城市土地集约利用程度的人工神经网络判定——以深圳市为例

研究目的:探索适合城市土地集约利用综合评价的非线性定量方法。研究方法:人工神经网络(ANN)模型及实证分析。研究结果:深圳市土地集约利用ANN模型评价结果显示,全市土地集约利用水平基本分4类。即集约利用-效益均衡型、集约/适度利用-效益偏好型、适度/低度利用-效益不均衡型和低度利用型。研究结论:基于ANN的城市土地集约利用评价模型,在处理不同量纲指标和减少人为主观因素上具有优势,其结果不仅能反映土地集约利用程度的区域差异,也能反映各指标对土地集约利用水平的贡献

Urban Land Use Intensity Assessment based on Artificial Neural Networks: A Case of Shenzhen

The purpose of the paper is to explore a nonlinear quantified method for urban land use intensity assessment which is integrated into the revision of comprehensive land use planning. Method of artificial neural network model (ANN) was employed. The results indicate the land use intensity can be classified into four types: intensive with benefit-balance, intensive/moderate with economic/ecological preference, moderate with non-benefit-balance and relative extensive. The ANN model takes advantages in treating indexes with different dimensions and in decreasing the subjective influence. The results of the model may reflect not only the regional differences but also the contributions of all indexes to land use intensity.