In this paper, an artificial neural network (ANN) based approach is proposed to estimate the capacity fade in lithium-ion (Li-ion) batteries for electric vehicles (EVs). Besides its robustness, stability, and high accuracy, technique can significantly improve state-of-charge (SOC) estimation accuracy over lifespan of battery, which leads more reliable battery operation prolonged lifetime. addition, allows accurate prediction remaining service time. Two identical 3.6-V/16.5-Ah Li-ion cells...

Battery performance prediction is crucial in many applications. A good mechanism of the battery has advantages. For example, it increases lifetime by preventing over (dis)charging battery, allows utilizing entire capacity and offers an access to user know amount energy pack. In this paper, some models are derived tested on a commercial Lithium cell. The results show capabilities these under different tests.

Electric vehicles (EVs) require reliable and very accurate battery state-of-charge (SOC) estimation to maximize their performance. A commonly used technique, the extended Kalman filter (EKF), provides an estimate of SOC. However, EKF has some limitations, such as it assumes knowledge statistics process noise measurement is available, which practically cannot be guaranteed. In this paper, adaptive equivalent-circuit model proposed for SOC estimation. The based on a common cell with parameters...

A novel method is proposed for forecasting the capacity of lithium-ion battery cells. The uses a Gaussian Process Regression model, machine learning framework. Besides high prediction accuracy and robustness possesses, offers other advantages, namely, it provides uncertainty information, has capability to cross-correlate trends between different These two merits make very reliable practical solution applications that use cell packs with large number interconnected derived, verified, compared...

The temperature of a battery cell is major parameter that must be continuously monitored to ensure safe operation. Most failures are linked thermal runaway due rise in the battery, which if not detected early can result destruction or fire hazard. This article proposes robust artificial neural network models with reduced complexity estimate surface different chemistries. proposed accurate, reliable, and use no sensor. Different architectures evaluated optimized. Derivation followed by...

For the safe and reliable operation of battery-driven machines, accurate state-of-charge (SOC) estimations are necessary. Unfortunately, existing methods often fail to identify patterns relevant long-term SOC estimation due complex battery cell characteristics such as aging. In this paper, we propose Uncorrelated Sparse Autoencoder with Long Short-Term Memory (USAL). USAL is a novel neural network that addresses challenging task given limited initial history cell's charge-discharge behavior....

This paper proposes a novel invariant extended Kalman filter (IEKF), modified version of the (EKF), for state-of-charge (SOC) estimation lithium-ion (Li-ion) battery cells. Unlike conventional EKF methods where correction term used to update state is linearly proportional output error, this employs IEKF independent resulting in significant reduction error and improving accuracy. In contrast classic method like more contemporary ones square root variant Cubature Filter (SCKF), can...

This paper covers some design and operation aspects of distributed battery micro-storage systems in a deregulated electricity market system. In this paper, the term "micro" refers to size energy storage (ES) system compared grid generation, with capacity from few kilowatt-hours up. Generally, ES enhances performance renewable generators (DGs) increases efficiency entire power Energy allows for leveling load, shaving peak demands, furthermore, transacting utility grid. different are covered...

Battery performance is strongly dependent on the ambient temperature. For example, at moderate temperatures, battery optimal, whereas extreme not optimized and sometimes unexpected. In order to predict behavior, a model that involves battery's underlying dynamics usually used. The majority of dynamic models are derived only one single temperature (room temperature), which can easily lead failure in predicting when varies. Therefore, adding some dependence those make management system more...

Accurate battery state-of-charge (SOC) estimation in real time is desired many applications. Among other methods, the extended Kalman filter (EKF) allows for high-accuracy real-time tracking of SOC. However, an accurate SOC model needed to guarantee convergence. Additionally, knowledge statistics process noise and measurement estimation. In this paper, two namely, multiple-model EKF (MM-EKF) autocovariance least squares technique, are proposed estimating lithium-ion (Li-ion) cells. The first...

A latent function decomposition method is proposed for forecasting the capacity of lithium-ion battery cells. The uses Multi-Output Gaussian Process, a generative machine learning framework multi-task and transfer learning. MCGP decomposes available trends from multiple cells into functions. functions are then convolved over kernel smoothers to reconstruct and/or forecast Besides high prediction accuracy possesses, it provides uncertainty information predictions captures nontrivial...

State-of-charge estimation is an essential part of a battery management system. Charging and discharging batteries involve complex chemical processes that could lead to undesired consequences, such as premature end life or fire hazards if the battery's state-of-charge not closely monitored. This work proposes method compensates for noisy measurements parametric uncertainties in cell improve accuracy. In this work, robust adaptive scheme based on Cubature Kalman filter proposed. The algorithm...

This paper investigates the role of artificial neural networks in enhancing accuracy instantaneous power estimation electric vehicles' batteries. In vehicles, a battery is used as main or complementary bidirectional source. To optimize energy management vehicle, sourced sinked by must be estimated real time under any condition. The function many variables including current, state charge, ambient temperature, and health. evaluates some existing equivalent circuit models for estimating...

Real-time battery state-of-charge (SOC) estimation is critical in many applications. The extended Kalman filter (EKF) has been successfully deployed SOC allowing real-time monitoring. However, modeling inaccuracies, measurement faults, and wrong initialization can cause the algorithm to diverge. precise knowledge of statistical information about process measurements noise crucial for accurate system estimation. This paper presents a novel approach based on maximum-likelihood (MLE). models...

This article proposes an adaptive sensorless measurement technique for internal temperature of lithium-ion (Li-ion) batteries. The proposed is based on measuring the impedance phase battery in real time by a simple search algorithm. then identifies zero-crossing frequency at which zero. From value frequency, cell's estimated. Details followed experimental verification using 3.8-V/2600-mAh Samsung Li-ion cell temperatures between -20 °C and +50 are presented.

This paper proposes an adaptive filter for estimating the surface temperature of lithium-ion battery cells in real time. The proposed sensorless method aims to achieve a highly accurate estimation at relatively low implementation cost. employs system dynamic and measurement models derived using polynomial curve fitting implemented autotuned extended Kalman (AA-EKF). Derivation technique followed by experimental verification are demonstrated.

Although several state-of-charge (SOC) estimation methods have been proposed at the battery cell level, limited work has done to identify effect of aging on SOC estimations. To address this challenge, article proposes a novel method for estimating Lithium-ion (Li-ion) cells by accurately modeling and degradation information. The method, termed as "NNGP," is deep neural network with Gaussian process feedback. feedback helps NNGP correlate trends over consecutive charge–discharge cycles....

In advanced battery management systems, it is critical to incorporate an accurate model that captures the cells' dynamics in order predict performance. this paper, a hysteresis for Lithium cell with improved transient response proposed. cell, loop exists charge/discharge cycle resulting increased complexity behavior. The proposed has advantages of being simple and fairly accurate. improvement response, which observed during after relaxation, been verified experimentally by applying different...

Battery management systems (BMS) must estimate the state-of-charge (SOC) of battery accurately to prolong its lifetime and ensure a reliable operation. Since batteries have wide range applications, SOC estimation requirements methods vary from an application another. This paper compares two methods, namely extended Kalman filters (EKF) artificial neural networks (ANN). EKF is nonlinear optimal estimator that used inner state dynamic system using state-space model. On other hand, ANN...

This paper presents two artificial neural network (ANN) based algorithms for battery state-of-charge (SOC) estimation. The SOC is an important quantity that must be estimated in real-time many applications. ANN a mathematical model consists of interconnected neurons inspired by biological networks and used to predict the output dynamic system on some historical data system. first algorithm presented this has open-loop structure known as nonlinear input (NIO) feed-forward algorithm, while...

Sensorless temperature estimation methods for batteries can be classified into three categories: analytical methods, observer-based and data-driven methods. In general, are easy to derive implement but have a limited performance due their open-loop nature. Observer-based high closed-loop nature demand accurate dynamic measurement models with up-to-date parameters estimation. Data-driven extremely stable robust massive parallel structure, they huge amount of data training. This paper presents...

An accurate state-of-charge (SOC) estimation is desired in most battery systems. It increases the reliability of system and extends lifetime battery. This paper proposes an Extended Kalman Filter (EKF) algorithm to estimate SOC a Lithium cell. To implement algorithm, improved cell model used. The results EKF show effectiveness ease implementation proposed technique.