基于CS-RBAPVS的超高次谐波检测算法

随着电力电子技术高占比应用于可再生能源发电系统，大量超高次谐波流入电网，从而引发新的电能质量问题。针对超高次谐波检测过程中计算量大、测量不精确等问题，提出一种基于压缩感知变步长正则化回溯自适应追踪超高次谐波检测算法。该方法利用狄利克雷核矩阵与离散傅里叶变换系数，并引入插值因子，建立基于压缩感知理论的超高次谐波模型；同时采用基于变步长的正则化回溯自适应追踪算法重构原信号，该算法通过结合正则化思想与子空间追踪算法的回溯思想，在无需预知原信号稀疏度的情况下，不仅实现了信号的精确重构，而且克服了固定步长所引起的问题。仿真测试验证了该算法的优越性与有效性。

Ultra high Harmonic Detection Algorithm Based on CS RBAPVS

With the high proportion of power electronics technology applied to renewable energy power generation systems, a large number of ultra high harmonics flow into the grid, causing new power quality problems. Aiming at the problems of large computational complexity and inaccurate measurement in the process of ultra high harmonic detection, a variable step size regularized backtracking adaptive tracking ultra high harmonic detection algorithm based on compressed sensing is proposed. The method uses Dirichlet kernel matrix and discrete Fourier transform coefficients, and introduces interpolation factors to establish a ultra high harmonic model based on compressed sensing theory. At the same time, a regularized backtracking adaptive tracking algorithm based on variable step size is used to reconstruct the original signal. When the sparsity of the original signal is unknown, the algorithm combines the regularization idea and the backtracking idea of the subspace tracking algorithm to realize the accurate reconstruction of the signal and overcome the problems caused by the fixed step size. Simulation tests verify the superiority and effectiveness of the algorithm.