摘要
传统的Prony算法抗干扰能力差,且无法给出其振荡模式辨识结果的准确度评价,因此提出一种基于多类型扰动轨迹的低频振荡分析方法,并对其准确度进行评价。利用各类型扰动信号与参考信号的平均能量之比对信号振幅进行预处理,避免振幅差异过大导致某些信号被掩盖。利用Prony算法对多类型的扰动轨迹进行分析获取主导振荡模式,计算各主导振荡模式的振幅偏差和相位偏差评价主导振荡模式的可信度。结合各主导振荡模式的振幅偏差、相位偏差、振荡能量百分比建立综合评价指标评价Prony算法的阶数选择是否合理。算例结果表明,所提方法及相应的评价方法具有良好的实用性。
Since the traditional Prony algorithm for oscillation mode identification has low anti-interference capability and could not evaluate the accuracy of the identification resuhs,a method of low-frequency oscillation analysis based on the multi-type perturbation trajectory is proposed,which pre-processes the amplitude of perturbation signals with the ratio of their average energy over the reference signal average energy for each perturbation type to avoid the signal masking due to the overlarge difference among signal amplitudes;applies the Prony algorithm to analyze the multi-type perturbation trajectory for recognizing the dominant oscillation modes;and calculates the amplitude deviation and phase deviation for each dominant oscillation mode to evaluate its credibility. A comprehensive index integrating the amplitude deviation,phase deviation and oscillation energy percentage is proposed to determine if the order of Prony algorithm is rationally selected. Results of case study show that,the proposed method including the corresponding evaluation method is quite practical.
出处
《电力自动化设备》
EI
CSCD
北大核心
2014年第12期52-57,共6页
Electric Power Automation Equipment
基金
国家自然科学基金资助项目(51077068
51107120)~~
关键词
电力系统
低频振荡
多类型信号
PRONY算法
主导振荡模式
信噪比
electric power systems
low-frequency oscillation
multi-type signal
Prony algorithm
dominant oscillation mode
signal-to-noise ratio