Real-Time EV Battery State Monitoring

Abstract Estimating the state of charge lithium-ion battery systems is important for efficient management systems. This work conducts a thorough evaluation multiple SOC estimate methods, including both classic approaches Coulomb Counting and extended Kalman filter machine learning techniques under different load profile on pouch cell. The assessment included variety experimental data collected from entire cycles, shallow dynamic operations utilizing Worldwide Harmonized Light Vehicles Test Procedure Hybrid Pulse Power Characterization tests done 100% to 10% SOC. While traditional performed well ordinary settings, they had severe limits during cycling. In contrast, technologies, notably random forest method, better across all testing conditions. approach showed outstanding accuracy while minimizing error metrics (RMSE: 0.0229, MSE: 0.0005, MAE: 0.0139) effectively handled typical issues such as drift ageing effects. These findings validate dependable robust real-time estimation in

Source

Accurate estimation of the state charge (SOC) lithium-ion batteries (LiBs) proportionally impacts efficiency battery management systems (BMS) considering dynamic and non-linear behavior LiBs. Changes in activities cathode anode materials internal resistance tend to impact capacity. When is operated at high or low temperatures under HWFET condition, capacity tends deteriorate drastically. Therefore, high-precision SOC required ensure safe stable operation. In this work, we propose a combined Improved Dung Beetle Optimization (IDBO) Extreme Learning Machine (ELM) framework for evaluate BMS. The novelty model stems from application IDBO algorithm, which incorporating Circle chaotic mapping, Golden sine strategy, Levy flight hyper-parameter optimization. This effectively resolves problems inconsistent performance instability arising randomly initialized hidden layer weights biases ELM, resulting enhanced prediction accuracy. proposed IDBO-ELM method validated context five parameters, namely, different ambient temperatures, operating conditions, materials, initial values, running time. ex... Read More

Source

Lithiumion batteries are the dominant battery type for emerging technologies in efforts to slow climate change. Accurate and quick estimations of state charge (SOC) internal cell temperature vital battery-management systems enable effective operation portable electronics electric vehicles. Therefore, a long short-term memory (LSTM) recurrent-neural network is proposed which completes estimation SOC average (Tavg) lithiumion under varying current loads. The trained evaluated using data compiled from newly developed extended single-particle model coupled with thermal dynamic model. Results promising, root mean square values typically 2% 1.2 K Tavg, while maintaining training testing times. In addition, we examined comparison single-feature versus multi-feature network, as well two different approaches partitioning.

Source

Abstract Battery lifetime and reliability depend on accurate state-of-health (SOH) estimation, while complex degradation mechanisms varying operating conditions strengthen this challenge. This study presents two physics-informed neural network (PINN) configurations, PINN-Parallel PINN-Series, designed to improve SOH prediction by combining an equivalent circuit model (ECM) with a long short-term memory (LSTM) network. process input data through parallel ECM LSTM modules combine their outputs for estimation. On the other hand, PINN-Series uses sequential approach that feeds ECM-derived parameters into supplement temporal analysis physics information. Both models utilize easily accessible voltage, current, temperature match realistic battery monitoring constraints. Experimental evaluations show outperforms baseline in accuracy robustness. It also adapts well different conditions. demonstrates simulated dynamic states from increase LSTM's ability capture patterns model's explain behavior. However, trade-off between robustness training efficiency of PINNs is discussed. The research findi... Read More

Source

Multisine electrochemical impedance spectroscopy (EIS) represents a highly promising technique for the online characterization of battery functional states, offering potential to monitor, in real-time, key degradation phenomena such as aging, internal resistance variation, and state health (SoH) evolution. However, its widespread adoption embedded systems is currently limited by need balance measurement accuracy with strict energy constraints requirement short acquisition times. This work proposes novel broadband EIS approach based on multiband multisine excitation strategy which signal spectrum divided into multiple sub-bands that are sequentially explored. enables available be concentrated portion at time, thereby significantly improving signal-to-noise ratio (SNR) without substantially increasing total time. The result more energy-efficient method maintains high diagnostic precision. We further investigated optimal design these sequences, taking account realistic imposed sensing hardware limitations amplitude noise level. effectiveness proposed was demonstrated within comprehensiv... Read More

Source

Real-time analysis of multi-dimensional time series data from sensors using a time series service that allows efficient and scalable computation and visualization of time series data from heterogeneous sources with different time units and sampling rates. The time series service receives queries, identifies the required data from the time series databases, computes transforms, and provides real-time output. It enables contextual analysis of multi-dimensional time series data for real-time monitoring and alerting using computations like correlation and regression.

Source

Battery management system for electric vehicles with a refined voltage measurement technique to optimize battery performance and durability. The system manages an electric battery device with multiple modules connected in series, each containing cells. Some cells are connected by a busbar. The measurement technique accounts for busbar voltage offsets. It uses a single slave element to gather data from two modules. Measurements from cells on the busbar are adjusted based on the busbar resistance. This prevents overestimation due to busbar voltage. The adjusted cell voltages are used for safety methods like derating charging power. The technique improves accuracy by avoiding erroneous voltage readings from busbar cells.

Source

Method to estimate battery health of electric vehicles accurately and online without removing the battery. The method involves estimating the state of charge (SOC) of the battery by relaxing it, discharging it, and checking the max charge. This allows precise SOC estimation even as the battery ages without full discharge. The SOC is used to estimate battery health. The steps are: 1. Relax battery at max charge, monitoring voltage and temperature. 2. Discharge to a target SOC while monitoring voltage and temperature. 3. Check if SOC reached max charge. If yes, proceed with health estimation, else repeat steps 1-2. By relaxing before discharge, the SOC error from aging resistance is avoided.

Source

Battery management system for accurately determining if a battery is defective in an electric vehicle. The system diagnoses the battery during normal operation to determine if it has abnormalities. Based on the results, it adjusts the diagnostic conditions for a second diagnosis to further evaluate the battery. This allows more accurate detection of defects like increased internal resistance, leakage current, or decreased state of health. It also prevents false positives by checking if the second diagnosis result is significantly different from the first. This provides a more reliable diagnosis of battery defects compared to single-diagnosis methods.

Source

Simplified and accurate method to measure battery resistance using existing vehicle components. It leverages the fact that when a battery is charging and discharging simultaneously, it outputs more current than it receives. By determining if this condition is met, the method can detect when the battery switches from a charged to discharged state. At this point, it starts measuring the resistance. This avoids the need for specialized equipment and allows using the vehicle's control system to enable battery resistance measurement during normal operation.

Source

Abstract The conventional method for assessing the health status of lithium batteries typically necessitates comprehensive data from complete charging and discharging cycles. prolonged duration required collection such may lead to issues including time inefficiency delays in battery state estimation processes. In response, this paper presents a rapid estimating based on local information short process. Additionally, address situation where correlation features is low specific regions entire voltage domain, complementary strategy proposed. This allows quick accurate using only process intervals across domain. First, that can represent full domain are extracted. Subsequently, multi-feature fusion approach combined with LightGBM algorithm employed construct model. Finally, effects various types, different operating conditions, diverse sampling window sizes accuracy analyzed through experiments, thereby demonstrating feasibility effectiveness proposed approach.

Source

Battery management system and method that provides real-time battery monitoring and predicts battery performance based on accumulated usage. The system connects a display panel to a battery pack to show battery status like voltage, current, and temperature in real time. It calculates cumulative usage by comparing actual battery data to a model. Using this, it predicts battery performance and adjusts charging/discharging to extend life. When a battery is nearing end-of-life, it determines recyclability based on the accumulated usage.

Source

Battery monitoring and management system for predicting battery degradation and making maintenance recommendations. The system uses big data analysis to evaluate battery health and predict degradation. It acquires battery voltage, current, temperature, capacity, discharge rate, etc. through sensors. The data is stored and analyzed to calculate health indices and degradation rates. A decision tree algorithm uses these to generate maintenance reminders, warnings, and charging strategies. A user interface displays battery status and alerts.

Source

Battery management system that estimates battery state of charge (SOC) more accurately by leveraging nonlinear characteristics of the battery's open circuit voltage (OCV) near full discharge. The system uses an extended Kalman filter to estimate SOC from battery measurements. But instead of just using the Kalman filter output, it also calculates OCV based on the provisional SOC estimate. Then it adjusts the SOC estimate using a measurement update with both the provisional SOC and OCV. This leverages the strong nonlinear OCV-SOC relationship near full discharge to improve SOC estimation accuracy.

Source

Battery monitoring and management system that uses compacted sensor data to generate insights and alarms for battery health and performance. The system receives sensor data from temperature, voltage, and current sensors of a battery. It compacts the data into coefficients and generates performance indicators based on those coefficients. Alarms are then generated when performance indicators exceed thresholds. This allows efficient monitoring and management of large amounts of battery sensor data by compacting and analyzing key performance metrics instead of storing and processing raw sensor data.

Source

Vehicle-pile collaborative state monitoring system for electric vehicles that accurately predicts battery life during charging and discharging. The system monitors parameters of the vehicle battery, charging pile, and charging/discharging state. It uses a model to predict battery life based on these parameters. Multiple influencing factors like charging cycles are combined to correct and improve the life prediction accuracy.

Source

An online adaptive method to estimate and control battery parameters like remaining useful life (RUL) and state of charge (SOC) for electric vehicles. It uses a combination of extended Kalman filtering and recurrent neural networks (GRU) to provide accurate and adaptive battery parameter estimation. The method involves predicting SOC using Kalman filters, then using the predicted SOC along with current battery and environmental data to predict RUL using a GRU neural network. The adaptive component lies in dynamically adjusting battery internal and external parameters based on the predicted RUL.

Source

A method for estimating the health status of batteries in energy storage systems that combines Kalman filtering with a neural network to accurately estimate battery health using only online sensor data. The method involves using Kalman filters to estimate key battery parameters like state of charge, then feeding that data into a long short-term memory neural network to predict battery health. This allows estimating internal battery degradation without requiring invasive measurements.

Source

Online monitoring and management of lithium-ion batteries using deep learning networks to improve efficiency, longevity, and safety of battery systems. The method involves real-time monitoring of battery parameters, generating balanced charging profiles, analyzing performance, predicting battery life, and detecting abnormalities using deep learning models. It leverages data processing, trend analysis, feature extraction, and machine learning techniques to provide accurate battery condition assessment and maintenance recommendations.

Source

Integrated battery management system for lithium batteries in electric vehicles that improves charging and discharging efficiency and protects battery health. The system monitors battery parameters during charging and discharging to enable pre-regulation. It determines charging power based on time and battery state, synchronizes with redundant power, and calculates charging time using user habit analysis. This allows lower charging currents to protect the battery while ensuring sufficient charging.

Source