基于数学形态学的电能质量扰动检测和定位

Detection and Location of Power Quality Disturbance Based on Mathematical Morphology

数学形态学因其在保留信号突变点信息方面有很好的效果,因此常用于短时电能质量扰动的检测和定位,但基于数学形态学的部分方法仍存在对某些过零点扰动检测失效的缺点,文章分析了3种基于数学形态学的扰动检测和定位方法,即基于1阶求导和形态梯度的方法、基于形态梯度和软阈值处理的方法、基于dq分解和高帽变换的方法,通过仿真比较了3种方法在分析电压暂降、电压暂升、电磁暂态振荡等信号方面的适应性,结果发现基于dq分解和高帽变换的方法在检测过零点扰动时具有很好的效果,因此选取这种方法对实测扰动数据进行了检测和定位分析。结果表明,基于dq分解和高帽变换的方法能正确检测与定位出任一时刻发生的扰动,具有较好的适应性与可行性。

Due to its good effect in the reservation of information of signal abrupt change, mathematical morphology is often applied in the detection and location of short-term power quality disturbance, however, there is still a defect of invalid detection existing in partial detection methods based on mathematical morphology while some zero-crossing disturbances are detected. In this paper three mathematical morphology based disturbance detection and location methods, i.e., the method based on first-order derivation and morphological gradient, the method based on morphological gradient and soft threshold processing and the method based on dq decomposition and top-hat transform, are analyzed, that is, by means of simulation the adaptability of these methods in the analysis on the signals of voltage sag, voltage swell and electromagnetic transient oscillation is compared. Analysis results show that the method based on dq decomposition and top-hat transform can obtain good effect in the detection of zero-crossing disturbances, thus this method is chosen to conduct the detection and location for the measured disturbance data, and it is shown that this method possesses good adaptability and feasibility as well as can detect and locate the disturbance occurred in arbitrary moment correctly.