Research On Frequency Measurement Algorithm In Power System

Frequency is not only an important indicator to measure power quality, but also an important feedback variable to implement security and stability control in the power system. And each process of power system operation asks for quick and accurate frequency measurement. However, computing speed or accuracy of the traditional frequency measurement algorithms cannot be guaranteed under the new trend of power system. Therefore, it is of great importance to amend the traditional algorithms or to study new algorithms so as to meet the development of the power system. In this paper, basic principle of traditional frequency measurement algorithms is studied, and three frequency measurement algorithms for different conditions are put forward. The main research results obtained are summarized as follows.Some commonly used traditional algorithms are time-consuming and computationally intensive. Especially when the frequency changes dramatically, the consumption of time is multiplied. These algorithms cannot fully meet the requirements of real-time occasion. Therefore, a fast frequency measurement algorithm based on numerical differential is proposed. With Newton interpolation method and Newton forward difference formula, multi-order derivative is obtained and system frequency is acquired. Simulation experiments proves that this algorithm has the characteristics of high accuracy, quick responds, and is immune to the change of voltage amplitude and phase angle. This algorithm is suitable for the occasion of real-time requirements.The phasor angle difference which applied in traditional Fourier frequency measurement algorithm cannot correctly reflect the real difference, resulting the principle error. Therefore, Fourier frequency measurement algorithm which is based on modified phasor angle difference is proposed. As the phasor angle, which acts as intermediate variable, is amended by correction factor, the theoretical error of traditional Fourier algorithm is eliminated. The algorithm is not sensitive to the noise and harmonics, which is inherited from Fourier algorithm. Besides, to overcome the paradox that traditional algorithm cannot have both accuracy and speed, quadratic interpolation iterative technology is adopted for revising the sample sequences. Simulation results show that, compared with the traditional Fourier algorithm, both computing speed and accuracy of the proposed algorithm were improved in the same hardware environment.Generally, the estimation of traditional frequency measurement algorithms is based on static models, which may produce large errors in the transient process. Therefore, a frequency measurement algorithm based on a dynamic model containing decaying direct-current component is proposed, and the least squares method is adopted to fit the dynamic model of Taylor polynomial expansion. Firstly, the dynamic model in the observation window is expanded to second-order Taylor polynomial. Then, the least squares method is applied to calculate Taylor coefficient. And the dynamic model of the frequency can be estimated by Taylor coefficient. The algorithm is not sensitive to noise and harmonics, and has the good features of high accuracy, low latency, and short not-applicable area, which makes the algorithm suitable for real-time frequency measurement in power system transient process. Simulation results show that, the proposed algorithm exhibits good performance under transient conditions, including signals with decaying direct-current component, signals with noise, signals with harmonics, frequency fluctuation, steps in phase and steps in amplitude.