Longzhang Ke

In order to improve image visual quality in high dynamic range (HDR) scenes while avoiding motion ghosting artifacts caused by exposure time differences, innovative image sensors captured two registered extreme-exposure-ratio (EER) image pairs with complementary and symmetric exposure configurations for high dynamic range imaging (HDRI). However, existing multi-exposure fusion (MEF) algorithms suffer from luminance inversion artifacts in overexposed and underexposed regions when directly combining such EER image pairs. This paper proposes a neural network-based framework for HDRI based on attention mechanisms and edge assistance to recover missing luminance information. The framework derives local luminance representations from a convolution kernel perspective, and subsequently refines the global luminance order in the fused image using a Transformer-based residual group. To support the two-stage process, multi-scale channel features are extracted from a double-attention mechanism, while edge cues are incorporated to enhance detail preservation in both highlight and shadow regions. The experimental results validate that the proposed framework can alleviate luminance inversion in HDRI when inputs are two EER images, and maintain fine structural details in complex HDR scenes.

Modular multilevel converters (MMCs) have been widely applied in high voltage direct current (HVDC) transmission project, and security is their fundamental issue. Submodule (SM) switching Devices fault of MMC in HVDC is the most common problem, nevertheless, traditional fault diagnosis methods are mostly based on single modality resulting in poor feature extraction ability. To solve this pain spot, A novel multimodal attention fusion (MAF) model is proposed in this paper. Firstly, the three-phase internal circulating currents of MMC are converted into two-dimensional (2D) time-frequency image by performing Synchronous Squeezing Transform (SST). To automatically focus on the discriminative regions most related to fault features, a time series attention model(TSAM) and a visual attention model(VAM) are proposed to learn the features of the 2D time-frequency image and the internal circulating current time series data, respectively. Then, a multimodal attention model(MAM) via intermediate fusion is proposed, which utilizes the internal correlation between visual features and time series features for joint fault feature extraction. Finally, a later fusion scheme is applied to combine the fault prediction results of three attention models for fault diagnosis. The fault diagnosis accuracy of the proposed MAF model are 98.4% and 97.3% on the 31-level and 61-level MMC datasets, respectively. Experiments demonstrate that the designed model boosts feature representation power and achieves better fault diagnosis performance than the state-of-the-art baseline methods.

When submodule (SM) compound fault occurs in the Modular multilevel converter (MMC), the time-domain waveform characteristics of output current and internal circuiting current of MMC are not obvious, especially on the condition of high-level MMC. To address this issue, A compound fault diagnosis method based on an improved capsule network (CapsNet) is proposed in this paper. Firstly, The front end of the network adopts the feature extraction structure of One-dimensional Convolutional Neural Network (1DCNN) combined with Long and short-term memory network (LSTM). Then, the original data of MMC three-phase output currents and three-phase circulating currents are used as fault detection signal. The back-end employs a main capsule layer and a digital capsule layer structure, and realizes the transfer of feature vectors through a dynamic routing algorithm. This feature extraction structure combines the light weight of 1DCNN and the sequential sensitivity of LSTM. While ensuring that the information is fully extracted, the computational cost of the model is greatly reduced. Furthermore, the overlapping sampling method is used to construct and expand the sample sets. Compared with the other three common deep learning methods, their fault diagnosis performance under the condition of different levels and levels changing are analyzed respectively. The experimental results demonstrate that the proposed method has excellent cross-domain learning ability and high fault recognition accuracy.