Wind power forecasting plays a crucial role in enhancing the stability and reliability of power system operations. However, accurately wind power prediction is challenging due to its randomness and discontinuity. Most current methods tend to offer a singular predicted value, which inevitably imbued with errors, the errors precipitate a cascade of stability issues within the power grid. In order to address this issue, this study proposed a novel model performed both point and interval forecasting of wind power, leveraging the synergistic capabilities of graph convolutional neural networks (GCN), long short-term memory neural networks (LSTM) and quantile regression. Firstly, the adjacency matrix was constructed by leveraging the pearson correlation between historical wind power data and meteorological variables. Secondly, the matrix and original data served as the input of GCN. Thirdly, the output of the GCN was fused with the original data, and then, the hybrid data was taken as the input of LSTM. Finally, pinball loss was used as the training objective to obtain multiple quantile values of wind power. The experimental findings indicated that the proposed model not only achieved the lowest error rates in point forecasting but also provided the most precise and dependable prediction intervals across various confidence levels.

The challenges of small target detection, inference speed, and detection accuracy are key research areas in intelligent waste classification recognition, particularly the incorporation of attention mechanisms into algorithms. Existing self-attention mechanisms often face speed limitations due to low efficiency, particularly in MHSA's tensor reshaping and element-by-element computation, which are computationally intensive and challenging for embedded device deployment. To address these issues, this paper introduces a Lightweight Multi-Scale Attention Model (LMSA) based on YOLOV5 and the self-attention mechanism. The model constructs a new lightweight multi-scale attention module, utilizing a single memory-constrained MHSA between effective FFN layers to enhance channel communication while improving access efficiency. Additionally, a cascaded grouped attention module is proposed to provide different divisions of full features to the attention head, saving computational overhead and enhancing attention diversity. Experimental results demonstrate that on the VisDrone2021 dataset, the AP75 of LMSA is 41.43%, achieving the state-of-the-art (SOTA) among mainstream models. On a self-constructed dataset, the AP50 of LMSA reaches 74.55%. With equivalent accuracy, LMSA has lower computational power consumption, and the inference speed reaches 140fps on an RTX3090 graphics card. The model's advantages include fewer parameters, high accuracy, and fast inference speed, making it more suitable for embedded detection equipment and contributing to the intelligent development of waste detection.

及时发现并诊断变电站运行中设备存在的问题,是保障电网安全运行的关键手段之一.基于深度网络与知识图谱技术,提出一种关联变电站设备多模态信息的故障智能诊断方法.利用深度学习技术和知识图谱方法对采集的多模态数据进行知识提取和融合,构建一个多模态信息融合的语义知识图谱;使用YOLOv4算法对故障样本聚类并提取先验框参数;将多模态知识图谱和YOLOv4视觉检测相结合,应用到变电站场景中,实现变电站设备的自主预警诊断.实验表明,该模型可以实现故障诊断决策智能化的目标,从而提高电网的日常运行、维护和管理效率.

Garbage classification is a social issue related to people’s livelihood and sustainable development, so letting service robots autonomously perform intelligent garbage classification has important research significance. Aiming at the problems of complex systems with data source and cloud service center data transmission delay and untimely response, at the same time, in order to realize the perception, storage, and analysis of massive multisource heterogeneous data, a garbage detection and classification method based on visual scene understanding is proposed. This method uses knowledge graphs to store and model items in the scene in the form of images, videos, texts, and other multimodal forms. The ESA attention mechanism is added to the backbone network part of the YOLOv5 network, aiming to improve the feature extraction ability of the network, combining with the built multimodal knowledge graph to form the YOLOv5-Attention-KG model, and deploying it to the service robot to perform real-time perception on the items in the scene. Finally, collaborative training is carried out on the cloud server side and deployed to the edge device side to reason and analyze the data in real time. The test results show that, compared with the original YOLOv5 model, the detection and classification accuracy of the proposed model is higher, and the real-time performance can also meet the actual use requirements. The model proposed in this paper can realize the intelligent decision-making of garbage classification for big data in the scene in a complex system and has certain conditions for promotion and landing.

目的 针对目前包装产业存在的产业链长、数据大而散、包装领域知识信息检索不精准等问题,设计一个基于知识图谱的包装领域智能问答系统.方法 采用知识图谱、智能问答、自然语言处理、深度学习和个性化推荐等技术,对互联网上各类包装领域数据进行采集汇聚、抽取知识、融合计算,最终形成一个包装领域混合型智能问答系统.结果 系统主要功能包括知识图谱、文本相似匹配、图像识别和自动问答,实现了包装领域大数据的知识卡片、语义搜索和沉浸问答.结论 该系统使用方便,通过问答方式,一站式快速按需获取多视图、多维度包装领域数据,提升了行业的数字化、信息化和智能化水平.

In order to enable Social Internet of Vehicles devices to achieve the purpose of intelligent and autonomous garbage classification in a public environment, while avoiding network congestion caused by a large amount of data accessing the cloud at the same time, it is therefore considered to combine mobile edge computing with Social Internet of Vehicles to give full play to mobile edge computing features of high bandwidth and low latency. At the same time, based on cutting-edge technologies such as deep learning, knowledge graph, and 5G transmission, the paper builds an intelligent garbage sorting system based on edge computing and visual understanding of Social Internet of Vehicles. First of all, for the massive multisource heterogeneous Social Internet of Vehicles big data in the public environment, different item modal data adopts different processing methods, aiming to obtain a visual understanding model. Secondly, using the 5G network, the model is deployed on the edge device and the cloud for cloud-side collaborative management, aiming to avoid the waste of edge node resources, while ensuring the data privacy of the edge node. Finally, the Social Internet of Vehicles devices is used to make intelligent decision-making on the big data of the items. First, the items are judged as garbage, and then the category is judged, and finally the task of grabbing and sorting is realized. The experimental results show that the system proposed in this paper can efficiently process the big data of Social Internet of Vehicles and make valuable intelligent decisions. At the same time, it also has a certain role in promoting the promotion of Social Internet of Vehicles devices.