Detecting, measuring, and classifying partial discharges (PDs) are important tasks for assessing the condition of insulation systems used in different electrical equipment. Owing to the implementation of the phase-resolved PD (PRPD) as a sequence input, an existing method that processes sequential data, e.g., the recurrent neural network, using a long short-term memory (LSTM) has been applied for fault classification. However, the model performance is not further improved because of the lack of supporting parallel computation and the inability to recognize the relevance of all inputs. To overcome these two drawbacks, we propose a novel deep-learning model in this study based on a self-attention mechanism to classify the PD patterns in a gas-insulated switchgear (GIS). The proposed model uses a self-attention block that offers the advantages of simultaneous computation and selective focusing on parts of the PRPD signals and a classification block to finally classify faults in the GIS. Moreover, the combination of LSTM and self-attention is considered for comparison purposes. The experimental results show that the proposed method achieves performance superiority compared with the previous neural networks, whereas the model complexity is significantly reduced.
We consider the ''harvest-then-transmit'' protocol in a wireless powered communication network (WPCN), where an energy-constrained access point (AP) harvests energy from the radio-frequency signals transmitted by a power beacon (PB) for assisting user data transmission. In the wireless information transfer (WIT) phase, AP employs the harvested energy to convey independent signals to multiple users through either time-division multiple access (TDMA) or orthogonal frequency-division multiple access (OFDMA). Aiming to maximize the sum rate (SR) of the WPCN, we jointly optimize the energy harvesting (EH) time and the AP power allocation, considering both the conventional linear and practical nonlinear EH models at the AP. The optimization problems of both TDMA-and OFDMA-enabled WPCNs are formulated as nonconvex programs, which are challenging to solve globally. To achieve an efficient optimal solution to the problem of TDMA-enabled WPCN, we first decompose the original nonconvex problem into three convex subproblems, and then propose a low-complexity iterative algorithm for its solution. For the OFDMA-enabled WPCN, the problem belongs to a difficult class of mixed-integer nonconvex programming due to the involvement of binary variables for subcarrier allocation. To overcome this issue, we convert the problem to a quasi-convex problem and then employ a bisection search to obtain the optimal solution. Simulation results are provided to confirm the benefit of jointly optimizing the EH time and the AP power allocation compared to baseline schemes. The performance of the proposed TDMAenabled WPCN is shown to be superior to that of the proposed OFDMA-enabled WPCN in terms of SR when the transmit power of PB and the number of antennas of AP are relatively large.INDEX TERMS Wireless powered communication networks, time-division multiple access, orthogonal frequency-division multiple access, non-linear EH, IoT, nonconvex optimization.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.