汽轮机状态估计建模与数据校验

State estimation modeling and data checking for steam turbines

提出了在汽轮机状态估计检测中应用结合Givens正交变换的Levenberg-Marquardt(LM)算法的思想。将LM算法应用于300 MW再热凝汽式汽轮机,应用结果表明,在大范围的测量误差和测量冗余比率下,结合Givens正交变换的LM算法进行汽轮机状态估计检测具有良好的稳定性。

It is proposed that the LM algorithm based on Givens orthogonal transform should be adopted for state estimation and checking for steam turbines. The LM algorithm was applied to a 300 MW reheat condensing steam turbine, and results show that the LM algorithm based on Givens orthogonal transform is effective when large measurement errors and redundancy exist.