• Energy Research

Accurately evaluating the health status of lithium-ion batteries (LIBs) is significant to enhance the safety, efficiency, and economy of LIBs deployment. However, the complex degradation processes inside the battery make it a thorny challenge. Data-driven methods are widely used to resolve the problem without exploring the complex aging mechanisms; however, random and incomplete charging-discharging processes in actual applications make the existing methods fail to work. Here, we develop three data-driven methods to estimate battery state of health (SOH) using a short random charging segment (RCS). Four types of commercial LIBs (75 cells), cycled under different temperatures and discharging rates, are employed to validate the methods. Trained on a nominal cycling condition, our models can achieve high-precision SOH estimation under other different conditions. We prove that an RCS with a 10mV voltage window can obtain an average error of less than 5%, and the error plunges as the voltage window increases.

Accurate health prognostics of lithium-ion battery packs play a crucial role in timely maintenance and avoiding potential safety accidents in energy storage. To rapidly evaluate the health of newly developed battery packs, a method for predicting the future health of the battery pack using the aging data of the battery cells for their entire lifecycles and with the early cycling data of the battery pack is proposed. Firstly, health indicators (HIs) are extracted from the experimental data, and high correlations between the extracted HIs and the capacity are verified by the Pearson correlation analysis method. To predict the future health of the battery pack based on the HIs, degradation models of HIs are constructed by using an exponential function, long short-term memory network, and their weighted fusion. The future HIs of the battery pack are predicted according to the fusion degradation model. Then, based on the Gaussian process regression algorithm and battery pack data, a data-driven model is constructed to predict the health of the battery pack. Finally, the proposed method is validated with a series-connected battery pack with fifteen 100 Ah lithium iron phosphate battery cells. The mean absolute error and root mean square error of the health prediction of the battery pack are 7.17% and 7.81%, respectively, indicating that the proposed method has satisfactory accuracy.

Abstract The practical application of electric vehicle needs an accurate and robust battery management system to monitor the battery state in real-time. The maximum available capacity (MAC) and maximum available energy (MAE) need to be derived before calculating state of charge and state of energy. However, the estimation of these two parameters is a difficult task due to the complicated and comprehensive influences of temperature, aging level and discharge rate. In this paper a data-driven algorithm, least squares support vector machine, is implemented to estimate the MAC and MAE, and the influences of temperature and degradation are taken into consideration. Meanwhile, a current correction term is proposed to compensate the effect of current rate. The experimental results verify the proposed methods have excellent estimation accuracy for LiFePO4 battery.

Abstract Photovoltaic (PV) power generation can help reduce households’ electricity from the power grid and thus reduce electricity bills. However, due to the intermittence and time-varying nature of PV power generation, part of the clean energy will be wasted. Especially in some places where PV power is allowed to be sold to the power grid, the PV power that exceeds the feed-in limit will be curtailed to reduce the pressure on the infrastructure of the power grid. Battery energy storage systems (BESSs) as energy buffers have attracted increasing attention to help improve the penetration of PV power to households. This paper presents an adaptive energy management method to minimize the energy cost of residential PV-battery systems. First, the uncertainty of the predictive electricity demand and PV power supply is modeled. Then a stochastic model predictive control (SMPC) strategy is used to determine the optimal power flow of the system. Due to the deviation between the predictive input values and the actual ones, the power flow from SMPC is adjusted based on the improved correction strategy (ICS) proposed in this paper. By comparing with the other two methods (one considers the uncertainty and the other does not), the proposed method can increase the economic benefits of the system by 18% and 63%, respectively. The wasted PV power that exceeds the feed-in limit can also be reduced by 24% and 31%. This verifies the effectiveness of the proposed method to improve the system's economic benefits and self-consumption of clean energy.

Abstract The state-of-health (SOH) of a lithium-ion battery is a key parameter in battery management systems. However, current approaches to estimating the SOH of a lithium-ion battery are mainly offline or have not solved the accuracy and efficiency problems. This paper attempts to solve these problems. A dynamic information extraction method based on a fast discrete wavelet transform is proposed to greatly improve the algorithm efficiency. Dimension reduction is performed on the battery current and voltage time series using the maximum entropy partition method to individually generate a symbolic time series. A cross D-Markov machine model is built based on the causal symbolic time series to extract the feature parameter and represent the lithium-ion battery SOH. An accelerated aging experiment using LiFePO4 batteries is conducted to identify different aging stages. The results show that the feature parameter is an accurate representation of the lithium-ion battery SOH, the maximum error of SOH can be within 0.113, and the average error can be within 0.0509 in the entire battery life cycle. The proposed method is more suitable for online application than the previous method because its computation time is 250–290 times shorter.

Abstract Physics-based model has been regarded as a promising alternative to equivalent circuit model due to its ability to describe internal electrochemical states of battery. However, the rigorous physics-based model, namely pseudo-two-dimensional (P2D) model, is too complicated for online application in embedded battery management system. In this paper, to simplify the P2D model, a series of polynomial functions are employed to approximate the electrolyte phase concentration profile, solid phase concentration profile, and non-uniform reaction flux profile, respectively. Especially, the accuracy of 2nd-order and 3rd-order polynomial approximations for reaction flux is compared, and the higher-order is validated with more strength. Benefit from the acquisition of above variables, the electrolyte potential is derived directly according to the conservation of charge at electrolyte phase; the accuracy of activation overpotential is also improved by using the non-uniform reaction flux rather than assuming the uniform current density in single particle (SP) model. Finally, the developed model is simulated by different constant current rates, hybrid pulse and driving cycles, and its outputs are compared with P2D model and original SP model. The results demonstrate that the model proposed in this paper could capture the battery characteristics efficiently, and also significantly reduce the computation complexity.

Lithium-ion battery state of health (SOH) accurate prediction is of great significance to ensure the safe reliable operation of electric vehicles and energy storage systems. However, safety issues arising from the inaccurate estimation and prediction of battery SOH have caused widespread concern in academic and industrial communities. In this paper, a method is proposed to build an accurate SOH prediction model for battery packs based on multi-output Gaussian process regression (MOGPR) by employing the initial cycle data of the battery pack and the entire life cycling data of battery cells. Firstly, a battery aging experimental platform is constructed to collect battery aging data, and health indicators (HIs) that characterize battery aging are extracted. Then, the correlation between the HIs and the battery capacity is evaluated by the Pearson correlation analysis method, and the HIs that own a strong correlation to the battery capacity are screened. Finally, two MOGPR models are constructed to predict the HIs and SOH of the battery pack. Based on the first MOGPR model and the early HIs of the battery pack, the future cycle HIs can be predicted. In addition, the predicted HIs and the second MOGPR model are used to predict the SOH of the battery pack. The experimental results verify that the approach has a competitive performance; the mean and maximum values of the mean absolute error (MAE) and root mean square error (RMSE) are 1.07% and 1.42%, and 1.77% and 2.45%, respectively.

Accurate, reliable, and robust prognosis of the state of health (SOH) and remaining useful life (RUL) plays a significant role in battery pack management for electric vehicles. However, there still exist challenges in computational cost, storage requirement, health indicators extraction, and algorithm design. This paper proposes a novel dual Gaussian process regression model for the SOH and RUL prognosis of battery packs. The multi-stage constant current charging method is used for aging tests. Health indicators are extracted from partial charging curves, in which capacity loss, resistance increase, and inconsistency variation are examined. A dual Gaussian process regression model is designed to predict SOH over the entire cycle life and RUL near the end of life. Experimental results show that the predictions of SOH and RUL are accurate, reliable, and robust. The maximum absolute errors and root mean square errors of SOH predictions are less than 1.3% and 0.5%, respectively, and the maximum absolute errors and root mean square errors of RUL predictions are 2 cycles and 1 cycle, respectively. The computation time for the entire training and testing process is less than 5 seconds. This article shows the prospect of health prognosis using multiple health indicators in automotive applications.