With the continuous development of the Internet of Things technology, the Internet of Things technology can be used in many aspects of the smart grid. Apply advanced Internet of Things technologies in substation GIS, including sensing technology, network communication technology, intelligent decision-making technology, automatic control technology, and energy and power technology, etc., and combine them closely to form a coordinated organic system. At present, partial discharge is the main cause of equipment insulation deterioration in substations. Partial discharge detection and evaluation has become an important means of monitoring insulation conditions. The monitoring of GIS partial discharge conditions in substations can help find insulation defects in GIS in time, avoid insulation failures, and improve the safe operation of GIS. It is possible to achieve GIS insulation condition maintenance, reduce power outage time and save maintenance costs. The effective application of the Internet of Things technology in the GIS partial discharge state detection system of intelligent substations can effectively improve the construction of modern intelligent substations in China.

Partial discharge (PD) surveillance constitutes a pivotal methodology for diagnosing insulation failures in electrical equipment. Enhancing comprehensively the precision of identifying PD anomalies in Gas Insulated Switchgear (GIS) is of paramount significance for ensuring the steady functioning of power grids. This study introduces a novel framework that integrates Phase-Resolved PD Graph Segmentation (PRPD-Grabcut) with a tailored MobileNets-based Convolutional Neural Network (MCNN) to classify GIS-related PD issues. Leveraging image segmentation via PRPD-Grabcut, crucial features are extracted from PRPD diagrams, which then facilitate the construction of the MCNN model. This model employs depth-wise separable convolutions alongside inverted residual architectures to tackle the vanishing gradient dilemma inherent in Deep Convolutional Neural Networks (DCNNs) during GIS PD pattern discernment. Upon the model's subsequent training and validation, empirical evidence illustrates that the PRPD-Grabcut-MCNN hybrid significantly alleviates the computational load and storage requisites of the model, concurrently enhancing the recognition precision and expediting the training process of the neural network. Relative to diverse established lightweight neural network architectures, MCNN manifests superior performance in terms of recognition accuracy, reduced cross-entropy loss, and expedited training duration.

Electrostatic electret generator has attracted a lot of attention in recent years, but their low power density cannot be ignored. Based on the existing research, this paper increases its power density from two aspects, including in-plane and in-space aspects. In the plane aspect, through simulation analysis, we can draw a conclusion that the output power can be increased effectively by using the bipolar charging method. From the perspective of space, the power density can be increased several times by connecting the output ends of the multi-layer generation structure. By integrating the method into the electrostatic electret generator, the power output can be effectively increased, which is suitable to the application of higher energy requirements in the specific field and a certain application prospect.