Based on Artificial Neural Network Modeling of White Birch Natural Forest at Daqing Mountain in Inner Mongolia

Abstract:

Make natural forest birch at Da Qingshan nature reserves in Inner Mongolia as the research object. The data is from the National Second-Class investigation data in 2006 by Inner Mongolia survey and design institute of forestry in 2006. Take 8 forest centre as study areas. All these datas would be sifted, and chosen the datas which the varieties of trees is white birch and the formation of the tree species is pure forest classes. The total of data is 4785. Use of Matlab software log-the sigmoid type function (logsig) and linear function (purelin) for the role of neurons. Based on the function of the concept of stand growth model, we choose age requirement (A), status level (N) and crown density (S) as input variables and the forest accumulation per hectare (M) as output variables to build and ttrain the stand growth BP artificial neural network model. And test the model fitting precision and inspection accuracy , the model fitting precision is 99.93%, inspection accuracy is 97.79%, these show that neural network modeling has better fitting precision and adaptability for the stand growth, and has good prediction ability.

• Periodical: Advanced Materials Research (Volumes 347-353)
• Pages: 16-21
• DOI: https://doi.org/10.4028/www.scientific.net/AMR.347-353.16
• Online since: October 2011
• Authors: Yang Xiao, Qiu Liang Zhang, Zhi Hua Zhao, Cheng Shuang Song
• Keywords: BPANN, Natural White Birch, Whole Stand Model

References

[1] Meng Xianyu. Forest Mensuration (the 3rd edition) [M].Beijing:Chinese Forestry Publishing House.2006.

[2] JiangZongli. Introduction to Artificial Neural Networks [m]. Higher Education Press, (2001)

[3] Guan，B．T．， Gertner，C．1991b． Modeling red pine tree survival with an artificial neural network．Forest Science，37：1429～1440.

[4] Hubert Hasenauer; Dieter Merkl; Martin Weingartner．2001．Estimating tree mortality of Norway spruce stands with neural networks．Advances in Environmental Research．Vo1．5, no 4，p．405～414.

DOI: 10.1016/s1093-0191(01)00092-2

[5] Foody，G．A．; Cutler， M．E．2002． Remote sensing of biodiversity： using neural networks to estimate the diversityand composition of a Bornean tropical rainforest from Landsat TM data．Geoscience and Remote Sensing Symposium，2002．IGARSS'02．2002 IEEE International．p．497～499.

DOI: 10.1109/igarss.2002.1025085

[6] Moisen G G;FreseinoTS．2002．ComParing five modelling teclmiques for Predieting forest characteristies．Eeologieal Modelling．Vol．157(2一3)：p．209～225.

[7] SusanL．King;KristinP．Bennett;ShannonList．2000．Modelingnoncatastro　 Phicindividual tree mortality using logistic regression，neural networks，and support vector methods．ComPuters and electronics in agriculture．Vol．27，No．l～3，P．401～406.

DOI: 10.1016/s0168-1699(00)00106-x

[8] DongChanghong.Matlab neural networks and Application: (the 2nd edition) [M].Beijing: National defence industry press.

[9] Jin Xingji,Jia Weiwei, Li Fengri. Whole Stand Growth Model for Natural Dahurian Larch Forests Based on BP ANN[J]. Bulletin of Botanical Research,2008,28(3):370-374.

[10] Alireza Givehchi,Gisbert Schneider.Impact of Descriptor Vector Scaling onthe Classification of Drugs and Nondrugs with Artificial Neural Networks.Journal of Molecular Modeling,2004，10(3):204一211.

DOI: 10.1007/s00894-004-0186-9

[11] Zhu Daqi,Shihui. Theory and Application of Artificial Neural Network[M].Beijing: Science Press,2006.