基于压缩感知理论的谐波畸变电压信号采集方法研究

Study on harmonic distortion of voltage signal acquisition method based on compressed sensing theory

针对电能质量信号参数变化快、数据采集量特别大的特点,论文对电能质量信号的压缩感知采集方法进行了研究,并根据电能质量信号检测的具体要求,改进了观测矩阵中轮换矩阵的构造方法,针对信号特征提高了低频段信号的权重,可以在较高压缩比条件下重构得到准确的信号。在不同压缩比条件下,对论文中提出的算法进行了仿真,并以谐波畸变电压信号为例,同采用高斯随机矩阵和传统轮换矩阵的算法进行了比较,仿真结果表明本方法压缩比高,信号重构效果好,性能稳定,优于传统轮换矩阵。

In view of the quick changing of power quality signal and large amount of data that need to be collected, we discussed application of compressed sensing theory to power quality signal acquisition in this paper. Also we demonstrated a new algorithm of constructing the rotation matrix according to the specific requirements for power quality signal, which we called weighted observation matrix. Regulating the weight could lead to reconstruct accura- cy change at different parts of signal independently. By adjusting the weight of different frequencies of signal ac- cording to signal characteristics known in advance and specific requirement to the measurement, this new algorithm has the ability to get more accurate reconstruction results of signal under higher compression ratio condition. The performance of this algorithm was evaluated under different compression ratios. Compared with classical observation matrix such as Gaussian random matrix and the traditional rotation matrix, our method demonstrated higher com- pression ratio, higher reconstruction accuracy and better robustness.