Title: Artificial neural network based adaptive control for plug-in hybrid electric vehicles
Authors: Kavya P. Divakarla; Sanjaka G. Wirasingha; Ali Emadi; Saiedeh Razavi
Addresses: Faculty of Engineering, Department of Electrical and Computer Engineering, McMaster University, Hamilton, ON L8S 4L7, Canada ' Faculty of Engineering, Department of Electrical and Computer Engineering, McMaster University, Hamilton, ON L8S 4L7, Canada ' Faculty of Engineering, Department of Electrical and Computer Engineering, McMaster University, Hamilton, ON L8S 4L7, Canada ' Faculty of Engineering, Department of Civil Engineering, McMaster University, Hamilton, ON L8S 4L7, Canada
Abstract: Plug-in hybrid electric vehicles (PHEV) have become vital for oil consumption reduction. They have not achieved their maximum potential due to control strategy limitations. Existing controllers are often tuned to achieve the best fuel economy for specific conditions. It is impractical to optimise a controller for every scenario. A control strategy for PHEVs using artificial neural networks (ANN) is presented. The advantages of implementing a controller using ANN include independence from drive cycle or user, precision and robustness, and updatable training set. Existing PHEV control strategies are used to model a base for city and highway driving. Simulation data was extracted to form an ANN training set, which was used to develop a new strategy that was better than existing ones. The controller was validated using different drive cycles. Furthermore, the sensitivity of ANN controllers is presented. The controller is also used to validate the charge depleting mode of PHEVs.
Keywords: adaptive control; ANN; artificial neural network; dynamic programming; intelligent control; PHEV; plug-in hybrid electric vehicles; power electronics; Markov dynamic programming; optimisation and propulsion systems.
DOI: 10.1504/IJEHV.2019.099985
International Journal of Electric and Hybrid Vehicles, 2019 Vol.11 No.2, pp.127 - 151
Received: 15 Feb 2018
Accepted: 06 Aug 2018
Published online: 03 Jun 2019 *