Fault arcs exhibit randomness, with current waveforms closely mirroring those of standard nonlinear load operations, posing challenges for traditional series fault arc detection methods. This study presents an improved detection approach using a lightweight convolutional neural network model, ShuffleNet V2. Current data from household loads were collected and preprocessed to establish a comprehensive database, leveraging one-dimensional convolution and channel attention mechanisms for precise analysis. Experimental results demonstrate a high fault arc detection accuracy of 97.8%, supporting real-time detection on the Jetson Nano embedded platform, with an efficient detection cycle time of 15.65 ms per sample. The proposed approach outperforms existing methods in both accuracy and speed, providing a robust foundation for developing advanced fault arc circuit breakers.
With the rapid development of high-voltage direct current (HVDC) transmission technology, the demand for monitoring the DC total electric field at ground level has increased significantly. The methods of the traditional electric field measurement are susceptible to noise interference, which affect the accuracy of measurement. This paper investigates the sources of noise interference in electric field measurements and proposes a signal processing method based on amplifier parallel optimisation of the signal conditioning circuit and Morlet wavelet decomposition. By improving the signal-to-noise ratio during the measurement of the device, the anti-interference capability and measurement accuracy of the field mill sensor are significantly enhanced. The experimental results show that the proposed signal processing methods of amplifier parallel optimisation signal conditioning circuit and Morlet wavelet decomposition can effectively eliminate the interfering signals during the measurement process, improve the signal-to-noise ratio of the measurement circuit, and retain the details and response speed of the signals, which improves the accuracy of the measurement of the ground electric field of the ultra-high-voltage direct current transmission.
In this paper, a family of a novel single-phase three-level PFC based on a switch-capacitor cell is proposed. The proposed PFC topologies have the characteristics of high power factor, low voltage stresses, and low power losses. Firstly, the derivation process of the novel PFC topologies is introduced in detail. Based on a representative circuit of the proposed novel PFC topologies, its operation principle is analyzed from the aspects of working current paths, key waveforms, and pulse distribution. Meanwhile, its equivalent circuit model is deduced. Secondly, the performance of the proposed PFC topologies is analyzed. Then the modulation technology based on capacitor voltage balancing is designed for the proposed topologies. Finally, an experimental prototype with a rated power of 800 W and a DC output voltage of 400 V is built. The experimental analysis is carried out from both the steady state and dynamic state. The experimental results verify the feasibility of the proposed novel three-level PFC topologies and the effectiveness of the modulation technology.
Traditional methods for evaluating power electronic converter efficiency heavily rely on simulations and often lack theoretical support, which can lead to inaccuracies and limit effective design optimization. To address these shortcomings, this paper proposes a modal segment integration method based on a device loss model. The analysis begins with the operating principles of the proposed circuit topology. A detailed power loss model is then established and applied to representative operating modes. Using the modal segment integration method, the total loss over a full operating cycle is calculated. Theoretical analysis estimates the system efficiency exceeds 98%. To validate the proposed method, a 1 kW experimental prototype with a 400 V DC output is built. The results show that the maximum error between the theoretical and experimental efficiency is less than 0.4%. This method offers a reliable theoretical basis for efficiency evaluation of three-level converter topologies and supports the structural design and performance optimization of power electronic systems.