乡村振兴视角下中国粮食安全水平的统计评价、动态演进与智能预警  

Statistical evaluation,dynamic evolution,and intelligent early warning of China’s food security level from the perspective of rural revitalization

作　　者：王慧玲 李松松 曹丹丹 WANG Huiling;LI Songsong;CAO Dandan(School of Finance,Nanjing Audit University,Nanjing 211815;School of Statistics and Data Science,Nanjing Audit University,Nanjing 211815,China)

机构地区：[1]南京审计大学金融学院,南京211815 [2]南京审计大学统计与数据科学学院,南京211815

出　　处：《干旱区资源与环境》2025年第4期1-12,共12页Journal of Arid Land Resources and Environment

基　　金：2024年度教育部人文社会科学研究青年基金项目(24YJC790169);江苏省2023年度高校哲学社会科学研究项目一般项目(2023SJYB0341)资助。

摘　　要：探究科学合理的粮食安全评价体系,刻画其动态演进趋势,并对风险点进行有效智能预警对于高质量保障国家和人民的粮食安全具有重要的政策意蕴和现实意义。文中利用2012-2021年31个省、自治区和直辖市的面板数据,从生产能力、基础支撑、绿色发展、国际贸易、风险隐患、民生福祉6个方面探索性构建粮食安全评价指标体系,采用实数编码的加速遗传算法优化的投影寻踪模型评估粮食安全水平,利用Kernel核密度方法、Markov链、热力图刻画粮食安全动态演进趋势,并基于粒子群优化后的BP神经网络开发了粮食安全智能预警模型。研究发现:1)粮食作物播种面积和粮食自给率是评估粮食安全的关键因素。2)中国粮食安全发展水平总体呈现上升态势,但动态演进过程缓慢、区域间发展不协调、区域差异扩大。3)西藏、青海、贵州、广西和海南被列入未来四年粮食安全高风险预警省份。Exploring a scientific and reasonable food security evaluation system,portraying its dynamic evolution trend,and providing effective and intelligent early warning of risk points have important policy implications and practical significance for guaranteeing the food security of the country and the people in high quality.Based on the panel data of 31 provinces,autonomous regions and municipalities directly under the central government,we construct an exploratory food security evaluation index system from six aspects,namely,production capacity,basic support,international trade,risks,and people's well-being,adopt the projection tracing model optimized by accelerated genetic algorithm with real number coding to assess the level of food security,and use Kernel kernel density method,Markov chain,and heat map to portray the dynamic and evolutionary trend of food security,and develop an intelligent early warning model for food security based on particle swarm optimized BP neural network.The study finds that:1)The sowing area of grain crops and the grain self-sufficiency rate are the key factors for assessing food security.2)China's food security development level generally shows an upward trend,but the dynamic evolution process is slow,inter-regional development is not coordinated,and the regional differences are widening.3)Tibet,Qinghai,Guizhou,Guangxi,and Hainan are included as food security high-risk early warning provinces in the next 4 years.

关 键 词：粮食安全 投影寻踪模型 BP神经网络 智能预警 空间演进 

分 类 号：F307.11[经济管理—产业经济]