HEV Optimal Battery State of Charge Prediction: A Time Series Inspired Approach

Abstract

Fuel efficiency in hybrid electric vehicles requires a fine balance between combustion engine usage and battery energy, using a carefully designed control algorithm. Owing to the transient nature of HEV dynamics, driving conditions prediction, have unavoidably become a vital part of HEV energy management. The use of vehicle onboard telematics for driving conditions prediction have been widely researched and documented in literature, with most of these studies identifying high equipment cost and lack of route information (for routes unfamiliar to the GPS) as factors currently impeding the commercialization of predictive HEV control using telematics. In view of this challenge, this study inspires a look-ahead HEV energy management approach, which uses time series predictors (neural networks or Markov chains), to forecast future battery state of charge, for a given horizon, along the optimal front (optimal battery state of charge trajectory). The primary contribution of this paper is a detailed theoretical appraisal and comparison of the neural network and Markov chain time series predictors over different driving scenarios (FTP72, SC03, ARTEMIS U130 and WLTC 3 driving cycles). Based on the analysis performed in this study, the following useful inferences are drawn: 1. Prediction accuracy decreases massively and disproportionately on average with increased prediction horizon for multi-input neural networks, 2. In a single-input/single-horizon prediction network, the performance of both the neural network and Markov chain predictors are similar and near optimal, with a mean absolute percentage error of less than 0.7% and a root mean square error of less than 0.6 for all driving cycles analysed, 3. Markov chains appeal as a promising time series predictor for online vehicular applications, as it impacts the relative advantage of high precision and moderate computation time.