Abstract
Hybrid electric vehicles are a compromise between traditional vehicles and pure electric vehicles and can be part of the solution to the energy shortage problem. Energy management strategies (EMSs) are highly related to energy utilization in HEVs’ fuel economy. In this research, we have employed a neuro-dynamic programming (NDP) method to simultaneously optimize fuel economy and battery state of charge (SOC). In this NDP method, the critic network is a multi-resolution wavelet neural network based on the Meyer wavelet function, and the action network is a conventional wavelet neural network based on the Morlet function. The weights and parameters of both networks are obtained by an algorithm of backpropagation type. The NDP-based EMS has been applied to a parallel HEV and compared with a previously reported NDP EMS and a stochastic dynamic programing-based method. Simulation results under ADVISOR2002 have shown that the proposed NDP approach achieves better performance than both the methods. These indicate that the proposed NDP EMS, and the CWNN and MRWNN, are effective in approximating a nonlinear system.
Highlights
Hybrid electric vehicles (HEVs) are regarded as an energy-saving solution in the vehicle industry.An energy management strategy (EMS) that coordinates the output power from an internal combustion engine (ICE) and an electric motor simultaneously is important for a HEV because Energy management strategies (EMSs) affect fuel economy and battery state of charge (SOC) directly.Many approaches have been taken to EMS, from effective but vehicle design sensitive rule-based strategies [1] to optimal control choices [2]
In the simulation test of the proposed neuro-dynamic programming (NDP)-based EMS, the detailed model of the parallel HEV is built in ADVISOR2002 (Advanced Vehicles Simulator) to accurately reflect the vehicle performance, which is based on Simulink
We have compared the proposed NDP method with the NDP method in [14] to test the proposed NDP based on common wavelet neural network (CWNN) and multi-resolution wavelet neural network (MRWNN)
Summary
Hybrid electric vehicles (HEVs) are regarded as an energy-saving solution in the vehicle industry. We can conclude that the studies mentioned above usually train one kind of neural network and use it online to predict some variables (e.g., vehicle velocity, optimal SOC curve, optimal value function, etc.). These methods are still not a real online EMS. This knowledge includes the structure of the studied parallel hybrid powertrain, the EMS optimal control problem to be solved, the topology and training method of the traditional wavelet neural network, the multi-resolution analysis theory for a function, and the topology and training approach for a MRWNN.
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
