土地利用配置的混沌蚁群优化算法研究
作者简介:陆军辉(1992-),男,湖南岳阳人,硕士生,主要从事土地利用模拟与优化、人工智能等研究。E-mail: lujunhui_gis@foxmail.com
收稿日期: 2017-03-15
要求修回日期: 2017-06-01
网络出版日期: 2017-08-20
基金资助
国家自然科学基金项目(41001078)
Land Use Optimization Allocation Based on Chaos Ant Colony Algorithm
Received date: 2017-03-15
Request revised date: 2017-06-01
Online published: 2017-08-20
Copyright
土地利用优化配置是促进土地可持续发展的重要举措,然而现有研究缺乏有效求解土地利用优化配置模型的新型混合式智能优化算法。本文结合蚁群算法和混沌模型,形成混沌蚁群优化(Chaos Ant Colony Optimization,CACO)算法,并以广州市增城区为研究区,对土地利用现状进行优化配置;然后在数量结构、目标函数值、空间布局等方面将优化结果与土地现状及标准蚁群算法优化结果进行对比分析。结果表明:① CACO算法能在满足多种约束条件下,有效解决多目标土地利用优化配置问题;② 与标准蚁群算法相比,CACO算法能增加土地利用的经济效益7.18亿元、生态效益0.33亿元、社会效益1.13%,同时降低地类转换成本1.15%;③ CACO算法能使土地利用现状空间分布多样性和均匀性的下降控制在1.30%以内,同时缩减地块数量8.86%,并使平均斑块大小增加9.77%,从而提升土地集约利用水平,更合理地配置各现状地类的空间分布,为研究区土地利用的科学规划与决策提供支持。
陆军辉 , 梅志雄 , 赵书芳 , 肖艳云 . 土地利用配置的混沌蚁群优化算法研究[J]. 地球信息科学学报, 2017 , 19(8) : 1026 -1035 . DOI: 10.3724/SP.J.1047.2017.01026
The optimal allocation of land use is an important and effective measures of promoting the sustainable development of the land. However, existing research was lack of efficient methods in the optimization allocation for the quantitative structure and spatial layout of land use by using original mixed algorithm. Therefore, this paper combined the ant colony optimization algorithm (ACO) with the chaos model, and proposed a hybrid and self-adapt chaos ant colony optimization algorithm (CACO). After that, in order to verify the feasibility and efficiency of the CACO, the Zengcheng district of Guangzhou was selected as the study case. The CACO was utilized to solve the model of land use optimization allocation based on the evaluation of actual land sustainable use. Finally, this study made some comparative analysis of the results of the CACO and the actual land use and the results of the ACO respectively in three main aspects: the quantitative structure of land use, the spatial layout of land use and the multiple objective functions. The results showed that: firstly, the CACO can effectively solve the complex problems of multi-objective land use optimization allocation under multiple constraint conditions; secondly, compared with the ACO, the coordination between economic benefit and ecological effectiveness in the CACO was weakened slightly. The CACO rose all others objective functions’ values. For example, economic benefits increased by 7.18 billion yuan, ecological effectiveness increased by 0.33 billion yuan, social benefits increased by 1.13%, while the land conversion costs shrank by 1.15%. Thirdly, compared with the ACO, the CACO decreased the diversity and evenness index of actual land use spatial distribution within 1.30%, made the number of total land patches reduced about 8.86%, and the average patches size increased 9.77%. The level of the land intensive use was improved. Therefore, the CACO could reasonably optimize actual various land use types to appropriate spatial layout, and supply useful technical support for scientific land use planning and decisions making.
Fig. 1 Study area and its actual land use types in 2014图1 研究区及其2014年土地利用现状 |
Fig. 2 Land use suitability assessment results in Zengcheng图2 增城区土地利用适宜性评价结果 |
Tab. 1 Economic benefit coefficients of farmland from 2005 to 2014表1 2005-2014年增城区耕地经济效益系数 |
年份 | 年末耕地面积/hm2 | 增加值/亿元 | 经济效益系数/(万元/hm2) |
---|---|---|---|
2005 | 37 112 | 172 407 | 4.646 |
2006 | 27 188 | 193 504 | 7.117 |
2007 | 27 082 | 203 177 | 7.502 |
2008 | 27 014 | 223 454 | 8.272 |
2009 | 27 120 | 240 506 | 8.868 |
2010 | 27 114 | 270 515 | 9.977 |
2011 | 27 060 | 278 903 | 10.307 |
2012 | 27 042 | 309 373 | 11.440 |
2013 | 27 020 | 345 337 | 12.781 |
2014 | 26 570 | 372 281 | 14.011 |
2020 | - | - | 24.160 |
Tab. 2 Economic and ecological benefit coefficients of different land use types in 2020表2 2020年增城区各地类经济效益系数和生态效益系数 |
变量 | 耕地 | 园地 | 林地 | 草地 | 水域 | 城镇 | 农村 | 交通 | 其他 |
---|---|---|---|---|---|---|---|---|---|
权重 | 1.00 | 1.54 | 1.50 | 1.54 | 0.55 | 7.02 | 0.67 | 0.67 | 0 |
经济效益系数/(万元/hm2) | 17.33 | 26.69 | 26.00 | 26.69 | 9.53 | 121.67 | 11.61 | 105.37 | 0.10 |
生态效益系数/(万元/ hm2) | 6.83 | 7.06 | 8.01 | 7.19 | 9.44 | 0 | 0 | 0 | 7.19 |
Fig. 3 Model framework of CACO algorithm图3 CACO算法模型框架图 |
Tab. 3 Optimization results for land use quantitative structures表3 增城区土地利用数量结构优化结果 |
统计类型 | 耕地 | 园地 | 林地 | 草地 | 水域 | 城镇 | 农村 | 交通 | 其他 | |
---|---|---|---|---|---|---|---|---|---|---|
优化下限 | 像元数/个 | 112 219 | 83 809 | 245 735 | 0 | 32 989 | 59 225 | 0 | 24 600 | 0 |
优化上限 | 像元数/个 | 646 671 | 646 671 | 646 671 | 26 190 | 32 989 | 96 756 | 14 664 | 24 600 | 15 938 |
地类现状 | 像元数/个 | 112 219 | 83 829 | 229 329 | 26 211 | 32 989 | 59 225 | 62 331 | 24 600 | 15 938 |
占比/% | 17.35 | 12.96 | 35.46 | 4.05 | 5.10 | 9.16 | 9.64 | 3.80 | 2.48 | |
斑块密度 | 0.106 | 0.236 | 0.027 | 0.548 | 0.244 | 0.069 | 0.580 | 0.606 | 0.788 | |
形状指数 | 5.59 | 3.20 | 20.76 | 2.12 | 4.34 | 8.96 | 1.81 | 3.03 | 1.71 | |
ACO | 像元数/个 | 112 505 | 74 405 | 260 600 | 21 006 | 32 989 | 94 993 | 14 879 | 24 600 | 10 694 |
占比/% | 17.40 | 11.51 | 40.30 | 3.25 | 5.10 | 14.69 | 2.30 | 3.80 | 1.65 | |
斑块密度 | 0.105 | 0.237 | 0.026 | 0.534 | 0.244 | 0.068 | 0.597 | 0.606 | 0.788 | |
形状指数 | 5.81 | 3.21 | 21.16 | 2.11 | 4.34 | 9.09 | 1.78 | 3.03 | 1.71 | |
CACO | 像元数/个 | 113 808 | 73 175 | 263 057 | 20 057 | 32 989 | 97 510 | 10789 | 24 600 | 10 686 |
占比/% | 17.60 | 11.32 | 40.68 | 3.10 | 5.10 | 15.08 | 1.67 | 3.80 | 1.65 | |
斑块密度 | 0.096 | 0.210 | 0.023 | 0.500 | 0.244 | 0.062 | 0.565 | 0.606 | 0.788 | |
形状指数 | 5.54 | 3.18 | 20.75 | 2.06 | 4.34 | 8.90 | 1.72 | 3.03 | 1.70 |
Tab. 4 The comparison of ACO and CACO on object functions optimization results表4 ACO和CACO的目标函数优化结果比较 |
统计类型 | 经济效益/亿元 | 生态效益/亿元 | 社会效益 | 经济生态协调性 |
---|---|---|---|---|
现状 | 542.048 | 95.242 | 466 812.60 | 1 293 318 |
ACO | 647.748 | 98.011 | 493 145.00 | 1 293 915 |
ACO增幅/% | 19.500 | 2.908 | 5.641 | 0.046 |
CACO | 654.925 | 98.337 | 498406.40 | 1 293 971 |
CACO增幅/% | 20.824 | 3.249 | 6.768 | 0.050 |
CACO-ACO | 7.1766 | 0.325 | 5261.400 | 56 |
Tab. 5 Comparison of landscape index for spatial optimization results between ACO and CACO表5 ACO和CACO优化空间结果的景观指数对比 |
统计类型 | 斑块数量 | 平均斑块大小 | 香农多样性指数 | 辛普森均匀度指数 | 香农均匀度指数 | 运算时间/h |
---|---|---|---|---|---|---|
现状 | 23 713 | 6.82 | 1.75 | 0.86 | 0.79 | - |
ACO | 23 567 | 6.86 | 1.74 | 0.86 | 0.79 | 9.66 |
CACO | 21 478 | 7.53 | 1.72 | 0.86 | 0.78 | 8.89 |
CACO比ACO增减幅度/% | -8.86 | 9.77 | -1.15 | 0.00 | -1.27 | -7.97 |
Fig. 4 Comparison of evolution curves between ACO and CACO图4 ACO和CACO进化曲线比较 |
Fig. 5 Comparison of land use spatial optimization patterns between ACO and CACO in 2020图5 ACO和CACO的2020年增城区土地利用空间优化布局结果对比 |
The authors have declared that no competing interests exist.
[1] |
[
|
[1] |
[
|
[3] |
[
|
[4] |
[
|
[5] |
[
|
[6] |
[
|
[7] |
[
|
[8] |
|
[9] |
[
|
[10] |
[
|
[11] |
[
|
[12] |
|
[13] |
[
|
[14] |
[
|
[15] |
[
|
[16] |
|
[17] |
[
|
[18] |
[
|
[19] |
[
|
[20] |
[ Zhang H H, Zeng Y N, Tan R, et al. A model for regional land use optimization allocation based on muti-agent system and its application[J]. Acta Geographica Sinica, 2011,66(7):972-984. ]
|
[21] |
[
|
[22] |
[
|
[23] |
|
[24] |
|
[25] |
|
[26] |
[
|
[27] |
[
|
[28] |
|
[29] |
|
[30] |
[
|
[31] |
|
[32] |
[
|
/
〈 | 〉 |