Khalid Abualsaud

Accurate estimation of State-of-Charge (SoC) and core temperature is fundamental to optimizing the performance, safety, and longevity of Lithium-Ion Batteries (LiBs), particularly in Electric Vehicles (EVs). Traditional estimation methods fail to account for the complex, non-linear interactions between thermal and electrical dynamics and the challenges posed by data uncertainty. This paper introduces a comprehensive framework to estimate core temperature and SoC, considering diverse charging levels and uncertainties. For the data generation phase, first, features are extracted from a control-oriented electro-thermal coupling model, offering a computationally efficient alternative to resource-intensive experiments and avoiding a lack of data. Then, a correlation analysis between the ambient temperature and each feature (e.g., internal resistances, thermal capacity, and time) is performed, with linear regression applied to generate features showing strong linear relationships, and a Gaussian Multivariate Copula model is used to generate features with weak or non-linear dependencies. For the estimation phase, hybrid tree-based models were employed due to their robustness in handling complex and noisy datasets, and computational efficiency while integrating the complementary strengths of individual models. Among the combinations tested, the Extra Trees Regressor-Random Forest (ETR-RF) model delivered the highest estimation accuracy, while the Decision Tree-LightGBM (DT-LGBM) model exhibited the fastest training time. The ETR-RF model consistently reduced estimation errors, achieving RMSE values of 0.047°C and 1.25°C for core temperature and 0.5% and 0.56% for SoC estimation across white Gaussian noise levels, with standard deviations of 0.02 and 0.2, respectively. In contrast, the DT-LGBM model prioritized computational efficiency, requiring 1 second (average training time) for SoC estimation and 0.66 seconds for core temperature estimation, performed on a system equipped with an Intel Core i7-7500U CPU (2.70GHz base, 2.90GHz boost).

The State-of-Charge (SoC) of Lithium-Ion Batteries (LIBs) is a crucial parameter for Battery Management Systems (BMSs) used in Electric Vehicles (EVs). This paper presents a comprehensive study on the SoC estimation of LIBs using advanced model-based methods. The practical implications of this research are significant, as they provide a reliable and efficient approach to SoC estimation, enhancing the performance and lifespan of LIBs in real-world applications, particularly EVs. A third-order equivalent circuit model is employed for the LIB based on electrochemical impedance spectra test results, with model parameters identified using a particle swarm optimization algorithm. Two real-time model-based estimation algorithms, Extended Kalman Filter (EKF) and Unscented Kalman Filter (UKF), are compared for SoC estimation. A hybrid approach based on UKF and EKF is presented. The results demonstrate that the UKF outperforms the EKF in SoC estimation, with the root mean squared error (RMSE) and maximum error for SoC estimation being 1.06% and 1.15%, respectively. The hybrid EKF-UKF approach provides the best performance for SoC estimation, achieving the lowest root mean squared error (RMSE) of 0.2% and a maximum error of 0.5% for SoC estimation. This approach leverages the strengths of EKF and UKF, offering superior accuracy and robustness in real-time battery monitoring in EV applications.

Accurate state-of-charge (SoC) estimation is crucial for enhancing the performance, longevity, safety, and reliability of lithium-ion batteries (LiBs) in electric vehicles (EVs). This study presents a comprehensive machine learning (ML)-based approach for SoC estimation of EV LiBs, addressing the challenges of model reliability, uncertainty, and real-world data variability. To ensure the model’s robustness and generalizability, preprocessing techniques, including normalization and scaling, were employed alongside rigorous cross-validation methods. A well-structuredMLpipeline was developed to integrate these processes, optimizing the entire model development cycle for efficiency and practical implementation. In the ML pipeline, we utilized Extra Trees Regressor (ETR) and Light Gradient Boosting Machine (LightGBM) and proposed an ensemble model, combining the strengths of ETR and LightGBM, namely ETR-GBM. We benchmarked the model’s performance against other ML models, such as CatBoost and Random Forest (RF). Under uncertain conditions, the proposed model emphasized its reliability and robustness, and its conclusions underscored the efficacy of the SoC estimation approach. The ETR-GBM consistently outperforms the individual models (ETR, LightGBM, XGBoost, CatBoost, Support Vector Regression (SVR), Random Forest (RF), and Bayesian) when noise is added to the training dataset.With a noise standard deviation of 0.1, the ETR-GBM demonstrated superior performance, achieving a Root Mean Square Error (RMSE) of 0.41%, surpassing the individual models, which exhibited RMSE values ranging from 0.85% to 0.91%.

Generating lightweight and efficient group secret keys (GSKs) by leveraging the inherent physical characteristics of the wireless communication channels for Internet of Things (IoT) devices has gained a lot of interest in recent times. GSKs play a crucial role in ensuring secure communication among IoT devices by encrypting broadcast control and alert messages. In this article, we propose a novel efficient approach to generating GSKs by exploiting the channel state information (CSI) of spatially correlated IoT nodes. We first implement a feedforward neural network (FNN) for identical channel feature generation for the nodes. We then employ a GSK scheme based on the generated channel features. We also propose a novel quantization technique named Adaptive Gaussian distribution-based quantization method with guard-band (AGDQG), that leverages discarded key bits from the GDQG quantization algorithm to improve various key metrics. Simulation results validate the performance of our proposed scheme, and the initial secret keys generated show excellent performance when assessed against key error rate, key generation rate, and entropy. The comparative analysis verifies the proposed quantization algorithm, and the results show that the GSK scheme demonstrates outstanding performance in the presence of an adversary.

Federated learning (FL) has gained attention for enabling efficient distributed learning while maintaining data privacy. However, the data privacy constraint reduces the transparency in the agents’ model update making the learning process vulnerable to Byzantine attacks. In this paper, a mathematical proof is provided to show that when the traditional model-combining scheme is used, the model will eventually diverge to non-useful solutions in the presence of Byzantine agents independently from their number or their contributions. A low complexity norm-control based aggregation approach is also proposed and shown to converge to the optimal and sub-optimal solutions in the absence or presence of Byzantine nodes, respectively. Monte-Carlo simulations are also conducted to verify and validate the mathematical derivations and the efficiency of the proposed approach in protecting the FL model.

Accurate state-of-charge (SoC) and state-of-health (SoH) estimation are crucial for optimizing electric vehicle (EV) performance and enhancing the battery lifetime, safety, and reliability, where machine learning (ML) plays a vital role in this regard. While existing surveys explore ML applications in EVs, they often need to address ML approaches for SoC and SoH estimation. This paper bridges this gap by comprehensively reviewing how ML is utilized for SoC and SoH estimation, analyzing their strengths and weaknesses across different battery chemistries. Our review offers a systematic breakdown of critical areas: fundamental concepts and functionalities of prominent ML algorithms for SoC and SoH estimation, a comparative evaluation of ML techniques applied to diverse EV battery types, an exploration of SoC and SoH estimation using modeling approaches within EV battery systems, and the critical role of dataset quality and model evaluation criteria. Moreover, this paper addresses ML tools developed for lithium-ion batteries (LiBs), image processing applications in EV batteries, and an in-depth investigation of the system model for ML-based SoC and SoH estimation. Furthermore, we present key concepts and methods for SoC and SoH estimation using ML, compare input features, metrics, hyperparameters, and datasets, and demonstrate ML-based system models for EV battery estimation. By conducting this thorough analysis, we aim to close the existing gap and stimulate future progress in ML for SoC and SoH estimation, primarily focusing on LiBs across different EV applications.

The COVID-19 pandemic has highlighted that effective early infection detection methods are essential, as they play a critical role in controlling the epidemic spread. In this work, we investigate the use of wearable sensors in conjunction with machine learning (ML) techniques for pandemic infection detection. We work on designing a wristband that measures various vital parameters such as temperature, heart rate, and SPO2, and transmits them to a mobile application using Bluetooth Low Energy. The accuracy of the wristband measurements is shown to be within 10% of the readings of existing commercial products. The measured data can be used and analyzed for various purposes. To benefit from the existing online datasets related to COVID-19, we use this pandemic as an example in our work. Hence, we also develop ML-based models that use the measured vital parameters along with cough sounds in order to determine whether a case is COVID-19 positive or not. The proposed models are shown to achieve remarkable results, exceeding 90% accuracy. One of our proposed models exceeds 96% performance in terms of accuracy, precision, recall, and F1-Score. The system lends itself reasonably for amendment to deal with future pandemics by considering their specific features and designing the ML models accordingly. Furthermore, we design and develop a mobile application that shows the data collected from the wristband, records cough sounds, runs the ML model, and provides feedback to the user about their health status in a user-friendly, intuitive manner. A successful deployment of such an approach would decrease the load on hospitals and prevent infection from overcrowded spaces inside the hospital.

Patients with hyperglycemia require routine glucose monitoring to effectively treat their condition. We have developed a lightweight wristband device to capture Photoplethysmography (PPG) signals. We collected PPG signals, demographic information, and blood pressure data from 139 diabetic (49.65%) and non-diabetic (50.35%) subjects. Blood glucose was estimated, and diabetic severity (normal, warning, and dangerous) was stratified using Mel frequency cepstral coefficients, time, frequency, and statistical features from PPG and their derivative signals along with physiological parameters. Bagged Ensemble Trees outperform other algorithms in estimating blood glucose level with a correlation coefficient of 0.90. The proposed model’s prediction was all in Zone A and B in the Clarke Error Grid analysis. The predictions are thus clinically acceptable. Furthermore, K-nearest neighbor model classified the severity levels with an accuracy of 98.12%. Furthermore, the proposed models were deployed in Amazon Web Server. The wristband is connected to an Android mobile application to collect real-time data and update the estimated glucose and diabetic severity every 10-seconds, which will allow the users to gain better control of their diabetic health.

In the burgeoning era of quantum communications, Internet of Things (IoT) devices in high-speed train (HST) environments encounter formidable challenges. These devices are constrained by limited power and computational capabilities while needing to safeguard their data and communications against adversaries equipped with quantum-grade computational power. To counter such threats, Quantum Key Distribution (QKD) emerges as a vital solution, facilitating secure communication between servers and IoT controllers, thereby shielding the more vulnerable IoT sensors. This paper delves into the application of QKD within the unique scenario of HSTs, employing Free Space Optics (FSO) to establish high data rate communication channels. Our experimental setup involves the integration of FSO links for photon exchange essential to QKD. We meticulously explore the QKD process in the context of HSTs, detailing our methodology that involves the alignment of FSO transceivers on ground base stations with those on the moving trains, thereby enabling efficient photon exchange. The study presents quantitative results demonstrating that this approach allows for the exchange of a substantial number of keys, with negligible impact on FSO data throughput. These findings highlight that our proposed method can significantly enhance IoT communication security in HSTs without compromising the Quality of Service (QoS) offered to train passengers. Furthermore, we assess the system’s performance under various visibility conditions, which is crucial for FSO viability. Our results indicate the robustness of the proposed QKD method in diverse operational scenarios, underlining its practical applicability in securing IoT communications within HST environments. Through this study, we provide a comprehensive understanding of implementing QKD in high-speed mobile settings, contributing valuable insights into its effectiveness and feasibility.

This paper investigated the use of unmanned aerial vehicles (UAVs) for the delivery of critical goods to remote areas in the absence of network connectivity. Under such conditions, it is important to track the delivery process and record the transactions in a delay-tolerant fashion so that this information can be recovered after the UAV’s return to base. We propose a novel framework that combines the strengths of cipher block chaining, physical layer security, and symmetric and asymmetric encryption techniques in order to safely encrypt the transaction logs of remote delivery operations. The proposed approach is shown to provide high security levels, making the keys undetectable, in addition to being robust to attacks. Thus, it is very useful in drone systems used for logistics and autonomous goods delivery to multiple destinations. This is particularly important in health applications, e.g., for vaccine transmissions, or in relief and rescue operations.