基于Prony算法和改进Budeanu理论的电力系统功率测量研究
发布时间：2018-11-09 21:35
【摘要】：传统基于傅里叶变换、小波变换等信号时频变换的电力系统功率测量方法大多没有考虑非整数次谐波(间谐波)的影响,采用一段时间窗的整体计算,不能得到电力系统具体时刻的瞬时功率,无法评估电力系统中的瞬态现象,如衰减谐波、冲击脉冲等。另外,传统方法的测量精度较低,受到采样数据的限制。本文提出了一种基于Prony算法和改进Budeanu理论的电力系统功率测量方法。借鉴Fryze理论和ShepherdZakikhani理论对经典频域功率理论Budeanu理论进行改进,将Budeanu理论推广到非整数次谐波情况。使用Prony算法可以将电压和电流分解为指数函数的线性组合,精确得到频率、振幅、初相和衰减因子等特征,进而可以计算电力系统的瞬时有功功率和无功功率。提出的方法充分考虑了电力系统中的间谐波功率以及衰减谐波、冲击脉冲等瞬态现象,可以计算得到电力系统的瞬时功率,同时具有动态性能好,测试精度高等优点。最后通过算例进行了仿真验证。本文提出的功率测量方法为电力系统非正弦复杂情况下的功率测量、瞬态现象研究及谐波治理提供了一种有效的思路和方法。
[Abstract]:The traditional power measurement methods based on Fourier transform, wavelet transform and other signal time-frequency transform do not take into account the influence of non-integer harmonics (interharmonics). The instantaneous power of power system at specific time can not be obtained, and the transient phenomena in power system can not be evaluated, such as attenuation harmonics, impulse pulses and so on. In addition, the measurement accuracy of the traditional method is low, which is limited by the sampling data. This paper presents a power measurement method for power system based on Prony algorithm and improved Budeanu theory. The classical frequency-domain power theory Budeanu theory is improved by using Fryze theory and ShepherdZakikhani theory. The Budeanu theory is extended to the case of non-integer harmonics. Using Prony algorithm, the voltage and current can be decomposed into linear combination of exponential function, and the characteristics of frequency, amplitude, initial phase and attenuation factor can be obtained accurately, and the instantaneous active and reactive power of power system can be calculated. The proposed method takes full account of the interharmonic power and the transient phenomena such as attenuation harmonics and impulse pulses in power system. It can be used to calculate the instantaneous power of power system and has the advantages of good dynamic performance and high measurement accuracy. Finally, a simulation example is given. The power measurement method proposed in this paper provides an effective method for power measurement, transient phenomenon research and harmonic control in power system with non-sinusoidal complex conditions.
【作者单位】： 电子科技大学电子科学技术研究院;
【分类号】：TM933.3
，
本文编号：2321622
[Abstract]:The traditional power measurement methods based on Fourier transform, wavelet transform and other signal time-frequency transform do not take into account the influence of non-integer harmonics (interharmonics). The instantaneous power of power system at specific time can not be obtained, and the transient phenomena in power system can not be evaluated, such as attenuation harmonics, impulse pulses and so on. In addition, the measurement accuracy of the traditional method is low, which is limited by the sampling data. This paper presents a power measurement method for power system based on Prony algorithm and improved Budeanu theory. The classical frequency-domain power theory Budeanu theory is improved by using Fryze theory and ShepherdZakikhani theory. The Budeanu theory is extended to the case of non-integer harmonics. Using Prony algorithm, the voltage and current can be decomposed into linear combination of exponential function, and the characteristics of frequency, amplitude, initial phase and attenuation factor can be obtained accurately, and the instantaneous active and reactive power of power system can be calculated. The proposed method takes full account of the interharmonic power and the transient phenomena such as attenuation harmonics and impulse pulses in power system. It can be used to calculate the instantaneous power of power system and has the advantages of good dynamic performance and high measurement accuracy. Finally, a simulation example is given. The power measurement method proposed in this paper provides an effective method for power measurement, transient phenomenon research and harmonic control in power system with non-sinusoidal complex conditions.
【作者单位】： 电子科技大学电子科学技术研究院;
【分类号】：TM933.3
，
本文编号：2321622
本文链接：https://www.wllwen.com/kejilunwen/dianlidianqilunwen/2321622.html
