Abstract:Fast charging has attracted increasing attention from the battery community for electrical vehicles (EVs) to alleviate range anxiety and reduce charging time for EVs. However, inappropriate charging strategies would cause severe degradation of batteries or even hazardous accidents. To optimize fast-charging strategies under various constraints, particularly safety limits, we propose a novel deep Bayesian optimization (BO) approach that utilizes Bayesian recurrent neural network (BRNN) as the surrogate model, given its capability in handling sequential data. In addition, a combined acquisition function of expected improvement (EI) and upper confidence bound (UCB) is developed to better balance the exploitation and exploration. The effectiveness of the proposed approach is demonstrated on the PETLION, a porous electrode theory-based battery simulator. Our method is also compared with the state-of-the-art BO methods that use Gaussian process (GP) and non-recurrent network as surrogate models. The results verify the superior performance of the proposed fast charging approaches, which mainly results from that: (i) the BRNN-based surrogate model provides a more precise prediction of battery lifetime than that based on GP or non-recurrent network; and (ii) the combined acquisition function outperforms traditional EI or UCB criteria in exploring the optimal charging protocol that maintains the longest battery lifetime.

What problem does this paper attempt to address?

The problem that this paper attempts to solve is the optimization problem of lithium - ion batteries during the fast - charging process. Specifically, the paper aims to develop an effective fast - charging strategy to reduce battery degradation while ensuring the safety of the charging process. Fast - charging is very important for electric vehicles (EVs) because it can alleviate range anxiety and shorten charging time. However, an inappropriate charging strategy may lead to severe battery degradation or even cause safety accidents. Therefore, the paper proposes a method based on Deep Bayesian Optimization (DBO), using the Bayesian Recurrent Neural Network (BRNN) as a surrogate model, combined with the acquisition function of the combination of Expected Improvement (EI) and Upper Confidence Bound (UCB) to balance exploration and exploitation, thereby optimizing the fast - charging strategy under various constraints. ### Main contributions of the paper: 1. **Surrogate model selection**: Using BRNN as a surrogate model, compared with the traditional Gaussian Process (GP) and non - recurrent neural networks, BRNN can predict battery life more accurately and can handle sequence data. 2. **Combined acquisition function**: Proposing a combined acquisition function that combines EI and UCB to better balance exploration and exploitation and improve the ability to find the optimal charging protocol. 3. **Experimental verification**: Experimental verification was carried out on the battery simulator PETLION based on the porous electrode theory. The results show that the proposed method is superior to existing methods in predicting battery life and optimizing charging strategies. ### Formula summary: - **State transition and output prediction**: \[ \mathbf{s}_t = f_s(\mathbf{x}_t, \mathbf{s}_{t - 1}|\theta_s), \quad y_t = W_y\mathbf{s}_t + b_y \] \[ \mathbf{s}_t=\phi(W\mathbf{x}_t + U\mathbf{s}_{t - 1}+b) \] - **Posterior distribution**: \[ p(y_t|D)=\int p(y_t|\theta_y)p(\theta_y|D)d\theta_y \] where $\theta_y = [W_y, b_y]$, $p(\theta_y|D)$ is the posterior distribution of the last - layer parameters under the given data set $D$, and $p(y_t|\theta_y)$ is the output distribution under the given fixed $\theta_y$. - **Expected Improvement (EI)**: \[ \alpha_{\text{EI}}(\mathbf{x})=\sigma_f(\mathbf{x})\left(Z_f(\mathbf{x})\Phi(Z_f(\mathbf{x}))+\phi(Z_f(\mathbf{x}))\right) \] \[ Z_f(\mathbf{x})=\frac{f(\mathbf{x}^+)-\mu_f(\mathbf{x})}{\sigma_f(\mathbf{x})} \] - **Upper Confidence Bound (UCB)**: \[ \alpha_{\text{UCB}}(\mathbf{x})=\mu_t(\mathbf{x})+\kappa\sigma_t(\mathbf{x}) \] Through these methods and formulas, the paper has successfully solved the challenges of battery life prediction and optimization during the fast - charging process.

Fast Charging of Lithium-Ion Batteries Using Deep Bayesian Optimization with Recurrent Neural Network

A Bayesian Mixture Neural Network for Remaining Useful Life Prediction of Lithium-Ion Batteries

Data-Driven Fast Charging Optimization for Lithium-Ion Battery Using Bayesian Optimization with Fast Convergence

Adaptive Safe Reinforcement Learning-Enabled Optimization of Battery Fast-Charging Protocols

Reinforcement Learning-based Fast Charging Control Strategy for Li-ion Batteries

Adaptive safe reinforcement learning‐enabled optimization of battery fast‐charging protocols

Health-aware Fast-charging Control of Lithium-Ion Battery Based on Reinforcement Learning

A sequence to sequence prediction model for remaining useful life of lithium-ion batteries with Bayesian optimisation process visualization

State‐of‐health estimation and remaining useful life for lithium‐ion battery based on deep learning with Bayesian hyperparameter optimization

Deep Deterministic Policy Gradient-DRL Enabled Multiphysics-Constrained Fast Charging of Lithium-Ion Battery

Specialized deep neural networks for battery health prognostics: Opportunities and challenges

A Bayesian Optimized Deep Learning Approach for Accurate State of Charge Estimation of Lithium Ion Batteries Used for Electric Vehicle Application

Online State-of-Health Estimation for Fast-Charging Lithium-Ion Batteries Based on a Transformer–Long Short-Term Memory Neural Network

Lithium Battery State-of-Charge Estimation Based on a Bayesian Optimization Bidirectional Long Short-Term Memory Neural Network

Efficient computation of robust, safe, fast charging protocols for lithium-ion batteries

An Improved Random Drift Particle Swarm Optimization-Feed Forward Backpropagation Neural Network for High-Precision State-of-charge Estimation of Lithium-Ion Batteries

A Novel Fusion Method for State-of-Charge Estimation of Lithium-Ion Batteries Based on Improved Genetic Algorithm BP and Adaptive Extended Kalman Filter

Optimal Fast Charging of Lithium Ion Batteries: Between Model-Based and Data-Driven Methods

Intelligent Learning Method for Capacity Estimation of Lithium-Ion Batteries Based on Partial Charging Curves

A Combined Data-Model Method for State-of-Charge Estimation of Lithium-Ion Batteries.

Efficient computation of safe, fast charging protocols for multiphase lithium-ion batteries: A lithium iron phosphate case study