CAJ | 학술논문

在分析测量数据的基础上，提取红树的平均基径、基径数、平均胸径、胸径数等特征参数，建立了预测红树株高的人工神经网络模型。采用Levenberg—Marquardt优化算法改进了BP神经网络算法；采用训练好的BP神经网络模型对距堤坝25，50，75m3个采样点的株高进行预测，预测值和实测值的均方根误差分别为0．0006，0．0022，0．0041，相关系数分别为0．99，0．95，0．94。结果表明利用BP神经网络对红树株高进行预测是可行的。
재분석측량수거적기출상，제취홍수적평균기경、기경수、평균흉경、흉경수등특정삼수，건립료예측홍수주고적인공신경망락모형。채용Levenberg—Marquardt우화산법개진료BP신경망락산법；채용훈련호적BP신경망락모형대거제패25，50，75m3개채양점적주고진행예측，예측치화실측치적균방근오차분별위0．0006，0．0022，0．0041，상관계수분별위0．99，0．95，0．94。결과표명이용BP신경망락대홍수주고진행예측시가행적。
Based on the test data analysis method, characteristic parameters of mangroves,i, e. average basal diameter, the number of basal diameter, average diameter at breast-high （DBH） , the number of DBH were extracted, and the artificial neural network model was developed to predict the height of mangroves. The back propagation（BP） algorithm was improved by using the Levenberg-Marquardt optimizing arithmetic. The heights of mangroves in the sampling points which are 25,50 and 75 m away from the dam were predicted using the trained BP neural network, standard errors are 0. 000 6,0. 002 2 and 0. 004 1 ,with the correlation coefficients of 0. 99,0. 95 and 0. 94. Results show that using BP neural network to predict the height of mangroves is feasible.