Synthetic Aperture Radar (SAR) scene classification is challenging but widely applied, in which deep learning can play a pivotal role because of its hierarchical feature learning ability. In the paper, we propose a new scene classification framework, named Feature Recalibration Network with Multi-scale Spatial Features (FRN-MSF), to achieve high accuracy in SAR-based scene classification. First, a Multi-Scale Omnidirectional Gaussian Derivative Filter (MSOGDF) is constructed. Then, Multi-scale Spatial Features (MSF) of SAR scenes are generated by weighting MSOGDF, a Gray Level Gradient Co-occurrence Matrix (GLGCM) and Gabor transformation. These features were processed by the Feature Recalibration Network (FRN) to learn high-level features. In the network, the Depthwise Separable Convolution (DSC), Squeeze-and-Excitation (SE) Block and Convolution Neural Network (CNN) are integrated. Finally, these learned features will be classified by the Softmax function. Eleven types of SAR scenes obtained from four systems combining different bands and resolutions were trained and tested, and a mean accuracy of 98.18% was obtained. To validate the generality of FRN-MSF, five types of SAR scenes sampled from two additional large-scale Gaofen-3 and TerraSAR-X images were evaluated for classification. The mean accuracy of the five types reached 94.56%; while the mean accuracy for the same five types of the former tested 11 types of scene was 96%. The high accuracy indicates that the FRN-MSF is promising for SAR scene classification without losing generality.

Accurate evaluation of the cable trench condition is the key to realizing its intelligent operation and maintenance, and the intelligent cable trench inspection robot is a reliable information acquisition method for the condition assessment of cable trench. According to task requirements and site’s environment, an intelligent inspection robot is designed. The crawler-type motion mechanism is determined based on the demand for sizes and obstacle-surmounting. Cartographer SLAM technology is used to map the underground cable trench environment, and a path planning method combined with the improved A^∗ algorithm and dynamic window method (DWA) is presented. Multiple sensors are equipped on this robot, and various kinds of information are obtained. Based on the information obtained, a condition assessment method of an underground cable trench is proposed. An extension neural network model is constructed to assess air quality. Channel capacity and ground flatness are calculated to evaluate the internal structure of the cable trench. The water depth and water surface area are comprehensively used to evaluate the hazard of water accumulation. The evaluation results are used to realize the linkage control of the cable trench’s operation and maintenance. A field test shows that the intelligent inspection robot can reliably obtain seven kinds of information from the underground cable trench, and the proposed condition evaluation method can assess the condition of the cable trench and provide four kinds of suggestions for potential hazards.

In aircraft feature detection, the difficulty of acquiring Synthetic Aperture Radar (SAR) images leads to the scarcity of some types of aircraft samples, and the high privacy makes the personal sample set have the characteristics of data silos. Existing data enhancement methods can alleviate the problem of data scarcity through feature reuse, but they are still powerless for data that are not involved in local training. To solve this problem, a new federated learning framework was proposed to solve the problem of data scarcity and data silos through multi-client joint training and model aggregation. The commonly used federal average algorithm is not effective for aircraft detection with unbalanced samples, so a federal distribution average deviation (FedDAD) algorithm, which is more suitable for aircraft detection in SAR images, was designed. Based on label distribution and client model quality, the contribution ratio of each client parameter is adaptively adjusted to optimize the global model. Client models trained through federated cooperation have an advantage in detecting aircraft with unknown scenarios or attitudes while remaining sensitive to local datasets. Based on the YOLOv5s algorithm, the feasibility of federated learning was verified on SAR image aircraft detection datasets and the portability of the FedDAD algorithm on public datasets. In tests based on the YOLOv5s algorithm, FedDAD outperformed FedAvg’s mAP0.5–0.95 on the total test set of two SAR image aircraft detection and far outperformed the local centralized training model.

In complex data environments, rational handling of unbalanced datasets is key to improving the reliability of early disease prediction. Early warning of disease risk in both temporal and spatial terms, contributes to disease prevention and treatment. To this end, a bi-dimensional substratum information mining model based on Association Rule Digging with Dynamic Thresholding and Weight Optimization (ARDdtwo) was proposed for the early diagnosis of thyroid cancer. It is an integrated assessment framework consisting of association rule digging by constructing a dynamic threshold model (ADRcdt) for qualitative analysis, and a self-optimizing component importance measurement model (SoCIM) for quantitative analysis. ARDcdt incorporates temporal and spatial features of sparse data to address the distributional bias problem. Moreover, new importance diagnostic calculations were designed to further identify high-risk low-frequency (HRLF). The SoCIM can determine the relative weight of each component by assessing its level of risk in the overall system based on the Risk Enhancement Level (REL) and Risk Reduction Level (RRL), realizing the self-adjustment and optimization of the weight setting. Finally, the model was validated through an empirical analysis. The evaluation of the research work shows that improved results were achieved, such as accuracy, f1-score, and precision, with optimized values of 36.04%,56.57%, and 53.89%, respectively. The overall area under the curve for the model was 0.882. This proves the validity of the proposed model for practical applications. For patients, it can simplify the pathological process and reduce the examination costs.

As a crucial component of power systems, distribution transformers are indispensable to ensure the sustainability of power supply. In addition, unhealthy transformers can lead to wasted energy and environmental pollution. Thus, accurate assessments and predictions of their health statuses have become a top priority. Unlike assumed ideal environments, however, some complex data distributions in practical scenarios lead to more difficulties in diagnosis. One challenge here is the potential imbalanced distribution of data factors since sparsely occurring factors along with some Unusual High-Risk (UHR) components, whose appearance may also damage transformer operations, can easily be neglected. Another is that the importance weight of data components is simply calculated according to their frequency or proportion, which may not always be reasonable in real nonlinear data scenes. With such motivations, this paper proposes a novel integrated method combining the Two-fold Conditional Connection Pattern Recognition (TCCPR) and Component Significance Diagnostic (CSD) models. Initially, the likely environmental factors that could result in distribution transformer heavy overloads were incorporated into an established comprehensive evaluation database. The TCCPR model included the UHR time series and factors that are associated with heavy overload in both spatial and temporal dimensions. The CSD model was constructed to calculate the risk impact weights of each risky component straightforwardly, in line with the total risk variation levels of the whole system caused by them. Finally, the results of one empirical case study demonstrated their adaptation capability and enhanced performance when applied in complex and imbalanced multi-source data scenes.