Accurate remaining useful life (RUL) prediction of lithium-ion batteries is critical for energy supply systems. In conventional data-driven RUL prediction approaches, the battery's degradation mechanism is difficult into incorporate in the RUL prediction. Furthermore, there are notable limitations in reflecting the significance of different time instances, and the uncertainty in the degradation process. Consequently, a novel data-model interactive RUL prediction approach based on particle filter-temporal attention mechanism-bidirectional gated recurrent unit (PF-BiGRU-TSAM) is proposed. Specifically, BiGRU-TSAM is trained offline through historical data, which assigns corresponding significance to battery capacities at different time instances. Moreover, regarding the interactive data-model for the online prediction phase based on PF-BiGRU-TSAM, the advantages of data-driven and model-based approaches are integrated, which accomplishes the purpose of modifying each other. The proposed PF-BiGRU-TSAM approach is validated with a real-world battery dataset. Experimental results demonstrate the proposed approach is better than some published approaches. Taking the 50th operational cycle of the four batteries B0005, B0006, B0007, and B0018 in the dataset as an instance, the absolute errors of the proposed PF-BiGRU-TSAM are 0, 1, 3, 3, respectively, which represents the proposed approach has an excellent performance.

Safe and reliable management of rechargeable lithium-ion batteries relies on accurate state-of-charge (SOC) estimation. SOC estimation algorithms developed based on the conventional state-space model can be inaccurate when the slope of the battery SOC to open-circuit voltage (V _OC ) profile is close to zero, where the observability matrix of the conventional battery state-space model is ill-conditioned. This ill-conditioned observability issue is usually overlooked and results in unreliable SOC estimations. This article proposes a robust state-space model using a set of carefully selected states, making the proposed model well-conditioned over the full SOC range and insesitive to the slope of the SOC-V _OC profile. The proposed model, therefore, solves the observability issue and improves the SOC estimation accuracy. To examine the performance of the proposed state-space model over the full SOC range, the battery is discharged using a standard electric vehicle driving profile. Experiments are further carried out in various environmental temperature conditions to validate the robustness of the proposed model. The results show that the proposed state-space model is robust against the change of the slope of the SOC-V _OC profile and significantly improves the SOC estimation accuracy, rendering more than 30% improvement in various temperature conditions.

Safely and efficiently managing a battery pack consisting of hundreds to thousands of battery cells is a critical but challenging task due to commonly observed uncertainties, e.g. temperature, battery degradation and SOC estimation inaccuracy. This paper proposes a robust and efficient most significant cell methodology that estimates the battery pack SOC depending on the determined most significant cells. The estimation adopting this methodology is robust to variations of temperature, battery degradation and battery cell SOC estimation inaccuracy. A battery pack simulator and a real battery pack designed for electric vehicles were used as prototypes to illustrate the high performance, robustness and effectiveness of the proposed methodology. Moreover, the proposed algorithm requires light computational effort, making it suitable for real-time operation.

Safely and reliably managing the lithium-ion batteries in electric vehicles for daily operations relies on accurate state-of-charge estimation. To pursue accurate state-of-charge estimation, battery electric circuit model-based algorithms demand minimized modeling error from accurate parameter identifications. Typical online parameter identification algorithms: least square fitting approach and recursive least square fitting are commonly adopted but have their limitations and issues with continuously providing accurate identification results. To overcome their problems and utilize their strengths, this paper proposes a condition-based parameter identification switching algorithm. The proposed algorithm accurately and robustly identifies the lithium-ion batteries' parameters based on the battery data condition and the adoption of Bollinger bands. By minimizing the modeling error, the accuracy of the state-of-charge estimation is thus improved. The electric vehicle simulation results indicate that the proposed algorithm can improve the parameter identification accuracy by an average of 70.80% and 44.95% reduction in the mean absolute parameter identification error, compared to typical least square fitting and recursive least square fitting approaches. With the help of the minimized modeling error, the state-of-charge estimation accuracy can thus be improved, rendering an average of 62.44% and 21.49% improvement.

Online state-of-charge estimations are critical to lithium-ion batteries in electric vehicles for safe and reliable operations. To ensure accurate online state estimations, an online battery parameter identification algorithm, such as the least square fitting approach, along with a shifting moving window, was developed and implemented, while the decision of the moving window size for general purpose is a technical issue not yet been well-addressed. This paper proposes an adaptive moving window size design methodology that adjusts the moving window size considering three important factors: load profile, state-of-charge to open-circuit voltage profile, and condition of data. The simulation results show that the proposed methodology helps to identify the parameters of the battery electric circuit model with high accuracy, where the results are compatible to the results using fine-tuned fixed window sizes. Also, the proposed methodology is general to driving profiles from moderate to severe without extra algorithm redesign. In the three standard driving profile cases, the parameter identification error is reduced by 59.15%, 69.02%, and 75.14%, respectively, compared to the worst case in experiments, which are compatible with the results after fine-tuning the fixed window sizes. With accurately identified parameters, the modeling error is minimized, and three typical linear SOC estimation algorithms thus render accurate SOC estimation, rendering an average of 13.4%, 64.43% 57.48% reduction from the worst case in experiments, respectively.