Wi-Fi deployed inside a building can be used for positioning indoor users. A commonly technology is weighted K-nearest neighbor (WKNN) fingerprint which positions user based on K nearest reference points measured beforehand. The challenge lies in how to configure the value of obtain best accuracy. In this paper, we propose self-adaptive WKNN (SAWKNN) algorithm with dynamic K. By adjusting signal strength, SAWKNN better accuracy than traditional WKNN. particular, significant percentage makes...

Visual inspection techniques for rail surface defects have become prevalent approaches to obtain information on damage. However, uneven illumination leads illegibility of local information, and the change wheel-rail area results in changeful background surface, both which pose challenges visual inspection. This paper proposes a novel algorithm that detects via partitioned edge features (PEF). PEF eliminates effect by effectively extracting building homogeneous surface. In process feature...

Rail surface defect inspection based on machine vision faces challenges against the complex background with interference and severe data imbalance. To meet these challenges, in this article, we regard detection as a key-point estimation problem present proposed attention neural network for rail via consistency of Intersection-over-Union(IoU)-guided center-point (CCEANN). The CCEANN contains two crucial components. components are stacked Hourglass backbone cross-stage fusion multiscale...

Driver distraction behavior is prone to induce traffic accidents. Therefore, it necessary detect them caution drivers in time for safety. In driver recognition, the variety of behaviors and diversity driving environment can have a certain effect on detection accuracy, information loss severe most existing methods. These make challenging improve real-time accuracy behavior. this paper, we propose an improved YOLOv7 based channel expansion attention mechanism detection, named CEAM-YOLOv7. The...

Considering the volatility and randomness of wind speed, this research suggests an improved hunter-prey optimization (IHPO) algorithm-based extreme learning machine (ELM) short-term power prediction model to increase accuracy. The original history data from farm are used in achieve feature extraction dimensionality reduction, using partial least squares’ variable importance projection (PLS-VIP) normalized mutual information (NMI) methods. Adaptive inertia weights added HPO algorithm’s search...

This paper presents a compact 0.92/2.45-GHz dual-band directional circularly polarized (CP) microstrip antenna for handheld radio-frequency identification (RFID) reader applications. The proposed comprises wideband dual-feed network and two stacked concentric patches assembled by orthogonally placed vertical probes. feeds quadrature signals across the RFID bands between 0.92 2.45 GHz. provide resonance frequencies in fundamental mode lower band higher band, respectively. Additionally,...

The dramatic development of Internet Things (IoT) is not only leading to the spectrum crunch, but also resulting in exorbitant energy consumption. It thus desirable liberate IoT from constraint scarcity and rein growing Wireless harvesting relay-assisted underlay cognitive networks (WEH-CRNs), which combine radio wireless techniques alleviate constraints by reusing spare for data transmissions ambient power supplies, are conceived as an efficient solution massive deployments. However, all...

Open-world object detection (OWOD), as a more general and challenging goal, requires the model trained from data on known objects to detect both unknown incrementally learn identify these objects. The existing works which employ standard framework fixed pseudo-labelling mechanism <tex xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">$(PLM)$</tex> have following problems: (i) inclusion of detecting substantially reduces model's ability ones. (ii)...

A dual-band 0.92/2.45 GHz circularly-polarized (CP) unidirectional antenna using the wideband dual-feed network, two orthogonally positioned asymmetric H-shape slots, and stacked concentric annular-ring patches is proposed for RF identification (RFID) applications. The measurement result shows that achieves impedance bandwidths of 15.4% 41.9%, 3-dB axial-ratio (AR) 4.3% 21.5%, peak gains 7.2 dBic 8.2 at 0.92 2.45 bands, respectively. Moreover, provides stable symmetrical radiation patterns...

Abstract. A companion paper developed a vehicle emission inventory with high temporal–spatial resolution (HTSVE) bottom-up methodology based on local factors, complemented the widely used factors of COPERT model and near-real-time (NRT) traffic data specific road segment for 2013 in urban Beijing (Jing et al., 2016), which is to investigate impact pollution air this study. Based sensitivity analysis method switching on/off pollutant emissions Chinese quality forecasting CUACE, modelling...

With the increasing demand of providing security for images/videos with private information, chaos-based cryptosystems have played an important role in image encryption because their excellent random properties and performance. However, existing systems defect due to small key space or other weakness. This paper introduces a new chaotic system using combination three conventional maps. The proposed shows behaviors. To demonstrate its application processing, scheme is also introduced....

A low-cost compact circularly polarized (CP) directional antenna is proposed for universal ultra-high frequency (UHF) RF identification (RFID) handheld reader applications. The consists of four sequentially rotated inverted-F antennas (IFAs) fed by a four-feed network with closed-form design formulas. prototype implemented FR-4 substrates and thick air layer the reduction cost. For return loss (RL) > 13 dB, 3-dB gain variation, axial ratio (AR) , achieves measured bandwidth 14.9% exhibits...

One major concern of existing wireless localization systems is the identification nonline-of-sight (NLOS) signal propagation, since NLOS can be considered dominant source error. Present methods usually assume that could make it not possible to perform in a consistent manner. However, validity foregoing assumption has been properly investigated. This paper presents theoretical analysis mobile user involving one or more beacons and shows given as being invalid when estimated location outside...

Dynamically measuring rail profile using the structured-light vision suffers from random vibrations on line laser projector, which would cause distorted profiles. This paper presents a simple and effective distortion rectifying method to address this issue. The is rectified by easily projecting it onto an auxiliary plane parallel cross section of rail. In order establish plane, three profiles formed radiating multiline structured light are used fit longitudinal axis. More importantly, only...

Rail section profile accurate measurement techniques are critical to guarantee the transport security and reduce maintenance cost. However, may make an affine transformation under vehicle vibration. It is a challenge identify calibrate distorted profiles in 2-D laser displacement system (LDS). By profound observation discussion on triple primitives including rail jaw, railhead inner line, waist curve, we found difference between normal ones. To measure quantitatively, proposed novel curve...

Heterogeneous networks (HetNets), which consist of traditional macro-cells overlaid with newly envisioned small cells (e.g., femtocells, picocells, microcells, and nanocells), are conceived as an appealing technology to satisfy the ever-increasing capacity requirements in future mobile networks. The cross-tier interference management is a challenging problem conventional HetNets due large-scale deployment random locations, lack complete coordination. However, cognitive HetNets, where...

Fastener needs to be monitored and inspected periodically ensure the rail's safety due its easily damaged accessory for railway infrastructure. Recently, Industrial Internet of Things (IIoT) artificial intelligence (AI)-based visual inspection techniques have been exploited realize online fastener's fault by using a detection IoT vehicle that is mounted with multitype sensors cameras according design our research team. However, instead traditional inspection, AI-based automatic fastener...

For the safety of railways, well-trained workers are required to check fastener constantly, which shows disadvantage large time cost, huge labor cost and might being dangerous workers. To address this achieve automatic detection, an inspection model based on deep convolutional neural network (DCNN) is adopted in paper. However, suffering from unbalanced training samples defective vs normal due fasteners far less than real railways. tackle problem, a novel sample generation method proposed...