基于GA-BP神经网络的树木生长模型研究

Research on Tree Growth Model Based on GA-BP Neural Network

在森林资源的实际调查中,树木胸径、树高和蓄积量的测量存在难度大、成本高、精度低等问题。基于树木生长与林分因子、生长环境之间的相关性,提出了一种利用遗传算法优化BP神经网络的树木生长模型来预测树木的胸径、树高和蓄积量,通过优化BP神经网络的权值和阈值,建立GA-BP模型,并与多元线性回归模型的预测结果进行了比较。结果表明:该模型的预测结果比多元线性回归模型更接近于实测值,GA-BP模型的最小R~2高于多元线性回归模型且更接近于1。采用遗传算法优化后的模型具有更高的预测精度,对景宁县树木生长预测具有指导意义。

In the actual survey of forest resources,the measurement of tree diameter,tree height and stock volume is difficult,costly and of low precision.Based on the correlation between tree growth and stand factors and its growing environment,a tree growth model based on genetic algorithm optimizing BP neural network was proposed to predict the DBH,tree height and accumulation of trees.By optimizing the weight and thresholds of BP neural network,a GA-BP model is established and compared with the prediction results of multivariate linear regression model.The experimental results show that the predicted results of this model are closer to the actual values than that of multiple linear regression model,and the minimum R2 of GA-BP model is higher than that of multiple linear regression model and closer to 1.The model optimized by genetic algorithm has higher prediction accuracy,which has guiding significance for tree growth prediction in Jingning County.