1,721,059 research outputs found
Distance Measurement Characterization for Ultra Wide Band Indoor Localization Systems
No full textLocalization system are widespread both outdoor and indoor and their objectives are various and of obvious importance: navigation, object and person localization and more. Indoor environments are particularly complex for position detection and localization; among several systems and the technologies used, the most commonly used is the Ultra Wide Band for its performance in terms of range and resolution. Any localization method strongly depends on a distance measurement and on its accuracy. In this paper after a short introduction on the state of art of indoor localization and on Ultra Wide Band technology, localization estimation technique is detailed in order to support the following sensor measurement characterization. The paper also report several measurements made on an indoor localization system based on UWB technology and extracts a noise model for the distance measurement. An exponential model for noise is extracted from measurements
An LSTM based soft sensor for rear motorcycle suspension
No full textThe increasing development of neural networks for classification and prediction of temporal sequences has opened the way for a new development of mathematical models for soft sensor design. In particular, Long Short-Term Memory (LSTM) networks have greatly improved execution time and reduced error in both single-step and multi-step prediction. In this context, it is therefore possible to improve on the current concept of Instrument Fault Detection and Isolation (IFDI), reducing costs and footprint by not using physical redundancies of sensitive elements but by employing virtual sensors themselves. Therefore, the work aims to develop a soft sensor for rear suspension stroke using an LSTM network. This new approach was trained on over 50000 samples acquired in a real-world environment, and the results were compared with ground truth on a total of over 100000 samples. The results of the analysis showed excellent potential of the method and wide room for improvement in future developments
A distributed measurement system for the estimation of air quality
No full textCurrently, one of the most serious environmental problems on the planet is air pollution by fine dust, including Particulate Matter with an aerodynamic diameter of about 10 μm (PM10) and 2.5 μm (PM2.5). The high concentration of these components, particularly in urban and industrial areas, leads to a worrying incidence of respiratory diseases. In fact, these dusts, which can be compared to a slow and silent killer, are so small that they can be inhaled and gradually accumulated in the respiratory system, which, according to the World Health Organization (WHO), has been the cause of death of 12 million people in the past five years [1]
Wireless Sensor Network for Low-cost Air Quality Measurement
No full textThe original proposal of Advanced Metering Infrastructure (AMI) based on short-range communication (wM-Bus) is suggested for the continuous monitoring of Particulate Matter within Smart Cities. A prototype of water meter equipped with a low cost off-the shelf PM sensor has been developed as remote node to be adopted in the radio Local Area Network. The simulation of a Smart Metering scenario based both on the result of the metrological characterization of the PM sensor (against the quality requirements of the PM measurement according to European regulations) and on the typical communication performance of the wM-Bus confirm the feasibility of the proposed AMI for the continuous analysis of the air pollution exposure within urban areas
Soft Sensors for Instrument Fault Accommodation in Semiactive Motorcycle Suspension Systems
No full textThis article describes the development and experimental verification of an instrument fault accommodation (IFA) scheme for front and rear suspension stroke sensors in motorcycles equipped with electronically controlled semiactive suspension systems. In particular, the IFA scheme is based on the use of nonlinear autoregressive with exogenous inputs (NARX) neural networks (NNs) employed as soft sensors for feeding the suspension control strategy back with measurement even in the presence of faults occurred on the sensors. Different NN architectures have been trained and tuned by considering real data acquired during several measurement campaigns. The performance has been compared with that of the well-known half-car model (HCM). Very satisfying results allow the soft sensor to be really integrated into fault-tolerant control systems. In experimental road tests, an implementation of the proposed IFA scheme on a low-cost microcontroller for automotive applications showed to be in real time. In this article, these experimental results are shown to prove the good performance of the IFA scheme in different motorcycle operating conditions
A Deep CNN Transformer Hybrid Model for Skin Lesion Classification of Dermoscopic Images Using Focal Loss
No full textSkin cancers are the most cancers diagnosed worldwide, with an estimated > 1.5 million new cases in 2020. Use of computer-aided diagnosis (CAD) systems for early detection and classification of skin lesions helps reduce skin cancer mortality rates. Inspired by the success of the transformer network in natural language processing (NLP) and the deep convolutional neural network (DCNN) in computer vision, we propose an end-to-end CNN transformer hybrid model with a focal loss (FL) function to classify skin lesion images. First, the CNN extracts low-level, local feature maps from the dermoscopic images. In the second stage, the vision transformer (ViT) globally models these features, then extracts abstract and high-level semantic information, and finally sends this to the multi-layer perceptron (MLP) head for classification. Based on an evaluation of three different loss functions, the FL-based algorithm is aimed to improve the extreme class imbalance that exists in the International Skin Imaging Collaboration (ISIC) 2018 dataset. The experimental analysis demonstrates that impressive results of skin lesion classification are achieved by employing the hybrid model and FL strategy, which shows significantly high performance and outperforms the existing work
Smart wearable devices for human exposure vibration measurements on two-wheel vehicles
No full textThe comfort experienced while driving a motorcycle is becoming a subject of great importance; indeed, the driver is exposed to vibrations that are typically caused by irregular profiles or wear of the road surface as well as by the aerodynamic influence and high-frequency rotation of the motorcycle engine. This paper discloses an original solution that allows the driver to monitor their exposure to vibration during a ride using a low-cost wearable device (smartwatch). A suitable measurement system has been designed and tested using a real motorcycle. The system captures acceleration signals in real time through Bluetooth communication and interfaces with a wearable device with a microcontroller unit that calculates vibrations transmitted through the driver's hands. Different indexes proposed in the literature are adopted for the comfort analysis in both time and frequency domains. The hand transmitted vibrations are also experimentally compared with those measured through a fixed accelerometer according to the prescription included in the standard ISO 5349 to show the feasibility of the proposed approach in typical application conditions
- …
