Search

In recent years, graph neural network is used to process graph data and has been successfully applied to graph node classification task. Due to the complexity of graph structure and the difficulty of obtaining node labels, node classification in datasets with fewer labels becomes a challenge. Existing node classification problems with few-label nodes datasets usually use neighbor aggregation schemes. However, these methods lack the “graph pooling mechanism,” which makes these models impossible to make full use of the global graph information. To solve the problem, we propose a Hierarchical Structure-Feature aware Graph Neural Network (HSFGNN) model. The model learns node features by a hierarchical mechanism that repeatedly utilizes coarsening and refining methods at different levels. Firstly, we use topological information, node characteristics and connectivity of the graph to construct an intuitive and effective coarsening part based on multivariable analysis technology. Secondly, we aggregate the unselected nodes’ features into the top-k nodes, so that the coarser nodes contain more abundant valid graph information. Finally, we propose an additional attention mechanism to obtain more accurate final representations of nodes. Experimental results show that the HSFGNN model achieves outstanding classification accuracy on the dataset with less labeled data.

Affected by electric field generated in power transmission line, live worker on the high-voltage transmission line is in dangerous working environment. In this paper, posture perception, numerical simulation, and risk assessment are adopted to realize the deep fusion of physical trajectory and spatial virtual electric field distribution of live worker, and digital twin bodies of live worker in different live working scenarios are constructed. Firstly, posture trajectories and 3D model of living worker in typical operation conditions are captured and reconstructed by full-body inertial action capture technology and this technology use motion capture unit which include 3-aixs accelerometer/gyroscope/magnetic field sensors. The length of complex gap and body workspace under various live work scenes are also calculated. By static electrical field finite element (FE) numerical modeling, surface electrical field distributions of live worker under different posture trajectories are then calculated. Finally, interactive mechanism between “body-information-physical environment-virtual environment” is introduced, and potential risk of live worker is on-line assessed by live work guide rules. This work develops a new assessment and protection method for virtual training and field operations of high-voltage live worker. Which is of great significance for enhancing operation level and ensuring living workers’ safety.

Hyperspectral data have been widely used in various fields due to its rich spectral and spatial information in recent years. Yet, hyperspectral images are always tainted by a variety of mixed noises. These noises seriously limit the accuracy of subsequent applications. To remove noise, this paper, based on low-rank tensor decomposition, combined with non-local self-similar prior information, proposes a tensor-based non-local low-rank denoising model, where non-local self-similarity uses mainly spatial correlation while low-rank tensor decomposition method uses mainly spectral correlation between bands. Traditional tensor-based methods are commonly NP-hard to compute and are sensitive to sparse noise. However, the method proposed in this paper can separate efficiently the low-rank clean image from Gaussian noise and sparse noise (pulses, deadlines, stripes, speckle, etc.) by using a new tensor singular value decomposition (T-SVD) and tensor nuclear norm (TNN). The NP-hard task was also achieved well by the alternating direction multiplier method. Due to the full use of spectral and spatial information of the data, Gaussian noise and sparse noise can be effectively removed. The effectiveness of our algorithm was verified through experiments using simulated and real data.

Based on the boost full bridge isolated converter (BFBIC) topology and considering the sudden changes in the external environment, a global maximum power point tracking (GMPPT) control strategy based on an improved gray wolf optimizer (IGWO) algorithm is proposed in this paper. In the strategy, a nonlinear tangent trigonometric function as a convergence factor is integrated into the gray wolf optimizer (GWO) algorithm. In addition, the active-clamp circuit and phase-shift are used to implement the soft switch technology for BFBIC converter in photovoltage (PV) system. Finally, the maximum power point tracking (MPPT) performance on PV system with the proposed IGWO algorithm under static and dynamic partial shading conditions (PSCs) was investigated and compared with other common perturb and observe(P&O), particle swarm optimization (PSO), artificial bee colony (ABC), adapt inertia weight salp swarm algorithm (WSSA), salp swarm algorithm with grey wolf optimizer (SSA-GWO), SSA with PSO (SSA-PSO), enhanced GWO (EGWO) MPPT algorithms. The effectiveness and stability of the proposed control strategy are validated, especially tracking speed under PSCs. Simulation results show that the BFBIC topology with the proposed IGWO algorithm outperforms other algorithms on most cases, especially only takes the tracking time of 0.24s and reaches the efficiency of 98.54% under the most severe PSCs.

In this paper, a multi-parameterized proximal point algorithm combining with a relaxation step is developed for solving convex minimization problem subject to linear constraints. We show its global convergence and sublinear convergence rate from the prospective of variational inequality. Preliminary numerical experiments on testing a sparse minimization problem from signal processing indicate that the proposed algorithm performs better than some well-established methods.

We investigated whether and how political misinformation is engineered using a dataset of four months worth of tweets related to the 2016 presidential election in the United States. The data contained tweets that achieved a significant level of exposure and was manually labelled into misinformation and regular information. We found that misinformation was produced by accounts that exhibit different characteristics and behaviour from regular accounts. Moreover, the content of misinformation is more novel, polarised and appears to change through coordination. Our findings suggest that engineering of political misinformation seems to exploit human traits such as reciprocity and confirmation bias. We argue that investigating how misinformation is created is essential to understand human biases, diffusion and ultimately better produce public policy.