【摘要】 风电变桨系统因为风速的变化而频繁变桨,长期处于恶劣的工作环境,使得风电场的运行与维护成为了最严峻的问题.现场对风电机组变桨系统的检测大多依赖于SCADA报警系统.由于SCADA系统没有综合考虑变桨系统各子系统以及风机运行参数间存在的强耦合性,每当故障发生时,检测系统总是出现一连串无序的报警信息,为故障停机后的维修造成了困难.针对报警数据的无序,基于风电SCADA系统的运行数据,通过Fisher判别法将数据进行识别分类,从而对故障源进行定位,并且用实验验证了分类器良好的分类效果.仿真和实验结果证明了方法的可行性,能够为后续的故障诊断和检修提供参考指导.更多还原

【Abstract】 Wind power variable propeller system changes frequently due to the change of wind speed, and on-site detection of wind turbine impeller system mostly relies on SCADA alarm system. Because the strong coupling between the operating parameters of the fan is not considered comprehensively in the SCADA system, a series of disordered alarm messages always appear in the detection system. In this paper, aiming at the disorder of alarm data, based on the operation data of the wind power SCADA system, Fisher’s discrimination method is used to identify and classify the data to locate the fault source. The experiment verifies the good classification effect of the classifier. The simulation and experimental results prove the feasibility of the method, which can provide reference for the subsequent fault diagnosis and maintenance.更多还原