计算机工程与应用
計算機工程與應用
계산궤공정여응용
COMPUTER ENGINEERING AND APPLICATIONS
2013年
17期
263-265
,共3页
神经网络%规范化方法%反向传播(BP)算法%固碳模型
神經網絡%規範化方法%反嚮傳播(BP)算法%固碳模型
신경망락%규범화방법%반향전파(BP)산법%고탄모형
neural network%normalization methods%Back Propagation(BP)algorithm%carbon fixation model
为了实现林木固碳释氧量的数字化估算，针对现有估算方法的不足，提出了基于BP神经网络的林木固碳释氧量的预测模型。基于对神经网络理论和固碳释氧量估算模型的研究，分析了林木在生长季节的CO2通量变化趋势，采用规范化方法对训练样本预处理，进行BP神经网络训练，并结合弛豫涡旋积累法和箱式法，建立了CO2通量神经网络模型。实验结果表明，所建模型具有较好的泛化性能，能够比较准确地估算出林木的固碳释氧量。
為瞭實現林木固碳釋氧量的數字化估算，針對現有估算方法的不足，提齣瞭基于BP神經網絡的林木固碳釋氧量的預測模型。基于對神經網絡理論和固碳釋氧量估算模型的研究，分析瞭林木在生長季節的CO2通量變化趨勢，採用規範化方法對訓練樣本預處理，進行BP神經網絡訓練，併結閤弛豫渦鏇積纍法和箱式法，建立瞭CO2通量神經網絡模型。實驗結果錶明，所建模型具有較好的汎化性能，能夠比較準確地估算齣林木的固碳釋氧量。
위료실현림목고탄석양량적수자화고산，침대현유고산방법적불족，제출료기우BP신경망락적림목고탄석양량적예측모형。기우대신경망락이론화고탄석양량고산모형적연구，분석료림목재생장계절적CO2통량변화추세，채용규범화방법대훈련양본예처리，진행BP신경망락훈련，병결합이예와선적루법화상식법，건립료CO2통량신경망락모형。실험결과표명，소건모형구유교호적범화성능，능구비교준학지고산출림목적고탄석양량。
In order to achieve the digital estimates of quantity of forest carbon fixation and oxygen release, a forest carbon fixa-tion release oxygen prediction model based on BP neural network is proposed according to the shortage of existing estimation methods. In addition to studying neural network theory and carbon fixation oxygen release quantity estimation model, it analyzes forest CO2 flux trends in the growing season, the training samples are pretreated using normalization methods, BP neural net-work training is conducted, and a neural network model of CO2 flux is established combining relaxation eddy accumulation method and chamber method. Experimental results show that the model has good generalization performance and can more accurately estimate the quantity of carbon fixation and oxygen release of forest.
