Wavelet transform‐based real‐time energy management strategy of hybrid energy storage system for electric vehicle

Jie Jiang,Chun Wang,Qiang Zhang,Yongzhi Zhang,Quanqing Yu

DOI: https://doi.org/10.1002/cta.3822

IF: 2.378

2023-10-02

International Journal of Circuit Theory and Applications

Abstract:In order to give full play to the characteristic of power sources in the HESS, a wavelet transform‐based real‐time EMS framework is proposed, which ensures the battery cannot be negatively affected by high frequency and peak currents to extend battery lifespan. A prediction model consisting of four neural networks is proposed to enable real‐time application of wavelet transform‐based EMS. Four neural networks are trained by different power component data obtained from wavelet transform decomposition, and they can predict demand power of different frequencies online. Based on the HESS models established at different temperatures, the WTFLCS is further verified through the simulation experiment at 10°C, 25°C, and 40°C to prove the robustness against temperature. Summary The peak and transient components of demand power caused by the complex and variable traffic environment could induce the accelerated degradation of the battery lifespan for electric vehicle (EV). This paper proposes a wavelet transform‐based real‐time energy management strategy (EMS) to fully exploit the advantages of the hybrid energy storage system (HESS). First, to adapt the characteristics of battery and ultracapacitor, wavelet transform is employed to decompose driving cycle into high frequency power and low frequency power. Second, since the wavelet transform is difficult to be directly implemented in real‐time control, a power prediction model including four neural network predictors is established, which are trained by the data obtained from wavelet transform to online predict the power of different frequencies. Third, the high frequency power and the low frequency power are distributed to the battery and ultracapacitor by fuzzy logic control (FLC) algorithm, which can significantly reduce the damage caused by current rapid changes into the battery. Accordingly, the battery lifespan is effectively extended because it significantly avoids the impact of rapidly changing and peak current. Finally, simulation results verify the effectiveness of wavelet transform‐based real‐time EMS, and the proposed EMS robustness against temperature is verified at different temperatures. Compared with the wavelet transform strategy (WTS), the peak currents of wavelet transform‐fuzzy logic control strategy (WTFLCS) are decreased by 26.07%, 25.66%, and 25.85% at 10°C, 25°C, and 40°C, respectively.

engineering, electrical & electronic