|
| 基于聚类算法与多智能体的细碎化耕地空间优化研究 |
| Spatial optimization of fragmented farmland based on clustering algorithm and multi-agent system |
| 投稿时间：2024-06-20 |
| DOI：10.13254/j.jare.2024.0438 |
| 中文关键词: 耕地细碎化，聚类算法，多智能体，耕地空间优化，新北区 |
| 英文关键词: farmland fragmentation, clustering algorithm, multi-agent system, farmland spatial optimization, Xinbei District |
| 基金项目:国家自然科学基金项目（41571082） |
| 作者 | 单位 | E-mail | | 黄炳元 | 南京大学地理与海洋科学学院, 南京 210023
自然资源部海岸带开发与保护重点实验室, 南京 210023
自然资源部国土卫星遥感应用重点实验室, 南京 210023
江苏省地理信息技术重点实验室, 南京 210023 | | | 黄秋昊 | 南京大学地理与海洋科学学院, 南京 210023
自然资源部海岸带开发与保护重点实验室, 南京 210023
自然资源部国土卫星遥感应用重点实验室, 南京 210023
江苏省地理信息资源开发与协同创新中心, 南京 210023 | qhhuang@nju.edu.cn | | 杨力虹 | 南京大学地理与海洋科学学院, 南京 210023
自然资源部海岸带开发与保护重点实验室, 南京 210023
自然资源部国土卫星遥感应用重点实验室, 南京 210023
江苏省地理信息技术重点实验室, 南京 210023 | | | 陈振杰 | 南京大学地理与海洋科学学院, 南京 210023
自然资源部海岸带开发与保护重点实验室, 南京 210023
自然资源部国土卫星遥感应用重点实验室, 南京 210023 | |
|
| 摘要点击次数: 207 |
| 全文下载次数: 258 |
| 中文摘要: |
| 通过微观地块调整实现耕地资源的规模利用，可以解决区域耕地细碎化治理难题，保障粮食安全和实现农业可持续发展。本探究运用景观格局指数和综合加权叠加法对常州市新北区耕地细碎化和土地资源宜耕性进行评价，使用聚类算法划分耕地细碎化类型，并构建多智能体耕地空间优化模型。结果表明：使用K-medoids聚类算法将现状耕地划分为连片规整类、形状复杂类和分散破碎类，其中非连片规整类耕地占比为53.68%，耕地细碎化较为严重；研究区土地资源宜耕性总体水平较高，高适宜性和较高适宜性地块占比分别为14.03%和44.83%，空间分布呈北高南低、西高东低的特征；多智能体模型能够有效实现细碎化耕地的空间优化，优化后耕地宜耕性提高，空间布局更加连续和聚集且形状更加规整；研究区内平衡情景可取得更好的优化效果，但会造成更大规模的耕地调整和宜耕性较高的乡镇承担更多的耕地恢复任务。本研究构建的耕地空间优化模型具有良好的优化效果，可为区域耕地细碎化治理提供科学决策依据。 |
| 英文摘要: |
| To achieve large-scale utilization of cultivated land resources through micro-parcel adjustments, addressing the challenges of fragmented cultivated land governance, ensuring food security, and promoting sustainable agricultural development, this study applied landscape pattern indices and a comprehensive weighted overlay method to evaluate the fragmentation and arable land resource suitability in the Xinbei District of Changzhou City. The K-medoids clustering algorithm was used to categorize types of farmland fragmentation, and a multi-agent model was developed to optimize the spatial layout of the cultivated land. The results showed that using the K-medoids clustering algorithm, cultivated land was classified into contiguous and regular, complex-shaped, and scattered and fragmented types. The proportion of non-contiguous and regular cultivated land was 53.68%, suggesting severe cultivated land fragmentation. The overall level of land resource suitability for cultivation in the study area was relatively high, with the proportion of highly and relatively highly suitable plots being 14.03% and 44.83%, respectively. The spatial distribution showed a pattern of higher in the north and west and lower in the south and east. The multi-agent model effectively optimized the spatial arrangement of fragmented farmlands, improving overall arable land suitability, creating a more continuous, clustered, and regular-shaped layout. A scenario involving intra-regional balance achieved better optimization results, but required larger-scale farmland adjustments, with areas of higher arable land resource suitability taking on more responsibility for land restoration. The farmland spatial optimization model developed in this study demonstrates strong optimization effectiveness and provides a scientific basis for addressing regional farmland fragmentation. |
| HTML
查看全文
查看/发表评论 下载PDF阅读器 |
| 关闭 |
|
|
|
