Real-time and hierarchical energy management-control framework for electric vehicles with dual-motor powertrain system

To achieve a real-time optimization of the economic and dynamic performance for electric vehicles equipped with the dual-motor powertrain system, this study proposed a hierarchical energy management-control framework to establish a collaborative relationship between the decision and the control layers. To be specific, the action-dependent heuristic dynamic programming is employed to obtain the optimal energy management strategy and control coefficient matrix for the control layer in real time. However, due to the unpredictability of the dynamic process, the modes shift costs are largely uncertain in the decision function which reduces the control precision. To improve the accuracy and efficacy of the proposed framework, the all-cost matrix for the dynamic process is collected by the complete experiment data. Intriguingly, the general shift regularity suitable for the multi-motor configuration has been discovered, revealing the energy cost distribution. As the test scenario, cycle following experiment and real-world cycle are employed the assess the performance of the various approach. Finally, actual vehicle experimental results demonstrate that the proposed framework significantly outperforms rule-based strategies in real-time applications, which can reduce the energy consumption and average shock by 7.7% and 12.6%. Furthermore, due to the all-cost matrix, the framework can effectively avoid 9.47% of calculation error.