基于多元时序分析的水华预测及因素分析方法

在水华防治工作中,水华预测一直都是一个难点,为了解决由于水华生成过程中多种特征因素间的交互影响建模困难,现有的水华预测方法预测结果还不够准确,以及不同影响因素与水华发生的相关性程度的判定等问题,采用多元时间序列分析技术,研究多特征因素的水华预测及因素分析方法,通过对水华生成过程中的特征因素时序建模分析,提出多重潜周期多元自回归模型,给出了基于多元周期平稳时序分析的水华预测以及因素分析结果。采用本文方法及传统方法分别对江苏太湖水华特征因素监测数据进行建模预测,结果表明,基于本文方法的水华特征因素预测结果与实测结果更相符、预测平均误差绝对值更小。     

OEDGE modeling of plasma contamination efficiency of Ar puffing from different divertor locations in EAST

Modeling with OEDGE was carried out to assess the initial and long-term plasma contamination efficiency of Ar puffing from different divertor locations,i.e.the inner divertor,the outer divertor and the dome,in the EAST superconducting tokamak for typical ohmic plasma conditions.It was found that the initial Ar contamination efficiency is dependent on the local plasma conditions at the different gas puff locations.However,it quickly approaches a similar steady state value for Ar recycling efficiency ＞0.9.OEDGE modeling shows that the final equilibrium Ar contamination efficiency is significantly lower for the more closed lower divertor than that for the upper divertor.