三种相量算法在变压器绕组参数辨识中的应用

Application of three phase algorithms in parameter identification of transformer winding

为了能够及时、准确地检测出由变压器绕组形变而导致的故障,需要选择一种更加精确、稳定地辨识变压器绕组参数的相量算法.对傅里叶、递推最小二乘和卡尔曼三种相量算法在变压器绕组参数辩识中的计算精度进行了理论上的分析,并且把三种算法运用于实际变压器绕组参数辨识中,进一步分析了三种算法的计算精度.对辨识结果的误差分析表明,用卡尔曼算法进行参数辨识不仅计算精度较高而且计算结果比较稳定,卡尔曼相量测量算法在变压器绕组参数辨识中的应用为准确检测变压器绕组故障提供了基础依据.

A more accurate and stable algorithm needs to be selected in the parameter identification of transformer winding in order to detect the fault caused by deformation of winding. The calculation precision of Fourier algorithm, recursive least square algorithm and Kalman algorithm applied in the parameter identification of transformer winding was analyzed. Three algorithms were applied in the parameter identification of actual transformer winding to perform the further analysis on the calculation precision. The error analysis of identification results indicates that in the parameter identification, Kalman algorithm has such advantages as high precision and high stability, which provides the basis for the fault detection of transformer winding.