This letter presents a hybrid transmitting coil (HBRTx) for angular-freedom wireless power transfer. The rotating magnetic field is analyzed and obtained in different charging planes by three coil configurations to adapt the Rx rotations. Furthermore, the CLC (capacitor-inductor-capacitor) impedance matching networks (IMN) are implemented on the Tx and Rx side to match the coil impedance to the source and load impedance and improve the power transfer efficiency (PTE). Experimental results show that our proposed WPT system can supply power effectively to a distance of up to 51 mm with a maximum PTE of 80.35%. Moreover, the PTE maintains above 70% regardless of the Rx’s rotations around Ox, Oy and Oz axes, which shows the system’s flexibility and mobility.

The purpose of single-channel speaker extraction is to distill the pure speech of the target speaker from a multi-talker speech mixture. Typically, an auxiliary reference network is employed to obtain the voiceprint features of the speaker’s speech as the input stimulus for the main speech extraction network, improving the robustness of the speech extraction. However, current research has paid little attention to the time-series characteristics of speech signals, resulting in a mismatch between the receptive field and signal features.  Additionally, commonly used speaker extraction models do not consider the spectral distribution characteristics of speaker speech. In this paper, we address these issues by extending the ConvNeXt model in image processing to TD-ConvNeXt model and combining it with TCN block to form the main body of our speech extraction network. We also revamp the ConvNeXt model to a new Spk block for the auxiliary network, which learns speaker identity information features from reference speech as embedding vectors. By utilizing this approach, we have achieved better results in speaker identification without compromising the quality of speech extraction. We also use multi-task learning to jointly train the extraction network and auxiliary reference network. Extensive experiments have been conducted to verify the significant single-channel target speech extraction performance improvements of our proposed model.

The use of a NPC-type multilevel converter as multiport inverter (MPI) has recently been considered a promising solution for the compact integration of multisource systems in various applications. This paper presents a broad analysis of a quasisingle stage propulsion system (QSS) that employs an NPC-type MPI and a downsized boost dc-dc converter for an urban tram vehicle with overhead line connection and onboard batteries. The control and dynamic performance are detailed and validated experimentally. Furthermore, the power losses, VA rating, weight, volume, and reliability of the QSS system are evaluated and benchmarked against a conventional architecture for a case-study catenary/battery tram model. The analysis reveals that the QSS system can achieve significant savings in volume, weight, and round-trip energy efficiency, higher reliability due to intrinsic redundancy, but with increased VA rating, part count, and complexity.This paper has been accepted for publication in the IEEE Transactions for Transportation Electrification. Please find the final peer-reviewed early-access version at: https://ieeexplore.ieee.org/document/10319775

Dengue fever (DF) is a mosquito-borne communicable disease, mostly manifest during South-East Asian monsoon. Approximately 2-5% people die due to severe hemorrhagic Dengue fever and, if left untreated, the number rises up to 20%. Although it is self-limiting, some cases, referred here as Doubtful Cases’ (DfC), may abruptly turn ‘fatal’ often causing mortality in the affected population. Diagnostically apprehending such ‘fatal’ cases is a key medical challenge as clinically it is nearly impossible to predict those. This article outlines a predictive model of dengue infection through a Deep Learning (DL) approach leading to deliver a predictive tool that can accurately preempt clinical ‘morbidity’ patterns. This tool is developed and validated on a set of prototype synthetic Dengue data and ratified against independent opinions from (10) clinicians to predict and grade the infection severity of such DfC case to ca 99% accuracy. The DL approach can be generically extended to other epidemic forms with DfCs that further extends the remit of this study.

In order to increase the precision of emotion identification in text, this research suggests a context-aware emotion recognition system employing transformer models, especially BERT. The model is able to comprehend complex emotions and context-dependent expressions since it was trained on a broad, emotion-labeled dataset. On a benchmark dataset, its efficacy is assessed in comparison to conventional techniques and standard transformer models. The system is proficient at gathering contextual information, and the findings demonstrate a considerable improvement in emotion recognition accuracy. This study improves textual emotion identification, opening the door to applications like chatbots that can recognize emotions and systems for tracking mental health. It also identifies potential areas for further study in developing transformer models for context-sensitive NLP applications.

The utilization of deep learning models has been widely recognized for its significant contribution to the enhancement of smart grid operations, particularly in the domain of power quality disturbance (PQD) classification. Nevertheless, the emergence of vulnerabilities like targeted universal adversarial attacks can significantly undermine the reliability and security of deep learning models. These attacks can exploit the model’s weaknesses, causing it to misclassify PQDs with potentially catastrophic consequences. In our previous research, we for the first time examined the vulnerability of deep learning models to targeted universal adversarial attacks on time series data in smart grids by introducing a novel algorithm that effectively attacks by maintaining a trade-off between fooling rate and imperceptibility. While this attack method demonstrated notable efficacy, it also emphasized the pressing need for robust defensive mechanisms to safeguard these critical systems. This paper provides a thorough examination and evaluation of different defense strategies, specifically adversarial training, defensive distillation, and feature squeezing, in order to identify the most effective method for mitigating targeted universal adversarial (TUA) attacks on time series data for three different types of imperceptibility (high, medium and low). Based on our analysis, adversarial training demonstrates a significant reduction in the success rate of attacks. Specifically, the technique reduced fooling rates by an average of 23.73% for high imperceptibility, 31.04% for medium imperceptibility, and a substantial 42.96% for low imperceptibility. These findings highlight the crucial role of adversarial training in enhancing the integrity of deep learning applications.

In this work we investigate a scenario in which protected and unprotected services coexist in an elastic optical network under dynamic traffic. In the investigated scenario, unprotected services can reuse the reserved idle bandwidth to provide protection to the protected services. Under this scenario, we propose a new heuristic algorithm that enables such reuse as well as we define and introduce the new assignment problem in elastic optical networks named as Transmission Spectrum Assignment (T-SA) problem. We evaluated the proposed algorithm through simulations and we observed that it was able to reduce the network blocking probability by at least 6%, 7% and 10% depending on the investigated topology.

This paper proposes a pre-processing method for the assimilation of microwave satellite observations over winter sea ice into a numerical weather prediction system.

Air pollution is a critical global concern, demanding precise air quality forecasting to mitigate its severe consequences. Our study introduces a novel Hyper Heuristic Multi-Chain Model (H2MaM) to project future air quality, considering various meteorological factors (MFs) and pollution-related variables like atmospheric pressure, temperature, humidity, and wind patterns. Leveraging 12 units of LSTMs, H2MaM accurately predicts forthcoming air pollutants (APs) concentrations such as PM2.5, CO, and NO2. Additionally, it accounts for spatiotemporal correlations between these APs and MFs, which significantly influence the air quality prediction for the next immediate time interval. H2MaM utilizes a multi-chain mechanism, employing 1-hour prediction models to forecast air quality hourly, enabling approximations for the next 12 hours. Furthermore, it demonstrates the ability to enhance the performance of any predictor. Experimental results substantiate H2MaM’s superiority over various models, including the Support Vector Regressor (SVR), Multi-Layer Perceptron (MLP), Recurrent Air Quality Predictor (RAQP), and Valchogianni models. H2MaM achieves impressive up to 75% better accuracy and consistency compared to SVR, 60% better than MLP, 38% better than RAQP, and 70% better than Valchogianni models.

Heating is a common problem in electric machines and cooling of small size machines is very challenging. Nowadays, the application of these machines in the actuation of weight-critical mechatronics systems is increasing. In this paper, a cooling mechanism for such electric machines with surface permanent magnet (SPM) is proposed. The mechanism is based on the fan-cooled air conditioning system except that no external impeller is needed. It has an air-inlet structure which comprises a set of six holes shrouded within a hollow cylindrical ventilation duct at the rear end of the rotor. The cooling mechanism is evaluated in a high-power-density SPM motor. The airflow rate is analyzed by developing a theoretical model using flow resistance network technique. Further verification of the airflow rate was performed with numerical computational fluid dynamics (CFD) simulation. With a discrepancy margin within 5.0%, both flow resistance network and CFD models show that the system generates a total flow rate of 118.0 L/min at maximum rated speed of the motor. Also, steady state thermal analysis of the motor using lumped parameter thermal network (LPTN) is verified in comparison with CFD simulation, which shows the cooling system maintains the motor temperature within operating range during high-speed operations.

Learning a fair and privacy-preserving graph neural network from private and limited sensitive attributes

There are cases, that models derived from first principles cannot accurately capture the system dynamics, which leads to uncertainty in model-based control. However, these nominal models are commonly used due to existence of mature methods to guaranty stability and feasibility of optimal control problems. Neural networks can be used as an alternative to approximate the dynamic model of a system directly from data. With uncertainty estimation in neural networks seeing increase in usage with regards to safety-critical applications, neural networks offer a natural way to quantify model uncertainty. Taking intuition from the above, in this work, we present an algorithm that estimates the future uncertainty of an MPC law, that combines a nominal model, derived from first principles, and a model learned with neural networks, that is employed to estimate the uncertainty of the nominal MPC horizon. The approach is based on the Monte Carlo dropout technique, aims in safety-critical applications and is general, as it can be tailored to any existing MPC strategy by gathering data from the plant. As an example, the method is applied to a simulated over- actuated marine platform that is controlled by an MPC law under realistic environmental disturbances for the task of navigation. The horizon uncertainty of the MPC, that employs Control Barrier Functions for obstacle avoidance, is used to shape the barrier of the safe set.

Even though achieving global connectivity represents one of the main goals of 5G and beyond wireless networks, exurban areas are still suffering frequent outages due to the lack of proper telecom infrastructures, which are often available only in urban areas. Indeed, cellular network design is usually capacity-driven, and thus base stations’ densities mostly follow population and especially revenue densities. Contextually, we focus on one of the most promising solutions to provide sufficient and reliable coverage in far-flung areas: aerial base stations, which consist of unmanned aerial vehicles carrying cellular base station equipment. In this paper, we extensively discuss the problem of bridging the so-called urban-rural digital divide (i.e., the connectivity gap between urban and rural areas) from various perspectives. First, we showcase various alternative solutions, and compare conventional terrestrial networks with aerial networks from a techno-economic point of view. Then, we highlight the topological aspects of rural environments and explain how they can affect the actual design of cellular networks. In addition, we investigate both the coverage probability and the reliability of the communication links via simulations, proving that the integration of aerial base stations can be quite promising in a 6G perspective. Finally, we propose two original extensions of our case study as open problems.

Regarding the passive WiFi sensing based crowd analysis, this paper first theoretically investigates its limitations, and then proposes a deep learning based scheme targeted for returning fine-grained crowd states in large surveillance areas. To this end, three key challenges are coped with: to relieve the influences of the randomness and sparsity induced by passive WiFi sensing, an attention-based deep convolutional autoencoder model is designed to recover accurate crowd density maps in a way similar to image reconstruction; to combat the anonymity caused by MAC randomization, following the identification of local high-density crowds (LHDCs) with the density clustering algorithm, i.e. DM-DBSCAN, a bidirectional convolutional LSTM based model is employed to infer LHDC speeds; to overcome the absence of passive WiFi sensing datasets for model training, three semi-synthetic datasets are produced by emulating passive WiFi sensing with practical pedestrian tracking datasets. Extensive experiments confirm that, the proposed scheme significantly outperforms existing WiFi-based methods in terms of crowd density estimation and provides superior crowd speed estimation. More importantly, the scheme can also produce consistent crowd states on a real-world dataset, revealing that it has the ability to support accurate, visualized and real-time crowd monitoring in large surveillance areas.

Medical Visual Question Answering (VQA) is a task that aims to provide answers to questions about medical images, which utilizes both visual and textual information in the reason?ing process. The absence of large-scale annotated medical VQA datasets presents a formidable obstacle to training a medical VQA model from scratch in an end-to-end manner. Existing works have been using image captioning dataset in the pre-training stage and fine-tuning to downstream VQA tasks. Following the same paradigm, we use a collection of public medical image captioning datasets to pre-train multimodality models in a self-supervised setup, and fine-tune to downstream medical VQA tasks. In the work, we propose a method that featured with Cross-Modal pre?training with Multiple Objectives (CMMO), which includes masked image modelling, masked language modelling, image-text match?ing, and image-text contrastive learning. The proposed method is designed to associate the visual features of medical images with corresponding medical concepts in captions, for learning aligned vision and language feature representations, and multi-modal interactions. The experimental results reveal that our proposed CMMO method outperforms state-of-the-art methods on three pub?lic medical VQA datasets, showing absolute improvements of 2.6%, 0.9%, and 4.0% on the VQA-RAD, PathVQA, and SLAKE dataset, respectively. We also conduct comprehensive ablation studies to validate our method, and visualize the attention maps which show a strong interpretability. The code and pre-trained weights will be released at https://github.com/pengfeiliHEU/CMMO.

In this manuscript, one of the higher educational institutions located at Lat. 13.62ON; Long. 79.29OE; Alt. 226 M is considered and the roof architectures and  orientations of seven sheds being utilized as six workshops (WS) and one central stores (CS) are studied. NASA-SSE satellite synthetic data is used for resource assessment and PV System 7.3 software is used to estimate the annual GHI at the concerned site. Physical inspection and measurement survey has been carried out to exactly obtain the shed orientation, azimuth angle and roof dimensions including slope angle. Based on these data, gross capacity of solar PV collectors that can be installed on the roof of sheds and corresponding energy yield is estimated. Different cases have been exercised to create real-time feasibility for erecting solar PV system on the roofs of the concerned sheds. The obtained results are so intensive that for the WS and CS sheds, the daily normalized production in kWh/kWp/Day and energy yield is more on the south facing roofs (CS & WS) followed by west (WS), east (WS) and north facing roofs (WS & CS). In comparison to the vertical installation, the specific production in kWh/kWp/Annum is more with the horizontal installation of PV modules on either side of the exposed roofs for WS and less for CS. The total maximum energy that can be generated on the roofs of total seven sheds is about 9,69,566 kWh/Annum which may reduce about 824.12 Tons of CO2 emissions per annum. It is an appreciable figure and could pave a path for establishing green electricity. This research has been carried out to inspire the enthusiastic higher educational institutions to think for installing considerable capacity of solar PV system on the roofs of their workshop sheds and contribute for establishing green electricity.  In this manuscript, one of the higher educational institutions located at Lat. 13.62ON; Long. 79.29OE; Alt. 226 M is considered and the roof architectures and  orientations of seven sheds being utilized as six workshops (WS) and one central stores (CS) are studied. NASA-SSE satellite synthetic data is used for resource assessment and PV System 7.3 software is used to estimate the annual GHI at the concerned site. Physical inspection and measurement survey has been carried out to exactly obtain the shed orientation, azimuth angle and roof dimensions including slope angle. Based on these data, gross capacity of solar PV collectors that can be installed on the roof of sheds and corresponding energy yield is estimated. Different cases have been exercised to create real-time feasibility for erecting solar PV system on the roofs of the concerned sheds. The obtained results are so intensive that for the WS and CS sheds, the daily normalized production in kWh/kWp/Day and energy yield is more on the south facing roofs (CS & WS) followed by west (WS), east (WS) and north facing roofs (WS & CS). In comparison to the vertical installation, the specific production in kWh/kWp/Annum is more with the horizontal installation of PV modules on either side of the exposed roofs for WS and less for CS. The total maximum energy that can be generated on the roofs of total seven sheds is about 9,69,566 kWh/Annum which may reduce about 824.12 Tons of CO2 emissions per annum. It is an appreciable figure and could pave a path for establishing green electricity. This research has been carried out to inspire the enthusiastic higher educational institutions to think for installing considerable capacity of solar PV system on the roofs of their workshop sheds and contribute for establishing green electricity.

Whilst Deep Neural Networks have been developing swiftly, most of the research has been focused on RGB image. This type of image has been traditionally optimised for human vision. However, RGB data is a highly re-elaborated and interpolated version of the collected raw data (i.e. the sensor collects one value per pixel), but an RGB image for human viewing contains 3 values, for red, green, and blue. This processing through the ISP (Image Signal Processing) requires computational resource, time, power and obviously increases by a factor of three the amount of output data. This work investigates Deep Neural Network based detection using (for training and evaluation) Bayer data, generated in different ways, from a benchmarking automotive dataset (i.e. KITTI dataset). A Deep Neural Network (DNN) is deployed in unmodified form, and also modified to accept only single field images, such as Bayer frames. Eleven different re-trained version of the DNN are produced, and cross-evaluated across different data formats. The results demonstrate that the selected DNN has the same accuracy when evaluating RGB or Bayer data, without significant degradation in the perception (the variation of the Average Precision is <1%). Moreover, the colour filter array position and the colour correction matrix do not seem to contribute significantly to the DNN performance. This work demonstrates that Bayer data can be used for object detection in automotive without significant performance loss, and that the processing currently used in ISP can be avoided, allowing for more efficient sensing-perception systems.

Laser cooling and trapping of atoms for neutral atom computing