地理研究 ›› 2017, Vol. 36 ›› Issue (3): 553-572.doi: 10.11821/dlyj201703013

• 研究论文 • 上一篇    下一篇

基于粒子群算法的大城市近郊区景观格局优化研究——以成都市龙泉驿区为例

欧定华(), 夏建国()   

  1. 四川农业大学资源学院,成都 611130
  • 收稿日期:2016-07-14 修回日期:2016-12-05 出版日期:2017-03-20 发布日期:2017-03-22
  • 作者简介:

    作者简介:欧定华(1984- ),男,四川宜宾人,博士研究生,主要从事景观生态规划与设计、土地利用规划与管理、“3S”技术应用研究。E-mail:357881550@qq.com

  • 基金资助:
    国家自然科学基金项目(31270498);四川省学术和技术带头人培养经费(2014);四川农业大学双支计划项目(2015)

Landscape pattern optimization in peri-urban areas based on the particle swarm optimization method: A case study in Longquanyi District of Chengdu

Dinghua OU(), Jianguo XIA()   

  1. College of Resources, Sichuan Agricultural University, Chengdu 611130, China
  • Received:2016-07-14 Revised:2016-12-05 Online:2017-03-20 Published:2017-03-22

摘要:

景观格局优化是实现区域生态安全的重要途径。以成都市龙泉驿为研究区,在景观适宜性评价、景观数量优化基础上,构建PSO景观格局空间优化模型与求解算法,对经济发展、生态保护、统筹兼顾情景景观空间布局进行优化。结果表明:基于PSO的景观格局空间优化模型与算法能利用粒子位置模拟景观分布进行空间格局优化,实现了数量与空间优化的有机耦合,是景观格局优化的有效方法。目标年,经济发展情景优势景观为城乡人居及工矿、果园,景观格局呈现出西部坝区以城乡人居及工矿、农田为主,东部山区以果园为主的分布特征;生态保护情景优势景观为森林、城乡人居及工矿,景观格局呈现出西部坝区以城乡人居及工矿、果园、农田为主,东部山区以森林为主的分布特征;统筹兼顾情景优势景观为森林、城乡人居及工矿和果园,景观格局呈现出西部坝区以城乡人居及工矿、农田为主,东部山区以森林、果园为主的分布特征。统筹兼顾情景方案未来潜在可能性最大,其经济、生态、综合效益均得以优化提升,是目标年研究区最佳景观格局空间布局方案。

关键词: 城市近郊区, 景观格局, 空间优化, 粒子群算法, 约束最优化方法, Logistic回归模型

Abstract:

Landscape pattern optimization is one of the important ways in achieving regional ecological security. In order to optimize landscape spatial layout of economic development, ecological protection and overall consideration scenario, a case study was carried out in Longquanyi District of Chengdu based on landscape suitability assessment and landscape quantity optimization. The method of particle swarm optimization (PSO) was used to establish landscape pattern spatial optimization model and algorithm. The results showed that, the landscape pattern spatial optimization model and algorithm based on PSO was the effective method of landscape pattern optimization. The model and algorithm could efficiently simulate landscape distribution by using particle space positions, conduct spatial pattern optimization, and realize united coupling of quantity structure and spatial distribution optimization. The dominant landscapes in economic development scenario were orchard, urban-rural residential and industrial-mining area in a target year. The landscape pattern showed that farmland, urban-rural residential and industrial-mining area dominated the western flatland region, while eastern mountainous area was mainly dominated by orchard. The dominant landscapes in ecological protection scenario were forest, urban-rural residential and industrial-mining area. The landscape pattern showed that the western flatland region was mainly dominated by farmland, orchard, urban-rural residential and industrial-mining area, while the eastern mountainous area was dominated by forest. The dominant landscapes in overall consideration scenario were forest, orchard, urban-rural residential and industrial-mining area. The landscape pattern showed that farmland, urban-rural residential and industrial-mining area dominated the western flatland region, while eastern mountainous area was dominated by forest and orchard. Compared with other scenarios, the overall consideration scenario could be the largest potential possibility in the future, since the economic, ecological and comprehensive benefits here would be the most optimized and promoted, implying the best spatial layout scheme for landscape pattern in the study area in the target year.

Key words: peri-urban areas, landscape pattern, spatial optimization, particle swarm optimization, constrained optimization method, Logistic regression model