Multi-power sources are included in hybrid electrical vehicles, which leads to multi-driving modes co-existing when driving the vehicle. However, the frequent driving mode switching (DMS) will probably need the engine to be started frequently, which can result in extra fuel consumption. So, avoiding unnecessary DMS should be fully considered when designing the control strategy. For solving this problem, a model predictive control (MPC) strategy integrating Markov chain driving intention identification is put forward. First, the component models of the powertrain system are established. Second, according to the real driving cycle data, a driving intention model based on the Markov chain is designed according to the real driving cycle data. Then the MPC-based control strategy aiming at reducing DMS times is proposed by integrating the cost of DMS. Finally, the proposed control strategy is contrasted with three other control strategies to verify its validity in reducing the mode switching times and improving the fuel economy.

中文翻译：

---

基于模型预测控制的控制策略减少并联混合动力汽车驾驶模式切换次数

混合动力汽车中包含多电源，这导致在驾驶车辆时多种驾驶模式并存。但是，频繁的驾驶模式切换（DMS）可能需要频繁启动发动机，这会导致额外的油耗。因此，在设计控制策略时应充分考虑避免不必要的DMS。针对这一问题，提出了一种结合马尔可夫链驾驶意图识别的模型预测控制（MPC）策略。首先，建立动力总成系统的部件模型。其次，根据真实驾驶工况数据，根据真实驾驶工况数据设计基于马尔可夫链的驾驶意图模型。然后通过整合DMS的成本，提出了以减少DMS次数为目标的基于MPC的控制策略。