Which methods can be used for rapid state of health estimation?
Best insight from top research papers
Rapid state of health (SOH) estimation methods for lithium-ion batteries have been proposed in the literature. One approach is to use a combination of the light gradient boosting machine (LightGBM) and weighted quantile regression (WQR) methods to learn a nonlinear mapping between measurable characteristics and SOH . Another method involves feature selection and machine learning techniques, such as linear regression (LR), to estimate SOH based on capacity-based curves . Additionally, an impedance calculation-based method using electrochemical impedance spectroscopy (EIS) has been proposed, where impedance features are selected and an extreme learning machine is used for SOH evaluation . Finally, a method based on incremental capacity (IC) analysis and Gaussian process regression (GPR) has been developed specifically for fast-charging batteries, achieving a significant reduction in mean absolute percentage error .
Answers from top 5 papers
More filters
| Papers (5) | Insight |
|---|---|
17 Citations | The paper proposes a novel method for state of health (SOH) estimation of fast-charging lithium-ion batteries based on incremental capacity (IC) analysis and Gaussian process regression (GPR). |
32 Citations | The paper proposes a fast impedance calculation-based method for state-of-health (SOH) estimation of lithium-ion batteries using electrochemical impedance spectroscopy (EIS). The method includes the use of impedance features called health factors and an improved fast Fourier transform (FFT) for online EIS acquisition. The proposed method allows for rapid SOH estimation within 35 seconds with estimation errors less than 2%. |
| The paper proposes a state of health (SOH) estimation method based on feature selection and machine learning methods, specifically using a linear regression model. | |
| The paper proposes the use of the LightGBM and weighted quantile regression (WQR) methods for rapid state of health (SOH) estimation of lithium-ion batteries. | |
| The paper does not mention any specific methods for rapid state of health estimation. The paper focuses on a method based on feature selection and machine learning for accurate and effective state of health estimation for lithium-ion batteries in electric vehicles. |
Related Questions
What is Rapid Health Need Assessment?5 answersA Rapid Health Need Assessment (RHNA) is a crucial tool used in disaster and humanitarian contexts to swiftly evaluate the health and safety needs of affected populations. RHNA involves rapid data collection methods, such as structured surveys or assessments, to identify immediate health concerns and prioritize response efforts. Techniques like the Rapid Needs Assessment and Response (RNR) facilitate active engagement among participants, enhancing learning and tailoring education to specific needs. RHNA models emphasize the importance of timely interventions, client involvement throughout the process, and the utilization of existing data sources for efficient decision-making. Combining different rapid assessment protocols, like the Rapid Assessment of Visual Impairment (RAVI) and Rapid Assessment of Refractive Errors (RARE), can provide a comprehensive overview of health needs across different age groups and ocular conditions.
How to estimate the state of health of an battery system?5 answersState of health (SOH) estimation of battery systems can be achieved using various methods. One approach is to use machine learning algorithms such as LightGBM and weighted quantile regression (WQR) to learn a nonlinear mapping between measurable characteristics and SOH. Another method involves feature selection and linear regression (LR) models to estimate SOH based on capacity-based curves. Mathematical model-driven approaches combine battery equivalent models with adaptive intelligent algorithms for real-time filtering. Gaussian process regression (GPR) can be used to estimate SOH by reconstructing charging voltage trajectories and predicting SOH based on charging duration. Additionally, the use of incremental capacity analysis (ICA), discrete wavelet transform (DWT), and deep learning models such as double-layer long short-term memory network (LSTM) can provide accurate SOH predictions.
What are the statistical and predictive modelling techniques to use for public health demand forecasting?5 answersStatistical and predictive modelling techniques for public health demand forecasting include combining statistical micro-simulation of demographic change with agent-based social simulation of individual interactions. Point and probabilistic forecasts are used to evaluate and compare the predictive performance across models, with methods such as quantifying the accuracy of epidemic forecasts. GM (1, 1) and TFGM (1, 1) models are applied to healthcare sector demand forecasting, with TFGM (1, 1) showing better forecasting performance than GM (1, 1) and Holt-Winters method. A hybrid simulation model using system dynamics and discrete event simulation is used to study the influence of long-term population changes on healthcare service demand. Multiple linear regression is used as a forecasting technique for demand in the healthcare domain, considering factors specific to healthcare and excluding outliers for more accurate forecasts.
What are some specific examples of how machine learning has been used in public health?4 answersMachine learning has been used in public health for various purposes. One specific example is the integration of machine learning techniques to improve decision-making processes in healthcare and public health, such as identifying hidden linkages or anomalies in critical medical data analysis. Another example is the use of machine learning in analyzing health data to develop computational models and AI systems for extracting insights, including patient data from multi-omics methods, clinical data, and data from the biomedical literature. Additionally, machine learning has been utilized to analyze the impact of air travel on the spread of diseases like coronavirus, with a focus on determining the effect of the number of flights on the number of cases and deaths reported in specific areas. Furthermore, machine learning has shown promise in disease detection and healthcare, outperforming medical teams in tasks such as pneumonia detection, skin cancer classification, hemorrhage detection, and arrhythmia detection.
What are the main tools that can estimate the healthy ageing?4 answersThere is lack of consensus about how to define healthy ageing and, unsurprisingly, diversity in the instruments for its measurement. However, progress is being made in describing and in devising tools to capture the healthy ageing phenotype. Attempts to measure healthy ageing have relied primarily on cross-sectional data collected in older people. More recent studies have assessed the healthy ageing phenotype using markers of multiple functional domains and have used longitudinal data to model the dynamics and trajectories of healthy ageing. Cardiovascular and lung function, glucose metabolism and musculoskeletal function are key subdomains of physiological function. Strength, locomotion, balance and dexterity are key physical capability subdomains. Memory, processing speed and executive function emerged as key subdomains of cognitive function. Markers of the HPA-axis, sex hormones and growth hormones were important biomarkers of endocrine function. Finally, inflammatory factors were identified as important biomarkers of immune function.
What are models used for time series data in health?2 answersTime series data in health can be analyzed using various models. One approach is to use recurrent neural network (RNN) models, such as the long short-term memory (LSTM) model, which have been applied to predict mortality and acute kidney injury (AKI) based on observational values. Another approach is to use time series models like the autoregressive integrated moving average (ARIMA) model, which has been used to analyze the spread of COVID-19 and predict the number of cases in the future. Additionally, semiparametric generalized linear models have been proposed for time-series data in population health and epidemiology, allowing for estimation of both finite-dimensional and infinite-dimensional parameters. These models provide valuable tools for analyzing and predicting health outcomes based on time series data.