Forest Growth Simulation Based on Artificial Neural Network

Abstract

The forest growth system is a complicated nonlinear system. The purpose of the paper is to solve the nonlinear system modeling of forest growth and to probe effective and feasible modeling technology of the forest growth for operating management of the forest resource. Based on survey data of 507 sample plots of pinus massoniana plantation, the forest growth model is established by using artificial neural network, which the network structure is 3:4:3.The total fitting precision of the model is 92% and the fitting precision of mean breast height diameter, mean height and volume is 90%, 95% and 92% respectively; while the total forecasting precision is 92% and the predicting precision of mean breast height diameter, mean height and volume is 87%, 95% and 92% successively, which can satisfy investigating and forecasting accuracy requirement of forest resource in forestry. The results showed that artificial neural network is effective and feasible modeling technology of the forest growth.