首页|国土空间规划实施监测网络(CSPON)算法模型体系构建及应用

国土空间规划实施监测网络(CSPON)算法模型体系构建及应用

扫码查看
国土空间规划实施监测网络(CSPON)是打造"可感知、能学习、善治理、自适应"的智慧规划,推进国土空间治理"数智化"转型的重要手段。针对当前CSPON建设亟需科学系统技术方法支撑的需要,系统回顾国内外CSPON算法模型相关研究的发展历程与现状问题,在此基础上,提出以"大算力""大数据""大模型"技术为支撑,以"大场景"应用为导向的CSPON算法模型技术框架,重点分析"算力、数据、模型、场景"四大核心要素在CSPON算法模型建设中的定位与作用及其相互耦合机理。结合实践案例阐述CSPON算法模型在安全底线管控、规划传导管控、规划实施评估、空间格局优化等国土空间规划实施监测典型场景中的应用。研究结果可为加快推进CSPON技术方法完善与实践应用落地提供科学依据。
Construction and application of algorithm model system for China Spatial Planning Observation Network(CSPON)
The construction of the China Spatial Planning Observation Network(CSPON)is a crucial means to promote intelligent planning,that is"perceptive,capable of learning,good at governance,and self-realizing."It is also a key support for building a green and smart digital ecological civilization.Addressing the urgent need for scientific and systematic technical meth-ods in CSPON construction,this paper explores the methodological approach for CSPON algo-rithm model construction and application.It proposes a CSPON algorithm model technical framework supported by"high-performance computing","big data",and"large model"technol-ogies,oriented towards"large-scale scenario"applications.The paper illustrates its application in typical spatial planning implementation monitoring scenarios such as safety baseline control,planning transmission control,planning implementation assessment,and spatial pattern optimi-zation through practical cases.The research indicates that:(1)"High-Performance Computing"is the foundation supporting CSPON's growing demands for high-frequency data processing and dynamic computational analysis.It should fully integrate modern computing architectures such as cloud computing,edge computing,and distributed computing to strengthen CSPON's computational foundation.(2)"Big Data"is the cornerstone for constructing and operating CSPON algorithm model systems.The integration and mining of multi-source big data from"space,aerial,ground,sea,and network"can support algorithm models in more comprehensive-ly and accurately characterizing complex spatial systems.(3)"Large Model"is the"intelligent core"of CSPON.The"multi-model parallel use"mode,integrating traditional"small models"like GIS spatial analysis and machine learning algorithms with generative"large models"such as language and vision models,will become the mainstream trend in CSPON algorithm model application and development.(4)"Large-scale Scenarios"serve as the application carrier and target of CSPON.They should establish application scenarios combining"common scenarios"and"personalized scenarios"based on local characteristics and business needs,fully leveraging the roles and advantages of different CSPON algorithm models.The research results provide a scientific basis for accelerating the improvement of CSPON technical methods and their practi-cal application.

CSPONalgorithm model"High-Performance Computing""Big Data""Large Model"smart planning

张鸿辉、钟镇涛、杨丽娅、崔学森、徐华键、李文静

展开 >

自然资源部碳中和与国土空间优化重点实验室,南京 210023

广东国地规划科技股份有限公司,广州 510650

CSPON 算法模型 "大算力" "大数据" "大模型" 智慧规划

2024

自然资源学报
中国自然资源学会 中国科学院地理科学与资源研究所

自然资源学报

CSTPCDCSSCICHSSCD北大核心
影响因子:2.073
ISSN:1000-3037
年,卷(期):2024.39(12)