With the emergence and development of digital technologies such as Internet, cloud computing, artificial intelligence, big data, economy has become an important tool for leading industrial transformation economic growth, is gradually becoming a key entry point enterprises to innovate enhance competitiveness. In "14th Five-Year Plan," China sets "accelerating building Digital China" long-term goal, while also prioritizing one its tasks. To respond national call times, various have accelerated...

Unsupervised domain adaptation (UDA) aims to transfer knowledge from the labeled source fully unlabeled target domain, thus improving classification performance of domain. Recently, self-training methods have shown their effectiveness on UDA. It iteratively trains data using generated pseudo-labels. However, feature space for generating pseudo-labels contains a large amount information, which traps model in making it challenging generator learn discriminative features In this article, we...

Online condition monitoring of power transmission belts is essential to keep industrial belt-driven equipment functioning smoothly and reliably. This paper presents a smart electrostatic sensor that monitors belt motion through detection static charge on the belt. A theoretical model established using method moments for calculation induced strip-shaped electrodes placed in vicinity moving both axially transversely. The unit converts into proportional voltage signals amplifiers measures speed...

Abstract Software project defect prediction can help developers allocate debugging resources. Existing software models are usually based on machine learning methods, especially deep learning. Deep learning‐based methods tend to build end‐to‐end that directly use source code‐based abstract syntax trees (ASTs) as input. They do not pay enough attention the front‐end data representation. In this paper, we propose a new framework represent code called multiperspective tree embedding...

In recent decades, many software reliability growth models (SRGMs) have been proposed for the engineers and testers in measuring precisely. Most of them is established based on non-homogeneous Poisson process (NHPP), it proved that prediction accuracy such could be improved by adding describing characterization testing effort. However, some research work indicates fault detection rate (FDR) another key factor affects final quality. early NHPP-based deal with FDR as constant or piecewise...

In order to efficiently use energy storage resources while meeting the power grid primary frequency modulation requirements, an adaptive droop coefficient and SOC balance-based control strategy for is proposed. Taking of battery as quantity, depth output adaptively adjusted prevent saturation or exhaustion SOC. The balanced introduced realize rational utilization fast balance in process under different charge states. Then, four evaluation indexes are proposed evaluate effect maintenance. a...

Astroviruses are a family of genetically diverse viruses associated with disease in humans and birds significant health effects economic burdens. Astrovirus taxonomic classification includes two genera, Avastrovirus Mamastrovirus. However, next-generation sequencing, broader interspecies transmission has been observed necessitating reexamination the current host-based approach. In this study, novel method is presented for emergent as yet unclassified astroviruses, based on whole genome...

With the continuous improvement of wind power penetration in system, volatility and unpredictability generation have increased burden system frequency regulation. its flexible control mode fast adjustment speed, energy storage has obvious advantages participating grid Therefore, this paper studies strategy combined with primary regulation proposes a method based on operation state turbines. This analyzes reserve rotor kinetic turbines operating different speed ranges, introduces fan...

Although software bug localization in maintenance and evolution is cumbersome time-consuming, it also very important, especially for large-scale projects. To lighten the workload of developers, researchers have developed various information retrieval (IR)-based models automated support. In this paper, we propose a new method that reduces time required localization. First, surface lexical similarity between report source code file calculated based on vector space model. Second, to address gap...

Software bug localization is very important in software engineering, but it also complicated and time consuming. To improve the efficiency of developers, researchers have developed various traditional machine learning methods. In this paper, we propose a novel method that improves performance. First, surface lexical correlation matching between reports source code files used to obtain features by deep neural network. Second, solve gap files, semantic them based on word embedding sentence...

Face on the problems of existed method substation ground grid corrosion searching, like blindness, heavy workload, low working efficiency. Through analysis fault diagnosis theory, using theory electric network and ideas matrix, we can construct equations. Introducing lowest energy principle advanced optimization algorithm, grip point have been effectively solved. Plus, worked out relevant software diagnosis. At same time, done some simulations, through simulation examples verified this is...

The measurement of the operating deflection shape (ODS) power transmission belts is great importance for fault diagnosis and prognosis industrial belt drive systems. This paper presents a novel method based on an electrostatic sensor array to measure ODS moving both axially transversely. integrates charge amplifier that converts induced strip-shaped electrode into voltage signal. Finite-element simulations are performed study sensing characteristics results reveal can respond vibration...

Multi grid-connected inverters islanding detection methods include two categories: passive and active detection. Firstly, an method based on negative sequence current disturbance is analyzed. Aiming at the shortcomings of this method, a novel proposed in paper. The uses wavelet transform to extract signal features, then according extracted features adopts neural network judge whether or not occurs. simulation results show that multi has good performance, does affect stability power quality system.

Predicting short-term urban traffic flow is a non- trivial task, for an intelligent transportation system could greatly facilitate infrastructure construction and enhances the efficiency of control. Unfortunately, influenced by numerous factors, which increases complexity prediction. In this paper, Multi-Grouped Least Squares Support Vector Machine (MLS-SVM) proposed MLS-SVM, spatiotemporal factors (e.g., time, geography, environment) are divided into different groups. Correlations between...

Reliability engineering implemented early in the development process has a significant impact on improving software quality. It can assist design of architecture and guide later testing, which is beyond scope traditional reliability analysis methods. Structural models work for this, but most them remain tested only simulation case studies due to lack actual data. Here we use metrics modeling are collected from source codes post versions. Through proposed strategy, redundant metric elements...

In recent years, machine translation has made outstanding achievements, but there are also the problems such as fuzzy rules, insufficient data and improper order in practice, which makes it still unsatisfactory.Machine learning can abstract features, establish complex mapping relationship between input output signals, effectively improve existing system.The concepts common models of deep given this paper.Based on neural model, paper analyses structure system.According to requirement system,...

The Basic Local Alignment Search Tool, BLAST, is an indispensable tool for genomic research. BLAST established itself as the canonical sequence similarity search in large part thanks to its meaningful statistical analysis. Specifically, reports E-value of each reported alignment, which defined expected number optimal local alignments that will score at least high observed alignment score, assuming query and database sequences are randomly generated.