Fast response flight multi-service nodes prediction based on one hidden layer neural networkLi, Baiqiang; yang, Xiaoming; Tang, Zhengjiang; Peng, Wentao; Zhou, Yang; Liao, Fangmin; Tan, Jing; Lin, Xuan
doi: 10.1088/1742-6596/2522/1/012006pmid: N/A
Airport scene multi-service process(ASMP) state prediction is one of the important links to optimize airport ground resource allocation, improve collaborative release scheduling, and improve airport operation efficiency. In order to solve the problem of the slow response of existing prediction models, a Gaussian kernel probability density model of multi-service nodes in airport scenes is established. Then the one hidden layer neural network(OHLNN) is used to predict the nodes of ASMP. A study on the data of more than 20,000 flights at a large airport in China shows that: The mean absolute error (MAE) of the mixed flight prediction result is only 2.62911min, the mean root mean square error (RMSE) is only 3.24564min, and the response speed of the prediction model in the verification set is about 0.48s. When the results of MAE and RMSE are good, the response speed of the prediction model is greatly improved.
A Research on Operator Design Method for Health Detection SoCLi, Qiuping; Wu, Fangke; Zhang, Xing; Wang, Xin’an; Ma, Jieru; Li, Xueqing
doi: 10.1088/1742-6596/2522/1/012026pmid: N/A
Improving the efficiency of System-on-Chip(SoC) design has always been a hot topic in SoC design methods. The quality of SoC architecture design directly affects the core elements of SoC performance and cost. However, architecture design usually requires highly experienced and skilled personnel, which are scarce resources. Operators in SoC, like instructions in a CPU, are extracted from algorithms and are both universal and flexible. SoC can be designed based on operators, and algorithmic functions can be implemented through operators’ interconnection. In this paper, we propose an operator design method to improve the efficiency of SoC design. Additionally, we design algorithmic operators using pulse signal processing as an example to enhance the efficiency of designing health detection SoCs.
Research and Application of Panoramic Data Visualization Platform for Stability Control Equipment Based on the Whole Life CycleGao, Qili; Huang, Ying; Li, Huaiqiang; He, Xiao; Sun, Yuyang
doi: 10.1088/1742-6596/2522/1/012027pmid: N/A
In recent years, with the mass access of new energy such as UHV commissioning projects and photovoltaic wind, the problems in safely and stably operating regional power grid are becoming more and more prominent. Mass power grid stability control system and equipment play a significant role in this aspect. However, there are a great variety of manufacturers and models of the stability control systems and equipment, as well as configuration schemes for the operation strategy and fixed value of the primary equipment. Due to insufficient technological support, it is relatively inconvenient to manage the information and data of stability and control system on the whole process from feasibility research, initial test, factory debugging, site debugging, commissioning acceptance, protection verification, daily operation and maintenance management to the decommission of the system, thus failing to achieve the objective of classified management according to the relevant requirements, nor to modify and improve in time. Based on the management requirements and status quo of the stability and control equipment, a technical research on the panoramic data visualization platform for the stability control equipment with full life-cycle information covering the feasibility research, release, debugging, commissioning, verification, operation, decommission and so on of the equipment, which is available for the panoramic informationized and digital management of the stability control equipment, so as to realize the whole life cycle and systematic closed-loop management of the stability control system and the equipment.
A Data-Driven Approach to Interdisciplinary Collaborative Discovery and RecommendationWang, Junxiang; Wang, Yizhe; Zhu, Zixuan; Hu, Yitong; Cao, Yuxuan; Xing, Zehao; Xie, Binzhu
doi: 10.1088/1742-6596/2522/1/012009pmid: N/A
The convergence of different fields is crucial for scientific advancement, and interdisciplinary collaborations among scholars are necessary for fostering this development. However, identifying and recommending potential interdisciplinary collaborators systematically is a challenging task. This paper proposes an interdisciplinary collaboration discovery and recommendation model for scholars in different fields. We utilized clustering algorithms, including K-Means, DBSCAN, and Affinity, to generate a scholarly interdisciplinary collaboration discovery graph. To recommend potential collaborators, we proposed an algorithm that considers both the suitability between scholars and the individual comprehensive influence of scholars. The effectiveness of the model was validated using data from 126 scholars at the Beijing University of Posts and Telecommunications.
Reliability Analysis for Programs with Redundancy Computation for Soft ErrorsMeng, Xiankai; Zhang, Zhuo; Xue, Jianxin; Chen, Fangshu; Wang, Jiahui
doi: 10.1088/1742-6596/2522/1/012022pmid: N/A
Soft error is one of the factors which may affect the reliability of computer programs. A common method to alleviate the impact of soft errors is redundancy computation, a classical data flow error detection mechanism. However, a program with redundancy computation may still have some vulnerable spots, which might be caused by the flaw during implementation or the instruction reordering given by compiler optimization. Finding the vulnerable spots of a program with redundancy computation is of great significance to evaluate the capability of the error detection mechanism. There are some conventional methods to analyze the reliability of a program under soft errors, such as the irradiation experiment, fault injection, and modeling analysis. However, the irradiation experiment is expensive, fault injection is very time-consuming, and the existing modeling analysis methods have not considered the error detection mechanism. This paper proposes a novel method of reliability analysis for programs with redundancy computation by analyzing the dynamic instruction sequence. Experimental results show that our approach has fairly high accuracy and a false negative rate of about 0.0545.
Key Points Trajectory and Predicted-Real Frames Distinction based Mirror and Glass Detection for Indoor 5G Signal AnalysisWang, Ziyue; Liu, Yanchao; Cheng, Xina; Ikenaga, Takeshi
doi: 10.1088/1742-6596/2522/1/012033pmid: N/A
Mirror and glass are ubiquitous materials that heavily influence the transmission of indoor 5G signals with extremely high frequency. The existing vision system always tends to neglect them or misdiagnose them, which is unsuitable for 5G signal analysis in the 3D indoor environment. This paper proposes key points trajectory distinction and predicted-real frames distinction to detect mirror and glass regions in video sequences. Firstly, key points trajectory is used to extract the special motion information of reflection in mirror and glass region. Secondly, predicted and real frames distinction is used to remove the wrong detection region at the pixel level. Extensive experiments demonstrate that the proposed method achieves 40 - 50 % accuracy higher than another related state-of-the-art method for mirror and glass detection in the general 3D living environment.
Pseudo-label self-training model for transfer learning algorithmChen, Zijie; Ling, Weixin
doi: 10.1088/1742-6596/2522/1/012008pmid: N/A
When aligning joint distributions between domains, the existing transfer learning algorithms usually assign pseudo labels due to the lack of labels in target domain. However, the noise in pseudo labels will affect the performance of transfer learning. Pseudo-label self-training for transfer learning (PST-TL) model is proposed to generate reliable pseudo labels for target domain and have a wide range of applications in existing algorithms. Pseudo labels are predicted by an ensemble classifier using absolute majority vote, and labels predicted successfully are considered to be high confidence. The training of ensemble classifier applies the self-training of joint pseudo labels strategy, adding strongly stable data to training set of the classifier. The semi-supervised and unsupervised transfer learning tasks in experiment show that the existing transfer learning algorithm can significantly improve the transfer performance after embedded by PST-TL model.
Research on Image Fusion in UAV Visual NavigationLi, Dongtao; Shao, Yuanzheng; Yu, Yao; Feng, Zhifang; Wen, Jiaolong
doi: 10.1088/1742-6596/2522/1/012007pmid: N/A
The accuracy of multi-sensor image fusion is the key to improve the accuracy of UAV visual navigation. In this paper, a two-step registration method combining region registration and feature registration is proposed. First, rough image registration is carried out based on improved Hough transform and Fourier-Mellin transform, and image precise registration is implemented based on Harris corner feature and cross-correlation function. Finally, wavelet transform is applied to image fusion and image edge enhancement to realize the coarse to accurate registration of SAR images.
Short-term Metallurgical Load Forecasting Based on Adaptive Ensemble LearningPengyu, Ma; Weijian, Kong; Zhiyong, Su
doi: 10.1088/1742-6596/2522/1/012002pmid: N/A
Accuracy and rapidity are the primary objectives of load forecasting, and also the necessary conditions for ensuring power supply and production schedule. However, in actual production, due to the variability of operating modes and the interference of production environment, the difficulties such as non-stable and high fluctuation are in the power load prediction. In light of this, we propose an adaptive hybrid prediction model based on Discrete Wavelet decomposition(DWT). It is well known that DWT can show local features such as mutation and fluctuation in the sequence, and has excellent multi-scale analysis ability. Adopt energy entropy evaluate the aggregation of wavelet coefficients in order to obtain the subsequence with the highest degree of preserving the main frequency of the original signal. Random Forest (RF) and Long Short-Term Memory (LSTM) were used to track low-frequency profile and high-frequency detail fluctuations respectively. Further, parameters of the heterogeneou models are optimized using the Particle Swarm Optimization(PSO) to improve the adaptability of hybrid models to different frequency components. Compared with other excellent metallurgical load forecasting techniques, the effectiveness and superiority of the proposed model are verified by experiments on the actual industrial data set of an electrical smelting furnaces for magnesia.
Study on Abnormal Heat of Polymeric Insulators on Coastal Power Transmission LineHu, Junhua; Yin, Jungang; Xu, Hua; Zhou, Yangyang; Wu, Huiling; Wang, Senlin; Pan, Shuyan
doi: 10.1088/1742-6596/2522/1/012004pmid: N/A
Heating at the high-voltage end has become one of the most frequent defects on polymeric insulators in service, particularly in humid and hot regions. Recent case surveys show that some highly overheated sample insulators survive almost all the standardized lab tests, which could not be fully explained by common related heating factors, including the fracture defects, superficial wet pollution, water absorption and moisture of polymeric housing, uneven sunlight, etc. In this paper, we try to find an alternative perspective to explain the temperature rise in the order of 10°C by studying the relationship between the heat at the high-voltage end and the degree of corrosion on a cast aluminum grading ring.