Omprehensive evaluation method of low-voltage transformer area operation state based on improved K-Means clustering and RBF neural network

To enhance the management level of low-voltage transformer area and guide the grass-roots operations and maintenance staffs found the transformer area problems in time, A comprehensive evaluation method of low-voltage transformer area operation state based on improved K-Means clustering and RBF neural network is proposed, to evaluate the low-voltage transformer area operation state accurately from multi-dimensions, eliminate line and metering equipment faults in time. This method the first use of K-Means algorithm to classify the low-voltage transformer area, then obtains the index system set of low-voltage staion area operation state, comprehensive order relation analysis and weighted value to calculate the evaluation value of the historical low-voltage transformer area, and finally takes the historical low-voltage transformer area evaluation value as the training sample to obtain the RBF neural network evaluation model, which is used to carry out the state evaluation in the operating low-voltage transformer area. An example shows that the comprehensive evaluation method can effectively carry out the state evaluation of the low-voltage transformer area, and provides technical support for the improvement of the management level of the low-voltage transformer area.