Lithium-Ion Battery Estimation in Online Framework Using Extreme Gradient Boosting Machine Learning Approach

Circulating white blood cell traits and prolonged night shifts: a cross-sectional study based on nurses in Guangxi

This study aimed to elucidate the effects of long day and night shifts on immune cells in a population of nurses. This cross-sectional study in December 2019 was based on a group of nurses. 1568 physically healthy caregivers were included, including 1540 women and 28 men. 1093 nurses had long-term shift work (working in a rotating system for > 1 year). The receiver operating characteristic curve, Ensemble Learning, and Logistic regression analyses were used to evaluate factors related to long-term shift work. The night shift group nurses had significantly higher MPV, PLCR, and WBC and significantly lower BASO%, ELR, MCHC, PLR, RDW-CV, and RDW-SD (P < 0.01). ROC curves showed that WBC, PLR, ELR, RDW_CV, and BASO% were more related to the night shift. Ensemble Learning, combined with the LASSO model, finally filtered out three indicators of night shifts related to ELR, WBC, and RDW_SD. Finally, logistic regression analysis showed that the nurses' night shift situation greatly influenced two peripheral blood ELR and WBC indicators (ELR: log (OR) =  - 3.9, 95% CI: - 5.8- - 2.0; WBC: log (OR) = 0.25, 95% CI: 0.18-0.32). Finally, we showed that, unlike WBC, the relative riskiness of ELR showed opposite results among junior nurses and middle-senior nurses (log (OR) 6.5 (95% CI: 1.2, 13) and - 7.1 (95% CI: - 10, - 3.8), respectively). Our study found that prolonged night shifts were associated with abnormal WBC and ELR, but after strict age matching, WBC remained significantly different. These findings help to confirm that COVID-19 and tumorigenesis (e.g., breast cancer) are significantly associated with circadian rhythm disruption. However, more detailed studies are needed to confirm this.

Data-driven interpretable analysis for polysaccharide yield prediction

Cornstalks show promise as a raw material for polysaccharide production through xylanase. Rapid and accurate prediction of polysaccharide yield can facilitate process optimization, eliminating the need for extensive experimentation in actual production to refine reaction conditions, thereby saving time and costs. However, the intricate interplay of enzymatic factors poses challenges in predicting and optimizing polysaccharide yield accurately. Here, we introduce an innovative data-driven approach leveraging multiple artificial intelligence techniques to enhance polysaccharide production. We propose a machine learning framework to identify highly accurate polysaccharide yield prediction modeling methods and uncover optimal enzymatic parameter combinations. Notably, Random Forest (RF) and eXtreme Gradient Boost (XGB) demonstrate robust performance, achieving prediction accuracies of 93.0% and 95.6%, respectively, while an independently developed deep neural network (DNN) model achieves 91.1% accuracy. A feature importance analysis of XGB reveals the enzyme solution volume's dominant role (43.7%), followed by time (20.7%), substrate concentration (15%), temperature (15%), and pH (5.6%). Further interpretability analysis unveils complex parameter interactions and potential optimization strategies. This data-driven approach, incorporating machine learning, deep learning, and interpretable analysis, offers a viable pathway for polysaccharide yield prediction and the potential recovery of various agricultural residues.

Data-driven direct diagnosis of Li-ion batteries connected to photovoltaics

Photovoltaics supply a growing share of power to the electric grid worldwide. To mitigate resource intermittency issues, these systems are increasingly being paired with electrochemical energy storage devices, e.g., Li-ion batteries, for which ensuring long and safe operation is critical. However, in this operation framework, secondary Li-ion batteries undergo sporadic usage, which prevents the application of standard diagnostic methods. Here, we propose a diagnostic methodology that uses machine learning algorithms trained directly on data obtained from photovoltaic charging of Li-ion batteries. The training is carried out on synthetic voltage data at various degradation conditions calculated from clear sky model irradiance data. The method is validated using synthetic voltage responses calculated from plane of array irradiance observations for a photovoltaic system located in Maui, HI, USA. We report an average root mean square error of 2.75% obtained for more than 10,000 different degradation paths with 25% or less degradation on the Li-ion cells.

Metaverse and Healthcare: Machine Learning-Enabled Digital Twins of Cancer

Medical digital twins, which represent medical assets, play a crucial role in connecting the physical world to the metaverse, enabling patients to access virtual medical services and experience immersive interactions with the real world. One serious disease that can be diagnosed and treated using this technology is cancer. However, the digitalization of such diseases for use in the metaverse is a highly complex process. To address this, this study aims to use machine learning (ML) techniques to create real-time and reliable digital twins of cancer for diagnostic and therapeutic purposes. The study focuses on four classical ML techniques that are simple and fast for medical specialists without extensive Artificial Intelligence (AI) knowledge, and meet the requirements of the Internet of Medical Things (IoMT) in terms of latency and cost. The case study focuses on breast cancer (BC), the second most prevalent form of cancer worldwide. The study also presents a comprehensive conceptual framework to illustrate the process of creating digital twins of cancer, and demonstrates the feasibility and reliability of these digital twins in monitoring, diagnosing, and predicting medical parameters.

XGBoost-Based Remaining Useful Life Estimation Model with Extended Kalman Particle Filter for Lithium-Ion Batteries

The instability and variable lifetime are the benefits of high efficiency and low-cost issues in lithium-ion batteries.An accurate equipment's remaining useful life prediction is essential for successful requirement-based maintenance to improve dependability and lower total maintenance costs. However, it is challenging to assess a battery's working capacity, and specific prediction methods are unable to represent the uncertainty. A scientific evaluation and prediction of a lithium-ion battery's state of health (SOH), mainly its remaining useful life (RUL), is crucial to ensuring the battery's safety and dependability over its entire life cycle and preventing as many catastrophic accidents as feasible. Many strategies have been developed to determine the prediction of the RUL and SOH of lithium-ion batteries, including particle filters (PFs). This paper develops a novel PF-based technique for lithium-ion battery RUL estimation, combining a Kalman filter (KF) with a PF to analyze battery operating data. The PF method is used as the core, and extreme gradient boosting (XGBoost) is used as the observation RUL battery prediction. Due to the powerful nonlinear fitting capabilities, XGBoost is used to map the connection between the retrieved features and the RUL. The life cycle testing aims to gather precise and trustworthy data for RUL prediction. RUL prediction results demonstrate the improved accuracy of our suggested strategy compared to that of other methods. The experiment findings show that the suggested technique can increase the accuracy of RUL prediction when applied to a lithium-ion battery's cycle life data set. The results demonstrate the benefit of the presented method in achieving a more accurate remaining useful life prediction.