mean height was created by using artificial neural network modeling technology, in Masson pine planted forest. Then, a site index model was created with the created equation as the directed equation, and the site index curved shape was drawn with the method of index level adjustment. Finally, the table of site index was worked out. The model structure is 1:3:1, the total fitting accuracy is 98.01 ％. Concretely, the mean topping height fitting accuracy of different age is 85.65 ％ to 99.67 ％ and the mean value is 97.92 ％. The corresponding testing accuracy is 93.88 ％ to 99.87 ％ and the mean value is 97.43％. The result stated clearly that the created model has very high fitting and testing accuracy and very strong prediction ability. The artificial neural network is an effective modeling technology of the stand growth process, and it was proposed to use it extensively in the woodland quality evaluation.

关键词： artificial neural network Pinus massoniana woodland quality evaluation site index

作者: Jiarong Huang Guangqin Gao Fang Guo

作者单位: Henan Agricultural University Zhengzhou 450002, P.R. China

会议类型: 国际会议

会议名称: 2011 Seventh International Conference on Natural Computation(第七届自然计算国际会议 ICNC 2011)

会议地点: 上海

会议语种:英文

页码: 699-902

在线出版日期: 2011-07-26（万方平台首次上网日期，不代表论文的发表时间）