海上直驱永磁风电系统非对称故障主动识别方法

Active identification method for asymmetric fault of offshore direct drive permanent magnet wind power system

传统风电系统故障识别方法容易受到线电压误差影响出现误判,导致故障数据识别结果精度不高,因此提出海上直驱永磁风电系统非对称故障主动识别方法。分析非对称故障下风电系统的运行特性。根据连续时间信号,采用主动学习方法消除信号基频。使用逆向扩散法调整主动学习参数,判断故障态、预警态事件。利用电压幅值和时间宽度双重标准分析线电压误差,采用数据互校验方法确认故障事件可信性,完成故障主动识别。实验结果表明,所提方法能够主动发现风电系统中的电流异常情况和谐波畸变程度,直流分量最大误差仅为0.05 A。

Traditional fault identification methods for wind power systems are prone to misjudgment due to line voltage errors,resulting in insufficient accuracy of fault data identification results.Therefore,an active identification method for asymmetric faults in offshore direct drive permanent magnet wind power systems is proposed.The operation characteristics of wind power system under asymmetric fault are analyzed.According to the continuous time signal,the active learning method is used to eliminate the fundamental frequency of the signal.The reverse diffusion method is used to adjust the active learning parameters and judge the events in fault status and early warning status.The double standards of voltage amplitude and time width are used to analyze line voltage error,and the reliability of fault events is confirmed by data mutual verification method to complete active fault identification.The experimental results show that the proposed method can actively detect the current anomaly and harmonic distortion in the wind power system,and the maximum error of the DC component is only 0.05 A.