Prediction of Mature Wood Anatomical Properties of Pinus banksiana Plantation Based on Support Vector Machines （SVM） | 基于支持向量机（SVM）的班克松人工林成熟材解剖性质的预测
2013
Zhang Yan, Northeast Forestry University Harbin (China),College of Material Science and Technology | Song Kuiyan, Northeast Forestry University Harbin (China),College of Material Science and Technology | Tong Da, Northeast Forestry University Harbin (China),College of Material Science and Technology
中国人. 成熟材含量的高低决定木材性质的优劣，合理界定幼龄材与成熟材的分界点，准确预测成熟材材质有利于木材高效加工利用。为了确定人工林班克松的成熟期和预测成熟材解剖性质，采用支持向量机（SVM）界定幼龄材与成熟材的分界点，在此基础上利用幼龄材解剖性质预测成熟材解剖性质。结果表明：人工林班克松幼龄材与成熟材的分界点在树木生长的第18年；成熟期解剖性质明显优于幼龄期，变化较幼龄期平缓；成熟预测误差低、相关性高；预测曲线能够体现解剖性质整体的变化趋势，但对解剖性质测试集突变点及其之后的变化情况表现不足。
显示更多 [+] 显示较少 [-]英语. Wood qualities and properties were depended on the mature wood content. Determining the demarcation between juvenile and mature wood and predicting mature wood properties accurately play an important role on wood efficient utilization. In order to demarcate juvenile and mature wood of Pinus banksiana plantation and predict the anatomical properties of mature wood based on the anatomical properties of juvenile wood, support vector machines （SVM） was used. Results showed that the dividing point of juvenile and mature wood of Pinus banksiana plantation is 18＇h years. The anatomical properties of mature wood were excellent with gently change than those of juvenile one. Mature wood prediction has low deviation, high correlation. The prediction curves can indicate the overall trend of anatomical properties, while have weak performance on catastrophe point and the following changing properties of the test set.
显示更多 [+] 显示较少 [-]