Vehicle performance such as fuel consumption and catalyst-out emissions is affected by a driving pattern, which is defined as a driving cycle with grades in this study. To optimize the vehicle performances on a temporary driving pattern, we developed a multi-mode driving control algorithm using driving pattern recognition and applied it to a parallel hybrid electric vehicle (parallel HEV). The multi-mode driving control is defined as the control strategy which switches a current driving control algorithm to the algorithm optimized in a recognized driving pattern. For this purpose, first, we selected six representative driving patterns, which are composed of three urban driving patterns, one expressway driving pattern, and two suburban driving patterns. A total of 24 parameters such as average cycle velocity, positive acceleration kinetic energy, stop time/total time, average acceleration, and average grade are chosen to characterize the driving patterns. Second, in each representative driving pattern, control parameters of a parallel HEV are optimized by Taguchi method though the fuel-consumption and emissions simulations. And these results are compared with those by parametric study. There are seven control parameters, six of them are weighting factors of performance measures for deciding the ratio of engine power to required power from driving load. And the other is the charging/discharging method of battery. Finally, in driving, a neural network (the Hamming network) decides periodically which representative driving pattern is closest to a current driving pattern by comparing the correlation related to 24 characteristic parameters. And then the current driving control algorithm is switched to the optimal one, assuming the driving pattern does not change in the next period.

The energy management strategy of a hybrid electric vehicle directly determines the fuel economy of the vehicle. As a supervisory control strategy to divide the required power into its multiple power sources, engines and batteries, many studies have been conducting using rule-based and optimization-based approaches for energy management strategy so far. Recently, studies using various machine learning techniques have been conducted. In this paper, a novel control framework implementing Model-based Q-learning is developed for the optimal control problem of hybrid electric vehicles. As an online energy management strategy, a new approach could learn the characteristics of a current given driving environment and adaptively change the control policy through learning. Especially, for the proposed algorithm, the internal powertrain environment and external driving environment are separated so they can be learned via the reinforcement learning framework, which results in a simpler and more intuitive control strategy that can be explained using the vehicle state approximation model. The proposed algorithm is tested and verified through simulations, and the simulation results present near optimal solution. The simulation results are compared with conventional rule-based strategies and optimal control solutions acquired from Dynamic Programming.INDEX TERMS Hybrid electric vehicle, optimal control, power management, Q-learning, reinforcement learning.

No abstract

The multimode electrically variable transmissions (EVTs) are an advancement of the conventional planetary gear hybrid powertrains (PGHPs). Using the analytical methodologies once applied to the single-mode PGHP, the dual-mode PGHP was investigated. The transmission efficiency and the system optimal operation point maximizing the system overall efficiency were analysed. With the system optimal operation points, the fuel economies of the dual-mode PGHP vehicle were calculated assuming constant-velocity cruising and NEDC mode driving. The top velocity and the acceleration performance were also estimated. To assign appropriate parameters for the new system, the design space and the optimization problem were defined using the vehicle performances from the analyses. An approach with various optimization techniques is proposed, and the optimal parametric design is obtained. The dual-mode PGHP showed improvements in efficiency and dynamic performances compared with the singlemode PGHP. In particular, the powertrain is expected to have advantages because it uses a smaller-sized motor/generator (MG) or two identical machines as two MGs.

No abstract