变压器类设备状态评价的回归建模研究

摘要

In the national power grid system,transformer equipment of power system plays an important role,and transformer equipment diagnosis directly guides the power transformer equipment maintenance. It proposes to use the nonlinear regression algorithm for equipment condition maintenance,build a transformer nonlinear regression model with a lot of samples on operation equipment state information. The approach can make up for the disadvantage of traditional multistep points in the evaluation system .thus provide a more accurate maintenance scheme. And experiments validate the effectiveness and correctness of the approach.%在国家电网系统中,电力系统变压器类设备起着至关重要的作用,而对变压器类设备的故障诊断又直接决定着电力变压器类设备的维护工作.文中提出在变压器类设备状态检修中,使用基于线性回归的建模方法,以大量运行中的设备的状态信息为样本,对变压器类设备状态建立回归模型,从而形成对变压器类设备更精确的状态评价,在很大程度上弥补了传统的阶梯性扣分评价体系的不足,以得出更精确的检修方案,并通过实验证明该方法的有效性和正确性.