基于自适应短时傅立叶变换的电频率跟踪测量算法

New Digital Algorithm for Measuring Electric Frequency Based on Adaptive Short-Time Fourier Transform

电频率快速、准确测量是电网及电气设备运行、控制、调节的基础。基于电气信号的异步采样数据,应用短时傅立叶变换(STFT)估计电气信号频率,选用矩形自卷积窗抑制谐波对测频的影响,并根据频率变化自适应调整时间窗宽度。算法实现简单、计算量小,测频范围大,在信号频率缓慢变化和快速变化时,均具有较好的测量精度和跟踪速度。仿真研究对测频算法在各种情况下的可行性和有效性进行了验证。

Accurate and fast estimation of electric frequency is the basis of operation, control, and regulation of power system and electric equipments. Based on adaptive short-time Fourier transform, a new digital algorithm is presented to evaluate the frequencies of electric signals. We apply a novel window called Rectangle Self-Convolution Window to eliminate the effect of harmonics, and the length of the data window is adjusted according to the variation of signal frequency. Using discrete time values of the input signal taken at a fixed sampling rate, the algorithm is simple and convenient to implement with high precision and short response time as well as wide measuring range. The algorithm is suitable for real-time application especially when the frequency changes are abrupt and the signal is corrupted with noise and other disturbances due to harmonics. Our numerical experiments demonstrated the feasibility and effectiveness of the proposed algorithm under several conditions.