A long term dynamic simulation method of power system based on variables grouping and seceding grouping strategy

YongJi Chang,Jilai Yu

DOI: https://doi.org/10.1109/DRPT.2008.4523567

2008-01-01

Abstract:A set of differential and algebraic equations is stiff in the long term dynamic simulation analysis for power system, which remains very computationally intensive, and decreases computational efficiency. According to space-time local characteristic of stiffness, this paper presents a long term dynamic simulation method based on strategies of integration variables grouping and integration variables seceding grouping orderly. On the one hand, this method makes the whole system divided into fast group and slow group adaptively depending on local truncation error (LTE), and each group is solved with different step size by adjusting relative step factor and minimal integral step size automatically. On the other hand, different variables are treated differently in different period of time. That is, one part of fast variables may secede grouping and integration operation automatically after attaining their decay time, and the shortcoming, which is to judge the grouping of variables by making use of LTE at every large step integration point, may be overcame. Thus numbers of variables participating in grouping and integration operation will be reduced gradually, and the computational efficiency is enhanced. Tests results show that this method assures numerical value stability effectively and adapt to solving stiff differential and algebraic equations in the long term dynamic simulation of power system.