Trip-Oriented Model Predictive Energy Management Strategy for Plug-in Hybrid Electric Vehicles

Abstract

Energy management strategies play a critical role in performance optimization of plug-in hybrid electric vehicles (PHEVs). In order to attain effective energy distribution of PHEVs, a predictive energy management strategy is proposed in this study based on real-time traffic information. First, an exponentially varied model for the velocity prediction is established, of which the tunable decay coefficient is regulated by the supported vector machine (SVM). In this manner, the prediction precision is improved. Then, by properly simplifying the powertrain model, the state of charge (SOC) reference trajectory is generated based on the fast dynamic programming (DP) with fast calculation speed and consideration of the traffic information. Moreover, the typical DP algorithm is leveraged to solve the nonlinear rolling optimization problem for minimizing the operating cost in a receding horizon. Simulation results demonstrate that the proposed algorithm can reach 92.83% operating savings, compared with that of the traditional DP; and save 6.18% cost compared with the MPC algorithm only with reference of the trip duration. INDEX TERMS Plug-in hybrid electric vehicles (PHEVs), energy management, model predictive control (MPC), traffic information, state of charge (SOC) reference.

A review on hybrid source energy management strategies for electric vehicle

“…Total operating fuel cost is minimized for the increase in traffic update frequency. 163 Zhai et al proposed a predictive model which predicts the velocity based on the pattern sequences. A power split optimization problem for the battery-SC hybrid sources is developed in which battery capacity loss and energy loss are considered.…”

A survey on key techniques and development perspectives of equivalent consumption minimisation strategy for hybrid electric vehicles

“…In terms of future driving conditions, it has been proven that the future road information, speed prediction, and driver styles are critical to the control performance of the EF online adaptation. These key parameters are obtainable by different methods, such as the support vector machine (SVM) for the vehicle speed prediction [20], geographic information system (GIS) for the terrain preview [21],…”

“…More complex methods, combining DP with simplified efficiency models of engine and motor, can be implemented to generate the SOC reference trajectory in each step of prediction horizon [20,64]. The proposed method can provide a SOC reference trajectory that highly correlates to the one generated by global optimal strategy.…”

“…The simulation results, shown in Figure 9, show that the predicted velocity sequences of NN-based speed predictor can capture the trend of the actual vehicle speed better than the others, thereby achieving the better fuel economy improvement. Lei et al [20] improved the performance of exponential prediction by using the support vector machine (SVM) according to the road type. It can dynamically regulate the key factor of the exponential prediction, referred to as the decay coefficient, to improve the prediction precision of the future speed.…”

An adaptive equivalent consumption minimization strategy for plug-in hybrid electric vehicles based on traffic information

“…The traffic information in the future mileage obtained by ITS is processed to build the approximate global driving condition. On this basis, the global reference SOC trajectory can be determined through the simplified DP, which is explained in detail in our previous research[47]. In short, the simplified DP is achieved by fitting the efficiency model of engine and…”