CAJ | 학술논문

将BP神经网络应用于城市土地集约利用评价研究中，可以排除人为设定权重的主观因素对评价结果的影响，能够准确评价土地集约利用类别，使评价结果更加客观、科学、合理。将评价等级分为集约利用、适度利用、低度利用和粗放利用，从土地利用、土地投入、土地产出三方面，选择9个评价指标构建评价指标体系，并利用MATI。AB中实现的人工神经网络模型对唐山市土地进行集约利用评价。结果显示：唐山市低度利用的土地占26．4％，适度利用的土地面积占22．7％，集约利用的土地占40．6％，过度利用的土地占10．3％，土地利用强度较低，还有一定的挖掘潜力。
장BP신경망락응용우성시토지집약이용평개연구중，가이배제인위설정권중적주관인소대평개결과적영향，능구준학평개토지집약이용유별，사평개결과경가객관、과학、합리。장평개등급분위집약이용、괄도이용、저도이용화조방이용，종토지이용、토지투입、토지산출삼방면，선택9개평개지표구건평개지표체계，병이용MATI。AB중실현적인공신경망락모형대당산시토지진행집약이용평개。결과현시：당산시저도이용적토지점26．4％，괄도이용적토지면적점22．7％，집약이용적토지점40．6％，과도이용적토지점10．3％，토지이용강도교저，환유일정적알굴잠력。
The application of BP neural network to the study of intensive use of urban land can exclude the influence on the evaluation results of the subjective factors, evaluate the intensive land use category accurately, and make the evaluation results more objective, scientific and reasonable. This paper selects 9 evaluation indexes from the aspects of land use, land investment and land output, and evaluates the intensive land use of Tangshan City, using BP neural network model achieved in MATLAB. The evaluation grades will be divided into intensive use, moderate use, low use and extensive use. The results show that 26.4% of land is low use, 22.7% moderate use, 40. 6% intensive use, and 10. 3% excessive use. As land use intensity is relatively low, there is a certain potential to be exploited.