基于BP神经网络的有载分接开关状态感知及故障诊断研究

Study on State Sense and Failure Diagnosis of Load Tap Switch Based on BP Neural Network

作为调节变压器的重要部件,有载分接开关在变压器工作可靠性以及电网稳定性起到了重要的作用,其不仅能够稳定负荷中心电压、调节无功潮流,同时能够有效地增加电力系统的调度灵活性,对于有载分接开关的状态感知以及故障诊断尤为重要。基于BP神经网络模型,利用有载分接开关调压时产生的机械振动信号数据,感知开关的状态,并对开关异常进行故障诊断,为电力系统的稳定运行提供支持。

As an essential part of regulating the transformer,the load tap switch plays a critical role in the operational reliability and stability of the power grid.It can stabilize the load center voltage and adjust the reactive power trend and effectively increase the scheduling flexibility of the power system,and it is essential for the state perception and fault diagnosis of the load tap switch.The paper uses the BP neural network model to use the mechanical vibration signal data generated by the voltage regulation to perceive the switch state and diagnose the switch abnormality to support the stable operation of the power system.