【摘要】 本文针对大型电站实时仿真中常见的热工流体网络的求解问题,提出大型稀疏矩阵的POSTD实时仿真算法。POSTD (Pre___Ordering Symbolic Triangular Decomposition) 算法是在RCM排序算法的基础上加以改正,而得到的整体近似最优排序算法。同时根据流体网络计算的特点和要求,将原位LDL~1分解为三个步骤进行,即符号模拟分解、数值分解及数值求解。并在自动建模系统的支持下,在网络模块自动生成之前完成模拟定序和符号分解工作。从而大大提高了动态仿真的计算速度,同时节省存储单元,达到实时仿真的要求。该算法已编制成通用软件包。该软件包具有通用性强、实时性好、使用灵活方便和仿真精度高的特点。能适合并满足电站各种流体网络的快速计算,已用于长春20万千瓦仿真系统中。更多还原

【Abstract】 This paper proposes a kind of effective real-time simulation algorithm for resolving thermohydraulic network: Pre-ordering Symbolic Triangular Decomposition Algorithm. POSTD algorithm adopts the technology of sub-optimal ordering sequence for sparse matrix. In order to satisfy the need of real-time simulation, the algorithm can be segmented three steps. In real-time computation, only the steps of numerical decomposition and calculation are needed. So if POSTD algorithm is used, the storage space can be saved and computing time can be reduced. The software package of POSTD algorithm has been developed.更多还原