面向新型电力系统的粗糙集和双流网络自动化物联设备故障诊断方法研究

Research on fault diagnosis method of rough set and dual-stream network automation IoT device for novel power system

在新型电力系统中,自动化物联设备类型多样,性能参数和状态指标众多,而目前主流基于深度学习的故障诊断方法,也存在参数较多、算法复杂等问题,难以满足实时应用。因此,文中提出一种基于粗糙集(rough set,RS)和双流网络(dual-stream network,DSN)的自动化物联设备故障快速诊断方法。首先构建物联设备故障判定函数,计算高压侧三相电流,针对连续属性和定性属性,采用离散化算法进行约简,输出故障特征集;然后利用卷积神经网络(convolutional neural network,CNN)和门控循环单元(gated recurrent unit,GRU)组成的DSN对历史故障信息进行训练,最终实现快速故障诊断。实验证明所提方法能够精准检测新型电力系统的自动化物联设备故障,具有良好的应用性能,和同类深度学习算法相比,精准率平均提升4.35%,时间平均下降14%。

In novel power system,there is a wide variety of automated IoT device types,numerous performance parameters,and status indicators.However,current mainstream fault diagnosis methods based on deep learning suffer from issues such as excessive parameters and complex algorithms,making real-time applications difficult.To this end,the paper proposes a rapid fault diagnosis method for automated IoT devices based on rough set(RS) and dual-stream network(DSN).Firstly,a fault determination function for IoT devices is constructed,computing the three-phase currents on the high-voltage side.Discretization algorithms are employed for both continuous and qualitative attributes to reduce complexity,yielding a set of fault features.Subsequently,a dual-stream network composed of convolutional neural networks(CNN) and gated recurrent units(GRU) is utilized to train historical fault information,achieving swift fault diagnosis.Experimental results demonstrate the effectiveness of the proposed method in accurately detecting faults in automated IoT devices of novel power system.It exhibits excellent application performance,with an average precision increase of 4.35% and an average time reduction of 14% compared to similar deep learning algorithms.