To address the issue of multipoint leakage detection in energy transportation systems, a multiscale convolutional neural network based on kurtosis and KL divergence (KKL-MSCNN) was proposed for multipoint leakage detection in energy transportation systems. Initially, the collected infrasound data undergoes complete ensemble empirical mode decomposition with adaptive noise (CEEMDAN). Furthermore, a hierarchical processing technique for intrinsic mode functions (IMFs) is proposed to reconstruct IMFs at two different feature levels. Following this, a lightweight multiscale convolutional neural network (MSCNN) is constructed, comprising two channels. The reconstructed IMF features are then extracted through serial and parallel convolution at varying scales, facilitating the completion of pipeline leakage level classification. In comparison to conventional methods, the proposed approach achieves a significant 20.26% increase in accuracy for multipoint leakage detection.

Aiming at the problem of insufficient label data in the pipeline leak detection field, this paper proposes a pseudolabel (PL) adaptive learning method based on multiscale convolutional neural network (MSCNN) with the idea of transfer learning for pipeline leak aperture identification. First, the convolutional and pooling layers for transfer learning feature extraction are improved by using a dual-channel MSCNN. Second, the KL divergence function after dimensionality reduction is used to calculate the distribution distance between the source domain and the target domain to improve the robustness of distribution alignment in high-noise environments. In addition, considering the interference of PL noise, this paper develops a pseudolabel (PL) dynamic threshold to achieve the purpose of PL adaptive updating. Compared with the fixed threshold, the improved PL learning (PLL) can effectively improve the prediction accuracy of the model. The effectiveness of the method proposed in this paper is verified by predicting pipeline leakage conditions at different distances and under different pressures. The comparative analysis results show that the method in this paper is superior to other transfer learning methods in terms of prediction accuracy, stability, and convergence speed.

In a dynamic threat environment, multisource track association is a major technical challenge for achieving large-scale real-time target tracking. However, existing methods focus mostly on matching low-dimensional kinematic features, hindering the capture of deep semantic coupling between different modal data, resulting in superficial association results and limited insight into complex behaviors. To overcome this limitation, we propose a multisource track association method based on a multiview weighted graph neural network, which aims to improve the association accuracy from the perspectives of multigranularity graph semantic modeling and robust deep feature learning. First, to mitigate the impact of noise present in the track data, we employ variational mode decomposition to smooth the data and reduce association-induced errors. We subsequently design a view distinguishing strategy that determines edge weight thresholds on the basis of weighted Euclidean distances. Specifically, if the distance between two nodes exceeds the threshold, a first-order graph is constructed to capture global physical differences; otherwise, a second-order graph is built to capture local similarities and reveal potential strong associations. Furthermore, we design a residual-aware attention graph convolutional network (GCN), in which residual connections are embedded into the GCN architecture to suppress the effect of oversmoothing. Finally, we construct a structural invariance constraint to facilitate the effective learning of subgraphs under structural inconsistencies. The experimental results demonstrate that, under the same testing conditions, the proposed method outperforms other comparative algorithms by effectively extracting high-dimensional features from tracks, with an average F1 score improvement of 12.5%, which has broad research prospects and application value.

Multisource track association (MTA) is a key measurement-fusion problem in maritime situational awareness, where tracks reported by heterogeneous sensors must be correctly matched under asynchronous sampling, sensor noise, and complex target maneuvers. Existing deep learning-based methods have improved association accuracy by learning discriminative track representations. However, many of them encode two tracks separately and compare only pooled track-level embeddings, so the fine-grained interactions between individual track points are introduced only implicitly. This may limit their ability to distinguish targets with similar motion patterns in dense or ambiguous scenarios. To address this issue, this paper proposes a multilevel interactive Siamese graph neural network, termed (MIS-GNN). First, each track is represented as a graph and encoded by a weight-sharing Siamese GraphSAGE encoder to preserve local spatiotemporal topology. Then, a global interaction layer based on cross-track self-attention is introduced to enable node-level message passing between two track graphs, providing a soft alignment mechanism for asynchronous observations. Finally, graph-to-node cross-attention and multiscale graph-to-graph similarity measurements are jointly employed to evaluate track similarity from complementary granularities. Experiments on the public MTAD benchmark demonstrate that MIS-GNN achieves an average F 1-score of 0.9264, outperforming state-of-the-art comparison methods. Additional sensitivity, significance, and efficiency analyses further verify the robustness and deployment feasibility of the proposed method.