生态学报2018，Vol.38Issue(1)：55-64,10.DOI:10.5846/stxb201612082530
基于微粒群-马尔科夫复合模型的生态空间预测模拟——以长株潭城市群为例
Predictive simulation of ecological space based on a particle swarm optimization-Markov composite model: A case study for Chang-Zhu-Tan urban agglomerations
摘要
Abstract
The distribution of ecological space has important functions,and its scientific prediction can provide a basis for making decisions regarding protecting the ecological security of a national landscape.A particle swarm optimization-Markov (PSO-Markov) composite model based on a risk assessment of ecological space was conducted using ArcGIS and MATLAB.The ecological space of the Chang-Zhu-Tan urban agglomeration in 2020 was predicted based on land-use data for 2013,and the resulting particle swarm was used to reconstruct the ecological space.The PSO-Markov composite model was established according to the following steps:First,a particle swam was selected and designed,and a 2000 m × 2000 m square unit was selected as the basic particle.Second,the particle was initialized from low to high risk based on the principle of ecological space.Third,a fitness function was established,and the value of ecological space risk was used to determine the spatial pattern of ecological space.Finally,the spatial position and speed of a particle swarm were updated according to the history of optimal value and the global optimal value of the particle swarm.The PSO-Markov composite model is a relatively new method for land-use pattern prediction.The quantitative scale of ecological space is predictable with an improved Markov model,and the pattern of ecological space is predictable with a PSO model.Thus,the PSO-Markov composite model has the following features compared with other models:First,this model can yield more reasonable quantitative predictions.Second,it utilizes a large search area and thoroughly considers local and global influence.Third,it is less influenced by the problem of dimension change,and it has a significant advantage regarding the solution of multiobjective problems.Finally,with short convergence time and high arithmetic speed,it is easy to realize.The ecological space of the Chang-Zhu-Tan urban agglomeration was predicted to decrease by 2020 with woodland and unused land changing the most dramatically and spatial variation primarily concentrating in the southwest region.The decrease in total ecological space area was shown to be primarily due to the expansion of land development.Thus,we must control the population density of urban agglomerations and optimize the structure and spatial distribution of agriculture/industry,human settlements,and the ecosystem in the urban agglomeration.In particular,we must reasonably plan and utilize urban developed land,as well as make full use of the ecological value of water bodies and undeveloped land,with emphasis on the protection of ecological resources,ecological corridors,and key ecological areas so as to establish a rationally structured,functioning ecological net system and to improve the ecological services of the ecosystem.Urban and rural spatial structure must be regulated rationally under the guidance of city planning.Moreover,measures should be carried out to protect and improve the quality of the environment and to enhance landscape diversity.These are our priorities for the future.关键词
生态空间/预测/微粒群-马尔科夫复合模型/长株潭城市群Key words
ecological space/prediction/particle swarm optimization (PSO)-Markov composite model/Chang-Zhu-Tan urban agglomeration引用本文复制引用
陈永林,谢炳庚,钟典,吴亮清,张爱明..基于微粒群-马尔科夫复合模型的生态空间预测模拟——以长株潭城市群为例[J].生态学报,2018,38(1):55-64,10.基金项目
国家重点研发计划项目(2016YFC0502406) （2016YFC0502406）
教育部人文社会科学重点研究基地重大项目(14JJD720016) （14JJD720016）
江西省社科规划项目(15SH10) （15SH10）
江西省高校人文社会科学项目(SH1401) （SH1401）
江西省教育厅科技项目(GJJ151018) （GJJ151018）
江西省教育厅教改项目(JXJG-15-14-9) （JXJG-15-14-9）
