A Reinforcement Learning Based Power Assisted Method with Comfort of Riding for Light Electric Vehicle

In this study, a reinforcement learning based power assisted method is proposed for pedelec, which is a Light Electric Vehicle (LEV) driven by a human's pedal force and electric motor. The proposed method adaptively chooses appropriate assisted power according to the environmental changes. By adaptively adjusting the assisted power after learning for pedelec, not only energy utilization is improved but comfort of riding for pedalist is also satisfied. Simulations of the proposed method on pedelec powered by battery are performed. Experimental results demonstrate that the proposed method satisfies the requirement of comfort of riding and achieve better energy utilization by comparing with other conventional power assisted methods.

[1]  H. Oman,et al.  Power control for battery-electric bicycles , 1993, Proceedings of the IEEE 1993 National Aerospace and Electronics Conference-NAECON 1993.

[2]  Kao-Shing Hwang,et al.  Reinforcement Learning in Strategy Selection for a Coordinated Multirobot System , 2007, IEEE Transactions on Systems, Man, and Cybernetics - Part A: Systems and Humans.

[3]  Ethem Alpaydin,et al.  Introduction to machine learning , 2004, Adaptive computation and machine learning.

[4]  Roy Chaoming Hsu,et al.  Adaptive Power Management Based on Reinforcement Learning for Embedded System , 2008, IEA/AIE.

[5]  Peng-Yeng Yin,et al.  Maximum entropy-based optimal threshold selection using deterministic reinforcement learning with controlled randomization , 2002, Signal Process..

[6]  W. C. Morchin.  Battery-powered electric bicycles , 1994, Proceedings of NORTHCON '94.

[7]  Richard S. Sutton,et al.  Reinforcement Learning: An Introduction , 1998, IEEE Trans. Neural Networks.

[8]  A. Muetze,et al.  Electric bicycles - A performance evaluation , 2007, IEEE Industry Applications Magazine.