基于BP神经网络的低丘红壤资源开发利用评价研究

Study on Investigation and Evaluation of Hilly Red Resources in Zhejiang Province by BP Neural Network

以遥感和GIS技术为支撑,对浙江省低丘红壤资源开发利用潜力进行了评价。以地形、土壤以及气候为基础,构建评价指标体系,采用BP神经网络法计算各评价单元综合得分,并将潜力值分为四个等级。评价结果表明:适宜开发、允许开发、限制开发和不适宜开发的低丘红壤资源分别为150916.31、183877.23、230796.59和1923219.13 hm2。评价结果与各指标的相关分析表明,开发潜力与土壤质地和坡度存在强相关。实地考察显示,评价结果与当地实际情况比较吻合。

The potential for exploiting and utilizing land resources in low hilly red soil region of Zhejiang Province was evaluated based on the technique of remote sensing and GIS.Terrain,soil and climate were considered in the assessment System.The integrated evaluation units were evaluated and computed by BP neural network,and the results were then grouped into four grades.The assessment results showed that the hilly red soil resources with feasible,permitted,limited,and disallowed exploiting and utilizing potential were 150916.31、183877.23、230796.59 和 1923219.13 hm^2,respectively.Soil texture and slope had high correlation coefficient with exploiting and utilizing potential.Compared with the ground truth,the assessment results were acceptable.