摘要
化学反应系统中的Leap算法可在获得较好精度的同时大幅提高模拟速度,最近提出的无偏Leap方法有效地克服了由Leap时间区间内的反应次数的近似均值与真实均值之间的偏差引起的Leap算法的误差的不足,本文讨论了一个基于物种相对改变估计真实均值的快速无偏τ-Leap算法,并将该算法推广到模拟时滞化学系统中.该快速算法具有易于编码、比前者更快等优点,当系统中的反应通道或物种的数目较大时,该方法具有更明显的速度优势。
The leap algorithms significantly accelerate stochastic simulation of chemically reacting systems with some acceptable losses by considering the leap condition. Recently, the unbiased leap methods are considered to overcome the deficiency appears in leap algorithms, in which the means of reaction numbers during a leap are not equal to the true ones. In this paper an efficient version of the unbiased γ-leap method is proposed by considering the relative changes of species to calculate true means. Further, this algorithm is extent to chemically reacting systems with delays. The new algorithms are easier to code, and faster to execute than the former. The speed-up in execution will be especially pronounced in systems that have many reaction channels or species.
出处
《计算数学》
CSCD
北大核心
2009年第3期309-322,共14页
Mathematica Numerica Sinica
基金
国家自然科学基金(30571059)
国家高科技研究发展计划(863)专项资助项目(2006AA02Z190)
上海市重点学科资助(S30405)
上海师范大学校级项目(SK200937)