The surrounding of transmission lines is complex, large-scale construction machinery, and smoke and fire pose threats to the circuit facilities and wires. When anomaly occurs, much time is required for the State Grid Corporation to fix it manually. To reduce the inspection burden, we propose a lightweight model running on embedded device to detect foreign objects of transmission lines. Based on the You Only Look Once (YOLO) v3, we use Mobilenetv2 instead of Darknet-53 as the backbone, and use depthwise separable convolution to replace 3 × 3 convolutional kernels in detection head, which greatly reduces the parameter size of the network. And the Fully Convolutional One-Stage Object Detection (FCOS)-like encoding and decode scheme is adopted to reduce network complexity. Meanwhile, in order to compensate for the degradation of accuracy, we have improved data augmentation, learning rate, and loss function. The experiments show that compared with other existed models, the improved YOLOv3 model has a smaller model size and higher detection speed without notably reducing detection accuracy, which has achieved the balance between detection speed and accuracy.
During the grouting operation in the underground coal mine, abnormal curing behaviors such as foaming and drainage often lead to the loss of reinforcement effect of the polyurethane/water glass (PU/WG) materials on coal walls and even cause safety accidents. Herein, three kinds of PU/WG grouting materials were successfully prepared by changing the type of catalysts, which were the normal sample (C7), the foaming sample (C14), and the sample with drainage (C17) during curing. The structure, thermal stability, and compressive strength of the three samples were characterized by scanning electron microscope (SEM), energy dispersive spectroscopy (EDS), X-ray diffraction (XRD), X-ray photoelectron spectroscopy (XPS), thermogravimetry-differential thermal analysis (TG-DTA), and pressure testing machine. The results showed that the abnormal curing behavior had little effect on the thermal stability of the material, but it had a significant effect on the microstructure and compressive strength of the consolidated body. C7 exhibited a typical three-phase distribution, in which the polysilicate microspheres encapsulated by acicular carbonate were embedded in the polymer continuum. The structure of C7 had high rigidity and hardness, and the compressive strength was up to 43 MPa. The three-phase structure of C14 disappeared gradually with the increase of catalyst content, the hard block material and matrix are porous, and the compressive strength was only 2.7 MPa. The organic polymer of C17 existed in the form of microsphere and distributed irregularly in the continuum composed of inorganic components, and the compressive strength was 4.9 MPa. The abnormal solidification behavior such as foaming and drainage made the water glass/polyurethane material lose its basic mechanical properties, which cannot meet the needs of grouting reinforcement in coal mines. Therefore, the type of catalyst had a significant impact on the stability of the system, and it is necessary to avoid selecting catalysts that are likely to cause abnormal solidification during formulation research.
Abstract. This paper describes the development status and main problems of load monitoring, introduces the key technologies of load monitoring, and by using a load state monitoring system, emphatically illustrates a detection algorithm for abnormal state of the secondary load of the current transformer on the three-phase line of the power grid. The algorithm mainly achieves the real-time detection of abnormal state such as disconnection, short connection and series connected semiconductor. In the light of this algorithm, the working principle is explained, the model formula is worked out, and the state criterion is given. The load condition monitoring system is debugged, put into operation and tested in the pilot operation, and the results show that the algorithm has a good effect.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.