中文摘要：

基于熵值法（EM）、粒子群算法（PSO）和支持向量机（SVM）的嵌套模型对中国石油安全进行评价及情景预测。首先,利用聚类和灰色关联分析方法构建出石油安全评价指标体系,并基于熵值法刻画出中国石油安全的历史演化规律。评价结果表明：（1）我国石油安全指数呈现出先降后升的趋势,且部分年份的石油安全指数波动较大;（2）国际原油价格波动、对外依存度不断上升以及市场流动性减弱的状况加剧了我国石油安全的脆弱程度,且地缘政治风险已成为我国石油安全的重要隐患,严重威胁着我国石油的供给安全。然后,在此基础上建立PSO-SVM石油安全预测模型,并通过设定三种情景方案,利用上述模型对我国2020年的石油安全状况进行情景分析。情景分析结果表明：（1）我国未来石油安全状况依然较为严峻,基准情景和低情景下石油安全指数仍旧处于很不安全区间;（2）不同的经济发展模式和政策取向对石油安全状况有着显著的影响。本文提供的＂情景—对策＂路径可成为相关能源政策制定的依据。

英文摘要：

This paper builds the nested model combining with Entropy method（EM）,Particle swarm optimization（PSO）and Support vector machine（SVM）to conduct evaluation and scenario prediction on the China＇s oil security.We first construct oil security evaluation indicator system by the methods of Cluster Analysis and Gray Correlation Analysis,and portray the history evolution law of China＇s oil security index.The evaluation results show that：1）China＇s oil security index decreased first and then increased,and there was a dramatic fluctuation in some years;2）the situation under which international crude oil price is highly volatile,oil external dependence degree is increasing gradually,market liquidity tends to weaken exacerbates the vulnerability of China＇s oil security;geopolitical risk has become an important risk of oil security in China,which seriously restricts China＇s supply security.Then we establish the forecasting model based on PSO-SVM model,and set three scenario schemes,conduct scenario analysis of China＇s oil security in 2020 based on the above model.The results of scenario analysis show that：1）the status of China＇s oil security will be much severe in the future;oil security index in benchmark scenario and low scenario will be in very insecure interval;2）different economic development pattern and policy orientation have a significant impact on the oil security.The scenario-countermeasure path presented in this paper is supposed to be the basis of energy policy formulation.