动态相量测量算法等效滤波器研究

Study of equivalent filters of dynamic phasor measurement algorithm

指出实现动态相量测量的DPEA、DPEFM和TFDPE算法本质上可视为线性非时变(LTI)系统,推导出各系统等效数字滤波器的表达式,并给出算法的扩展形式。基于滤波器频响特性定义一组技术指标,并建立其与基波扰动、谐波、噪声和衰减直流各因素产生测量误差的对应数量关系。由此,一方面基于滤波器频响特性对测量算法综合性能进行便捷有效的评估,另一方面基于技术指标对扩展算法进行优化,设计出具有最佳频响特性的测量算法。针对P类性能要求,给出了数据窗长为2个和3个周波优化设计结果。技术指标和仿真实验结果表明,在相同数据长度下,优化算法相比原算法在综合性能上有大幅提升,特别是具有了非常强的抗直流干扰能力。

A group of dynamic phasor measurement algorithms,including dynamic synchrophasor estimation algorithm(DPEA),dynamic phasor estimator based on frequency-domain model(DPEFM)and time-frequency information-based dynamic phasor estimator(TFDPE),are revealed to be linear time-invariant(LTI)systems,and the expressions of their equivalent filters are derived.In addition,these algorithms are extended to more generalized forms in the sense of filter equivalence.A group of filter specifications are defined based on the frequency response of the equivalent filter,and the quantitative relations are built between these filter specifications and the measurement errors caused by non-nominal fundamental tone,harmonics,decaying DC,and wideband noise.In terms of these relations,on the one hand,the measurement performance of these algorithms are conveniently evaluated and compared based on their frequency response;on the other hand,filter optimization is formulated to design optimum filters so that the algorithms achieve the best overall measurement performance.Optimization results of P-class algorithm are presented in the cases of 2-cycle and 3-cycle data length.Filter specifications and simulation results verify that,the overall measurement performance of the optimized algorithms are significantly improved as compared to their original versions,where excellent performance of rejecting decaying DC disturbance is achieved in particular.