With the tremendous revival of artificial intelligence, predictive maintenance (PdM) based on data-driven methods has become most effective solution to address smart manufacturing and industrial big data, especially for performing health perception (e.g., fault diagnosis remaining life assessment). Moreover, because existing PdM research is still in primary experimental stage, works are conducted utilizing several open-datasets, combination with specific applications such as rotating...

Industrial wireless sensor networks (IWSNs) have the potential to contribute significantly in areas such as cable replacement, mobility, flexibility, and cost reduction. Nevertheless, industrial environment that IWSNs operate is very challenging because of dust, heat, water, electromagnetic interference, interference from other devices, which make it difficult for current guarantee reliable real-time communication. In this paper, we present a novel method based on segmented slot assignment,...

With the great evolution of human society from information era to smart automation era, intelligent production and maintenance have become core orientation Industry 4.0. Data-driven predictive (IPdM) offers potential solutions satisfy urgent demand predicting impending failures mitigating unplanned downtime industrial machinery, is envisioned as a critical enabler for future Internet Things. In this article, we provide comprehensive survey deep learning (DL)-based IPdM researchers...

Network slicing technology and the Open Radio Access (O-RAN) architecture provide solutions for resource isolation dynamic optimization in allocation models heterogeneous services. However, existing employ uniformly sized radio Resource Blocks (RBs) fixed-length time slots, which incur low utilization a high packet loss rate, especially when arrival rate is high. In this paper, we propose multi-cell, multi-dimensional resource, multi-timescale O-RAN framework, named M3O-RANS, meets Quality...

In the process of food drying, heat exchangers commonly face issues low transfer efficiency and poor energy performance. This study focuses on a novel gas-phase rotary shell-and-tube exchanger, analyzing effects operational parameters (shell-side rotational speed hot air velocity) its shell-side performance pressure drop losses through series univariate experiments. Furthermore, an improved NSGA-II algorithm is proposed, incorporating adaptive mechanisms fuzzy penalty mechanism. method,...

High order spectrum is a powerful tool for processing the nonlinear and non-Gaussian signals of rotating machinery. As one typical representative high spectrum, bispectrum analysis method has been used widely due to its advantages low effective algorithm. However, traditional based on single channel information often results in inconsistent fault feature extraction while analyzing phase coupling complex vibration collected from two different measurement directions at same point, which will...

Software-defined satellite networking (SDSN) has recently gained unprecedented attention due to its great controllability for traffic delivery. However, it is still facing several fundamental challenges in routing, which mainly involves effective network maintenance, heterogeneous convergence, and customized flow steering. Hence, this article, we propose an elastic routing framework SDSN, aiming first build a robust control path signalling exchanges, then offer transparent channel IP...

As the signal-to-noise ratio (SNR) of surface microseismic monitoring data is generally low and large, traditional detection picking algorithms cannot satisfy real-time high accuracy to processing. Therefore, a Fast Akaike information criterion (Fast-AIC) algorithm proposed for event automatic first arrival time picking. First, an method events based on multitrace energy stacking proposed, avoid missed false in conventional methods. Second, Fast-AIC developed by mathematical derivation from...

Localization has emerged as an attractive solution to enable new business models that rely on location-based services in wireless networks for communication, sensing, and control. In particular, time difference-of-arrival (TDOA) is one of the widely used localization models. However, conventional TDOA requires precise synchronization between a target node anchor nodes measuring difference, which leads large number packets communication. To reduce packet transmission, we propose model...

The demand for accurate localization in complex environments continues to increase despite the difficulty extracting positional information from measurements are not yet complete. Like any other process, also has security requirements. use of ultra-wideband (UWB) indoor systems have recently grown quickly industries, with a reliable, fast, and high accuracy performances. In particular, time difference-of-arrival (TDOA) is one widely used models. However, as TDOA measurement errors increase,...

Abstract Deep neural network (DNN) inference task offloading is an essential problem of edge intelligence, which faces the challenges limited computing resources shortage devices and dynamics networks. In this article, DNN in queue‐based multi‐device multi‐server collaborative investigated. To support efficient inference, we formulate a multi‐objective optimization that minimizes average delay maximizes accuracy. Due to time‐varying queue load states random arrival, it challenging solve...

Recently, Cyclic Queuing and Forwarding (CQF) based shaper has been proposed by IEEE 802.1Qch to satisfy the demands for deterministic industrial services. However, CQF only defines queuing model flow features, investigations of scheduling Time-Triggered flows are still at early stage. How make more efficient intelligent is an open issue be solved. In this paper, we propose a Deep Reinforcement Learning (DRL) assisted two-stage algorithm, namely DeepCQF, solve temporal problem on model,...

As a key component of rotating machinery, the health perception bearings is essential to ensure safe and reliable operation industrial equipment. In recent years, research on equipment based data-driven methods has received extensive attention. Overall, most studies focus several public datasets verify effectiveness their algorithms. However, scale these cannot completely satisfy representation learning deep models. Therefore, this paper proposes novel method, called CarNet, obtain more...

Increasing availability of multi-resolution synthetic aperture radar (SAR) data enabled us to demonstrate the benefits existing imaging solutions offered by satellites for purposes archaeological prospection. In this study, mid-resolution L-band phased array type (PALSAR) high-to-very-high-resolution X-band TerraSAR-X/TanDEM-X TerraSAR-X add-on digital elevation measurements (TanDEM-X) (Stripmap/Staring Spotlight) images were synergistically applied prospection at Han Hangu Pass, in Xin'an,...

The fatty acid desaturase (FADS9‐III) gene from M. alpina ATCC 32222 was cloned and overexpressed in a homologous system. Lipid analysis via GC‐MS revealed that the Δ9‐III had moderate activity with palmitic (C16:0) stearic (C18:0) as substrates. Palmitoleic (C16:1) oleic (C18:1) contents increased 4.0‐ 1.94‐fold FADS9‐III overexpression clones compared to wild‐type strain. Low temperature (12°C) significantly improved conversion ratio of C16:0 C16:1 (9.12‐fold increase). In addition,...