摘要
蒙特卡罗模拟方法是进行核电参数不确定性分析、得到统计意义楼层谱的基本方法之一,但该方法要求大样本循环,计算成本较高。为提高计算效率,基于并行理论改进蒙特卡罗模拟样本循环模式,在MATLAB软件平台上完成了楼层反应谱概率分析并行程序的开发。数值算例表明,该程序实现了核电结构的随机参数大样本空间分析,能够有效获取不同保证率的楼层反应谱,极大提高了运算效率,同时也保证了楼层谱统计结果的计算精度。
Generally,uncertainty analysis is carried out by applying Monte Carlo method,but it is time-consuming by employing the thread of serial programming and make deterministic seismic analysis on a large quantity of random samples.This paper analyzes the reasons for low computational efficiency of conventional methods based on parameter uncertainty of nuclear power plant structures.Program was made by MATLAB to achieve parallel computing ability.Therefore,the operation time of Monte Carlo process was sharply shortened.Numerical examples indicate that not only adequate sample space for accurate calculation is guaranteed,but also computing efficiency is greatly improved as well as the statistical precision of floor response spectra is ensured.
出处
《水电能源科学》
北大核心
2014年第9期112-114,85,共4页
Water Resources and Power
基金
国家重大科技专项项目(2011ZX06002-10)
国家自然科学基金重点项目(51138001)
中央高校基本科研业务费专项资金项目(DUT13LK16)
关键词
核电楼层反应谱
参数不确定性
蒙特卡罗模拟
并行运算
nuclear power floor response spectra
parameter uncertainty
Monte Carlo simulation
parallel calculation
