Search

Due to the diversity of power consumption patterns, the false positive rate (FPR) of data-driven electricity theft detection (ETD) methods is too high to meet practical needs, which severely restricts the engineering application of data-based methods. To reduce FPR of ETD methods based on advanced metering infrastructure (AMI), a deep neural network with low FPR (LFPR-DNN) is proposed in this paper. First, a deep model is constructed based on one-dimensional convolution and residual network, which can automatically extract features from consumption data. Then, a two-stage training scheme is used to train the network. In the first stage, the conventional gradient descent algorithm is used to update the network weights. To minimize the impact of data imbalance on detection performance, focal loss is used. Besides, grid search is used to optimize the hyper-parameters of the model. In the second stage, with FPR as the optimization objective, the particle swarm optimization (PSO) algorithm is used to train the network. Finally, the proposed LFPR-DNN is verified by using the open Irish data set. Compared to other state-of-the-art classifiers, LFPR-DNN has the lowest FPR with 0.29% and the highest AUC with 99.42%. The FPR is reduced by an order of magnitude, which verifies the effectiveness of the proposed method. [ABSTRACT FROM AUTHOR]

• Use the improved adaptive wavelet threshold to remove PQD noise. • Use the improved deep belief network to fully extract the PQD feature information. • The classification model named DBN-ELM has high classification accuracy and training efficiency. To solve the problems of noise interference and artificial feature extraction in power quality disturbance (PQD) classification, a new method combining adaptive wavelet threshold denoising and deep belief network fusion extreme learning machine (DBN-ELM) is proposed. Firstly, the noise content of the layer is determined by calculating the energy ratio of the wavelet coefficients of each layer, and an adaptive wavelet threshold is constructed based on the energy ratio to denoise the PQD signals. Secondly, the feature extraction capability of DBN is used to extract the feature from the PQD signals after denoising. Finally, a novel PQD classifier called DBN-ELM is constructed by integrating an ELM into a DBN, which avoids global fine-tuning of DBN and improves PQD classification efficiency. The simulation result and experimental verification show that the proposed method can effectively suppress PQD noise and performs well on DBN-ELM classification. [ABSTRACT FROM AUTHOR]