基于FDA贡献图的风电机组变桨系统故障定位

Fault location of pitch system based on FDA contribution diagram

电动变桨系统是风电机组的重要组成部分,其故障率一直偏高。在现场,每当SCADA系统检测到故障时,总是出现一连串的报警信息,随机且无序,无法确认真正故障源。针对SCADA系统的连锁报警问题,将Fisher判别分析法引入到变桨系统故障诊断中,形成一种基于FDA贡献图的故障源分离方法。首先计算出故障数据的偏离方向以及各变量相对于该方向偏离的贡献率,然后生成FDA贡献图,由此甄别出引起故障的主要参变量,实现故障源分离及定位。通过模拟数据和实例数据分析表明,该方法可用于指导后续故障诊断和制定检修预案。

The electric pitch control system is the important part of wind turbines, and its failure rate is always high. Generally, the true fault location cannot be confirmed by the ambiguous alarm information appeared in the supervisory control and data acquisition（SCADA） system of a wind farm when abnormal data are detected. Therefore, the Fisher Discriminant Analysis（FDA） method is introduced to the fault diagnosis of pitch-controlled systems, and a fault source separation method based on the FDA contribution diagram is proposed to identify the alarm information of SCADA. The deviation direction of fault data is calculated as well as the deviation contribution rate of each variable to the direction; then the FDA contribution diagram is generated for recognizing the primary parameter that caused alarm, separating the fault source, and locating the faulty component in the electric pitch system. The effectiveness of this method is verified through the analysis of the simulation data and industrial instances and it can be used in the fault diagnosis and maintenance decision of wind turbines.