基于加权平均粗糙度的配电网故障诊断分层模型

针对配电网发生故障后故障诊断警报信息存在不确定性和不完整性导致难以得出准确结论的问题，提出了利用改进可辨识矩阵对故障样本进行约简，在约简的基础上根据加权平均粗糙度的大小对故障样本进行分层从而建立模型的方法。该方法避免了分层的盲目性和冗余性，大大缩小了诊断模型空间，且加权平均粗糙度计算简单，易于实现。按加权平均粗糙集的大小区分层内各特征信息的优先级，具有很强的容错能力。仿真实例证明了该模型的有效性。

Fault Diagnosis Layer Model of Distribution Network Based on Weighted Mean Roughness

To improve the indeterminacy and imperfection of electric power fault information that cause the difficulty to obtain accuracy fault diagnosis results, a method is proposed in which by use of improved discernible matrix the fault samples are reduced and the reduced samples are layered according to the value of weighted mean roughness, thereby a fault diagnosis layer model is built. With the proposed method the blindness and redundancy of layering can be avoided and the space for diagnosis model is evidently diminished. In addition the weighted mean roughness is easy to be calculated and implemented. According to the value of weighted mean roughness the priority of different characteristic infatuation can be distinguished, so the proposed method possesses strong tolerant ability. Simulation results show that the built model is effective.