Abstract:Equivalent Circuit Model(ECM)has been widelyused in battery modeling and state estimation because of itssimplicity, stability and interpretability.However, ECM maygenerate large estimation errors in extreme working conditionssuch as freezing environmenttemperature andcomplexcharging/discharging behaviors,in whichscenariostheelectrochemical characteristics of the battery become extremelycomplex and <a class="link-external link-http" href="http://nonlinear.In" rel="external noopener nofollow">this http URL</a> this paper,we propose a hybridbattery model by embeddingneural networks as 'virtualelectronic components' into the classical ECM to enhance themodel nonlinear-fitting ability and adaptability. First, thestructure of the proposed hybrid model is introduced, where theembedded neural networks are targeted to fit the residuals of theclassical ECM,Second, an iterative offline training strategy isdesigned to train the hybrid model by merging the battery statespace equation into the neural network loss function. Last, thebattery online state of charge (SOC)estimation is achieved basedon the proposed hybrid model to demonstrate its applicationvalue,Simulation results based on a real-world battery datasetshow that the proposed hybrid model can achieve 29%-64%error reduction for $OC estimation under different operatingconditions at varying environment temperatures.

What problem does this paper attempt to address?

The problem that this paper attempts to solve is that under extreme working conditions (such as freezing ambient temperatures and complex charge - discharge behaviors), the traditional equivalent circuit model (ECM) has a large error in the state estimation of lithium - ion batteries. These extreme conditions make the electrochemical characteristics of the battery extremely complex and nonlinear, resulting in the inability of the traditional ECM to model accurately. Therefore, the paper proposes a new hybrid model. By embedding a neural network into the traditional ECM, the nonlinear fitting ability and adaptability of the model are enhanced, thereby improving the accuracy of battery state estimation. Specifically, the main contributions of the paper include: 1. **Proposing a hybrid model structure**: By embedding a neural network as a "virtual electronic component" into the classical ECM, these neural network modules are trained to fit the residuals of the classical ECM, thereby increasing the nonlinear fitting ability and adaptability of the model and reducing the fitting error under extreme working conditions. 2. **Designing an iterative off - line training strategy**: In order to solve the indirect training problem of the embedded FNN module, the paper proposes an iterative off - line training strategy. The battery state - space equation is incorporated into the traditional mean - square error (MSE) to form a physical - information - loss function to guide the training of the FNN module. Through the above methods, the paper aims to improve the state - estimation accuracy of lithium - ion batteries under different working conditions, especially under extreme ambient temperatures. The experimental results show that the SOC estimation error of the proposed hybrid model under different working conditions is reduced by 29% to 64% compared with the traditional ECM.

A Neural-Network-Embedded Equivalent Circuit Model for Lithium-ion Battery State Estimation

Electrochemical Model Based State of Charge Estimation for Lithium Batteries with Backstepping Techniques

Estimation of lithium-ion battery electrochemical properties from equivalent circuit model parameters using machine learning

A comparative study of lumped equivalent circuit models of a lithium battery for state of charge prediction

Hybrid Equivalent Circuit Model of Lithium Ion Battery Considering Nonlinear Capacity Effects

A comparative study of equivalent circuit models and enhanced equivalent circuit models of lithium-ion batteries with different model structures

ICNCS: Internal Cascaded Neuromorphic Computing System for Fast Electric Vehicle State-of-Charge Estimation

An improvement of equivalent circuit model for state of health estimation of lithium-ion batteries based on mid-frequency and low-frequency electrochemical impedance spectroscopy

Joint State Estimation of Lithium-Ion Batteries Combining Improved Equivalent Circuit Model with Electrochemical Mechanism and Diffusion Process

Enhancing the estimation accuracy in low state-of-charge area: A novel onboard battery model through surface state of charge determination

Real-time estimation of negative electrode potential and state of charge of lithium-ion battery based on a half-cell-level equivalent circuit model

A Hybrid Battery Equivalent Circuit Model, Deep Learning, and Transfer Learning for Battery State Monitoring

State-of-Charge Estimation of Li-ion Batteries Based on A Hybrid Model Using Nonlinear Autoregressive Exogenous Neural Networks

A Comprehensive Physics-Based Equivalent-Circuit Model and State of Charge Estimation for Lithium-Ion Batteries

A novel method for identification of lithium-ion battery equivalent circuit model parameters considering electrochemical properties

State Estimation of Lithium-ion Batteries Using Adaptive Extended Kalman Filter and Long Short-Term Memory Networks

A Lithium Battery SOC Estimation Method Based on AEKF-Composite Neural Network Hybrid Driven Model

Hybrid Modeling of Lithium-Ion Battery: Physics-Informed Neural Network for Battery State Estimation

Co-Estimation of State-of-Charge and State-of- Health for Lithium-Ion Batteries Using an Enhanced Electrochemical Model

A Novel Observer for Lithium-Ion Battery State of Charge Estimation in Electric Vehicles Based on A Second-Order Equivalent Circuit Model

Detection of Voltage Fault in Lithium-Ion Battery Based on Equivalent Circuit Model-Informed Neural Network