Search

Aiming at the problems of high cost of manual investigation, low accuracy and difficult side-to-end information interaction in the verification of the relationship between distribution transformer terminal and household transformer in low-voltage substation area, a low-voltage substation area topology identification verification method based on DTW algorithm is proposed. According to the time series voltage data of smart meter and distribution transformer terminal, this paper uses DTW algorithm to calculate the similarity of time and voltage series, and checks the user by setting threshold. Then KNN algorithm is used to determine the station area, and the household transformer relationship is determined. Finally, according to the time series voltage data collected by the meter, the DTW algorithm is used to calculate the voltage sequence similarity between users, and the feeder position of users is located combined with the topology. Experiments show that the method proposed in this paper has the characteristics of high accuracy and high efficiency, and it can accurately verify the relationship between household and transformer of distribution transformer terminal equipment in low voltage substation area.

This paper proposes a critical service optical network layer recovery mechanism based on fault prediction and fault model assistance and deep learning technology, improving the success rate of critical service recovery.

Redundant parameters in robot kinematic error model may affect the accuracy and robustness of identification results, and should be removed before calibration. In this paper, the pose error model of end-effector was derived, and redundant parameters of robot with arbitrary structure were obtained by analyzing the linear correlation columns of Jacobian matrix. To verify the correctness of the proposed method, a self-calibration method based on plane constraint was proposed. The simulation results show that non-redundant parameter identification model can improve the position accuracy regardless of the presence of noise, while the full-parameter identification model is susceptible to noise and the identification results do not converge. The correctness of the parameter redundancy analysis is proved by the experiments on Franka Emika serial robot. After removing redundant parameters, the position errors reduced from 2.41mm to 1.20mm.

In this paper, we propose a differentiable light-weight architecture search method, called DLWAS. It can search for CNNs with fewer parameters and floating point operations (FLOP-s) while maintaining the SOTA performance. Concretely, we first build a new light-weight search space, which contains the latest and effective light-weight operations, limiting the parameters and FLOPs from the source of neural architecture search (NAS). Secondly, we propose an effective neural architecture optimization method, which results in a more sparse and robust topology for differentiable NAS. Experimental results show that DLWAS achieves an error rate of 2.76% on CIFAR10 comparable to the state-of-the-art methods, using only 2.4M params and 336M FLOPs. The parameters and FLOPs are reduced by 30% and 36% compared with the closest counterpart DARTS, respectively. On ImageNet, our model achieves 3.2% better top-1 accuracy than the SOTA MobileNet, while using fewer parameters and FLOPs.

In recent years, with the development of power communication network, the research on the networking mode and intelligent routing reconstruction of power grid backbone network has gradually become a hot direction, the intelligent reconstruction and optimization method of communication network have also become the focus of public network and power communication network. Focusing on the network survivability, this paper puts forward the application strategy of self-optimization technology based on deep learning in power grid backbone communication network from three aspects: the basic model of power grid backbone optical network, the data acquisition and fault diagnosis technology of power grid backbone communication network, and the intelligent reconfiguration method of communication network for power production business.

The Data Distribution Service (DDS) standard enables integration and scalability of common distributed simulation systems, and its flexible data communication mechanism can improve the communication efficiency. We propose an overall architecture for a distributed simulation system based on DDS, including the communication interface adapted to the EXata platform and the system clock synchronization module. The proposed simulation system can simulate a large-scale network and satisfies the demands of communications simulations.

Target detection and tracking is a key technology in intelligent monitoring system. However, it is still difficult to design a robust, accurate and real-time target detection and tracking algorithm. Taking target Tracking as the research focus, this paper focuses on one of the current mainstream algorithms, namely, TLD (tracking-learning-detection) target Tracking algorithm. Based on the theory of TLD target tracking and corner detection, a new target tracking method was proposed to solve the existing problems. A classifier is built online to track and learn the target in real time, and a target contour detector is added in the system to complement the random forest detector to solve the problem of tracking failure. Experimental results show that the proposed algorithm can effectively overcome the influence of fast change or depth occlusion, and improve the tracking efficiency of moving targets.

Today’s video service providers all desire to deeply understand the ever changing factors on user QoE to attract more users. In this paper, we study the user engagement with respect to video quality metrics and improve user engagement in adaptive video streaming systems. We conduct a comprehensive study of the real data from iQIYI, covering 700K users and 150K videos. We find bitrate switch becomes the new dominant factor on user engagement instead of rebuffering events. We also observe the impact of rate of rebuffering is more dominant than rebuffering time. We examine novel interdependencies between quality metrics in the system, e.g., the positive correlation between bitrate switch and average bitrate, which is due to the context system strategy, i.e., conservative bitrate enhancing strategy adopted by iQIYI. To improve user engagement, we propose a new engagement centric QoE function based on real data and design server side ABR algorithm which leverages our new QoE function. We evaluate our method for online test in iQIYI, with 490K real users viewing 666K streams. The results show our approach outperforms existing approaches by significantly improving the viewing time, i.e., 2.8 minutes longer viewing time per user.

This digest paper reports on the demonstration of a hybrid system consisting of a silicon carbide (SiC) microdisk cavity and a hexagonal boron nitride (h-BN) nanomechanical resonator. Strong mechanical coupling between the constituents is revealed in both optical interferometry measurements and finite element method (FEM) simulations, even with the thickness of h-BN scaled down to a few atomic layers. Numerical computations assess that the hybrid device will lead to an improvement of vacuum coupling rate by 454 Hz (∼47%) at the h-BN thickness of 36 nm, compared to the bare SiC microdisk. Understanding of the nanomechanical properties facilitates design and implementation of SiC/h-BN hybrid systems and opens a new door to coherent transduction between optics and mechanics on this platform.

A frequency selective rasober (FSR) with high-temperature-resistance, wideband radar cross section reduction (RCSR), and flat-broad transmission band is presented in this paper. It consists of three functional layers: a piece of insulation tile, scattering cancellation metasurface based on interdigitated element, and a triple-layer-coupled-resonator spatial filter frequency selective surface. At low-frequency band, the incident energy is diffused to various directions, resulting in 10 dB RCSR over 2.52-11.68 GHz with a 129% fractional bandwidth. Besides, there is a highly efficient transmission band (insertion loss < 0.595 dB) over 16.53-19.83 GHz. It is also worth mentioning that our FSR can maintain such good properties at 500°C for 5 seconds with the help of insulation tile.

The detection of containerized containers is the core of intelligent de-palletizing system, which is lack of robustness due to factors such as the gaps between the containers, the patterns on the surfaces of containers, noises and ambient light. For this reason, a containers detection algorithm based on computer vision is proposed in this paper. This method used line merging operations to compensate for the broken lines in the Canny edge detection and LSD line extraction process. Secondly, based on the geometric characteristics of the containerized containers, four general rectangular structure elements are defined, which are used to construct the rectangles by a specific matching search algorithm. Then, based on the improved mean-shift clustering algorithm, cluster centers are used to replace a cluster of rectangles, which are used for x-axis and y-axis clustering operations, so as to eliminate the false detections caused by the factors such as patterns and straps. Finally, the non-maximum suppression algorithm is used to screen out the accurate rectangular contours. Experiments have verified that the detection method has an accuracy rate of about 99% for the identification of solid-colored containers. It also has good robustness for containers with patterns, straps, stains, and deformations. The recognition accuracy rate is about 95%.

It is known to all that Coronavirus disease 2019 (COVID-19) is an infectious disease, which has spread to more than 180 countries and territories around the world since the end of 2019.lt is clear that this disease results in a moment of panic. However, some people used the data of Tieba to analyze public opinion and describe the intensity of people’s emotions, but Tieba is not representative. Because whether it is in terms of popularity, number of users or age spread, the data of Weibo is more representative compared to Tieba. Therefore, this article carries out the public opinion analysis by using Python crawler to obtain data on the topic of COVID-19 in Weibo. The methods we use include word frequency analysis tools, plug-ins in the ROST content mining system, and analysis interfaces in the Baidu Brain Al open platform. The purpose of this article is to analyze the mental health level of the public, such as nervousness, fear, panic, and anxiety. By experimental results on the relevant Weibo in February and March, we find that positive emotions stood at 14.64%,neutral emotion blog posts still accounted for 66.64% and negative emotions accounted for 18.72%.In the negative segment, the general negative emotions accounted for 12.485%,the moderate negative emotions accounted for 4.9%,and the users showed a high degree of negative emotions accounted for 0.455%.And then we roughly give the relevant strategies and suggestions for some groups which have the intense negative emotions. The limitation is that we can’t give them some specific suggestions for different users for the time being, which needs further research in the future.