摘要
采用人工神经网络的建模技术建立了林分蓄积与立木密度、林分年龄和地位指数间的全林分生长模型,并以所建全林分生长模型为基础构造目标函数,采用改进单纯形法对建立的目标函数进行搜索寻优,以找到最佳的经营密度及对应的间伐时间。结果表明,拟合结果符合林分生长规律,可以用BP神经网络建模技术建立相应的全林分生长模型;用改进单纯形法对目标函数进行优化,不仅可以求得最优的保留密度,而且可以求得对应的间伐时间。
By adopting the artificial neutral network,the whole-stand growth BP model was established between stand stock and stand density,stand age and site index,then the objective function was constructed based on the whole-stand growth BP model,and finally the improved simplex method was applied to searching the best reserved density and thinning time.The results showed that BP neural network could be used to construct whole-stand growth model,and the fitting results met the stand growth law;through optimizing the objective function with the improved simplex method,the paper could obtain not only the best reserved density,but also the best thinning time.
出处
《安徽农业科学》
CAS
北大核心
2011年第23期14078-14081,共4页
Journal of Anhui Agricultural Sciences
基金
河南省科技攻关基金课题(0624050007)
关键词
马尾松
人工神经网络
全林分生长BP模型
改进单纯形
Pinus massoniana
Artificial neural network
Whole stand growth BP model
Improved simplex method
